Fluent Essays - An annotated bibliography of at least five additional resources related to the research topic by Day 7 of Week 2.***No plagiarism*** It must be less than 30\% match on authenticity report.***PLEASE DO

AN ANNOTATED BIBLIOGRAPHY OF AT LEAST FIVE ADDITIONAL RESOURCES RELATED TO THE RESEARCH TOPIC BY DAY 7 OF WEEK 2.NO PLAGIARISM IT MUST BE LESS THAN 30\% MATCH ON AUTHENTICITY REPORT.***PLEASE DO

An annotated bibliography of at least five additional resources related to the research topic by Day 7 of Week 2.No plagiarism It must be less than 30\% match on authenticity report.***PLEASE DO - Business & Finance

PLACE ORDER

An annotated bibliography of at least five additional resources related to the research topic by Day 7 of Week 2.***No plagiarism*** It must be less than 30\% match on authenticity report.***PLEASE DO NOT ACCEPT THIS ASSIGNMENT IF YOU CANNOT MEET THE DEALINE OF 11:00 PM ON 9/17/2018 CST.***Assignment: Conducting a***No plagiarism*** It must be less than 30\% match on authenticity report.Assignment must look exactly like the attached sample templatearticles and sample are attached An annotated bibliography of at least five additional resources related to the research topic by Day 7 of Week 2.***No plagiarism*** It must be less than 30\% match on authenticity report.***PLEASE DO 1 Sample Annotated Bibliography Student Name Here University Sample Annotated Bibliography Autism research continues to grapple with activities that best serve the purpose of fostering positive interpersonal relationships for children who struggle with autism. Children have benefited from therapy sessions that provide ongoing activities to aid autistic children’s ability to engage in healthy social interactions. However, less is known about how K–12 schools might implement programs for this group of individuals to provide additional opportunities for growth, or even if and how school programs would be of assistance in the end. There is a gap, then, in understanding the possibilities of implementing such programs in schools to foster the social and thus mental health of children with autism. Annotated Bibliography Kenny , M. C., Dinehart, L. H., & Winick, C. B. (2016). Child-centered play therapy for children with autism spectrum disorder. In A. A. Drewes & C. E. Schaefer (Eds.), Play therapy in middle childhood (pp. 103–147). Washington, DC: American Psychological Association. In this chapter, Kenny, Dinehart, and Winick provided a case study of the treatment of a 10-year-old boy diagnosed with autism spectrum disorder (ADS). Kenny et al. described the rationale and theory behind the use of child-centered play therapy (CCPT) in the treatment of a child with ASD. Specifically, children with ADS often have sociobehavioral problems that can be improved when they have a safe therapy space for expressing themselves emotionally through play that assists in their interpersonal development. The authors outlined the progress made by the patient in addressing the social and communicative impairments associated with ASD. Additionally, the authors explained the role that parents have in implementing CCPT in the patient’s treatment. Their research on the success of CCPT used qualitative data collected by observing the patient in multiple therapy sessions . CCPT follows research carried out by other theorists who have identified the role of play in supporting cognition and interpersonal relationships. This case study is relevant to the current conversation surrounding the emerging trend toward CCPT treatment in adolescents with ASD as it illustrates how CCPT can be successfully implemented in a therapeutic setting to improve the patient’s communication and socialization skills. However, Kenny et al. acknowledged that CCPT has limitations—children with ADS, who are not highly functioning and or are more severely emotionally underdeveloped, are likely not suited for this type of therapy . Kenny et al.’s explanation of this treatments’s implementation is useful for professionals in the psychology field who work with adolescents with ASD. This piece is also useful to parents of adolescents with ASD, as it discusses the role that parents can play in successfully implementing the treatment. However, more information is needed to determine if this program would be suitable as part of a K–12 school program focused on the needs of children with ASD . Stagmitti, K. (2016). Play therapy for school-age children with high-functioning autism. In A.A. Drewes and C. E. Schaefer (Eds.), Play therapy in middle cildhood (pp. 237–255). Washington, DC: American Psychological Association. Stagmitti discussed how the Learn to Play program fosters the social and personal development of children who have high functioning autism. The program is designed as a series of play sessions carried out over time, each session aiming to help children with high functioning autism learn to engage in complex play activities with their therapist and on their own. The program is beneficial for children who are 1- to 8-years old if they are already communicating with others both nonverbally and verbally. Through this program, the therapist works with autistic children by initiating play activities, helping children direct their attention to the activity, eventually helping them begin to initiate play on their own by moving past the play narrative created by the therapist and adding new, logical steps in the play scenario themselves. The underlying rationale for the program is that there is a link between the ability of children with autism to create imaginary play scenarios that are increasingly more complex and the development of emotional well-being and social skills in these children. Study results from the program have shown that the program is successful: Children have developed personal and social skills of several increment levels in a short time. While Stagmitti provided evidence that the Learn to Play program was successful, she also acknowledged that more research was needed to fully understand the long-term benefits of the program. Stagmitti offered an insightful overview of the program; however, her discussion was focused on children identified as having high-functioning autism, and, therefore, it is not clear if and how this program works for those not identified as high-functioning. Additionally, Stagmitti noted that the program is already initiated in some schools but did not provide discussion on whether there were differences or similarities in the success of this program in that setting. Although Stagmitti’s overview of the Learn to Play program was helpful for understanding the possibility for this program to be a supplementary addition in the K–12 school system, more research is needed to understand exactly how the program might be implemented, the benefits of implementation, and the drawbacks. Without this additional information, it would be difficult for a researcher to use Stigmitti’s research as a basis for changes in other programs. However, it does provide useful context and ideas that researchers can use to develop additional research programs. Wimpory, D. C., & Nash, S. (1999). Musical interaction therapy–Therapeutic play for children with autism. Child Language and Teaching Therapy, 15(1), 17–28. doi:10.1037/14776-014 Wimpory and Nash provided a case study for implementing music interaction therapy as part of play therapy aimed at cultivating communication skills in infants with ASD. The researchers based their argument on films taken of play-based therapy sessions that introduced music interaction therapy. To assess the success of music play, Wimpory and Nash filmed the follow-up play-based interaction between the parent and the child. The follow-up interactions revealed that 20 months after the introduction of music play, the patient developed prolonged playful interaction with both the psychologist and the parent. The follow-up films also revealed that children initiated spontaneously pretend play during these later sessions. After the introduction of music, the patient began to develop appropriate language skills. Since the publication date for this case study is 1999, the results are dated. Although this technique is useful, emerging research in the field has undoubtedly changed in the time since the article was published. Wimpory and Nash wrote this article for a specific audience, including psychologists and researchers working with infants diagnosed with ASD. This focus also means that other researchers beyond these fields may not find the researcher’s findings applicable. This research is useful to those looking for background information on the implementation of music into play-based therapy in infants with ASD. Wimpory and Nash presented a basis for this technique and outlined its initial development. Thus, this case study can be useful in further trials when paired with more recent research. An annotated bibliography of at least five additional resources related to the research topic by Day 7 of Week 2.***No plagiarism*** It must be less than 30\% match on authenticity report.***PLEASE DO Available online at http://www.anpad.org.br/bar BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46, Jan./Mar. 2014 Productivity Spillovers from Foreign Direct Investment in the Brazilian Processing Industry Nádia Campos Pere ira Bru hn E-mail address: [email protected] Universidade Federal de Goiás - UFG Departamento de Administração -UFG, Campus Catalão, 75704 -020, Catalão, GO, Brazil. Cristina Lelis Leal Calegario E-mail add ress: [email protected] Universidade Federal de Lavras – DAE/UFLA Universidade Federal de Lavras, Campus UFLA, Caixa Postal 3037, 37200 -000, Lavras, MG, Brazil. Received 27 September 2012; received in revised form 30 April 2013 (this p aper has been with the authors for two revisions); accepted 7 May 2013 ; published online 2 nd January 2014. Productivity Spillovers from Foreign Direct Investment 23 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar Abstract The increasing importance of foreign direct investment (FDI) to international production has prompted considerable interest in its real ef fects on host economies all over the world. The aim of this study was investigate whether the presence of FDI produces productivity spillovers in Brazilian processing industries . We conduct our analysis using a panel database on twenty -three Brazilian proc essing industries and applied Moderated Multiple Regression (MMR) and Generalized Linear Models (GLM) analysis of variance to address potential spillover effects from foreign presence. This paper finds evidences of the coexistence of both positive and nega tive effects arising from FDI on the productivity of Brazilian industries. We found n egative effects for FDI presence in labor -intensive industries . Furthermore, FDI benefits depend on the absorptive capacity of industries , confirming the hypothesis that a minimum level of absorptive capacity is required so that locally owned enterprises (LOEs) can benefit from foreign presence . Key words : foreign direct investment ; productivity spillovers ; manufacturing Industry ; Generalized Linear Models ; absorption cap acity . N. C. P. Bruhn, C. L. L. Calegario 24 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar Introduction Although the attitudes towards the impacts foreign direct investment (FDI) has on the host economy have been mixed , many governments around the world actively attempt to attract FDI to their countries using substantial fiscal and fi nancial incentives. One of the reasons for these policy interventions is the belief that locally owned enterprises (LOEs) can benefit from the foreign owned enterprises (FOEs) through productivity spillovers (Görg & Gre enaway, 2004 ). Productivity spillover discussion is part of a broad debate on the effects arising from FDI’s inflows and the presence of multinational corporations (MNCs) in the host economies (Blomström & Kokko, 1998 ; Buckley, Clegg , & Wang, 2010 ; Findlay, 1978 ; Markusen & Venables, 1999 ). FDI and multinational corporations (MNCs) are subtly different facets of the international phenomenon, but are not perfect ly synonymous ( Cohen, 2007 , p. 36). A widely accepted concept of MNCs in academic and business circles consider s them an enterprise that engages in FDI activities and “owns or, in some way, contr ols value -added activities in more than one country”. MNCs account for almost all FDI flows ( Kupfer & Hasenclever, 2002 , p. 391 ) and are the main cause of the major changes in the way that business is conducted throughout the wor ld ( Cohen, 2007 ). Productivity spillover theory is based on the argument that FDI occurrence requires MNCs to be more efficient than their indigenous counterparts operating in the same location ( Buckley, Clegg , & Wang, 2010 ). So, firm -specific assets, such as marketing and management capabilities , technological know -how and reputation, that play important role s in Dunning ’s traditional Eclectic FDI theory (2000, 2008) are fundamental to th e argument that MNC ownership advantages should lead to relativity higher performances than their counterparts. This notion of “performance differentials is the basis for the general hypothesis that FDI generate s productivity spillovers ” (Buckley, Clegg , & Wang, 2010 , p. 217). Blomström and Kokko (1998) comments that, when MNCs establish a subsidiary in certain countries, they bring a series of new knowledge and technologies that can s pill over for LOEs, resulting in competitiveness increases and productivity gains, known as productivity spillovers. This study aims to answer the following research question: Are there productivity spillover effects from FDI received by the Brazilian proc essing industries? Our intention is to contribute to the discussion on FDI ’s real impacts , assessing whether the FDI received by the Brazilian processing industries contributed to their performance. In other words, the main objective of this research was to investigate if there were productivity spillovers effects from FDI received by Brazilian processing industries. Specifically we aimed to: (a) investigate the effects of a set of variables representing the characteristics of the industry and characteristi cs of the country on industry productivity; and ( b) investigate the moderating effect of FDI on industry productivity. We also conduct our analysis using a panel database of twenty -three Brazilian processing industries. We applied the Moderated Multiple Re gression (MMR) and Generalized Linear Models (GLM) analysis of variance to address potential spillover effects from foreign presence. Our main purpose in this study is not to identify with either the pro or con schools of thought on the subject, but to sh are our belief that both sides have made valid points on the subject. The paper highlights, as argued by Cohen (2007) , that each corporation and each industry or country is a special case and that FDI is an extremely complex an d heterogeneous phenomenon. It’s not our intention to reach a conclusion to the question of whether these phenomena are good or bad, but to add to our far from comprehensive knowledge of what are FDI effects and how they really affect the productivity of Brazilian industries. Only a few studies have considered how the relationship between foreign presence and spillover benefits change as inward FDI to Brazilian industries rises , and little attention has been given to the conditions under which spillover mig ht be l arger, non -existent or negative (Cohen, 2007 ). The solid understanding of the FDI role on the host economy is vital not only for researchers but also policymakers and managers interested in understanding how FDI inflows influence industry Productivity Spillovers from Foreign Direct Investment 25 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar performance. Given their dynamicity and complexity, FDI and MNC phenomena have introduced extraordinary and perhaps revolutionary changes that have profoundly altered the global economy. Policymakers in particular need information that provide s them with the necessary tools for the decision making process. They influence the regulatory regime in which MNCs and LOEs are embedded and , therefore , need to understand how policy instruments can induce or control MNC action s so that they can o ffer benefits to LOEs. MNC effects on host economies is also relevant for managers. First of all , positive effects of spillovers can be used to build a reputation , since compan ies are concerned with stakeholders. Second, recognizing complementary interests and areas of conflict helps identify strategies that benefit both MNCs and LOEs in host economies. Our main contribution is not providing a definite explanation on the issue. It is, however, just a step in a long journey to a more accurate understanding o n the subject. There still remains a need for continuing the research and data accumulation in this field ( Cohen, 2007 ). This paper ’s structure is organized as follows: the next section presents the evolution of FDI inflows to the Brazilian economy. Third s ection presents the theory of spillovers from FDI , while fourth section clarifies the conceptualization and operationalization of FDI spillover determinant factors , which is followed by a presentation of our methods and data in fifth s ection . Sixth s ection presents the empirical results and seventh s ection presents the discussions. The last section offers final considerations. Foreign Direct Investment Trajectory Until the Second World War, only a small portion of capital movement was related to FDI. FDI global flows had suffered a slump in the 1970s related mainly to the oil shock and macroeconomic crises. The decline in FDI global flows in the 1970s and 1980s was interrupted by a reaction in the early 1990s, when FDI flow s became really significant. In 1990, FDI flows worldwide were approximately $200 billion, reaching the ir maximum level at the end of the decade, in 2000, when flows reached $ 1.4 trillion (Figure 1). After this global boom, FDI flows fell to $651 billion in 2002, approximately half the value that was reached in the peak period in 2000 ( United Nations Confe rence on Trade and Development [UNCTAD ], 2003 ). N. C. P. Bruhn, C. L. L. Calegario 26 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar Figure 1. FDI Flows Worldwide and by Groups of Economies , 1980 -2006 (billions of dollars). Source : United Nations Conference on Trade and Development. (2007 , outubro) . Crescimento pulverizado do investimento direto estr angeiro em 2006. Anais da Conférence des Nations Unies Pour le Commerce et le Développement , Geneva, Switzerland . FDI global flows began to increase again in 2003 and in 2006 accumulated $1 .306 billion, an increase of over 38\% compared to the previous year , reaching levels close to those reached in 2000 (UNCTAD, 2007 ). FDI inflows increased in all three groups of economies , reflecting the propensity of growth of multinational corporations (MNCs) and f avorable economic performance in many parts of the world. While in developed countries FDI reached $857 billion in 2006 ( a growth of 45\% over the previous year), the flow reached its highest levels in developing countries and economies in transition. The f lows in developing countries accumulated $379 billion, representing an increase of 21\% over 2005, while the flows to transition economies reached $69 billion , representing an increase of 68\% compared to 2005 ( UNCTAD, 2007 ). The major sources of FDI were MNCs from developed countries, especially the European Union (EU). MNCs from developing countries and economies in transition continued their international expansion , led by China (UNCTAD, 2007 ). The flows of FDI in to Brazil started primarily during the 1955 – 1960 period, when specific governmental programs were created to attract foreign capital as strategy for industrial development through imp ort -substitution industrialization. In the 1970 s, the amount of capital went down , mainly due to the oil shock associated with macroeconomic crisis. The 1970s were characterized by a large FDI inflow in to the Brazilian economy . The main determinants of the FDI supply abundance were related to economic growth orientation and a non - discriminatory foreign capital police consolidation . D uring the 1980s , there was a reversal of capital flows , essentially from the lack of credibility due to non -accomplishment of external obligations , economic instability and increased uncertainty associated with anti -inflationary plans . Starting in the 1990s there was an extraordinary recovery of FDI flow growth , reflecting the financial globalization effects and mergers and acqu isitions (M&A) possibilities due to the opening and privatization of the Brazilian economy (Fernandes & Campos, 2008 ). The Brazilian economy experienced a boom in FDI flows in 2000 . After this period , FDI flows to the Brazilia n economy decreased , following the world’s FDI behavior , but also reflecting the inexpressive Brazilian economy ’s growth and the end of the privatization phase that marked the 90s. In 2004, there was a reaction to FDI inflows and, ac cording to the United N ations Con ference on World Developed Economies Developin g Economies Western Europe and Commonwealth Countries Billions of dollars 0 200 400 600 800 1000 1200 1400 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Bilhões de dólares Productivity Spillovers from Foreign Direct Investment 27 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar Trade and Developmen (UNCTAD, 2007 ), FDI to the Brazilian economy ha d the highest increase rate in the world in 2007 (from $18.8 billion in 2006 to $37.4 billion , representing an increase of 99.3 \%). This new record surpassed 2000, when FDI inflows reached $32.8 billion, and 22 \% of the total amount of FDI inflows were related to privatization operations . The new record occurred even without privatization operations , reinforcing the significance of the record reached in 2007 . The upward movement of FDI in the Brazilian economy in recent years occurred simultaneously with appreciation of the Real (R$, the Brazilian currency) , which might seem paradoxical, since domestic currency appre ciation makes Brazilian assets more expensive, as well as the cost of inputs, raw materials and components needed for multinational subsidiaries operation maintenance. However, w hat is observed is that Real appreciation didn’t affect FDI inflows as one mig ht imagine . In fact , Real appreciation in recent years is not a result of passing situational factors , but rather sustained improvement in Brazilian fundamentals that fosters predictability over the longer - term horizon (Sociedade Brasileira de Estudos de Empresas Transnacionais e da Globalização Econômica [Sobeet ], 2007 ). Theory of Spillover from Foreign Direct Investment Productivity spillover effects from inward FDI are usually defined as “the in fluence of the presence of foreign -owned enterprises (FOEs) on productivity of LOEs” (Buckley, Clegg , & Wang, 2010 , p. 193). They are generated by non -market transactions involving MNCs, in particular when knowledge is spread to LOEs of the host country without a contractual relationship ( Buckley, Clegg , & Wang, 2010 ; Meyer, 2004 ). In the literature on FDI spillovers, it is argued that MNCs that establis h subsidiaries in other countries are different from LOEs in the receiving economy market for two main reasons. The first is that they bring with them superior knowledge about foreign markets (Caves, 1971 ) and certain technolog ical properties that constitute their specific advantages that allow them to compete with other MNCs and local firms that usually have better knowledge of local market and consumer preference s (Aitken, Harrisson , & Lipsey 1996 ; Blomström & Kokko, 1998 ). The second reason is that the entry or presence of MNCs alters existing market equilibrium s, forcing local firms to become more efficient to protect their market shares and profits ( Blomström & Kokko, 1998 ). Firm -specific assets, such as marketing and management capabilities , technological know -how and reputation, that play important role s in Dunning ’s tradition al Eclectic theory of FDI (Dunning, 2000 ; Dunning & Lundan, 2008 ), are fundamental to the argument that MNC ownership advantages should lead to relativity higher performances than their counterparts. This notion of performance differentials is th e basis for the general hypothesis that FDI generate s productivity spillovers ( Buckley, Clegg , & Wang, 2010 ; Thang, 2011 ). According to Caves (1974) , the benefic ial effects of FOEs can be summarized in terms of : (a) allocative efficiency gains that arise from pro -competitive effects; ( b) technical efficiency improvements from demonstration of superior practices; and ( c) technology transfer when the presence of FOE s furnishes LOEs with access to advanced technology. Therefore, spillover effects can be reflected in improved productivity and other benefits in LOEs ( Buckley, Clegg , & Wang, 2010 ). Spillover theory states that indu stry and country specificities have a strong relationship with spillover occurrence. It suggests that there is a high degree of heterogeneity across industries due to learning capabilities and technologies absorbing capacity differences. Moreover, changes in macroeconomic policies have different impacts across industries. In this sense, best performing industries offer better innovations and new knowledge absorption conditions that make LOEs in a host country more competitive. If LOEs are able to offer stro ng competitive relationships with MNCs, then N. C. P. Bruhn, C. L. L. Calegario 28 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar they will find themselves forced to constantly change their techniques, transferring their technologies more frequently than could be imitated by LOEs (Gachino, 2010 ). However, a num ber of negative spillovers effects have been put forward recently with the argument that negative effects start to become apparent when greater levels of foreign presence begin to counteract the positive effects on LOE productivity. According to Aitken and Harrisson (1999) , this may occur since FOE affiliates may draw demand away from their counterparts through the introduction of new products and innovation processes leading to price reductions ( Ai tken & Harrison, 1999 ). For this reason, there are some potential costs associated with FDI , such as the emergence of concentrated market structures that may compromise the development of competitive markets (Appleyard & Field, 1998 ; Blomström & Kokko, 1998 ). Conceptual Model and Hypothesis Pioneer s tudies on FDI spillovers came from Caves (1974) for Australia, Globerman (1979) for Can ada , and Blomström (1986) for Mexico. They found a positive relationship between foreign presence and labor productivity, implying that foreign presence had a positive influence on productivity (Gachino, 2010 ). Th ese studies have since been developed and refined, but the basic approach remains ( Görg & Greenaway, 2004 ). Most econometric analysis use a framework that regresses the labor productivity or the total factor productivity of firms on a range of independent variables. To measure productivity fro m multinational firms , a variable is included that proxies the exten t of foreign’ firms penetration… . In other words, the regression allows for an effect of FDI o n the productivity of firms in the same industry. If the regression analysis yields a positive and statistically significant coefficient of the foreign presence, this is taken as a evidence spillovers have occurred ( Gör g & Greenaway, 2004 , p. 176). Based on Gachino ’s (2010) model for determinants of spillover and studies using meta -analysis methodology developed by Görg and Strobl (2001) , Görg and Greenaway (2004) , and Wooster and Diebel (2010) , we outline the determinants of spillover occurrence in a broad conceptual model, as presented in Figure 2. Theory on FDI spillovers states that, especially in developing countries, their occurrence depends on a number of factors related not only to MNC characteristics, but also to specific characteristics of LOEs in host countries. Furthermore, factors related to industry , region and country characteristics determine FDI spillover occurrence ( Gachino, 2010 ; Görg & Greenaway, 2004 ). According to Gachino (2010) , the occurrence of spillovers does not only depend on the p resence of MNCs, but also on absorptive capacity, presence of support structures and institutions, presence of interactions and trade orientation. Other factors include ﬁrm size, age, ownership structure, performance, ﬁrm strategy and industry structure (Gachino, 2010 , p. 203) . The theory suggests some possible mechanisms through which spillovers may occur, such as imitation, worker mobility, competition and linkages, as presented in Figure 2 ( Görg & Greenaway, 2004 ). However, although presented in the conceptual model, they do not constitute object s of this study and will not be analyzed. Furthermore , analysis of spillover occurrence through the se channels require s in-depth analysis and , ther efore , qualitative methodologies for collecting and analyzing data would be required . Productivity Spillovers from Foreign Direct Investment 29 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar Figure 2 . Conceptual Model of Spillover Occurrence . Aiming to analyze the contributions FDI has on LOE productivity in host countries we have chosen to an alyze the following characteristics of the Brazilian processing industries and Brazil itself . Industry c haracteristics Following previous studies, (Blomström & Per rson, 1983 ; Buckle y, Clegg, & Wang, 2010 ; Caves, 1974 ; Globerman, 1979 ; Hale & Long, 2011 ; Kokko, 1994 ), we estimate d our function with labor productivity as our depe ndent variable. According to Buckley, Clegg, Zheng, Siler and Giorgioni (2010 , p. 295) “traditional models of economic growth pred ict that capital accumulation will raise the level of output per worker up to a point of diminishing returns”. Until the 1980s, the neoclassical model for economic growth developed by Solow (1956) were used as a reference for analytical models of per capita income growth determinants over the long ter m (Marin ho & Bittencourt, 2007 ). Solow ’s (1956) main contribution was to formulate a measure of technical progress for per capita output growth, known as total factor productivity (TFP). The measure had as its starting formulating point a Cobb -Douglas production function structure in which the author found the occurrence of significant residuals, measured by the difference between real output growth rates and weighted capital and labor production factor g rowth rates ( Marin ho & Bittencourt, 2007 ). Later studies based on Romer ’s (1986) pioneer ing contributions, dedicated efforts to include other factors into the production function that could reduce the re sidual values and suggest a greater contribution by capital, including human capital, to economic growth. One of Romer ’s (1986) important contribution s was to emphasize the simplicity of aggregate growth models and the belief that the “rate of return on investment and the rate of growth of per capita output … were … expected to be decreasing functions of the level of the per capita capital stock ” (p. 1002 ). Romer (1986) offered Productivity Spillovers DETERMINANT FACTORS FOR PRODUCTIVITY SPILLOVERS Industry Foreign Presence Industry Characteristics - Capital input - Labor input - Qualific ation - Return on assets - Technological intensity - Absorption capacity Mechanisms Competition Imitation Worker mobility Linkages Country Characteristics - Capital cost - Economic instability N. C. P. Bruhn, C. L. L. Calegario 30 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar an alternative view of ’ long -term prospects for growth where the rate of investment and the rate of return on capital might increase rather than decrease with increases in the capital stock. He suggested that endogenous technological change was driven primarily by the accumulation of k nowledge by forward -looking and profit -maximizing agents (Romer, 1986 , p. 1003). Another Romer contribution is presented in the author’s belief that “in contrast to models in which capital exhibits diminishing marginal productivit y, knowledge will grow without bound ” ( Romer, 1986 , p. 1003). Capital input importance (including physical and human capital) comes from its ability to generate positive externalities that raise firm productive capacity. It can be used as an accumulated knowledge indicator and experience in the form of learning -by -doing, in which externalities result in increasing returns ( Marinho & Bittencourt, 2007 ; Romer, 1986 ). Thus, human ca pital is an important driver of productivity and efficiency in business. Just as physical capital, human capital also increases a company ’s ability to produce goods and services. In fact, organization al efficiency depends on both material capital, accumula ted in the form of machinery and equipment investments, and human capital, acquired through investments in education. To Blomström and Kokko (1998) , the existing level of capabilities in an industry is essential to knowledg e exchange. The authors argue that worker training is considered a determining factor of how local businesses can benefit from new knowledge to become more competitive. Human capital is considered an important productivity and efficiency stimulant in enter prises. Just as physical capital, human capital also increases an enterprise ’s ability to produce goods and services. According to Jajri (2007) , education and training of a workforce in order to upgrade capabilities and knowledge w ill result in higher -skilled and more efficient workers, thus leading to better productivity levels. Capital intensity indicates investments made by firms in machinery, equipments and infrastructure that leads to increases in capital stock and enhance d pro duction capacity. The higher the capital intensity is, the higher level of firm automation and the higher the expected productivity (Buckley, Clegg, Zheng, Siler, & Giorgioni, 2010 ). Labor quality, in this study represented by the qualification variable, indicates the level of labor force capabilities or education. Improvements in a labor force can lead to increases in productivity (Buckley, Clegg, Zheng et al. , 2010 ). Our discussion suggests the following: H1: Everything else constant, the higher the capital input, the higher the productivity of the industry. H2: Everything else constant, the higher the labor input, the higher the productivity of t he industry. H3: Everything else constant, the higher the qualification, the higher the productivity of the industry. As presented in our previous discussion , the studies that followed Solow’s (1956 ) pioneer contribution to TFP for mulation were dedicated to investigate other determinant factors that could reduce the residuals in productivity functions. In our study, we introduce the return on assets, technological intensity, foreign presence and absorption capacity as variable s of a nalyses for industry characteristics and capital cost and economic instability for country characteristics. Return on assets (ROA), also known as return on investment (ROI) is a profitability index and measures management effectiveness in terms of generat ing profits from assets (Gitman, 2004 ). To Assaf (2003) , return on assets represents the total return produced by applications of assets. Thus , industries that have higher levels of return on assets may i ndicate that investments in assets have been used efficiently to generate sales growth and productivity gains. Productivity Spillovers from Foreign Direct Investment 31 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar We, then, developed the following hypothesis: H4: Everything else constant, the higher return on assets index, the higher the productivity of the industry. Aiming to control the heterogeneity between industries in terms of dynamism, technological sophistication and investment in research and development, we incorporated industry technological intensity in our productivity model. According to Jajri (2007) , the development of new products or new technologies allows production methods that result in a shifting of the production frontier upwards . Innovativeness and technological capability are factors that differentiate the va rious types of organizations and industries and, consequently, affect their economic performances ( Reichert, Beltrame, Corso, Trevisan , & Zawislak, 2011 ). According to Buckley, Clegg and Wa ng (2010 , p. 197), “the low technology nature of the host industry is thought to exacerbate the severity of negative competitive impact of FOEs presence. Under such conditions, the growth of negative spillover effects can rapidly became dominant when fore ign presence ” increases beyond some level. Therefore, the higher the technological intensity, the higher the long -term profitability. The technological intensity dummy variable was built from a reform in the Economic Activities National Classification (CNA E), based on the methodology proposed by the Organisation for Economic Co -Operation and Development (OECD , 2008 ), a ccording to the industry le vel of technological intensity. Thus, industries were classif ied into four categories: (a) low -tech intensity industries; ( b) medium -high technological intensity industries; ( c) medium -low technological intensity industries; ( d) and high -tech intensity industries. The sectors as classified into categor ies are presen ted in the appendix. Since technological effort is a critical determinant of productivity growth and international competitiveness, we will consider that: H5: Everything else constant, the higher the technological intensity, the higher the productivity of the industry. Productivity spillovers theory affirm that MNCs have productive assets, management and market capabilities , coordinated relationships with suppliers and consumers and reputation s that make them superior in terms of knowledge production and ma nagement and market techniques. If MNCs have such advantages, then it is expected that the ir presence can positively influence LOE productivity (Aitken & Harrison, 1999 ). Such gains in productivity caused by foreign influence are called productivity spillovers ( Blomström & Kokko, 1998 ). As discussed above, “FDI not only transfers capital but also technologies, managerial capabilities and advanced production functions. Therefore, the greater the foreign investments inflows, the higher productivity will be” ( Buckley, Clegg, Zheng et al. , 2010 , p. 296 ). However, spillover effects don’t occur automatically ( Narula, 2002 ). When a firm sets up a plant overseas or acquires a foreign plant, it does so with the expectation of realizing a higher rate of return than a given home country firm with an equivalent investment. The source of the higher return is the technological ad vantage, including innovative management and organizational processes as well new production methods and technologies. Therefore, multinational firms will not simply hand over the source of their advantage (Görg & Green away, 2004 ). In addition, MNCs are profit -driven and, therefore, are not interested in creating a knowledge transfer environment without receiving a good reward in exchange. For this reason, there are some potential costs associated with FDI inflows, such as an increase in unemployment and the emergence of more concentrated market structures, especially in economies in transition and developing countries (Appleyard & Field, 1998 ). N. C. P. Bruhn, C. L. L. Calegario 32 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar Empirical research all over the world ( Görg & Strobl, 200 1) has shown that evidence on positive spillovers are contradictory or mixed and both empirical and theoretical studies have focused on explaining these mixed results ( Thang, 2011 ). T he author argues that an important conclusion of such studies is that the signal and magnitude of productivity spillovers depend on the nature of firms and industries and the conditions of host countries. Cohen (2007) argu ment s that FDI is a complex and heterogeneous phenomena and that different kinds of businesses produce different kind of corporate activity and, consequently, different results. The nature, objectives and effects of specific kinds of firms in one industry are not applicable to others. This discussion brings the following hypothesis: H6: Everything else constant, the higher the foreign presence, the higher the productivity of the industry if there are spillover effects from FDI; or the higher the foreign presenc e the lower the productivity of the industry. Studies point out absorption capacity as an important FDI spillover determinant factor. For Cohen and Levinthal (1990) , the absorpti on capacity is the ability to recognize the value of a new knowledge, the capacity to assimilate and apply it, based on business purposes. The authors add that absorptive capacity is fundamental to firms innovative capacity development, which is cumulative and depends on various firm characteristics, suc h as employees ’ individual skills, internal organization and investments in Research and Development (R&D). Narula (2002) no tes that MNC FDI and operations do not automatically generate positive externalities. MNCs may disseminate a large number of externalities that can easily be assimilated or not, depending on LOE capacity. When MNCs establish a plant overseas or acquire a foreign plant, they do so with the expectation of higher rates of return that could be receive d if compared to their home country with an equivalent investment. The largest source of return is the technological advantage, including new management processes and new production methods. Thus, MNCs will not simply undo these benefits sources ( Görg & Greenaway, 2004 ). However, theory suggests that even if a MNC has as main motivation the internalization of technology and its use, this can spread or overflow to firms in the host economy. Thus, Görg and Greenaway (200 4) argue that FDI benefits only occur when LOEs have the ability to learn, as well as the capabilities and abilities to imitate MNCs and the internal infrastructure to provide such development conditions. Buckley, Wang and Clegg (2010) argue that positive spillovers are expected for LOEs that have superior absorptive capacity. Our discussion suggest the following hypothesis: H7: Everything else constant, the higher the absorption capacity, the h igher the productivity of the industry. Country c haracteristics This group is represented by capital cost and economic instability as variables of analyses to represent country characteristics. Capital cost represents the price paid to borrow money for a certain period of time. Interest rate behavior affects consumption and investment decisions, external resource flows, exchange rate value and as a consequence, the competitiveness of products from that country. In this study, it is represented by the Se lic domestic interest rate , set by the Monetary Policy Committee (Copom) since it has effects on the production structure as it determines the cost of investments (Galeano & Mata, 2007 ; Nakabashi, Cru z, & Scatolin , 2008 ; Sonaglio, Zamberlan, Lima , & Campos, 2010 ). Interest rates are crucial for industries ’ productivity level , since they play a decisive role in productive investment decisions . While classical economists emph asized that investment s were Productivity Spillovers from Foreign Direct Investment 33 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar determined solely by the interest rate (Omar, 2008 ), Keynes (1983) suggested two factors as key determinants of investment spending : the cost of capital , which is the interest r ate , and the expected future profitability of investment projects . The higher future profitability is, the higher the level of investment s. Thus , when interest rates rise , certain investment projects become impractical (Omar, 2008 ). According to the Brazilian Central Bank (Banco Central do Brasil [BACEN ], n.d. a), the Selic rate is obtained by calculating the weighted and adjusted average rate of financing transactions for a day , backed by federal government bonds and routed in the referred system or in clearing and settlement systems in the form of repurchase agreements. The origin of the Selic rate is in the interest rates observed in the market and basically reflect s the liquidity conditio ns in the monetary market (supply versus demand for resources ) (BACEN , n.d. a). Then, H8: Everything else constant, the higher the capital cost, the lower the productivity of the industry. Economic stability is one of the necessary conditions for entrepreneurs to feel secure in their investment decision s (Galeano & Mata, 2007 ). More stable economies with low inflation rates provide industries with better economic performance levels ( Damasceno, 2008 ; Galeano & Mata, 2007 ; Lima, 2005 ; Nonnenberg & Mendonça, 2005 ). To represent economic instability, we use the inflation rate mea sured by the National Broad National Index of Consumer Price ( Índice Nacional de Preços ao Consumidor Amplo [IPCA ]). The inflation rate is used as a proxy for the degree of economic instability, given that the classic symptom of loss of control in an econo my, both fiscal and monetary, is uncontrolled inflation ( Nonnenberg & Mendonca, 2005 ). Hence , the following hypothesis is suggested: H9: Everything else constant, the higher the economic instability, the lower the productivi ty of the industry. Methods and Data Econometric model The econometric model presented was based on previous studies (Aitken & Harrison, 1999 ; Blomström & Pe rrson, 1983 ; Buckley, Clegg & Wang, 2010 ; Buckley, Clegg, Zheng et al. , 2010 ; Girma & Gorg, 2005 ; Globerman, 1979 ; Hale & Long, 2011 ; Kokko, 1994 ) aiming to identify FDI effects on the productivity of Brazilian industries. Productivity spillover is usually analyzed in an econometric function in which a number of covariates are assumed to have an effect on productivity, one of which is the foreign presence. The econometric specification identifies spillover effects varying across industries according to their level of absorptive capacity (ABC) . Our assumption is that industries with higher absor ption capacity will be able to obtain advantages of foreign presence given their abilities to recognize and assimilate new knowledge. Aiming to investigate the spillover effects of FDI depending on the absorptive capacity of industries we developed our mai n total factor productivity (TFP) function as follows: TFP it = 1Kit + 2Lit + 3FDI it + 4ABC*F DI it + 5 Kit*FDI it + 6Lit*FDI it + Zit +  it + it (1) in which 1, 2, 3 , 4 5 e 6 are the parameters to be estimated; K it is the capital input in industry i at time t; L it is the labor input in industry i at time t; 3, 4, 5 and 6 are FDI spillover effects; FDI it is the N. C. P. Bruhn, C. L. L. Calegario 34 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar measure of foreign presence in industry i at time t;  = set of parameters related to other characteristics of the industry to be estimated; Z it = set of variables that capture other specifi c characteristics of the industry in time i;  = set of parameters related to characteristics of the country to be estimated; it = set of variables that capture specific characteristics of the country at time i; itc = the random term error. The 4 coef ficient measures the spillover effects arising from FDI depending on the absorption capacity of industr ies while 5 and 6 coefficients represents respectively, spillover effects arising from FDI depending on the capital and labor -intensity of industries. If the estimate d coefficients turn out to have a positive and statistically significant sign, this is taken as spillover evidence. Industry a bsorptive capacity ( ABC ) is defined as: ABC it = TFP it / max industry (TFP t) (2) in which TFP is based on the indi vidual TFP compared to the larger TFP (Kathuria, 2000 ) with index measures between 0 and 1. The closer the index is to one, the greater the industry absorption capacity. The TFP estimation is constructed based on the basic pro duction function described as follows : Yit = 0 + 1Kit + 2Lit + it (3) in wich Yit is industry productivity, measured by the industrial production sector value i at time t; 0 is the intercept; 1 e 2 are parameters to be estimated; Kit is capital inp ut, in industry i at time t; Lit is labor input in industry i at time t; itc is the random term error. TFP is the share of the dependent variable not explained by the physical quantities of two traditional factors (capital and labor inputs ), in other word s, the residuals. So, what is not explained by the inputs accumulation, particularly capital and labor, will be explained by the growth of TFP. The effects of variables related to ( a) industry characteristics; ( b) country characteristics; ( c) foreign prese nce; and ( d) interactions with foreign presence were analyzed using Moderated Multiple Regression (MMR) and Generalized Linear Models (GLM) analysis of variance. We performed the analysis undertaken in this study using SAS statistical software, version 8. The Moderated Multiple Regression (MMR) involved hierarchical regression to test: (a) Equation 1: the relationship of the primary predictors of interest ( industry and country characteristic variables) on the dependent variable; (b) Equation 2: the relation ship of the primary predictors of interest (as in Equation 1) plus foreign presence on dependent variable; and (c) Equation 3: the relationship of the primary predictors of interest , foreign presence (as in Equation 2) plus the relationship of foreign pres ence interaction terms on the dependent variable. The interaction variables incorporated in the model were : (a) c apital input * foreign presence ; (b) labo r input * foreign presence ; (c) and absorption capacity * foreign presence . They are used to demonstra te the effect of a given variable depending on the moderating effect of an other. Before estimating the regressions, we conducted a correlation test in order to verify the relationship degree between variables and if there were problems associated with mult icollinearity. The method used to measure the association degree between variables in this study was Pearson correlation coefficient. We used the tolerance (TOL) and variance inflation factor (VIF) as complementary measure to detect multicollinearity. We verified the autocorrelation presence in error terms through scatter plot of predicted values in relation to waste diversion. According to Gujarati (2006) , residual graphical analysis offers a simple summary to understand a comp lex problem. They allow a simultaneous examination of individual cases, while showing data behavior as an aggregate. Autocorrelation premise is related to population error terms, which cannot be directly observed. What we usually have are residuals, which are proxies that can provide evidence about autocorrelation in error terms presence (Gujarati, 2006 ). Productivity Spillovers from Foreign Direct Investment 35 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar Adjustment analysis in Generalized Linerar Models is based on statistical deviance. To do so, we followed Allison ’s (2001) instructions. In general, the smaller the deviation value, the better the model fits the data. Two other measures commonly used in model adjustment analysis were used in this study: Akaike Information Criterion (AIC) and Schwartz. The statistical AIC is obtained by adding 2k to the deviation, where k represents the number of model parameters (Allison, 2001 ). The statistical Schwartz is obtained by adding k log n to the deviations (n represents the sample s ize). In general, when comparing two or more models, the best one is the on e that shows the lower values of these statistics. Description of variables The series used in the analysis have panel database form and contain aggregate data on twenty - three int entionally chosen Brazilian processing industries, defined by the Na tional Code of Economic Activities (CNAE). A description of variables and their expected signs are summarized in Table 1. Table 1 List of Selected Variables Descr iption and Their Expect ed Signs Variable Description Expected signs Dependent variable Productivity (PROD) Industrial transformation value divided by the number of people employed in the industry. Characteristics of the industry Capital input (CAP) Total assets of the ind ustry. + Labor input (TRAB) Number of people employed in the industry , including salaried people with or without employment + Qualification (QUAL) Ratio of total gross wages paid by the industry to the number of people employed in the same industry + Return on assets index (IROA) Ratio of sales operating revenue in the industry in relation to its total assets + Technological intensity (INTEC) Dummy variable representing industry technological intensity based on four categories: (1) low -tech intensity industries; (2) medium -high technological intensity industries; (3) medium -low technological intensity industries; (4) and high -tech intensity industries Foreign presence (FDI) Flows of foreign direct investment (FDI) received by the industry +/- Absorp tion capacity (CAPAB) The individual TFP compared to the larger TFP + Country Characteristics Capital cost (CUSTOCAP) Internal Selic interest rate proxy variable - Economic instability (INSTEC) Inflation rate, measured by the National Consumer Price Index (IPCA) - Productivity , capital input , labor input , qualification and rate of return on assets variables were obtained from the Annual Industrial Research (PIA) database, released by Instituto Bras ileiro de Geografia e Estatística (IBGE , 2005) . The scope of the PIA sectoral aggregated database includes companies that meet the following requirements in December 31 of the reference year: ( a) be in active N. C. P. Bruhn, C. L. L. Calegario 36 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar status in Enterprises Central Register (CEMPRE ) from IBGE; ( b) have a main economic activity classified in C (Extractive industries) or D (Processing Industries) sections from CNAE classification ; (c) be headquartered anywhere in Brazil ; (d) have five or more employees. The sample is obtained by simpl e stratified sampling. Surveyed companies are part of a census selection form for the universe of companies with 30 or more employees. A random stratum identifies companies with 5 to 29 people employed , randomly selected without replacement. FDI inflows, c lassified according to the CNAE classification, were obtained from the Foreign Capital Census of Bra zil ’s Central Bank (BACEN , n.d. b). The Selic rate and IPCA inflation rate were obtained from Institute of Applied Research (Instituto de Pesquisa Econômica Aplicada [IPEA ], 2011 ) database. Since not all industries presented available data for all years, it became necessary to exclude some observations from the sample. Furthermore, as some continuous variables showed significant amplitude, we used logarithmic transformation in order to reduce their amplitudes. Thus, these variables will be incorporated into the analysis in logarithmic form. Results Pearson co rrelation results showed that some variables such as capital input, labor input and foreign presence were correlated (Table 2). However, as these variables were essential to the productivity function, we could not simply exclude these variables from the mo dels. So, we used the tolerance (TOL) and variance inflation factor s (VIF) as complementary measure s to detect multicollinearity. Table 2 Pearson Correlation Test Selected Variables 1 2 3 4 5 6 7 8 1a 1,00 0,6094 <,0001 0,7225 <,0001 0,1498 0,1562 -0,470 <,0001 -0,386 0,0002 0,005 0,996 0,185 0,078 2b 1,00 0,6667 <,0001 -0,237 0,0235 -0,062 0,5548 -0,629 <,0001 0,002 0,998 0,004 0,970 3c 1,00 0,2856 0,0064 0,0024 0,9817 -0,254 0,0155 -0,104 0,325 -0,094 0,376 4d 1,00 0,0852 0,4216 0,5618 <,0 001 0,081 0,445 0,107 0,309 5e 1,00 0,0976 0,3573 0,0666 0,529 -0,333 0,001 6f 1,00 -0,002 0,979 0,020 0,843 7g 1,00 0,630 <,0001 8hi 1,00 Note . aCapital input; bLabor input; cforeign presence; dqualification; return on assets; ftechnological intensity; gcost of capital; heconomic instability. Productivity Spillovers from Foreign Direct Investment 37 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar The results showed that for all t hree analyzed equations, TOL and VIF values were within acceptable limits and do not denote collinearity , indicat ing that the results are reliable. Then, we verified the autocorrelation presence in error terms through scatter plot of predicted values in relation to waste diversion (Figure 3). Since there was no observed trend pattern from predicted observations related to residuals deviation for the proposed m odels, we can say that there is no residuals ’ heteroscedasticity evidence and, therefore, the models are appropriate. (a) Equation 1 (b) Equation 2 (c) Equation 3 Figure 3. Predicted Values for Waste Deviations for Proposed Equations . Table 3 presents t he results for the proposed equations containing the individual parameters, deviations, Akaike Information Criterion (AIC) ,Schwartz, R -Square and F -Value statistics. Based on results, we can say that Equation 3, which includes industry and country characteristics, as well as foreign presence and interactions with foreign presence , presented the best adjustment. Table 3 Estimates for Productivity Models Parameters Equation (1) Equation (2) Equation (3) INTERCEPTO -5.5474 *** -6.0530 *** 3.7388 *** ICAP 0.6438 *** 0.6912 *** 0.0503 ITRAB -0.5879 *** -0.5919 *** 0.1479 ** QUAL 0.1295 0.1264 0.0947 IROA 0.9599 *** 1.0035 *** 0.01 10*** INTEC 0 0.1136 0.0778 0.1870** INTEC 1 0.0 342 0.0 317 0.0 806*** CUSTOCAP -0.2839 -0.3301 -0.1434 INSTEC 0.0048 0.0065 0.00 19 FDI -0.03 61 0.4355 FDI*ICAP 0.00 222 FDI*ITRAB -0.0233 ** FDI*CAPAB 0.1797 *** Deviance 15,38 15,21 2,32 AIC 31,38 33,21 26,32 Schwartz 31,05 32,84 25,82 R- Square 0,6082 0,6054 0,9342 F-Value 15.91*** 13.54*** 100.69*** Note . *** Significant at 1\% , ** significant at 5\% , * significant at 10\% P r e d -3 -2 -1 0 1 R e s d e v -2 -1 0 1 2 3 P r e d -3 -2 -1 0 1 R e s d e v -2 -1 0 1 2 3 P r e d -3 -2 -1 0 1 R e s d e v -2 -1 0 1 2 3 N. C. P. Bruhn, C. L. L. Calegario 38 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar The results for equation 1 showed that the positive relationship between the capital input variable (ICAP) and depe ndent variable productivity ( PROD) is consistent with proposition 1 (P1), indicating that the higher the levels of capital input, the greater the productivity of industries . Labo r input variable (ITRAB) was negatively related to the dependent variable and inconsistent with proposition 2 (P2) indicating that the higher the labor input the less productivity is the industry. Rate of return on assets variable (IROA) was statistically significant at 1 \% and positive related to dependent variable, as expected (See Proposition 4) . The result is consistent with theory , indicating that industries that optimize their assets generate more productive profits . The result for the technological intensity variable ( INTEC) is consistent with literature and with H5, suggesting that higher levels of productivity is related with high -technology intensity industries, suggesting that a minimum level of technological sophistication is required so that industries can provide significant levels of productivity. The next step consisted of estimating Equation 2, which includes, besides variables presented in Equation 1, the foreign presence variable. Results presented for the estimated equation indicate that, for the period analyzed, there is no evidence of productivity spillover occurre nce resulting from foreign presence in analyzed industries. Furthermore, the variable incorporation demonstrates no significant changes in direction and significance levels of other variables. Finally, we incorporated the interaction variables with foreign presence in Equation 3. Results presented for Equation 3 show that Labor input (ITRAB) became positive and coherent with our expectations presented in H2 showing that the higher the labor input, the higher the industry productivity. When analyzing results obtained from Labo r input (LNITRAB) with foreign presence interaction (FDI), we found an inverse relationship , indicating that FDI entry in labor intensive industries provides fewer productivity gains . Interaction of the foreign presence variable (FDI) wi th absorption capacity (CAPAB) was incorporated into the model in order to test the effect of foreign presence depending up on the moderating effect of industry absorption capacity. The results found for this interaction indicated a positive relationship in the order of 0. 1797 and statistically significant at 1\% , confirming the hypothesis that FDI presence benefits depends on industry absorptive capacity . Discussions Results for proposed equations showed a positive relationship between capital input (ICAP), and rate of return on assets (IROA) variables with the dependent variable productivity ( PROD). Those results are consistent with our propositions and demonstrated that these variables are directly related with productivity of studied industries. Howe ver, the results obtained for labor input (ITRAB) variable differ from what we expected in H2 in Equation 1 and 2 and can be justified by the fact that labor intensive industries may use their labor input investments to generate higher levels of productivi ty inefficiently . This means that, although the new investments in labor input may create or expand their production capacity, they do not necessarily expand their capacity to generate outputs and, thus, become more productive. The t echnological intensity (INTEC) variable suggests that high -technology intensity industries are more productive. The results indicate that the higher the technological sophistication and dynamism, the higher the productivity of a given industry. Results for Equation 2 showed tha t when incorporating foreign presence (FDI) in Equation 2 we find no evidence of productivity spillover from foreign presence . Evidence show s that, although it is undeniable that foreign presence can lead to technology diffusion, it is noteworthy that the transfer of Productivity Spillovers from Foreign Direct Investment 39 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar this technology to LOEs does not happen automatically. Moreover, one cannot lose sight of the difference between learning technology and operational creation of a new technology. FDI can be effective in transferring the result of innovation, bu t not necessarily the ability to innovate. Furthermore, when a MNC establishes a plant overseas or acquires a foreign plant, it does so in the expectation of higher rates of return that could receive, compared to their home country, with an investment equ ivalent ( Narula, 2002 ). The largest source of return is in technological advantages, including new management processes and new production methods. Thus, ETNs will not simply get rid of their benefits sources ( Görg & Greenaway, 2004 ). The result is not necessarily inconsistent with theory since there may be negative spillover effects in some industries, while others present positive effects. To answer this question, we included FDI interactio n variables aiming to understand if there are any variation of FDI effect s depending on specific industry characteristics. Equation 3 showed that, w hen analyzing results from labor input we found a positive relation as expected in H2 but different from res ults for Equation 1 and 2. They s uggest that Brazilian labor - intensive industries can use their labor input investments to generate higher levels of productivity , but when ITRAB interacts with FDI we found negative effects of labor inputs on industry produ ctivity. One plausible explanation for negative results is presented by Görg and Greenaway (2004) , Aitken and Harrison (1999) and Buckley, Clegg and W ang (2010) . The y argue that, at a greater level of foreign presence, MNC negative effects start to become apparent as foreign firms may reduce LOE productivity through competition effects. According to them, MNCs usually have lower marginal costs due to f irm specific advantage s, which allows them to attract demand away from LOEs through the introduction of innovation processes and differentiated products . L abor -intensive industries may not be prepared to compete with them . Furthermore , competing with MNCs may require technologies and organizational processes that are very specific and require higher levels of investment s, which may include, in addition to the purchase of machinery and latest equipment, staff training and research and development (R&D) inves tments , that are not practical for LOEs in labor -intensive industries . Another explanation for negative spillover effect s of FDI in labor -intensive industries is presented by Blomström and Kokko (1998) an d emphasizes that, if the home country’s labor force is well educated and wages are relatively higher, the structural shift is likely to bring emphasis on production in advanced industries with high labor productivity in the home country. Thus, the simple production process es requiring lots of unskilled labor may be moved to foreign affiliates , leading to lower levels of productivity in the industry. Thus , industries with higher foreign presence will have lower productivity levels. It is evident that MNC presence can also i nduce a reduction in the number of companies in the industry, since when competing with MNCs , less efficient domestic firms may be forced to cease operation (Blomström & Kokko, 1998 ). However , Caves (1971) adds that , whatever the market structure resulting from FDI influence , it is argued that a MNC ’s entry tends to induce more active competitive behavior than would the entry of a domestic company with the same initial scale. Even if a MNC has as its main motivation the technology use internationalization, it can spread or spillover for firms in the host economy. Our results are consistent with theory and show that firms ’ absorptive capacity have shown to be a determinant factor for FDI spillov er occurrence, evidence that can be confirmed by the results of foreign presence ( FDI) and absorption capacity (CAPAB) interaction . The evidences show that FDI benefits depend on industry absorptive capacity and that not all industries should be expected t o benefit equally from foreign presence spillovers. According to the literature, host industries must possess high absorptive capacity to obtain the advantages from foreign presence (Blomström , Globerman, & Kokko, 1999 ; Castellani & Zanfei, 2003 ; Girma & Gorg, 2005 ; Haddad & Harrison, 1993 ; Hale & Long, 2011 ; Kathuria, 2000 ; Kokko, 1994 ; Malik, Rehman, Ashraf, & Abbas, 2012 ). The entry of MNCs in industries with high absorption capacity may lead to an increase in the N. C. P. Bruhn, C. L. L. Calegario 40 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar competition level between companies , forcing them to become more efficient . Inc reased competition that occurs in the presence of MNCs is considered a knowledge intensifier diffusion mechanism by increasing competitive pressure , in particular if it induces industries to use existing resources more efficiently . Final Considerations This paper finds evidence of the coexistence of both positive and negative effects arising from FDI on Brazilian industries ’ productivity. Our findings suggest that inward FDI leads to positive spillover effects in high absorption capacity industries and negative effects in labor -intensive industries. Finding that spillovers do not occur equally in all industries leads to questions related to unconditional or unrestricted FDI liberalization policies. The result s confirm Buckley, Clegg and Wang (2010) findings that the “complexity of spillover effects challenges the lais sez -faire view that all inward FDI into all types of industries is equally valuable in terms of productivity spillover benefits ” ( Buckley, Clegg , & Wang, 2010 , p. 192 ). The results are similar to Cohen ’s (2007) findings that FDI sometimes has positive effect, but sometimes negative , neutral or irrelevant effects. According to the author , subsidiaries of MNCs operating in different national and regional environments generate effects that ranges from highly deleterious to high beneficial. To the author, no two MNCs are organized alike and share the same production profile, even subsidiarie s of the same company will never be identical in their output and impact on the host economy. They have different business culture s and produce different effects on host economies. Based on the following arguments presented by Cohen (2007) , o ur comprehension is that: ( a) different kinds of businesses produce different kinds of corporate activity; and ( b) the nature, objectives and effects of specific kinds of firms in different industries and countries are not applicable to othe rs. So, based on Cohen (2007) argu ments, we suggest not using the good -versus -bad approach, since it can be superficial at best and inaccurate at worst. We defend the thesis of the heterogeneity in FDI effects and the imperativ e of disaggregation in studies in the various levels of analysis, including the individual manager, the firm, the industry and the environment. In each category there is a vast heterogeneity of issues to be discussed and comprehended ( Buckley & Lessard, 2010 , p . 7). Our findings yield original insights into the complexity of spillover effects FDI into Brazilian industries and brings into discussion the need for a deeper understanding of its possible causes (Buckley, Clegg , & Wang, 2010 , p. 192) and the necessity of well defined and structured sectoral policies seeking to attract higher quality FDI that can effectively contribute to Brazilian industry competitiveness and national in dustry development. Our evidence ’s implication is that policymakers should appreciate both the critical need to preserve and maximize competition among MNCs and LOEs and the need to achieve at least a minimum level of domestic technological capability and technical education (Buckley, Wang, & Clegg, 2010 ; Cohen, 2007 ); comprehending that there is no way to stop MNCs from growing and increasing their market shares as they respond to market competitiveness. This doesn’t mean that a passive government compliance is desirable, but they should consider looking at the ways in which initially disadvantaged LOEs could be helped in acquiring the necessary capabilities to compensate their disadvantag es. This might occur, for example, through and public -private partnership seeking a closer approximation with MNCs and through an increase in vigorous enforcement laws. The analyses bring implications for policymakers, as suggested by Buckley and Ruane (2010) : (a) it is important that policy makers understand that MNC strategies are not only local or regional, but global. Host countries need to focus on what immobile resources they can offer to combine with MNC Productivity Spillovers from Foreign Direct Investment 41 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar resources to achieve synergy that can benefit industries in the host country; ( b) sectoral directions require selectivity projects , considering a careful cost -benefit analysis and strategic bargaining; ( c) combining both financial and fiscal performance -based incent ives designed to ensure outcome benefits to host country LOEs; ( d) monitoring project outcomes. Study limitations can be highlighted due to the fact that, given the nature of the object of the study and database unavailability, the channels and other impor tant variables determining spillover occurrence could not be incorporate d in the model , such as systemic interactions among agents arising from networks and other forms of informal linkages, infrastructural and institutional support structure and some spec ific firm characteristics, such as firm size –scale, R&D investments, firm strategy, proximity with MNCs and regional characteristics could not be analyzed in this study. As suggestion for future studies we recommend a research agenda in bo th quantitative and qualitative terms such as: ( a) the relationship between different kinds of foreign investment strategies by certain kinds of FOEs and the increases in LOE economic performance and innovation capacity; (b) how human capital, natural, economic and politi cal environment s are related to the success or failure of FDI benefits on LOEs of host countries; ( c) which variables can determine whether a FDI project is beneficial to host economies or not , and which benefits can be achieved through public policies. We expect that the disaggregation and in -depth appreciation of these heterogeneous, complex and dynamic phenomena can lead to many other accurate insights into the subject. References Aitken, B ., & Harrison, A. (1999) . Do domestic firms benefit from direct foreign investment? Evidence from Venezuela. American Economic Review , 89 (3), 605 -618. doi: 10.1257/aer.89.3.605 Aitken, B., Harrison, A., & Lipsey, R. E. (1996) . Wages and foreign ownership: a comparative study of Mexico, Ve nezuela, and the United States . Journal of International Economics, 40 (3/4), 345 - 371. doi: 10.1016/0022 -1996(95)01410 -1 Allison, P. D. (2001). Logistic regression using SAS Systems : theory and applications (2nd ed.) . Cary, NC: SAS Institute. Appleyard, D. R., & Field, A. J., Jr. (1998). International economics . Boston , EUA: Irwin/McGraw - Hill. Assaf, A. , Neto (2003). Finanças corporativas e valor . São Paulo: Atlas. Banco Central do Brasil. (n.d.a) . Descri ção da taxa Selic . Retrieved from http://www.bcb.gov.br/?SELICDESCRICAO Banco Central do Brasil . (n.d.b ). Investimento estrangeiro direto. Estoque de investimento estrangeiro direto (1950 - Junho/1995) . Retrie ved http://www.bcb.gov.br/?INVEDIR Blomström , M. (1986) . Foreign investment and productive efficiency: the case of Mexico. Journal of Industrial Economics , 35 (1), 97 -110. Blomström, M., Globerman, S., & Kokko, A. (1999) . The determinants of host country spillovers from foreign direct investment: review and synthesis o f the literature [ Working Paper Nº 76]. The European Institute of Japanese Studies , Strockholm, Sweden. Blomström, M., & Kokko, A. (1998) . Multinational corporations and sp illovers. Journal of Economic Survey, 12 (3), 247 -277. doi: 10.1111/14 67 -6419.00056 N. C. P. Bruhn, C. L. L. Calegario 42 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar Blomstr öm, M. , & Pear sson , H. (1983) . Foreign investment and spillover efficiency in an underdeve loped economy: evidence from the Mexican manufacturing industry. World Development, 11 (6), 493 -501. doi: 10.1016/0305 -750X(83)90016 -5 Buckley, P. J., Clegg, J., & Wang, C. (20 10) . Is the relationship between inward FDI spillover effects linear? An empirical examination of the case of China. In P. J. Buckley (Org.) , Foreign direct investment , China and the world economy (pp. 192 -215 ). London: Palgrave Macmillan. Buckley, P. J., Clegg, J., Zheng, P., Siler, P. A., & Giorgioni, G. (2010) . The impact of foreign direct investment on the productivity of Chinas automotive industry. In P. J. Buckley (Org. ), Foreign direct investment , China and the world economy (pp. 284 -304 ). London: Palgrave Macmillan. Buckley, P. J., & Lessard, D. R. (2010) . Regaining the Edge for Internati onal business research. In P. J. Buckley (Org.), Foreign direct investment , China and the world economy (pp. 7-24 ). London: Palgrave Macmillan. Buckley, P. J., & Ruane, F. (20 10) .Foreign direct investment in Ireland: policy implications for emerging economie s. In P. J. Buckley (Org.), Foreign direct investment , China and the world economy (pp. 365 -385 ). London: Palgrave Macmillan. Buckley, P. J., Wang, C., & Clegg, J. (2010) . The impact of foreign ownership, local ownership and industry characteristics on spillover benefits from foreign direct investment in China. In P. J. Buckley (Org.), Foreign direct investment , China and the world economy (pp. 305 -326 ). London: Palgrave Macmillan. Castellani, D., & Zanfei, A. (2003) . Tec hnology gaps, inward investments and pr oductivity of European firms. Economics of Innovation and New Technology, 12 (6), 555 -576. doi: 10.2139/ssrn.276869 Caves, R. E. (1971) . Internation al corporations: the industrial economics of foreign investment. Economica , 38 (149), 1 -27. Caves, R. E. (1974) . Multinational firms, competition, a nd productivity in host -country mark ets . Economica, 41 (162) , 176 -193 . Cohen, S. D. (2007) . Multinational corporations and forei gn direct investmen t: avoiding simplicity, embracing complexity . Oxford: Oxford University Press. Cohen, W. M., & Levinthal, D. A. (1990) . Absorptiv e capacity: a new perspective on learning and innovation. Administrative Science Quarterly , 35 (1), 128 -152. Damasceno, A. O. (2008) . Liberalização da conta de capitais e crescimento econômico : evidências de dados em painel para a América Latina. Revista de Economia Política , 28 (4), 595 -611. doi: 10.1590/S0101 -31572008000400004 Dunning, J. H. (2000) . The eclectic paradigm as an envelope for economic and business theories of MNE activity. International Business Review, 9 (2), 163 -190. doi: 10.1016/S0969 - 5931(99)00035 -9 Dunning, J. H., & Lundan, S. M. (2008) . Multinational enterprises and the global economy (2a ed. ). Cheltenham: Edward Elgar. Fernandes, E. A., & Campos, A. C. (2008) . Investimento direto estrangeiro e o desempenho das exportações brasileiras . Revista de Economia Política, 28 (3), 490 -509. doi: 10.1590/S0101 - 31572008000300008 Findlay, R. (1978) . Some aspects of technol ogy transfer and direct foreign investment. American Economic Review, 68 (2), 275 -279. Productivity Spillovers from Foreign Direct Investment 43 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar Gachino, G. (2010) . Technological spillovers from multinational presence : towards a conceptual framework. Progress in Development Studies, 10 (3), 193 -210. doi: 10.1177/146499340901000301 Galeano, E. A., & Mata, H. T. C. (2007) . O comportamento do investimento no Brasil no período de 1995 a 2005. Perspectivas Contemporâneas, 2(2), 2 -25 . Girma , S., & Gorg, H. (2005) . Foreign direct investment, spillovers and absorptive capacity : evidence from quantile regressions. Paper Discussion , Kiel Inst itute for World Economics , 1 (1248), 1 -31. Gitman , L. J. (2004). Princípios de administração financeira . São Paulo: Pearson. Globerman, S. (1979) . Foreign direct investment and spillo ver efficiency benefits in Canadian manufacturing industries. Canadian Journal of Economics, 12 (1) , 42 -56. Görg, H., & Greenaway, D. (2004) . Much ado about nothing? Do domestic firms re ally benefit from foreign direct investment? World Bank Research Observer, 19 (2), 171 -197. doi: 10.1093/wbro/lkh019 Görg , H., & Strobl, E. (200 1). Multinational companies a nd productivity spillovers: a meta -analys is. The Economic Journal , 111 (475), 723 -739. doi: 10.1111/1468 -0297.00669 Gujarati, D. (2006) . Econometria b ásica (4th ed .). Rio de Janeiro: Elsevier. Haddad, M., & Harrison, A. (1993) . Are there positive spillovers from direct foreign investment? Evidence from panel data for morocco. Jour nal of Development Economics, 42 (1), 51 -74. doi: 10.1016/0304 -3878(93)90072 -U Hale, G. , & Long, C. (2011) . Are there productivity spillovers from foreign direct in vestment in China? Pacific Economic Review, 16 (2), 135 -153 . doi: 10.1111/j.1468 -0106.2011.00539.x Instituto Brasileiro de Geografia e Estatístic a. (2005) . Pesquisa industrial anual . Retrieved from http://www.ibge.gov.br/home/estatistica/economia/industria/pia/produtos/produto2005/default.s htm Instituto de Pesquisa Econômica Aplicada. (2011) . Ipeadata macr oeconômico . Retrieved from http://www.ipeadata.gov.br Jajri, I. (2007) . Determinants of total factor productivity growth in Malaysia. Journal of Economic Cooperation, 28 (3), 41 -58. Kathuria, V. (2000) . Productivity spillovers from technology transfer to Indian manufacturing firms. Journal of International Dev elopment , 12 (3), 343 -369. doi: 10.1002/(SICI)1099 - 1328(200004)12:3<343::AID -JID639>3.0.CO;2 -R Keynes , J. M. (1983). Teoria geral do emprego, do juro e do dinheiro . São Paulo: Abril Cultural . Kokko, A. (1994) . Technology, market characteristics, and spillovers. Journal of Development Economics , 43 (2), 279 -293. doi: 10.1016/0304 -3878(94)90008 -6 Kupfer, D., & Hasenclever, L. (2002). Economia industrial: fundamentos teóricos e práticas no Brasil . Rio de Janeiro: Campus . Lima, A. , Jr. (2005) . Determinantes do investimento direto estrangeiro no Brasil (Dissertação de me strado ). U niversidade Fed eral de Minas Gerais , Belo Horizonte, MG, Bras il. Malik, A. R. A., Rehman, C. A., Ashraf, M., & Abbas, R. Z. (2012) . Exp loring the link between fore ign direct investment, multinational enterprises and spillover effects in developing N. C. P. Bruhn, C. L. L. Calegario 44 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar economie s. International Journal of Business and Management, 7 (1), 230 -240. doi: 10.5539/ijbm.v7n1p230 Marinho, E., & Bittencourt , A. (2007) . Produtividade e crescimento econômico na América Latina: abordagem da fronteira de produção estocástica. Estudos Econômicos, 37 (1), 5 -33. doi: 10.1590/S0101 -41612007000100001 Markusen, J. R., & Venables, A. J. (1999) . Foreign direct investment as a catalyst for industrial development. European Economic Review, 43 (2), 335 -356. Meyer, K. E. (2004) . Perspectives on multinational enterprises in emerging economies. Journal of International Business Studies , 35 (4) , 2 5 9 -276. doi: 10.1057/palgrave.jibs.8400084 Nakabashi, L., Cruz, M. J. V., & Scatolin, F. (2008) . Efeitos do câmbio e juros sobre as exportações da indústria brasileira. Revista de Economia Contemporânea, 12 (3), 433 -461. doi: 10.1590/S1415 -98482008000300002 Narula, R. (2002) . Innovation systems and inertia in R&D location: Norwegian firms and the role of systemic lock -in. Research Policy , 31 (5), 795 -816. doi: 10.1016/S0048 -7333(01)00148 -2 Nonnenberg, M. J. B., & Mendonça, M. J. C. (2005) . Determinantes dos investimentos diretos externos em países em des envolvimento. Estudos Econômicos , 35 (4), 631 -655. doi: 10.1590/S0101 -41612005000400002 Omar, J. H. D. (2008) . Taxa de juros: comportamento, determinação e implic ações para a economia brasileira. Revista de Economia Contemporânea , 12 (3), 463 -490. doi: 10.1590/S1415 - 98482008000300003 Organisation for Economic Co -Operation and Dev elopment . (2008) . Benchmark definition of foreign direct investment . Retrieved f rom http://www.oecd.org/daf/inv/investmentstatisticsandanalysis/40193734.pdf Rei chert, F., Beltrame, R., Corso, K., Trevisan, M., & Zawislak , P. (2011) . Technological capability ’s predictor variable s. Journal of Technology Management & Innovation , 6(1), 14 -25. doi: 10.4067/S0718 -27242011000100002 Romer, P. (1986) . Increasing returns and long -run growth . Journal of Political Economy, 94 (5), 1002 - 1037. Sociedade Brasileira de Estudos de Empresas Transnacio nais e da Globalização Econômica. (2007, agosto) . Fortalecimento da autonomia das agências reguladoras favorece fluxos de investimento direto estrangeiro. Boletim Sobeet, Ano 6(48), 1 -3. Solow, R. (1956) . A contribution to the theory of economic growth. The Quarterly Journal of Economics, 70 (1), 65 -94 . doi: 10.2307/1884513 Sonagli o, C. M., Zamberlan, C. O., Lima, J. E., & Campos, A. C. (2010) . Evidências de desindustrialização no Brasil: uma análise com dados em painel. Economia Aplicada , 14 (4), 347 -372. doi: 10.1590/S1413 -80502010000400005 Thang, T. T. (2011) . Productivity spillovers from foreign direct investment: what if productivity is no longer a black b ox? South East Asian Journa l of Management, 5(1) 1 -18. United Nations Conference on Trade and Development . (2003) . FDI Policies for development: national and international perspectives . New York, NY: United Nations. Retrieved from http://unctad.org/en/Docs/wir2003overview_en.pdf Productivity Spillovers from Foreign Direct Investment 45 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar United Nations Conference on Trade and Development . (2007, outubro ). Crescimento pulverizado do investimento direto estrange iro em 2006. Anais da Conférence des Nations Unies Pour le Commerce et le Développement , Geneva, Switzerland. Wooster, R. B., & Diebel, D. S. (2010) . Productivity spillovers from foreign direct investment in developing countries: a meta -regression ana lysis. Review of Development Economics, 14 (3), 640 -655. doi: 10.1111/j.1467 -9361.2010.00579.x N. C. P. Bruhn, C. L. L. Calegario 46 BAR, Rio de Janeiro, v. 11, n. 1, art. 2, pp. 22-46 , Jan ./Mar . 201 4 www.anpad.org.br/bar APPENDIX Technological intensity of i ndustries was classified into fo ur categories: 1. Low -tech intensity industries: food product and beverage manufacture, tobacco product manufacture, textile manufacture, apparel and accessor y article manufacture; leather preparation and leather goods manufacture, travel article, wood prod uct manufacture, pulp manufacture, paper products and recycling. 2. Medium -low technological intensity industries: rubber and plastic product manuf acture, coke manufacture, refined petroleum, nuclear fuel development and alcohol production; metallurgy, metal product manufacture - except for machinery and equipment, and other transport equipment construction and vessel repair manufacture. 3. Medium -high technological intensity industries: electrical machinery and equipment manufacture, motor vehicle manufacture, trailers and bodies, chemical and pharmaceutical manufacture, railroad equipment and transport equipment manufacture. 4. High -tech intensity industr ies: pharmaceuticals, office machinery and computer equipment manufacture, electronic material and communication equipment manufacture, instrumentation equipment and medical manufacture, precision instruments and industrial automation optical equipment, ti mers and clocks, electronic and communication equipment and apparatus manufacture. An annotated bibliography of at least five additional resources related to the research topic by Day 7 of Week 2.***No plagiarism*** It must be less than 30\% match on authenticity report.***PLEASE DO International Journal of Forecasting 32 (2016) 1317–1339 Contents lists available atScienceDirect International Journal of Forecasting journal homepage:www.elsevier.com/locate/ijforecast Global equity market volatility spillovers: A broader role for the United States Daniel Buncic a, ∗ ,Katja I.M. Gisler b a Institute of Mathematics & Statistics, University of St. Gallen, Switzerland b School of Economics & Political Science, University of St. Gallen, Switzerland a r t i c l e i n f o Keywords: Realized volatility HAR modelling and forecasting Augmented HAR model US volatility information VIX International volatility spillovers a b s t r a c t Rapach et al. (2013) recently showed that U.S. equity market returns contain valuable information for improving return forecasts in global equity markets. In this study, we extend the work of Rapach et al. (2013) and examine whether U.S.-based equity market information can be used to improve realized volatility forecasts in a large cross-section of international equity markets. We use volatility data for the U.S. and 17 foreign equity markets from the Oxford Man Institute’s realized library, and augment our benchmark HAR model with U.S. equity market volatility information for each foreign equity market. We show that U.S. equity market volatility information improves the out-of-sample forecasts of realized volatility substantially in all 17 foreign equity markets that we consider. Not only are these forecast gains highly significant, they also produce out-of-sample R2 values of between 4.56\% and 14.48\%, with 9 being greater than 10\%. The improvements in out- of-sample forecasts remain statistically significant for horizons up to one month ahead. A substantial part of these predictive gains is driven by forward-looking volatility, as captured by the VIX. © 2016 International Institute of Forecasters. Published by Elsevier B.V. All rights reserved. ‘ . . . since the US equity market is the world’s largest, investors likely focus more intently on this market, so that information on macroeconomic fundamentals relevant for equity markets worldwide diffuses gradually from the US market to other countries’ markets .’ [Rapach,Strauss,&Zhou,2013, p. 1635] 1. Introduction In a recent influential paper,Rapachet al.(2013) showed that the equity market returns of the United States ∗ Correspondence to: Institute of Mathematics and Statistics, Bo- danstrasse 6, 9000 St.Gallen, Switzerland. E-mail addresses: [email protected](D. Buncic), [email protected](K.I.M. Gisler). URL: http://www.danielbuncic.com(D. Buncic). (US) have significant predictive power for forecasting eq- uity returns in a large cross-section of international eq- uity markets. This predictive power is attributed to the leading role played by the US in generating relevant macroeconomic and financial information for both US and non-US investors.Rapachet al.(2013) argue that information frictions cause information to diffuse only gradually from the US to other equity markets around the world, leading to lagged US returns having predic- tive content. The US is the world’s largest economy, is a large and important trading partner for many coun- tries, and has the world’s largest equity market in terms of market capitalization. Thus, when forming investment decisions, investors who take a global investment per- spective are focused intently on not only developments in macroeconomic and financial fundamentals in the http://dx.doi.org/10.1016/j.ijforecast.2016.05.001 0169-2070/ ©2016 International Institute of Forecasters. Published by Elsevier B.V. All rights reserved. 1318 D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 US, but also the formation of expectations and risk premia that arise from this process. 1 The objective of this study is to provide a first compre- hensive analysis of the predictive content of US-based eq- uity market volatility information for volatility forecasts in a large cross-section of 17 international (non-US) equity markets. For this purpose, we use daily realized volatility data from the Oxford Man Institute’s realized library and augment the well-known and widely used heterogeneous autoregressive (HAR) model ofCorsi(2009) with (lagged) daily, weekly and monthly US realized volatility and VIX HAR components. We utilize the HAR model ofCorsi (2009) as our benchmark realized volatility model for mea- suring the contribution of US-based volatility information to realized volatility forecasts in international equity mar- kets, employing standard in-sample and out-of-sample evaluation criteria. In this context, our study can be viewed as an extension of the work ofRapachet al.(2013), but with our analysis focussing on the role of the US as a source of relevant volatility information. We find US-based volatility information to play an overwhelmingly strong role for all 17 international equity markets that we consider. Our study is related to a growing body of volatility spillover literature. The literature on spillovers in interna- tional equity markets goes back to the research ofEunand Shim(1989),Hamaoet al.(1990) andLinet al.(1994). More recent studies include those ofBaurandJung(2006) and Savva,Osborn,andGill(2009), among others. The contri- butions to this body of literature typically differ in their definitions of the interdependence measure adopted and the modelling approaches used. For example,Hamaoet al. (1990) use a generalized autoregressive conditional het- eroscedastic (GARCH) type of model to analyze spillovers across three major equity markets. They define spillovers as impacts from foreign stock markets on the conditional mean and variance of daytime returns in the subsequently traded markets. Their results show evidence of volatility spillovers across these markets. Similarly,Linet al.(1994) employ a signal extraction model with GARCH innovations in order to analyze spillover effects between two major eq- uity markets. In contrast toHamaoet al.(1990),Linet al. (1994) do not find any evidence of spillover effects be- tween the two stock markets, but rather attribute their 1 The New York Stock Exchange (NYSE) is by far the largest equity market in the world, with a market capitalization of over 21 Trillion US Dollars (as of the end of 2014). The second is the NASDAQ, with a market capitalization of around 7 Trillion. Tokyo and London are the next biggest, with market capitalizations of around 4 Trillion. Moreover, an abundance of economic and financial data are released every day. These range from soft survey data related to durable goods, inventories, employment reports, the ISM (manufacturing) index, PMIs (purchasing manager indices) and the like, to hard data releases related to jobless claims, home sales, residential construction, personal income and outlays, PPI, CPI, employment and GDP figures. International financial agents and the financial media focus on these releases intently. Also, in terms of a calender (or trading) day timeline, it is the last (or one of the last) equity markets to close. As market participants begin work on a given day, they naturally look at important financial and economic developments in the US first. The dominant role of the US market as a source of both return and volatility transmission in international equity markets has been documented in numerous multi-country studies (see for exampleBecker, Finnerty,&Friedman,1995;Engle,1990;Hamao,Masulis,&Ng,1990; King&Wadhwani,1990;Lin,Engle,&Ito,1994, and others). contradictory results to non-synchronous trading and stale quotes at opening time. On the other hand,EunandShim (1989) apply a vector autoregressive model (VAR) to stock market returns from nine international markets, and use simulated responses to trace the spillover effects of inter- national stock market shocks. They find that US stock mar- ket shocks are transmitted to the other markets quickly, but not vice versa, highlighting the dominant role of the US. More recently,Savvaet al.(2009) use a dynamic con- ditional correlation (DCC) model to analyze return and volatility spillovers across four major stock markets. Their results show that both domestic stock prices and volatili- ties are subject to spillover effects. In fact, they find more evidence of spillovers from Europe to the US than the other way around. However, this finding might be attributable to the pseudo-closing approach that they use in order to avoid synchronous trading. With the increasing availability of high-frequency data, the literature on volatility spillovers has again gained mo- mentum (seeBonato,Caporin,&Ranaldo,2013;Diebold &Yilmaz,2014,2016;Dimpfl&Jung,2012;Fengler& Gisler,2015, among others).DieboldandYilmaz(2014) model realized volatility as a vector autoregressive process and define volatility spillovers based on a multiple-step- ahead forecast-error variance decomposition. Their results suggest that there are strong realized volatility spillovers across financial institutions, particularly during crises. Us- ing a similar approach,FenglerandGisler(2015) extend the results ofDieboldandYilmaz(2012,2014)by including realized covariances in the spillover transmission mech- anism. They show that realized covariance spillovers are substantial, and allow for an earlier detection of the recent financial and debt-ceiling crises that are attributable to a flight-to-quality phenomenon.Bonatoet al.(2013), on the other hand, define spillovers as the dependence of real- ized covariance on cross-lag realized covariances. They model realized covariance matrix as a Wishart autoregres- sive process, and find that sector and currency covariance spillovers improve the forecasting performance. Similarly, DimpflandJung(2012) model realized volatility and re- turn spillovers around the globe in a structural VAR frame- work. They find significant return and realized volatility spillovers that also result in forecast improvements. 2 In summary, the spillover literature has analyzed return and volatility spillovers in international stock markets exten- sively. However, the literature to date has not analyzed the predictive content of US realized volatility information for realised volatility forecasts in a large cross-section of inter- national equity markets. Our study aims to fill this gap. Although our study is related to the volatility spillover literature, we intentionally avoid the use of (structural) VAR approaches for modelling the information flow from the US to international equity markets. Standard structural VAR models require assumptions on the causal ordering 2 The modelling of spillover effects also plays a much broader role in the financial stability literature. For instance, given the role of US volatility and an interconnected world, it may be important to account for US-based information when designing macro-prudential stress tests, especially for Eastern European countries. See for instanceBuncicandMelecky(2013) for a recent study as to how this could be implemented. D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 1319 if impulse responses or forecast error variance decompo- sitions are used as measures of spillovers. This may not be a problem for smaller VARs, where the causal order- ing is known from the underlying assumption as to which equity market generates the most important information (i.e., the US), for instance, or in cases where the causal or- dering is based on the chronological structure of the mar- kets that are analysed (see for exampleDimpfl&Jung, 2012;Knaus,2014). Nevertheless, since we are considering realized volatility data for 17 international equity markets, it becomes much more difficult to justify the ordering of the countries in a structural VAR. Also, overlapping trading hours mean that one can only analyze up to three different international stock markets (i.e., over three different trad- ing/time zones). Moreover, estimating unrestricted VARs with large numbers of variables is highly inefficient, lead- ing to poor out-of-sample forecast performances. Thus, we prefer to examine the role of the US as a source of volatility information within our proposed simple augmented HAR modelling framework. Using daily realized volatility data for the US and 17 international equity markets, covering the period from January 3, 2000, to November 13, 2015, we find the US eq- uity market to play a strong role internationally as a source of volatility information. Our in-sample results show that US-based volatility information is jointly highly significant. The daily and weekly HAR components of the log VIX, to- gether with the daily and monthly HAR components of the US realized volatility, are the most important sources of volatility information from the US. For some equity mar- kets, such as the All Ordinaries and the EURO STOXX 50, the parameter estimate on the daily US HAR component has a larger magnitude than its own daily HAR component, suggesting that the previous day’s high frequency volatil- ity information from the US is more important than its own lagged volatility. Moreover, our in-sample analysis shows that the low frequency volatility component from the US has a negative effect on the realized volatility in non-US equity markets. This finding is consistent across all 17 of the international equity markets that we consider. Our out-of-sample analysis shows that one-step-ahead forecasts from the augmented HAR model with US volatility information are highly significant, yieldingClark andWest(2007) adjusted t-statistics of at least 8.4 and as high as 15.7, indicating rather strong rejections of the null hypothesis of no forecast improvement. The one-step- ahead out-of-sample R2 values range from 4.56\% (Hang Seng) to 14.84\% (All Ordinaries), and are above 10\% for nine of the 17 equity markets that we analyse. Thus, the forecast improvements are not only highly statistically significant, but also sizeable economically. To put these magnitudes into perspective,PattonandSheppard(2015) recently documented improvements in out-of-sample R2 values of in the order of 2.5\%–3\% by splitting the volatility into bad and good volatility states (in addition to various other considerations related to leverage and signed jumps). Thus, improvements in excess of 10\% are substantial. Our out- of-sample analysis also shows that forecast improvements are experienced consistently over the full out-of-sample period, and are not driven purely by individual episodes. The forecast improvements for the multiple-step-ahead horizon remain highly significant (at the 1\% level) for all 17 international equity markets at the five-day-ahead (one week) and 10-day-ahead (two week) horizons, and start to deteriorate at the 22-day-ahead (one month) horizon. The improvements in the 22-day-ahead forecasts remain significant and produce positive out-of-sample R2 values for 12 of the 17 equity markets that we analyse. Overall, our results show that US-based volatility data are most informative for forecasts of realized volatility for the All Ordinaries index and all of the European equity markets that are included in our comparison. The remainder of the paper is organised as follows. In Section2, we outline how realized volatility is mod- elled and how we extend the standard HAR model ofCorsi (2009) by augmenting it with US-based information about equity market volatility. The data that we use in the study are described in detail in Section3. In Section4, we eval- uate the importance of US-based volatility information for the determination of volatility in 17 international (non-US) equity markets, by means of in-sample and out-of-sample evaluations. In Section5, we provide an analysis of the ro- bustness of our findings. Lastly, we conclude the study. 2. Modelling the volatility This section outlines the modelling approach that we use to assess the role played by US equity market volatility information in improving realized volatility forecasts in a large cross-section of international equity markets. Before describing the empirical model that we employ for modelling and forecasting realized volatility in international equity markets, we first briefly describe the background that links empirical realized volatility to its theoretical counterpart, integrated volatility. 2.1. Theoretical framework Let p t denote the logarithm (log) of an asset price at time t. The log asset price is assumed to be a continuous- time diffusion process that is driven by Brownian motion, with the dynamics described by the following stochastic differential equation: dp t= µ tdt +σ tdW t, (1) where µ tis a predictable and locally bounded drift term, σ t is a càdlàg volatility process that is bounded away from zero, and W tis a standard Brownian motion. The quadratic variation (QV) process of p t is given by 3 : QV t=  t 0 σ 2 s ds . (2) In the absence of jumps, as is the case in our setting in Eq. (1), the term  t 0 σ2 s ds in Eq.(2)is known as the integrated variance (IV) of the process p t. 3 The quadratic variation process of p t is defined as [p t] = plim m→∞  m k = 1( p (t k ) − p(t k − 1)) 2 , where plim denotes convergence in probability, and 0 =t 0 ≤ t 1 < · · · RV t= m  i = 1 r 2 t ,i, (3) and its square root √ RV tis known as the realized volatility . The general properties of the estimator in Eq.(3)are summarised byAndersen,Bollerslev,Diebold,andLabys (2003). 2.2. Empirical volatility model There exist three broad classes of empirical models for RV. The first belongs to the traditional ARMA and fraction- ally integrated ARMA (ARFIMA) classes of long-memory time series models for RV (seeAndersenet al.,2003;Bail- lie,1996;Baillie,Bollerslev,&Mikkelsen,1996;Comte &Renault,1996,1998, among many others). The second class considers nonlinear time series models, where long- memory patterns in RV are generated spuriously from nonlinear short-memory models with structural breaks or regime switches (see for instance the papers byChen,Här- dle,&Pigorsch,2010;Fengler,Mammen,&Vogt,2015; McAleer&Medeiros,2008, and others). The third belongs to the class of heterogeneous autoregressive (HAR) mod- els for RV, as initially introduced into the realized variance modelling literature byCorsi(2009). We use the HAR model ofCorsi(2009) as our bench- mark RV model for each of the foreign equity markets that we consider. The HAR model has a cascade-type structure, where the volatility at any point in time is constructed as a linear combination of daily, weekly and monthly volatility components. This temporal cascade structure is motivated by the so-called heterogeneous market hypothesis (HMH) ofMülleret al.(1993), where it is assumed that finan- cial markets are populated by heterogeneous agents, each with different endowments, risk profiles, institutional con- straints and information processing capabilities, as well as various other characteristics (seeCorsi,2009, for a more detailed discussion). The defining feature of the HAR model is that each agent has a different time horizon for trading. The intuition is that the short-term volatility does not mat- ter to a long-term investor, whereas the long-term volatil- ity is still of importance to short-term investors because of its impact on the investment opportunity set. To formalise the structure of the HAR model for RV, let log RV ( d ) t = log RV t, log RV (w) t =1 5  5 i = 1 log RV t+ 1− iand log RV ( m ) t =1 22  22 i = 1 log RV t+ 1− ibe the daily, weekly, and monthly HAR components. The HAR model is then defined as 4 : log RV t+ 1 = b 0 + b( d ) log RV ( d ) t + b(w) log RV (w) t + b( m ) log RV ( m ) t + ϵ t+ 1, (4)4 Note here that the original formulation of the HAR model byCorsi (2009) used RV instead of log RV in the HAR specification in Eq.(4). where ϵ t+ 1 is an innovation term. One of the main at- tractions of the HAR model in Eq.(4)is its simplicity. Once the daily, weekly, and monthly volatility components have been constructed, the HAR model can be estimated by ordinary least squares (OLS) regression. Moreover, the HAR model is an extremely difficult benchmark model to beat in out-of-sample forecast evaluations, due to its par- simonious setup (seeCorsi,Audrino,&Renó,2012, for a recent survey of different types of models for RV that have been evaluated against the HAR model). Since we are in- terested primarily in a real time out-of-sample compari- son of the predictive content of US equity market volatility information on the volatility in other global equity mar- kets, it is necessary to update the model parameters of interest recursively when constructing the forecasts. Un- like AR(FI)MA and other more general nonlinear time se- ries models, which require a numerical optimisation of the likelihood function, and therefore are time consuming to estimate, as well as frequently being numerically unstable, the HAR model in Eq.(4)can be estimated efficiently and accurately by standard OLS procedures. At this point, we should also highlight the fact that the HAR model ofCorsi(2009) has undergone numerous re- finements since its initial introduction. For instance, some recent evidence suggests that separating the quadratic variation process in Eq.(2)into continuous and jump com- ponent parts can lead to better out-of-sample forecasts (see for instanceAndersenet al.,2007;Corsiet al.,2010; Corsi&Renó,2012). Moreover, allowing for nonlinear and asymmetric effects in the HAR model, such as the leverage effect, also seems to be beneficial for out-of-sample fore- casting (seeBollerslev,Litvinova,&Tauchen,2006;Chen &Ghysels,2011;Corsi&Renó,2012;Patton&Sheppard, 2015, among others). Nevertheless, in spite of these find- ings, we want to abstract from the inclusion of such refine- ments of the HAR model in this study, and instead focus our attention solely on the role of the US as a source of information in relation to international asset price volatil- ity, and, most importantly, on the question of whether this information can be exploited in order to improve fore- casts of the realized volatility in other global equity mar- kets. 5 In studies using S&P 500 RV data, the improvements Nevertheless, there has been a shift toward modelling the log RV series. In the words ofAndersen,Bollerslev,andDiebold(2007,p. 704): ‘ from a modeling perspective, the logarithmic realized volatilities are more amenable to the use of standard time series procedures ’. Moreover, log transformed RV data are much closer to being normally distributed, and there is also no need to impose any non-negativity restrictions on the fitted and forecasted volatilities. We will therefore followCorsi,Pirino,andRenó (2010),CorsiandRenó(2012) and many others in the empirical RV literature and use log RV in the HAR model. 5 Evidently, as the number of regressors grows, one could also make the modelling of the HAR more flexible by using either a time-varying parameter model, like those that were used byBuncicandMoretto(2015), BuncicandPiras(2016) andGrassi,Nonejad,anddeMagistritis(2014), or a shrinkage estimator such as the lasso for variable selection, as was done byBuncicandMelecky(2014) in a cross-sectional setting. Nevertheless, despite the fact that our econometric modelling could be extended to address additional, potentially important features in the data, this would abstract further from our consideration of the information contained in US volatility data for forecasting the volatility in international equity markets. D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 1321 in out-of-sample forecast performances seem to be rather marginal relative to the magnitudes that we find at the in- ternational level. For instance, the models considered by PattonandSheppard(2015), which allow the volatility to be split into bad and good volatility components (in ad- dition to various other considerations related to leverage and signed jumps), lead to improvements in the out-of- sample R2 values of around 2.5\%–3\% points at best (see Ta- ble 6 ofPatton&Sheppard,2015). Another recent study byProkopczuk,Symeonidis,andWeseSimen(in press), which analysed the impacts of jumps on energy-related asset prices, found that jumps do not improve the out-of- sample forecasts of the volatility. Their findings are robust to a number of different jump detection procedures, and also to varying the size of the estimation window. We assess the value of US equity market volatility data for forecasts of the log RV in other equity markets around the world by augmenting each individual foreign (local) equity market’s benchmark HAR model with US volatility information. This is achieved by adding the log RV and log VIX HAR components from the US as predictor variables. Since the VIX displays a rather strong long-range dependence, it appears to be desirable to apply a HAR- type structure. In fact,Fernandes,Medeiros,andScharth (2014) recently assessed the success of different HAR type models in forecasting the VIX, and found that a pure HAR specification as perCorsi(2009) is very difficult to beat out-of-sample. 6 We specify the following augmentedHAR model for each of the 17 international equity markets that we consider: log RV t+ 1 = benchmark (local) HAR components of each foreign equity market    β 0 + β( d ) log RV ( d ) t + β(w) log RV (w) t + β( m ) log RV ( m ) t + β( d ) VIX log VIX ( d ) t + β(w) VIX log VIX (w) t + β( m ) VIX log VIX ( m ) t    US volatility information: VIX HAR components + β( d ) US log RV ( d ) t ,US + β(w) US log RV (w) t ,US + β( m ) US log RV ( m ) t ,US    US volatility information: RV HAR components + ϵUS t + 1, (5) where the daily, weekly, and monthly HAR components for US log RV and log VIX, denoted by log RV ( ·) t ,US and log VIX ( ·) t , are defined analogously to the local HAR components used above in Eq.(4). The log VIX tseries is the log of the Chicago Board Options Exchange (CBOE) volatility index (henceforth, VIX for short), β 0 is a standard regression intercept, and ϵUS t + 1 is again a random disturbance term. Our motivation for including the VIX as an additional source of volatility information in the augmented HAR model in Eq.(5)is as follows. Recall that the VIX measures the volatility implied by option prices on the S&P 500, thus 6 We thank an anonymous referee for suggesting that we also use a HAR structure for the log VIX process. In an earlier version of the paper, we used only log VIX ( d ) t = log VIX tas an additional control variable in the augmented HAR specification in Eq.(5). Adding the weekly and monthly HAR VIX components improved both the overall in-sample fit and the out- of-sample forecast performance of the augmented HAR model in Eq.(5). reflecting investors’ expectations about the stock market volatility over the next month. 7 Thus, the VIX is meant to provide not only a forward looking view on expectedUS eq- uity market volatility, but also a general sense of the risk aversion in the market. A higher value in the VIX is gen- erally taken as an indication that market participants an- ticipate an overall negative economic or financial outlook, and hence have an increased aversion to risk (seeBrun- nermeier,Nagel,&Pedersen,2009, for a discussion). This increased risk aversion is likely to spill over into other in- ternational equity markets, given the dominant position of the US in the world economy as a source of economic and financial information. Moreover, in a recent study,Grassi et al.(2014) documented that the VIX has some predictive power for S&P 500 realized volatility forecasts. We there- fore expect the VIX likewise to contain predictive informa- tion that can be exploited to improve the realized volatility forecasts in other international equity markets. 3. Data We obtain daily volatility data from the publicly- available Oxford-Man Institute’s Quantitative Finance Realized Library ofHeber,Lunde,Shephard,andShep- pard(2009). The Oxford-Man Realized Library uses high- frequency tick data from Reuters DataScope Tick History 8 to construct a whole suite of daily realized measures of the asset price variability, as well as providing the num- ber of transactions, the time span between the first and last observations, the close-to-open return, the local open- ing time, the high–low range, the high–open range, and the opening and closing prices for each series. 9 The library con- tains realized measures for four US and 17 foreign (non-US) equity price indices from January 3, 2000, to the present. Our sample ends on November 13, 2015. As our preferred estimator of asset price variation, we use the realized variance sampled at equally-spaced five minute intervals (simply ‘five minute RV’ henceforth). This is the estimator given under the heading ‘*.rv’in the 7 The VIX is computed as the weighted average of the implied volatilities of options on the S&P 500 index for a wide range of strikes, and mainly first and second month expirations. Note here thatChow,Jiang, andLi(2014) recently showed the VIX to be a biased measure of market expectations about the future volatility. Nevertheless, we include the VIX as a regressor in the HAR model in order to account for the potential predictive information that it may have for the volatility in other global equity markets, rather than trying to gauge whether it is an appropriate measure of volatility expectations in the US. 8 http://www2.reuters.com/productinfo/tickhistory/material/ DataScopeTickHistoryBrochure_260707.pdf. 9 The term ‘realized measures’ was coined byLiu,Patton,andSheppard (2015). The various types of realized measures that are included in the library are listed athttp://realized.oxford- man.ox.ac.uk/documentation/ estimators. With regard to the quality of the tick data,Heberet al.(2009) point out that the raw data from Reuters DataScope Tick History are already of a high quality. Nevertheless,Heberet al.(2009) still employ the high frequency data cleaning procedure described in detail athttp: //realized.oxford- man.ox.ac.uk/documentation/data- cleaningand in the references therein, in order to make the data suitable for econometric analysis. Also, our data are from Library Version 0.2. The url link to the data source ishttp://realized.oxford- man.ox.ac.uk/media/1366/ oxfordmanrealizedvolatilityindices.zip. 1322 D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 Oxford-Man Realized Library for each equity market data block. The choice of the five minute RV is due partly to simplicity and partly to robustness. In a recent extensive study of realized measures,Liuet al.(2015) highlighted the fact that there is little evidence to suggest that the five minute RV is outperformed significantly by any of the other realized measures that are considered in the benchmark comparison. In particular, when working with international equity market data,Liuet al.(2015,p. 294) pointed out that ‘‘ more sophisticated realized measures generally perform significantly worse ’’ than the five minute RV. We therefore use the five minute RV estimator ofHeber et al.(2009) – henceforth simply ‘RV’, to avoid cumbersome and repetitive language – throughout this study. Moreover, to make the magnitudes of the RV measure comparable to those of the VIX, we transform all RV series to annualised volatilities. 10 In total, we have access to realized measures data for 17 international equity markets that are included in the Oxford-Man Realized Library. These are the FTSE 100 (United Kingdom), the Nikkei 225 (Japan), the DAX (Germany), the All Ordinaries (Australia), the CAC 40 (France), the Hang Seng (Hong Kong), the KOSPI (South Korea), the AEX (The Netherlands), the Swiss Market Index (Switzerland), the IBEX 35 (Spain), the S&P CNX Nifty (India), the IPC Mexico (Mexico), the Bovespa (Brazil), the S&P TSX (Canada), the Euro STOXX 50 (Euro area), the FT Straits Times (Singapore), and the FTSE MIB (Italy). For the US, the library contains realized measures for four different equity market indices. These are the Dow Jones Industrial Average (DJIA), the Russel 2000, the Nasdaq 100 and the S&P 500. We use the S&P 500 as our key headline US equity market index. The Nasdaq 100 is a specialized technology industry index, and thus is defined too narrowly to be considered as a valid headline US equity market index. The Russel 2000, on the other hand, is likely to be too sensitive to volatility movements induced by the small cap nature of the index. From our point of view, only the DJIA qualifies as a viable alternative to the S&P 500, as it is an index that is focused on widely by the financial media, thus providing broad headline information about the performance of US equities. Nevertheless, one evident shortcoming of the DJIA is that it is composed of only 30 blue chip stocks, and we therefore find it to be too narrowly focused as well. Thus, our preference is to use the S&P 500 as our key equity market index for the US. 11 The VIX data that we include in the augmented HAR model in Eq.(5)are obtained from the St. Louis Fed FRED2 database. 1210 This is done by taking the five minute RV series, re-scaling it by 100 2 × 252, and taking the square root to be interpreted as the annualised volatility (in percentage terms). That is, the annualized realized volatility is equal to: (‘ ∗ .rv ′ × 100 2 × 252 )1 /2 . 11 However, we would like to stress that, while we have chosen the S&P 500 here, the results that we obtain change very little if we use the DJIA as the US headline index instead. The results based on the DJIA are available from the authors upon request. 12 The url of the database ishttp://research.stlouisfed.org/fred2/. The FRED mnemonic for the VIX is VIXCLS, and it contains daily closing prices (16:15 EST) of the Chicago Board Options Exchanges (CBOEs) volatility index. Table 1provides standard summary statistics on all of the (log transformed) RV and VIX data that are used in our study. In addition to the summary statistics inTable 1, we also show time series and autocorrelation function (ACF) and partial ACF (PACF) plots inFigs. 1and2, to provide further information about the data series that we use. The first to fifth columns ofTable 1show the equity index, the corresponding country, the full sample period, the number of observations T, and the percentage of missing entries (\%Miss). The percentages of missing entries were obtained by matching the dates from the Oxford-Man Realized Library to official trading dates data from Bloomberg. In columns six to twelve, we list standard sample statistics such as the mean, median (Med), standard deviation (Std.dev), skewness (Skew) and kurtosis (Kurt), as well as the minimum (Min) and maximum (Max) of each series. The last six columns (grouped in threes) provide the first to third order ACF and PACF (ACF(1–3) and PACF(1–3), respectively). We can see from the third column ofTable 1 that there are some differences with respect to the actual first data points across the various equity markets that are available. For all but two series, the first data point is on either the 3rd or the 4th of January 2000. For the S&P TSX (Canada), the sample starts on May 2, 2002, and for the S&P CNX Nifty (India) it starts on July 8, 2002. 13 The end of the sample is either the 12th or 13th of November 2015 for all series except for FT Straits Times (Singapore), which ends on the 18th of September 2015, due to data availability in the Oxford-Man Realized Library. Looking over the summary statistics inTable 1, one sees that the log RV data are distributed fairly symmetrically, with the means and medians lining up reasonably closely, the skewness being between 0 and 1, and the kurtosis being around 3 for all but four markets. 14 Interestingly, Bovespa and FT Straits Times have the lowest variations, with the standard deviations of log RV being only around 0.37, while those for the remaining series are closer to 0.5. The ACF and PACF entries inTable 1highlight the well- known long-memory property of volatility data. The most persistent log RV series are the KOSPI (South Korea) and the Swiss Market Index, with first order ACFs of 0.86 and 0.85, respectively, while the least persistent ones are the All Ordinaries and the Bovespa, with values of around 0.67. The VIX is the most persistent series overall, with an ACF(1) of 0.98. The long-memory property of realized volatility 13 Before July 8, 2002, the availability of realized measures data for the S&P CNX Nifty was extremely sparse. That is, only 100 data entries were available for the 653 entries before July 8, 2002 (553 missing entries). We therefore decided to delete all entries before July 8, 2002, and start the effective sample for the S&P CNX Nifty from July 8, 2002. There are three other equity markets with unusual missing data patterns that deserve mentioning: (1) the All Ordinaries (Australia), where data are missing for 15 consecutive days from July 4, 2014, to July 25, 2014; (2) the FT Straits Times (Singapore), where 43 consecutive entries are missing from January 2, 2008, to March 3, 2008; and (3) the Hang Seng (Hong Kong), which had 168 entries missing out of 300 between September 5, 2008 to November 3, 2009. All missing entries were deleted from the final data set used in the analysis. 14 These exceptions are the log RV series of Bovespa (Brazil) and Euro STOXX 50, which are close to 5, and S&P CNX Nifty (India) and IPC Mexico (Mexico), with kurtosis values of 4.1 and 3.6, respectively, thus showing somewhat heavier tails than a Gaussian random variable. D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 1323Table 1 Summary statistics of (log) RV and (log) VIX data. Equity index Country Full sample period T\%Miss Mean Med Std.dev Skew Kurt Min Max ACF(1–3) PACF(1–3) FTSE 100 United Kingdom 04.01.2000–13.11.2015 3986 0.63 2.41 2.37 0.51 0.47 3.14 1.13 4.68 0.84 0.81 0.79 0.84 0.34 0.21 Nikkei 225 Japan 04.01.2000–13.11.2015 3844 1.38 2.61 2.60 0.43 0.30 3.53 1.29 4.50 0.78 0.72 0.69 0.78 0.30 0.17 DAX Germany 03.01.2000–13.11.2015 4017 0.47 2.78 2.75 0.51 0.36 3.13 1.27 4.80 0.84 0.80 0.77 0.84 0.33 0.20 All Ordinaries Australia 04.01.2000–13.11.2015 3966 1.29 2.13 2.09 0.48 0.51 3.51 0.57 4.13 0.67 0.66 0.64 0.68 0.38 0.23 CAC 40 France 03.01.2000–13.11.2015 4040 0.45 2.70 2.70 0.48 0.31 3.18 1.16 4.73 0.83 0.79 0.77 0.83 0.33 0.19 Hang Seng Hong Kong 03.01.2000–13.11.2015 3643 7.47 2.54 2.51 0.41 0.60 3.88 1.27 4.65 0.72 0.70 0.68 0.72 0.38 0.23 KOSPI South Korea 04.01.2000–13.11.2015 3907 0.46 2.65 2.63 0.51 0.32 2.93 1.33 4.81 0.86 0.83 0.81 0.86 0.35 0.19 AEX The Netherlands 03.01.2000–13.11.2015 4039 0.50 2.60 2.55 0.50 0.48 3.18 0.87 4.56 0.84 0.81 0.78 0.84 0.33 0.19 Swiss Market Index Switzerland 04.01.2000–13.11.2015 3972 0.48 2.45 2.36 0.46 0.89 3.82 1.45 4.63 0.85 0.82 0.80 0.85 0.34 0.20 IBEX 35 Spain 03.01.2000–13.11.2015 4005 0.47 2.74 2.78 0.49 −0.04 2.88 1.17 4.61 0.84 0.80 0.78 0.84 0.32 0.21 S&P CNX Nifty India 08.07.2002–13.11.2015 3295 0.97 2.73 2.67 0.48 0.75 4.09 1.20 5.38 0.75 0.70 0.67 0.75 0.31 0.19 IPC Mexico Mexico 03.01.2000–13.11.2015 3967 0.76 2.42 2.36 0.49 0.62 3.58 1.07 4.74 0.68 0.64 0.61 0.68 0.33 0.20 Bovespa Brazil 03.01.2000–12.11.2015 3879 1.31 3.01 2.98 0.37 0.70 4.92 1.61 4.87 0.67 0.61 0.56 0.67 0.29 0.15 S&P TSX Canada 02.05.2002–13.11.2015 3379 0.68 2.15 2.08 0.52 0.84 4.07 0.75 4.56 0.79 0.76 0.73 0.79 0.35 0.20 Euro STOXX 50 Euro Area 03.01.2000–13.11.2015 4017 1.27 2.76 2.73 0.50 0.10 5.09 −1.07 5.11 0.77 0.72 0.70 0.77 0.32 0.21 FT Straits Times Singapore 03.01.2000–18.09.2015 3839 2.71 2.33 2.30 0.38 0.60 3.70 1.45 4.29 0.81 0.77 0.75 0.81 0.33 0.22 FTSE MIB Italy 03.01.2000–12.11.2015 4000 0.68 2.65 2.63 0.49 0.30 2.97 1.42 4.75 0.83 0.79 0.76 0.83 0.33 0.19 S&P 500 United States 03.01.2000–13.11.2015 3964 0.73 2.53 2.50 0.53 0.47 3.36 1.02 4.94 0.80 0.76 0.73 0.80 0.34 0.18 VIX (log) United States 03.01.2000–13.11.2015 3992 0.00 2.96 2.92 0.37 0.65 3.30 2.29 4.39 0.98 0.97 0.96 0.98 0.08 0.08 Notes: The table reports standard summary statistics of the 18 annualised (log) realized volatility series of the various equity markets and the (log) VIX. Columns one to five show the equity index, the corresponding country, the full sample period, the sample size T, and the percentage of missing trading days (\%Miss). Columns six to twelve show common sample statistics, namely the mean, median (Med), standard deviation (Std.dev), skewness (Skew) and kurtosis (Kurt), as well as the minimum (Min) and maximum (Max) of the series. The last six columns (grouped into blocks of three) provide the first- to third-order autocorrelation function (ACF) and partial ACF (PACF). The full available sample is from January 3, 2000, to November 13, 2015. 1324 D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339Fig. 1. Time series evolution of (log) RV and (log) VIX over the full available sample period for each series. data is also visible from the ACF and PACF plots inFig. 2. For instance, it is evident that the log RV series of Bovespa displays the shortest memory, while the Nikkei 225 has the most hyperbolic-looking ACF decay pattern. Finally, the time series plots of the log RV and log VIX inFig. 1show that the movement in volatility across equity markets is rather homogenous for major events such as the Lehman Brothers collapse in September 2008. One last, and potentially important, point that we would like to stress here is that the Oxford-Man Realized Library only uses intraday data collected over the official (local) trading hours of the respective equity markets of in- terest. That is, no variation due to overnight price changes is considered in the construction of the realized measures. Since we are using information from the US at time tto forecast the (log) realized volatility in all other foreign eq- uity markets at time t+ 1 (and further ahead), there is no overlap in the official trading hours between the US mar- ket’s previous day closing and the foreign market’s current day opening. 1515 The official trading hours of the New York Stock Exchange (NYSE) are from 9:30 to 16:00 Eastern Standard Time (EST), which is 14:30–21:00 Coordinated Universal Time (UTC) in (northern hemisphere) winter. Of 4. Assessing the value of US volatility information We begin our assessment of the importance of US eq- uity market volatility data and its usefulness for improv- ing the modelling and forecasting of the realized volatility in other international equity markets by looking at the in- sample contribution of US-based volatility information to the model. We then extend the analysis by using standard forecast evaluation techniques to determine whether these in-sample gains carry over into the out-of-sample forecast environment. Before evaluating the in-sample fit of the augmented HAR model in Eq.(5), it will be convenient to condense the representation of the model somewhat. For this purpose, the foreign equity markets that we include, the first one to open the next day is the Australian Securities Exchange (ASX) in Sydney at 10:00 Australian Eastern Standard Time (AEST), which is 00:00 UTC. During (northern hemisphere) summer, the UTC closing time for the US market is 20:00 UTC, while the ASX in Sydney opens at 23:00 UTC. Hence, there is a three-hour gap between New York closing and Sydney opening. Also, the switches to and from Daylight Saving Time (DST) do not occur on the same days. For the US, DST is ‘ on’ from March to November, while for Australia, DST is ‘ off’ from April to October. However, this is immaterial for our discussion, as it does not cause any overlap in trading hours. D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 1325Fig. 2. Autocorrelation function (ACF) and partial ACF (PACF) plots of all (log) RV series. let us define x t =  1 log RV ( d ) t log RV (w) t log RV ( m ) t  as the ( 1 × 4 ) vector of HAR components (including an intercept term) of the foreign equity market of interest, and let the two (1 × 3 ) vectors containing US equity volatility information be denoted by xVIX t = log VIX ( d ) t log VIX (w) t log VIX ( m ) t  and xUS t = log RV ( d ) t ,US log RV (w) t ,US log RV ( m ) t ,US  . We further define y t+ 1 = log RV t+ 1. Then, we can express the augmented HAR model in Eq.(5)in the following compact form: y t+ 1 = local volatility info  x tβ +xVIX t β VIX + xUS t β US    US volatility info + ϵUS t + 1, (6) where β= β 0 β( d ) β (w) β( m ) ′ , and β VIX = β ( d ) VIX β(w) VIX β( m ) VIX  ′ and β US = β ( d ) US β(w) US β( m ) US  ′ are the corresponding (4 × 1), ( 3 × 1) and (3 × 1) dimensional foreign and US parameter vectors, respectively. Similarly, the local equity market’s 1326 D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 HAR model in Eq.(4)can be written compactly as: y t+ 1 = x tb + ϵ t+ 1, (7) where x t is as defined above, b=  b 0 b( d ) b(w) b( m ) ′ , and ϵ t+ 1 is an error term. 4.1. In-sample evaluation We fit the HAR model in Eq.(6)to three sample periods, in order to gauge the magnitude and significance of the es- timated parameters in Eq.(6). We first estimate the model over the full data set available, then also consider the two sub-periods leading up to and following the Lehman Broth- ers collapse on September 15, 2008. Estimation results for the full period are shown inTable 2, while estimation re- sults for the two sub-periods are provided in Table A.1 and Table A.2 in theAppendix. In each table, the first column shows the foreign equity index of interest and the second the time period over which the model in Eq.(6)was fit- ted, while the remaining columns show the set of 10 point estimates of the augmented HAR model parameters, and two χ2 test statistics of a joint test of significance of all US parameters being different from zero (null hypothe- sis is H 0 : [ β VIX ; β US ] = 0 6× 1) and only those of the US log RV series being different from zero (null hypothesis is H 0: β US = 0 3× 1). In square brackets below the parame- ter estimates ( χ2 -test statistics), we show two-sided (one- sided) p-values computed using a heteroskedasticity and autocorrelation consistent (HAC) variance/covariance ma- trix estimator. 16 In addition to the tabulated in-sample estimation results, we also provide graphical representations of the β parameter estimates (excluding the intercept) over the three sub-periods inFig. 3. Each plot inFig. 3shows point estimates (very thin blue line) and the corresponding 95\% confidence intervals (light blue shading) for the full sample period. We then superimpose the point estimates obtained from the pre and post Lehman Brothers collapse periods (thick red and thin black lines, respectively) on these plots. The most notable in-sample fitting results can be summarised as follows. 17 First, US equity market volatility data from the previous trading day are highly informative. A formal test of the null hypothesis H 0 : [ β VIX ; β US ] = 0 6× 1 is rejected strongly by the data for all equity markets of interest. The values of the χ2 US - test statistic are between 95.09 (lowest) for the Hang Seng and 511.43 (highest) for the AEX over the full sample period. As the 1\% upper tail critical value of a χ2 random variable with six degrees of freedom is 16.81, we can see that these are fairly strong rejections. When assessing the significance of the log RV US predictors separately from the log VIX, as is shown in the last column ofTable 2, we can see that these remain highly significant, with p-values of less than 1\% in general, 16 We use a standard Bartlett kernel and aNeweyandWest(1994) rule of thumb bandwidth set equal to 4 (T /100 )2 /9 . 17 We would like to emphasize here that we offer only a descriptive assessment of the in-sample results, and are not interested in a structural interpretation of these estimates per se. As was pointed out by a referee, the parameter estimates that we report in the tables are reduced form estimates, and care must be taken to not interpret them as structural ones. with the only exceptions being the Bovespa and KOSPI series, which are not affected significantly by lagged US RV information. Second, it is evident from the plots inFig. 3that the estimates are quite stable over the three sample periods, generally remaining inside (or at least close to) the 95\% confidence intervals (CI) of the full sample period esti- mates. 18 Looking at the magnitudes of the parameter es- timates, we can see that the daily and weekly log VIX t components are highly significant. Moreover, the daily log VIX tcomponent is positive, while the weekly compo- nent is negative. The negative sign on the ˆ β (w) VIX coefficient is somewhat surprising, as it suggests that the weekly VIX component has a negative effect on the high frequency daily volatility component of the foreign equity market of interest. 19 Furthermore, we find a mixed picture in terms of significance for the US HAR components. That is, we find the daily US HAR component to be highly significant for all of the foreign equity indices except for the three North and South American indices and the KOSPI. These results are consistent with the χ2 US - and χ2 RV -test statistic results, which show that the Bovespa and KOSPI in partic- ular are not affected significantly by lagged US RV infor- mation. Moreover, due to the general trading hour overlap between the three North and South American equity mar- kets and the NYSE, most of the US-based equity market volatility information is probably transferred to the three North and South American equity indices on the same trad- ing day, potentially being responsible for the insignificant daily US HAR component. In addition to the daily compo- nent, we also find the monthly US HAR component to be significantly different from zero at the 1\% level for all Eu- ropean indices and the All Ordinaries. 4.2. Out-of-sample forecast evaluation Given the strong in-sample evidence of the importance of lagged US-based equity market volatility information for the determination of the international equity market volatility, we now assess the value of this information within an out-of-sample forecast environment. Below, we begin by outlining the general prediction setting and evaluation criteria that we use, then proceed to present the out-of-sample forecast evaluation results. 4.2.1. Prediction setting We follow the standard literature on realized volatil- ity forecasting (seeAndersenet al.,2007;Corsi&Renó, 18 Note here that we have plotted the confidence interval for the full sample period, which will contain much tighter intervals than the smaller pre and post Lehman Brothers collapse periods, due to the larger number of observations in the absence of any severe structural breaks. Thus, if these intervals include the point estimates of the two subperiods most of the time, we can take this as indicating that no substantial structural breaks have influenced the parameter estimates. 19 It should be clear here that the two weekly components are correlated, due to the cumulative construction of these series. Although it may seem that the negative sign could be attributed to this correlation, one would also expect to see highly inflated standard errors with multi- collinearity issues, resulting in largely insignificant point estimates. However, this is not the case here. Thus, we do not believe that the opposite sign structure is driven purely by the correlatedness between these two components. D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 1327Table 2 Augmented HAR model parameter estimates over the full sample period. Equity index Sample period ˆ β ˆ β VIX ˆ β US χ2 US -stat χ2 RV -stat FTSE 100 04.02.2000–13.11.2015 − 0.1103 0.1319 0.4250 0.3176 1.1338 −0.7812 −0.1645 0.0796 −0.0071 −0.1289 468.78 29.49 United Kingdom [0.0165] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.0248] [0.0000] [0.8676] [0.0027] [0.0000] [0.0000] Nikkei 225 07.02.2000–13.11.2015 0.1301 0.2971 0.3302 0.2779 0.8267 −0.6297 −0.1994 0.1023 −0.0566 0.0033 252.75 30.85 Japan [0.0078] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.0085] [0.0000] [0.1349] [0.9364] [0.0000] [0.0000] DAX 03.02.2000–13.11.2015 − 0.0252 0.1723 0.4360 0.3063 1.1279 −0.8449 −0.1554 0.0915 −0.0364 −0.1002 393.11 23.93 Germany [0.5620] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.0387] [0.0000] [0.4017] [0.0339] [0.0000] [0.0000] All Ordinaries 07.02.2000–13.11.2015 0.0838 0.0368 0.3720 0.5229 0.6772 −0.3721 −0.2950 0.2708 −0.0646 −0.1933 452.67 180.45 Australia [0.0766] [0.0603] [0.0000] [0.0000] [0.0000] [0.0028] [0.0002] [0.0000] [0.1530] [0.0001] [0.0000] [0.0000] CAC 40 03.02.2000–13.11.2015 − 0.0391 0.1431 0.4785 0.2525 1.2060 −0.9286 −0.1060 0.0866 −0.0203 −0.1160 484.65 30.08 France [0.3208] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.1089] [0.0000] [0.6045] [0.0033] [0.0000] [0.0000] Hang Seng 03.02.2000–13.11.2015 0.0970 0.1322 0.4103 0.3765 0.6160 −0.5222 −0.0708 0.0470 0.0429 −0.0737 95.09 13.77 Hong Kong [0.0349] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.3784] [0.0120] [0.2621] [0.0726] [0.0000] [0.0032] KOSPI 07.02.2000–13.11.2015 0.0734 0.2893 0.4128 0.2499 0.8755 −0.8524 −0.0339 0.0102 0.0333 −0.0102 207.64 3.50 South Korea [0.0944] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.6525] [0.5439] [0.3432] [0.8033] [0.0000] [0.3210] AEX 03.02.2000–13.11.2015 − 0.0576 0.1460 0.4822 0.2636 1.2354 −0.9707 −0.1128 0.1020 −0.0124 −0.1327 511.43 38.96 The Netherlands [0.1557] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.0884] [0.0000] [0.7501] [0.0011] [0.0000] [0.0000] Swiss Market Index 04.02.2000–13.11.2015 − 0.0089 0.1783 0.4877 0.2555 1.0473 −0.8159 −0.1286 0.0463 0.0194 −0.1058 422.33 16.04 Switzerland [0.7973] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.0422] [0.0039] [0.5450] [0.0038] [0.0000] [0.0011] IBEX 35 04.02.2000–13.11.2015 − 0.0290 0.1889 0.4884 0.2546 1.0647 −0.8797 −0.0357 0.0634 −0.0185 −0.1336 382.94 25.34 Spain [0.4776] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.5867] [0.0015] [0.6204] [0.0004] [0.0000] [0.0000] S&P CNX Nifty 07.08.2002–13.11.2015 0.1624 0.2663 0.3800 0.2817 0.5777 −0.5863 −0.0159 0.1513 −0.0210 −0.0873 200.38 58.48 India [0.0065] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.8650] [0.0000] [0.6721] [0.1279] [0.0000] [0.0000] IPC Mexico 03.02.2000–13.11.2015 0.0470 0.1420 0.2148 0.5839 1.1118 −0.6908 −0.3059 −0.0024 0.1191 −0.2130 269.58 17.41 Mexico [0.4040] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.0006] [0.9273] [0.0124] [0.0000] [0.0000] [0.0006] Bovespa 04.02.2000–12.11.2015 0.3182 0.2504 0.3621 0.2467 0.7810 −0.4710 −0.2896 −0.0054 −0.0134 0.0367 165.31 0.73 Brazil [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.0003] [0.7893] [0.7403] [0.4167] [0.0000] [0.8664] S&P TSX 05.06.2002–13.11.2015 − 0.0835 0.1531 0.3783 0.4155 1.1951 −0.8073 −0.1860 −0.0080 0.0150 −0.1636 273.36 20.91 Canada [0.0741] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.0310] [0.7308] [0.7640] [0.0021] [0.0000] [0.0001] Euro STOXX 50 03.02.2000–13.11.2015 − 0.0223 0.0790 0.4335 0.3447 1.2180 −0.8929 −0.1431 0.1798 −0.0364 −0.1910 393.92 60.33 Euro Area [0.6539] [0.0054] [0.0000] [0.0000] [0.0000] [0.0000] [0.0885] [0.0000] [0.4811] [0.0007] [0.0000] [0.0000] FT Straits Times 03.02.2000–18.09.2015 0.1254 0.2262 0.3737 0.3219 0.5266 −0.3673 −0.1998 0.0849 −0.0390 0.0236 163.59 40.46 Singapore [0.0035] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.0016] [0.0000] [0.2057] [0.4809] [0.0000] [0.0000] FTSE MIB 03.02.2000–12.11.2015 − 0.0271 0.1790 0.4731 0.2643 1.1828 −1.0343 −0.0089 0.0635 0.0012 −0.1292 437.33 22.45 Italy [0.5304] [0.0000] [0.0000] [0.0000] [0.0000] [0.0000] [0.9004] [0.0006] [0.9750] [0.0014] [0.0000] [0.0001] Notes: The table reports OLS regression estimates of the augmented HAR model parameters in Eq.(6)for each foreign equity index. Columns one and two show the equity indices and the corresponding full sample fitting periods. Columns three to 12 show the OLS parameter estimates, together with (two-sided) p-values, computed using heteroskedasticity and autocorrelation (HAC) robust standard errors, in square brackets below the estimates. The last two columns show the χ2 -test statistics ( χ2 US -stat and χ2 RV -stat) of a joint significance test with null hypotheses of H 0: [ β VIX ; β US ] = 0 6× 1 and H 0: β US = 0 3× 1, respectively, with corresponding (HAC-based) p-values in brackets below. 1328 D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339Fig. 3. Plots of all parameter estimates from the augmented HAR model over the full period and the pre and post Lehman Brothers collapse periods. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 2012, and others), and implement a ‘ direct’forecasting ap- proach. 20 That is, we define the (normalised) h-period log 20 SeeChevillon(2007),ChevillonandHendry(2005),Clementsand Hendry(1996),Marcellino,Stock,andWatson(2006), andPesaran,Pick, andTimmermann(2011), among others, for a motivation, evaluation and comparison of the direct forecasting approach to iterated forecasts. RV series as: y ( h ) t =1 h h  j = 1 y t− j+ 1 = 1 h h  j = 1 log RV t− j+ 1, (8) and re-formulate the predictive relations in Eqs.(6)and (7)for the general h-step-ahead long-horizon regression D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 1329 setting as: y ( h ) t = x t− hβ ( h ) + xVIX t − hβ ( h ) VIX + xUS t − hβ ( h ) US + ϵUS t (9) y ( h ) t = x t− hb ( h ) + ϵ t, (10) then compute h-step-ahead forecasts as ˆ y US t + h|t = x tˆ β ( h ) + xVIX t ˆ β ( h ) VIX + xUS t ˆ β ( h ) US (11) ˆ y t+ h|t = x tˆ b ( h ) . (12) The hsuperscripts on the β( h ) • and b( h ) terms (and their estimates) indicate that these are from the h-period off- set (or long-horizon predictive) regressions in Eqs.(9) and(10). 21 The forecast errors that correspond to the predictions in Eqs.(11)and(12)are defined as: ˆ e US t + h|t = y( h ) t + h − ˆ y US t + h|t (13) ˆ e t+ h|t = y( h ) t + h − ˆ y t+ h|t . (14) Mean squared forecast errors (MSFEs) are computed as: MSFE =1 T os T  t = T is ( ˆ e t+ h|t ) 2 (15) MSFE (US )= 1 T os T  t = T is ( ˆ e US t + h|t ) 2 , (16) respectively, for the two models. The terms T os and T is de- note the numbers of out-of-sample and in-sample obser- vations, where T os = T− T is − h+ 1, and Tis the full sample size. We use the first 500 observations, corresponding to approximately two years of data, as the in-sample fitting period. We consider 500 observations to be large enough to give reasonably precise estimates of all of the parameters required to initialise the out-of-sample forecasts. FollowingCorsiandRenó(2012),Neely,Rapach, Tu,andZhou(2014),Rapachet al.(2013) and others, we then use an expanding window (or recursive) forecasting scheme, where we add an extra observation to the 500 in-sample data points and then re-estimate the models to produce recursively updated parameter estimates and forecasts. Overall, this gives us a minimum of around 2600 data points that can be used to conduct a statistically meaningful out-of-sample forecast evaluation. We should stress again here that we use rather large in-sample fitting and out-of-sample evaluation periods, in order to ensure that our general conclusions regarding the improvements in forecast performances are not sensitive to the choices of these two windows. 4.2.2. Evaluation criteria We assess the out-of-sample forecast performance of the augmented HAR model in Eq.(6)by following the approaches ofCorsiandRenó(2012) and the recent lit- erature on forecasting the equity premium (seeCamp- bell&Thompson,2008;Neelyet al.,2014;Rapach 21 We report in-sample estimation results from the long-horizon regressions on the full sample in Table A.3, Table A.4, and Table A.5 in theAppendix. et al.,2013, and many others), and evaluate the fore- casts in terms of theClarkandWest(2007) mean squared forecast error (MSFE) adjusted t-statistic (denoted CW- statistic) and theCampbellandThompson(2008) out-of- sample R2 (denoted R2 os henceforth). 22 Since the augmented HAR model in Eq.(6)nests the standard HAR model in Eq.(7), we utilize theClarkandWest(2007) MSFE-adjusted t -statistic, which corrects for the bias that arises when the DM test is used to compare nested models. Following the suggestion byClarkandWest(2007, p. 294), the simplest way to compute the MSFE-adjusted t -statistic is to form the sequence: CW t+ h = (ˆ e t+ h|t ) 2 − (ˆ e US t + h|t ) 2    DM t+ h +  ˆ y t+ h|t − ˆ y US t + h|t  2    adj t+ h . (17) The DM t+ h term in the first part of Eq.(17)is the standard DieboldandMariano(1995) sequence that is computed to test for (unconditional) superior predictive ability. The second term, adj t+ h, is an adjustment term that arises due to the nested nature of the models being compared, and performs a bias correction (seeClark&West,2007, for more details). The CW-statistic (for horizon h) is then computed as: CW-statistic = CW  V ar ( CW ), (18) where CW =T− 1 os  T t = T is CW t+ h and Var ( CW )is the variance of the sample mean, which can be obtained simply as the HAC-robust t-statistic on the intercept term from a regression of CW t+ h on a constant. 23 The CW-statistic implements a test of the null hypoth- esis that the MSFE of the benchmark HAR model, which does not include US equity market volatility information, is equal to that of the augmented HAR model’s forecast in Eq. (6), against the one-sided alternative hypothesis that the benchmark’s MSFE is greater than that of the augmented HAR model. Hence, a rejection of the null hypothesis sug- gests that the forecasts from the augmented HAR model are significantly smaller than those from the benchmark HAR model on average. It should be highlighted again here that the CW-statistic is particularly suitable in the given con- text, as it is designed for the comparison of nested(fore- casting) models. Our benchmark model yields the standard HAR model, which can be obtained from the augmented HAR model by restricting [β VIX ; β US ] in Eq.(6)to 0 6× 1. The R2 os ofCampbellandThompson(2008) is computed as follows. Let MSFE (US )be the MSFE from the augmented HAR model including US volatility information, and let 22 Note here that we are performing simple pairwise forecast compar- isons between the augmented and benchmark HAR models for each for- eign equity market’s log RV series, rather than comparing forecasts from many models. Thus, aDieboldandMariano(1995) (DM) type test of un- conditional predictive ability is sufficient for our purpose of assessing the contribution of US-based volatility information to each foreign equity market’s volatility forecasts. 23 See also the discussion in Section 2.1 ofDiebold(2015) for more background on this in the context of the traditional Diebold–Mariano (DM) statistic. 1330 D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 Table 3 One-step-ahead out-of-sample forecast evaluation results (expanding window). Equity index Country Out-of-sample period T os MSFE Rel-MSFE R2 os CW-stat p-value FTSE 100 United Kingdom 14.03.2002–13.11.2015 3356 0.0582 0.8772 0.1228 14.7187 0.0000 Nikkei 225 Japan 22.04.2002–13.11.2015 3171 0.0656 0.9071 0.0929 11.3231 0.0000 DAX Germany 12.03.2002–13.11.2015 3374 0.0612 0.8905 0.1095 15.3556 0.0000 All Ordinaries Australia 21.03.2002–13.11.2015 3315 0.0899 0.8552 0.1448 15.4533 0.0000 CAC 40 France 15.03.2002–13.11.2015 3389 0.0554 0.8598 0.1402 15.5369 0.0000 Hang Seng Hong Kong 19.04.2002–13.11.2015 3005 0.0621 0.9544 0.0456 8.4199 0.0000 KOSPI South Korea 25.04.2002–13.11.2015 3239 0.0539 0.9277 0.0723 11.9394 0.0000 AEX The Netherlands 18.03.2002–13.11.2015 3389 0.0564 0.8585 0.1415 15.5860 0.0000 Swiss Market Index Switzerland 22.03.2002–13.11.2015 3333 0.0440 0.8763 0.1237 14.5667 0.0000 IBEX 35 Spain 27.03.2002–13.11.2015 3357 0.0550 0.8935 0.1065 14.6317 0.0000 S&P CNX Nifty India 15.09.2004–13.11.2015 2659 0.0810 0.9336 0.0664 10.0383 0.0000 IPC Mexico Mexico 19.03.2002–13.11.2015 3336 0.0928 0.9210 0.0790 11.4184 0.0000 Bovespa Brazil 17.04.2002–12.11.2015 3250 0.0617 0.9521 0.0479 11.1674 0.0000 S&P TSX Canada 24.06.2004–13.11.2015 2790 0.0800 0.9153 0.0847 10.0294 0.0000 Euro STOXX 50 Euro Area 11.03.2002–13.11.2015 3373 0.0822 0.8645 0.1355 13.0201 0.0000 FT Straits Times Singapore 01.04.2002–18.09.2015 3237 0.0377 0.9317 0.0683 10.1945 0.0000 FTSE MIB Italy 20.03.2002–12.11.2015 3356 0.0599 0.8894 0.1106 15.7027 0.0000 Notes: The table reports the one-step-ahead out-of-sample forecast evaluation results using an expanding estimation window for the 17 foreign equity markets that we consider. We produce the first out-of-sample forecast using an initial 500 (in-sample) data points, then expand this window. Columns one to four show the equity markets of interest, the corresponding country, the out-of-sample evaluation period and the number of out-of-sample observations T os . Columns five to seven then show the mean squared forecast errors (MSFEs) of the benchmark HAR model (without US volatility information), the relative MSFE (Rel-MSFE), computed as MSFE (US )/MSFE, where MSFE (US )and MSFE are from the augmented and benchmark HAR models respectively, and the CampbellandThompson(2008) out-of-sample R2 (R 2 os ), computed as R2 os = 1− MSFE (US )/MSFE. The last two columns report the Clark–West (CW) test statistics and the corresponding one-sided asymptotic p-values. MSFE denote the mean squared forecast error from the benchmark HAR model. Then, the R2 os comparing the performances of the two forecasts is defined as: R 2 os = 1− MSFE (US ) MSFE . (19) Intuitively, the R2 os statistic in Eq.(19)measures the pro- posed model’s MSFE reduction relative to the benchmark model. If R2 os > 0, the proposed model performs better than the benchmark model, while R2 os < 0 suggests that the benchmark model performs better. In addition to the CW-statistic ofClarkandWest(2007) and the out-of-sample R2 ofCampbellandThompson (2008), we also compute the cumulative difference be- tween the squared forecast errors of the two HAR models over the out-of-sample period. This cumulative difference (denoted cumSFE) is used commonly in the forecasting literature as a tool for highlighting the predictive perfor- mance over time of the proposed model relative to the benchmark model (seeGoyal&Welch,2008;Rapachet al., 2013, among many others). In our setting, this difference is defined as cumSFE t+ h = t  τ = T is  (ˆ e τ+ h|τ )2 − (ˆ e US τ + h|τ )2  , ∀ t = T is , . . . , T− G. (20) The cumSFE sequence allows us to analyse the changes over time in the forecast performances of the two models. A value of the cumSFE series that is above zero indicates that the cumulative sum of the squared forecast errors of the benchmark model is larger than that of the proposed augmented HAR forecasts, indicating that the benchmark’s forecasts are less accurate. Moreover, an upward-sloping cumSFE sequence means that the proposed augmented HAR model produces consistentlybetter predictions than the benchmark HAR model (i.e., without US volatility information). 4.2.3. Forecast evaluation results One-step-ahead results .Table 3presents the one-step- ahead out-of-sample forecast evaluation results for all 17 international equity markets that we consider, using an ex- panding (recursive) estimation window, with the initial in- sample period consisting of T is = 500 observations. The first four columns inTable 3show the foreign equity in- dex of interest, the corresponding country, the actual out- of-sample evaluation period, and the effective number of out-of-sample observations T os that are used. Columns five to seven show the MSFE of the benchmark HAR model, the relative MSFE (denoted Rel-MSFE and computed as MSFE (US )/ MSFE), and the R2 os ofCampbellandThompson (2008). The last two columns then show theClarkand West(2007) MSFE-adjusted t-statistic (CW-statistic) and the corresponding one-sided p-values. The evaluation results inTable 3show the strong positive effect of information about US equity market volatility on the out-of-sample forecasts of log RV in global equity markets. The CW-statistic is in excess of 8.4 for all 17 international equity markets that we consider, resulting in p-values that are effectively zero. The out-of-sample R2 values ofCampbellandThompson(2008) are as high as 14.48\%, 14.15\%, 14.02\% and 13.55\% for the All Ordinaries, AEX, CAC 40 and Euro STOXX 50, respectively, with the two lowest values, 4.79\% and 4.56\%, being recorded for the Bovespa and Hang Seng. Note here that these R2 os magnitudes are considerable. To put them in perspective, compare them to those reported byPattonandSheppard (2015), which allow the volatility to be split into bad and good volatility components (in addition to various other considerations related to leverage and signed jumps) and which yield R2 os improvements of (only) about 2.5\%–3\% D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 1331 points at best, at the one-step-ahead horizon (see Table 6 ofPatton&Sheppard,2015). Although it is difficult to compare our findings to theirs directly, since we consider different information sets, it should still be clear that the forecast improvements from augmenting the benchmark local HAR model with US-based volatility information are substantial. Note here also that we are using sample sizes of at least 2600 observations for the out-of-sample evaluation periods, and up to 3300. Thus, our test results are not sensitive to, or driven by, ‘ small sample issues’. We provide additional evidence of the strength of our out-of-sample forecast results by examining the evolution over time of the cumulative difference between the squared forecast errors from the augmented HAR model and those of the benchmark HAR model. This cumSFE series (at the one-step-ahead horizon, h= 1), as defined in Eq.(20), is plotted as the thin blue line inFig. 4. Recall that the cumSFE series is defined such that an increasing value indicates an improvement in the augmented HAR model’s predictive performance relative to the benchmark HAR model (i.e., the benchmark HAR model produces larger one-step-ahead out-of-sample forecast errors). In addition to the expanding (recursive) window based cumSFE series shown inFig. 4, we also compute the cumSFE series based on forecasts from a rolling window scheme, i.e., one that constructs the forecasts using a fixed-length estimation window of 500 observations when rolling through the out- of-sample period. This series is plotted as the thick orange line inFig. 4. Our intention here is to provide a visual indication that our expanding (recursive) window based out-of-sample forecast evaluation results are broadly similar to those obtained from a rolling window based set- up, and therefore are not sensitive to this choice. Examining the cumSFE series shown inFig. 4, we can summarise the most interesting results from these plots as follows. First, the cumSFE is (nearly) uniformly above zero for the entire out-of-sample evaluation period and for all 17 foreign equity markets that we consider. The main exceptions are the Bovespa index for Brazil and the Hang Seng index for Hong Kong, which do not appear to be above zero consistently until about October 2007, but which increase steadily thereafter. Second, the cumSFE series increases (nearly) monotonically for all series over the full out-of-sample period. There are occasional instances of ‘ flattening off ’ for some of the 17 equity markets, occurring largely around the time period between September 2008 and June 2010. Nonetheless, if one was to draw a hypothetical straight line from the beginning of the out- of-sample period until its end in November 13, 2015, one would find the cumSFE series to line up to such a straight line fairly closely. This highlights the consistent and steady improvement over time that is provided by the inclusion of US volatility information when forecasting the volatility in other international equity markets. Third, it is interesting to observe that the cumulative improvement of the augmented HAR model over the benchmark HAR is strongest for the All Ordinaries, the Euro STOXX 50, the CAC 40, the DAX and the AEX, and weakest for the Hang Seng, the FT Straits Times and the Bovespa equity indices. Overall, it is clear that, apart from the All Ordinaries, the European equity indices benefit the most from the inclusion of US-based equity market volatility information. The strong improvement in the log RV forecasts of the All Ordinaries makes sense because of the narrow time gap between NYSE closing in the US and ASX opening in Australia. Nonetheless, it is somewhat surprising to see here that the predictive improvement is much weaker for the other four Asian equity indices, namely the Nikkei 225, the Hang Seng, the FT Straits Times, and the KOSPI, where the trading gap is similarly short as for the All Ordinaries. 24 Of these four equity indices, the Nikkei 225 shows the largest forecast improvements when US volatility information is included, though the improvements are considerably smaller than for the All Ordinaries index. Looking at the one-step-ahead out-of-sample forecast evaluation results presented inTable 3, we can see that, in general, the predictability pattern in the European equity markets is fairly homogenous across the eight indices that we include. The improvement in the out-of-sample R2 of CampbellandThompson(2008) is between 10.65\% (IBEX 35) and 14.15\% (AEX). The improvements for the three North and South American equity indices are smaller than for the European equity markets overall, with the Brazilian Bovespa showing the smallest gain (4.79\%). In regard to this result, we conjecture that the general trading hour overlap between these markets and the NYSE means that most of the US-based equity market volatility information is transferred on the same trading day. The NYSE is open from 14:30 to 21:00 UTC (during winter). The IPC Mexico and S&P TSX trade over the same hours as the NYSE, while the Bovespa is open from 13:00 to 20:00 UTC. The HAR components of the respective foreign equity markets seem to absorb and carry most of the relevant volatility information in real time, thereby reducing the importance of lagged US volatility information. 25 Overall, our results highlight the strong out-of-sample predictive content of US volatility information for volatility forecasts in a broad range of international equity markets. Multi-step-ahead results . Our multi-step-ahead out-of- sample forecast evaluation results are presented inTa- ble 4. We followCorsiandRenó(2012) and construct (nor- malised) multi-period log RV forecasts for horizons of h= 5, 10 and 22 steps ahead, as defined in Eq.(8).Table 4is split into three parts, with each part corresponding to one of the three forecast horizons that we consider. The column entries inTable 4contain the same information as the one- step-ahead evaluation results reported inTable 3. Before discussing the multi-step-ahead forecast evalu- ation results, we would like to stress that we take partic- ular care when computing the HAC standard errors that 24 Both the Nikkei 225 and the KOSPI open at 00:00 UTC during summer, the same as the All Ordinaries, while the FT Straits Times and Hang Seng open at 01:00 and 01:20, respectively. 25 In a somewhat different context,Nikkinen,Mohammed,Petri,and Äjiö(2006) found that Latin American countries are not affected by US news announcements, which highlights the fact that they are less integrated with the US. Also, in the news effect and announcement literature,Brand,Buncic,andTurunen(2010) showed that European equity and bond markets react less to news from the US, such as initial unemployment claims, after conditioning on ECB announcements. 1332 D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339Fig. 4. Time series evolution of the cumulative difference between the squared one-step-ahead forecast errors from the benchmark HAR model and those from the augmented HAR model (cumSFE). The thin (blue) lines show the results computed on an expanding estimation window, using an initial in-sample fitting period of 500 observations. The thick (orange) lines show the corresponding rolling window (fixed T is = 500) results. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) are needed to construct the p-values of the CW-statistic. It is well known that h-step-ahead forecast errors follow at least an MA( h− 1) process. When computing the dif- ferences of the squared forecast errors from the two com- peting models in order to construct the CW-statistic, the CW t+ h sequence itself will be autocorrelated for h> 1. This autocorrelation can be sizable for large values of h. We employ a pre-whitening step, using an ARMA (1 ,1 ) as the approximating model for the CW t+ h sequence so as to re- duce the initial autocorrelation in the series, then apply a quadratic spectral (QS) kernel-based non-parametric HAC estimator to the residuals from the ARMA (1 ,1 ) model. FollowingAndrewsandMonahan(1992), we choose the bandwidth optimally, with an AR(1) as the approximating model for the ARMA (1 ,1 ) (pre-whitened) residuals, then re-colour to obtain the required HAC standard errors. 2626 That is, using the notation ofAndrewsandMonahan(1992), the bandwidth parameter is set to 1 .3221  ˆ α (2 ) T os  1/5 , where the constant ˆ α (2 ) = 4ˆ ρ 2 /( 1− ˆ ρ )4 , and ˆ ρ is the AR(1) parameter estimate obtained from an AR(1) regression of the (pre-whitened) residual series obtained We can see from the multi-step-ahead forecast evalua- tion results inTable 4that the forecast improvements rela- tive to the benchmark HAR model remain highly significant for all 17 international equity markets at the 5-day-ahead (one week), 10-day-ahead (two week), and 22-day-ahead (one month) horizons. At the 22-day horizon, the forecast improvements are only insignificant for the KOSPI and the S&P CNX Nifty. 27 To summarise the out-of-sample forecast evaluating results that we have presented in this section, it is clear that including lagged US equity market volatility information leads to substantial improvements in the out-of-sample predictions of volatility in all 17 of the international equity markets that we analyze. Moreover, this improvement has from the ARMA (1 ,1 ) model fitted to the CW t+ h sequence. We then ‘ re- colour ’ again to obtain the HAC variance, using the ratio of the square of the ARMA lag polynomials (seeAndrews&Monahan,1992for more details of the exact computations). 27 For more information about the influences of the different predictor variables, consult the long-horizon predictive regression results reported in Table A.3, Table A.4, and Table A.5 in theAppendix. D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 1333 Table 4 Multiple-step-ahead out-of-sample forecast evaluation results (expanding window). Equity index Country Out-of-sample period T os MSFE Rel-MSFE R2 os CW-stat p-value Forecast horizon h= 5 FTSE 100 United Kingdom 26.03.2002–13.11.2015 3348 0.0377 0.8826 0.1174 10.2683 0.0000 Nikkei 225 Japan 07.05.2002–13.11.2015 3163 0.0438 0.9123 0.0877 6.9091 0.0000 DAX Germany 22.03.2002–13.11.2015 3366 0.0412 0.9105 0.0895 9.7591 0.0000 All Ordinaries Australia 04.04.2002–13.11.2015 3307 0.0397 0.8515 0.1485 9.6919 0.0000 CAC 40 France 27.03.2002–13.11.2015 3381 0.0382 0.8768 0.1232 10.3148 0.0000 Hang Seng Hong Kong 02.05.2002–13.11.2015 2997 0.0311 0.9440 0.0560 4.8934 0.0000 KOSPI South Korea 08.05.2002–13.11.2015 3231 0.0350 0.9618 0.0382 5.5161 0.0000 AEX The Netherlands 28.03.2002–13.11.2015 3381 0.0403 0.8853 0.1147 10.7585 0.0000 Swiss Market Index Switzerland 05.04.2002–13.11.2015 3325 0.0316 0.9031 0.0969 9.5770 0.0000 IBEX 35 Spain 10.04.2002–13.11.2015 3349 0.0377 0.9084 0.0916 9.3822 0.0000 S&P CNX Nifty India 27.09.2004–13.11.2015 2651 0.0472 0.9512 0.0488 7.2642 0.0000 IPC Mexico Mexico 03.04.2002–13.11.2015 3328 0.0459 0.8873 0.1127 9.1160 0.0000 Bovespa Brazil 29.04.2002–12.11.2015 3242 0.0371 0.9539 0.0461 5.7047 0.0000 S&P TSX Canada 08.07.2004–13.11.2015 2782 0.0465 0.9191 0.0809 7.7913 0.0000 Euro STOXX 50 Euro Area 21.03.2002–13.11.2015 3365 0.0494 0.8696 0.1304 9.6366 0.0000 FT Straits Times Singapore 11.04.2002–18.09.2015 3229 0.0211 0.9166 0.0834 6.4696 0.0000 FTSE MIB Italy 03.04.2002–12.11.2015 3348 0.0394 0.9247 0.0753 9.5552 0.0000 Forecast horizon h= 10 FTSE 100 United Kingdom 11.04.2002–13.11.2015 3338 0.0390 0.9311 0.0689 8.0319 0.0000 Nikkei 225 Japan 21.05.2002–13.11.2015 3153 0.0429 0.9433 0.0567 5.4364 0.0000 DAX Germany 09.04.2002–13.11.2015 3356 0.0427 0.9455 0.0545 7.2399 0.0000 All Ordinaries Australia 18.04.2002–13.11.2015 3297 0.0382 0.8972 0.1028 7.8199 0.0000 CAC 40 France 12.04.2002–13.11.2015 3371 0.0407 0.9205 0.0795 8.0689 0.0000 Hang Seng Hong Kong 16.05.2002–13.11.2015 2987 0.0287 0.9559 0.0441 4.5394 0.0000 KOSPI South Korea 22.05.2002–13.11.2015 3221 0.0349 0.9879 0.0121 3.4735 0.0003 AEX The Netherlands 15.04.2002–13.11.2015 3371 0.0436 0.9236 0.0764 8.4435 0.0000 Swiss Market Index Switzerland 19.04.2002–13.11.2015 3315 0.0351 0.9463 0.0537 7.7335 0.0000 IBEX 35 Spain 24.04.2002–13.11.2015 3339 0.0391 0.9491 0.0509 7.1849 0.0000 S&P CNX Nifty India 11.10.2004–13.11.2015 2641 0.0454 1.0046 −0.0046 4.2106 0.0000 IPC Mexico Mexico 17.04.2002–13.11.2015 3318 0.0418 0.9262 0.0738 7.6535 0.0000 Bovespa Brazil 14.05.2002–12.11.2015 3232 0.0362 0.9812 0.0188 3.9384 0.0000 S&P TSX Canada 22.07.2004–13.11.2015 2772 0.0456 0.9407 0.0593 6.6146 0.0000 Euro STOXX 50 Euro Area 08.04.2002–13.11.2015 3355 0.0498 0.9221 0.0779 7.7728 0.0000 FT Straits Times Singapore 25.04.2002–18.09.2015 3219 0.0204 0.9540 0.0460 5.3990 0.0000 FTSE MIB Italy 17.04.2002–12.11.2015 3338 0.0412 0.9637 0.0363 6.8724 0.0000 Forecast horizon h= 22 FTSE 100 United Kingdom 16.05.2002–13.11.2015 3314 0.0461 0.9628 0.0372 5.9062 0.0000 Nikkei 225 Japan 25.06.2002–13.11.2015 3129 0.0486 0.9782 0.0218 4.3789 0.0000 DAX Germany 14.05.2002–13.11.2015 3332 0.0500 0.9783 0.0217 5.5171 0.0000 All Ordinaries Australia 23.05.2002–13.11.2015 3273 0.0410 0.9381 0.0619 6.1348 0.0000 CAC 40 France 17.05.2002–13.11.2015 3347 0.0478 0.9485 0.0515 6.3290 0.0000 Hang Seng Hong Kong 21.06.2002–13.11.2015 2963 0.0281 0.9978 0.0022 3.4927 0.0002 KOSPI South Korea 28.06.2002–13.11.2015 3197 0.0390 1.0362 −0.0362 0.8022 0.2112 AEX The Netherlands 20.05.2002–13.11.2015 3347 0.0524 0.9518 0.0482 6.3770 0.0000 Swiss Market Index Switzerland 24.05.2002–13.11.2015 3291 0.0437 1.0065 −0.0065 4.4579 0.0000 IBEX 35 Spain 30.05.2002–13.11.2015 3315 0.0437 0.9686 0.0314 5.9966 0.0000 S&P CNX Nifty India 22.11.2004–13.11.2015 2617 0.0494 1.0602 −0.0602 −0.2516 0.5993 IPC Mexico Mexico 22.05.2002–13.11.2015 3294 0.0424 0.9858 0.0142 5.4811 0.0000 Bovespa Brazil 19.06.2002–12.11.2015 3208 0.0399 1.0095 −0.0095 2.9363 0.0017 S&P TSX Canada 26.08.2004–13.11.2015 2748 0.0490 0.9675 0.0325 5.1390 0.0000 Euro STOXX 50 Euro Area 13.05.2002–13.11.2015 3331 0.0553 0.9515 0.0485 6.5706 0.0000 FT Straits Times Singapore 31.05.2002–18.09.2015 3195 0.0242 1.0026 −0.0026 4.4795 0.0000 FTSE MIB Italy 22.05.2002–12.11.2015 3314 0.0461 0.9785 0.0215 4.8580 0.0000 Notes: The table reports the multi-step-ahead out-of-sample forecast evaluation results for the 17 international equity markets that we consider. Forecasts for horizons h= 5,10 and 22 are shown in the top, middle and bottom panels, respectively. The target variable is the (normalised) multi-period log RV, as defined in Eq.(8). The columns are the same as those described inTable 3. The p-values corresponding to the CW-statistic are computed from HAC robust standard errors, where we conduct a pre-whiteningstep using an ARMA (1 ,1 ) model for the CW t+ h sequence in order to reduce the initial autocorrelation in the series, then apply a quadratic spectral (QS) kernel based non-parametric HAC estimator to the ARMA (1 ,1 ) residuals. We followAndrewsandMonahan (1992) and choose the bandwidth optimally with an AR(1) as the approximating model, then re-colourto obtain the HAC standard errors of the CW t+ h sequence. a lasting impact, affecting forecasts as far as one month ahead. The equity markets that are impacted most by the US volatility information are the Australian All Ordinaries index and all of the European equity indices in our sample. The weakest results are obtained for the South American equity markets, and some of the Asian markets. 1334 D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 5. Robustness checks In this section, we address some pertinent concerns in relation to the robustness of our out-of-sample forecast evaluation results. 28 In particular, we address questions related to: (i)most of the out-of-sample forecasting power coming from the log VIX, (ii)the role of each foreign equity market’s own forward- looking volatility information, and (iii)the importance of other European and/or Asian equity markets’ realized volatilities. All of the tables and figures required to support our dis- cussion below are provided in theAppendix. Here, we sim- ply summarize the main findings of the robustness checks. In all of the evaluations that we present, our reference model is the augmented HAR model, which uses only US volatility information. To conserve space and to improve the readability of the supporting figures and tables that are provided, we only report out-of-sample evaluation results based on an expanding (recursive) estimation window, us- ing the first 500 data points for the initial in-sample fitting. In all figures, we draw the reference results from the aug- mented HAR model in Eq.(6)with a blue line, to facilitate the comparison to previously plotted results. 5.1. Does the VIX drive all of the forecast improvement results? It is evident from the results reported inTable 2that the log VIX HAR components capture a substantial part of the overall in-sample improvement when both forward- looking and backward-looking US volatility information are included in the prediction model. This can be seen from the relative magnitude of the ˆ β VIX and ˆ β US coefficients, as well as the χ2 US and χ2 RV -statistics. We determine how much of the out-of-sample forecast improvement is driven by the VIX predictor alone by removing xVIX t from the augmented HAR in Eq.(6)and repeating the out-of-sample forecast evaluations, again against the benchmark HAR model in Eq.(7)as before. These evaluation results are reported in Table A.6, Figure A.1 and Table A.7. Overall, we can see that the VIX plays an important role in predicting the volatility in all 17 of the interna- tional equity markets that we consider. Nevertheless, the predictive performance is heterogeneous, and depends on both the forecast horizon and the foreign equity market being analysed. For instance, at the one-step-ahead hori- zon, we can see from Table A.6 that the forecast improve- ments remain highly significant for all 17 international equity markets. The lowest CW-statistic recorded now drops to around four (S&P TSX), while the largest one is still over 16 (All Ordinaries). However, the out-of-sample 28 An earlier version of this paper also assessed the impact of increasing the size of the in-sample fitting period to 1000 observations and using the Dow Jones Industrial Average as the headline US equity index on the out- of-sample results. Overall, our findings are not affected by these choices. These additional results are available upon request. R 2 values are uniformly lower, with some of them being as low as 0.69\% and 0.84\% for the Bovespa and S&P TSX one- step-ahead forecasts, though that for the All Ordinaries is still rather high, at 12.59\%. Comparing the cumSFE series of the full augmented HAR model (blue line) to that of the one that only includes US RV HAR components as regres- sors (brown line), plotted in Figure A.1, we can see that, apart from the All Ordinaries, and also the FT Straits Times, the S&P CNX Nifty and the Hang Seng to a lesser extent, the slopes of the cumSFE series are subdued considerably, with those of the Bovespa and S&P TSX in particular remaining rather flat over the entire out-of-sample period. For most of the other international equity market indices, the VIX HAR components account for approximately half of the cumu- lative predictive gains. One can see from the longer forecast horizon evaluation results reported in Table A.7 that the performance of the augmented HAR model without the VIX HAR components diminishes quickly. Although the CW-statistic remains sig- nificant at the 1\% level for all 17 equity markets at the five-day-ahead horizon, the overall improvement in the forecasts is noticeably weaker, resulting in much smaller R 2 os values. Again, the only exception here is the All Ordi- naries series, which yields an R2 os of 9.53\%. The improve- ments deteriorate further for the 10- and 22-day-ahead prediction horizons. Nevertheless, 14 of the 17 forecast improvements remain significant at the 1\% level for the 10-day-ahead horizon, though some of the R2 os values are rather small and/or negative. The R2 os for the All Ordinaries stays sizeable, at 7.00\%, followed by the Nikkei 225 and the IPC Mexico, with R2 os values of around 3.3\%. At the 22-day- ahead horizon, only the All Ordinaries and the Nikkei 225 retain significant and sizable predictive improvements in terms of out-of-sample R2 values. In summary, we can conclude that the improvements in forecasts up to one week ahead are significant and sizeable, and, with the exception of the Bovespa and S&P TSX equity indices, are notdriven solely by the VIX HAR components. Nevertheless, it is clear that the amount of predictive in- formation contained in the VIX when forecasting volatil- ity in international equity markets is large, and becomes increasingly important when constructing longer horizon predictions. 5.2. Controlling for other forward-looking volatility We have seen that a substantial part of the out-of- sample predictive gains for some of the 17 foreign equity markets is due to forward-looking volatility information, which we capture by including the (US) VIX in the aug- mented HAR model in Eq.(6). Two questions that arise are whether the S&P 500 option implied volatility index (VIX) captures allof the relevant forward-looking volatil- ity information for all markets, and how informative a for- eign equity market’s own option implied forward-looking volatility information is. 29 To assess the importance of forward-looking volatility information, as contained in the option implied volatility indices of each foreign equity 29 We thank an anonymous referee for pointing this out to us. D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 1335 market’s own VIX series, we obtain VIX data for 13 eq- uity indices from Bloomberg and add local VIX HAR com- ponents to Eq.(6)as additional predictors. 30 The 13 equity markets for which VIX data are available are listed below. Entry Region Equity market’s volatility indexCountry Data begins1 Oceania ASX200 VOL INDEXAustralia 03.01.2008 2 Asia HSI VOL INDEX Hong Kong03.01.2001 3 NIKKEI VOL INDEXJapan 05.01.2001 4 KOSPI 200 VOL INDEXSouth Korea 03.01.2003 5 INDIA NSE VOL INDEXIndia 02.11.2007 6 Europe VDAX NEW Germany 03.01.1992 7 VAEX AEX VOL The Netherlands04.01.2000 8 CAC40 VOL INDEXFrance 04.01.2000 9 FTSE100 VOL INDEXUnited Kingdom 05.01.2000 10 EURO50 VIX Euro Area 04.01.1999 11 Americas SP TSX60 VOL INDEX Canada 02.10.2009 12 MEXICO VOL INDEXMexico 29.03.2004 13 CBOE BRAZIL ETF VOL INDEXBrazil 17.03.2011To clarify what we do, let xVIX t ,FC =  log VIX ( d ) t ,FC log VIX (w) t ,FC log VIX ( m ) t ,FC  be a (1 × 3) vector of local VIX HAR com- ponents, where the standard daily, weekly and monthly components are computed as before. We assess the value added by including local forward-looking volatility in- formation in addition to the US predictors by modifying Eq.(6)to: y t+ 1 = local volatility info  x tβ +xVIX t β VIX + xUS t β US    US volatility info + forward-looking local volatility info    x VIX t ,FC βVIX FC + ϵUS t + 1, (21) where βVIX FC = β VIX (d ) FC βVIX (w) FC βVIX (m ) FC  ′ is a (3 × 1) vector of parameters that captures the impact of the foreign country’s local VIX information. All of the other terms in Eq.(21)are as defined previously. As is evident from the list of VIX indices above, we do not have any option implied volatility data available for the Swiss Market Index, the IBEX 35, the FT Straits Times and the FTSE MIB. Also, the available (local) VIX data for some of the equity markets that we include do not go as far back as our RV data (i.e., to the beginning of 2000). 30 Bloomberg also has VIX indices for Russia and South Africa, but these are not listed here because they are not used in our analysis. Also, there are two VDAX indices: an old version, with the mnemonic VDAX VSMI, and a new version. We use the new version, with mnemonic VDAX NEW.Rather than shortening the out-of-sample evaluation pe- riod or excluding these equity markets from the robust- ness analysis, we decided to replace the local market’s VIX HAR predictor vector xVIX t ,FC in Eq.(21)with a European (or Asian) VIX HAR ‘ factor’ predictor vector, which we de- note by fVIX t ,EU (or fVIX t ,ASIA for Asia). That is, let X(EU )= log {[(VDAX NEW ) (VAEX AEX VOL ) (CAC40 VOL INDEX ) ( FTSE100 VOL INDEX ) (EURO50 VIX )]} be the (T × 5 ) log- transformed data matrix consisting of all of the European VIX indices listed under entries 6–10 above. Then, the Eu- ropean VIX HAR factor is defined as the (1 × 3 ) vector f VIX t ,EU = [ fVIX t ,EU ( d ) fVIX t ,EU (w) fVIX t ,EU ( m ) ] , where fVIX t ,EU is the first principal component of X(EU ), with the daily, weekly and monthly HAR components (i.e., fVIX t ,EU ( d ) , f VIX t ,EU (w) , and fVIX t ,EU ( m ) ) computed as before. Similarly, for fVIX t ,ASIA , the first principal component is extracted from X(ASIA )= log{[(HSI VOL INDEX ) (NIKKEI VOL INDEX ) (KOSPI 200 VOL INDEX )]} . We then construct forecasts from Eq.(21)for all 17 in- ternational equity markets, using fVIX t ,EU in place of xVIX t ,FC for the Swiss Market Index and the IBEX 35, FT Straits Times, FTSE MIB, All Ordinaries, S&P CNX Nifty, Bovespa and S&P TSX indices. 31 These are then compared to the forecasts constructed from the augmented HAR model in Eq.(6), which only includes US volatility information in addition to the local HAR RV components. Before presenting and discussing these results, we would like to emphasise here that, since we are inter- ested chiefly in the out-of-sample predictive performance of US volatility information for each of the 17 international equity markets that we consider, we only report the out- of-sample evaluation results. Also, when constructing fore- casts from factor-based regression models, it is common to extract the factors recursively when rolling through the out-of-sample period so as to avoid concerns related to look-ahead biases; that is, using future data when con- structing the factors at time t. In order to tilt the out-of- sample prediction results using the European (or Asian) VIX HAR factor in favour of the local VIX model in Eq.(21), we use the full sample data to compute fVIX t ,EU (or fVIX t ,ASIA ) once and then roll through the out-of-sample data, instead of extracting the factor recursively. This should work in favour of the eight equity markets listed above for which no VIX data are available and the factor-based approach is used. Initially, we again present a visual assessment of the out-of-sample forecast gains by plotting the cumSFE se- quences of our augmented HAR model of Eq.(6)and the model that adds local VIX information (or a European VIX factor) to the predictor set, as defined in Eq.(21)in Figure A.2. 32 As before, both sequences are again computed rela- tive to the local HAR model given in Eq.(7), with the fore- cast horizon being one step ahead. That is, the blue lines in Figure A.2 are the same as the blue lines plotted inFig. 4. 31 Since the beginning date of the KOSPI 200 VOL INDEX is in 2003, the sample and forecasting periods for the Asian countries are shortened correspondingly. 32 Using fVIX t ,ASIA in place of fVIX t ,EU produces consistently worse forecasts; to conserve space, the results are not reported here, but they are available upon request. 1336 D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 The green lines in Figure A.2 are the cumSFE sequences of our augmented HAR model that adds local forward-looking volatility information to the augmented HAR model as de- fined in Eq.(21)(the legend entry is + FC VIX HAR). 33 Recall that the model in Eq.(21), which adds local forward- looking volatility information to the predictor set, pro- duces consistently better out-of-sample forecasts if the green line in Figure A.2 is consistently above the blue one. As we can see from Figure A.2, this is only the case for the Nikkei 225, Hang Seng, KOSPI and Euro STOXX 50, and very slightly for the DAX. Visually, the improvement seems to be strongest for the Hang Seng series, which is driven largely by a single episode that occurred at around the time of the Lehman Brothers’ collapse in September/October 2008. For the DAX, KOSPI and Euro STOXX 50, the improvement appears to be rather marginal, while it is noticeable from about the end of 2011 onwards for the Nikkei 225. To formally gauge the magnitude of any potential out- of-sample forecast improvements, we compute the gain in out-of-sample R2 from adding local VIX information to the augmented HAR model. That is, we define 1R2 os ( h ) = [ R 2 os computed from Eq.(21)— R2 os computed from Eq.(6) ] , where hdenotes the forecast horizon that is being evaluated, i.e., h= 1,5 ,10 ,22. When 1R2 os ( h ) > 0, there is an increase in R2 os from adding local VIX information to the predictor set in the augmented HAR model. The statistical significance is examined again within aClarkand West(2007) MSFE-adjusted t-test environment, since we are examining the predictive gain from adding local VIX information to the augmented HAR model in Eq.(6); that is, the augmented HAR model of Eq.(6)is nested in the model with local VIX information in Eq.(21). Table A.8 reports the predictive gains in terms of 1R2 os ( h ) for h= 1 ,5 ,10 ,22 in the last four columns. To avoid cluttering the table with extra columns showing the magnitudes of the CW-statistics, we have merely added asterisks next to the 1 R2 os ( h ) entries that yield significant CW-statistics. We use standard asterisk notation to denote significance at the 1\%, 5\%, and 10\% levels, respectively. 34 We can see from the evaluation results that are reported in Table A.8 that the change in the out-of-sample R2 values as a result of including local VIX HAR components in the augmented HAR model of(6)is negative for six of the 17 equity markets at the one-step-ahead horizon that we consider. The importance of this predictor variable deteriorates further with an increasing h, producing only three positive 1R2 os values out of 17 at the 22-day-ahead horizon. Most of the non-negative increases in 1R2 os are rather small in magnitude, with notable exceptions being the improvements recorded for the Hang Seng (3.13\%), the Euro STOXX 50 (1.43\%), the Nikkei 225 and KOSPI (each around 1.4\%) and the DAX (0.9\%), at the one-step-ahead 33 Note that everything in these plots is kept as in previous figures in order to facilitate comparisons. Some of the countries have different beginning dates for the out-of-sample evaluation, due to the lack of available VIX data, so the plots for the Nikkei 225, the Hang Seng, the KOSPI and the IPC Mexico are shifted somewhat, due to the later out-of- sample starting periods. 34 That is, ∗ , ∗∗ and ∗∗∗ denote significance at the 10\%, 5\% and 1\% levels, respectively. horizon. Moreover, these improvements are statistically significant. From the long-horizon evaluations, it is evident that the improvements remain consistently significant, though small at times, up to 22 days ahead for the Euro STOXX 50, and up to 10 steps ahead for the KOSPI index, the DAX, Hang Seng and the IBEX 35, with the later two being only weakly significant at the 10-day-ahead horizon. To summarise our results as to the robustness to local volatility information, we conclude that for most of the equity markets there is only a very slight improvement in out-of-sample performances, with six (for h= 1) or more (for h> 1) markets in fact producing negative 1R2 os values. At the one-step-ahead horizon, only around three to four equity markets improve significantly and achieve sizeable gains, with the improvements in R2 os being 1\%–3\%. However, these gains deteriorate with the forecast horizon, and are unimportant at h= 22. Overall, we conclude that adding local VIX data to the augmented HAR model in Eq.(6)yields small or no forecast gains, with the most notable exception of the Euro STOXX 50, and, at shorter horizons, the Nikkei 225, the DAX and the Hang Seng VIX information. 35 5.3. Controlling for RV information in other European and/or Asian equity markets As a last robustness check, we analyze whether the realized volatility in other European and/or Asian equity markets includes relevant information that could be utilised to improve out-of-sample forecast performance. Rather than selecting a few dominant European or Asian equity markets and then adding their HAR component vec- tors into Eq.(21)one at a time as extra control variables for each of the international equity markets that we consider, we again prefer to extract a common RV factor from the European and Asian realized volatility information using principal components. 36 To formalise this, let Y(EU )= log {[ RV(DAX) RV(CAC 40) RV(AEX) RV(Swiss Market Index) RV(IBEX 35) RV(Euro STOXX 50) RV(FTSE MIB) ]}denote 35 The Euro STOXX 50 and the DAX seem to have the most liquid and mature VIX indices. At this point, it is not clear why the other equity markets’ VIX indices are not informative for long horizon forecasts at least, even if short horizons are affected the most by spillover effects. 36 Specifying, say, 11 European and Asian markets to be used one at a time as controls for the potentially relevant RV information contained in these other equity markets would seem feasible here. However, this raises the possibility of too many statistical tests being carried out, an issue which is known as the ‘ multiple comparisons’ problem in the statistics literature. As a solution, one could use a Bonferroni type of correction when evaluating the out-of-sample performance; that is, adjust the significance level of the test based on the number of additional tests that are constructed. However, it is not clear to us whether this is justified when implementing the MSFE-adjusted t-test on the nested model comparisons ofClarkandWest(2007). Moreover, it would be cumbersome to report the evaluation results in an informative way, as one would have 11 prediction evaluations for each international equity market and forecast horizon. In order to stay within the same testing environment and simplify the presentation of the results, we extract a European and an Asian RV factor, rather than including additional RV predictors one at a time as extra controls. As was done with the VIX series in Section5.2, we again extract the factors from the full sample period only once, rather than recursively as new information becomes available. D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 1337 the (T × 8) vector of log-transformed RV data for all Eu- ropean equity indices that are available to us. The Euro- pean RV HAR factor is then defined as the (1 × 3) vector f RV t ,EU = [ fRV t ,EU ( d ) fRV t ,EU (w) fRV t ,EU ( m ) ] , where fRV t ,EU is the first principal component of Y(EU ), with the daily, weekly and monthly HAR components computed as be- fore. The Asian RV factor (denoted by fRV t ,ASIA hence- forth) is computed as the first principal component from Y (ASIA )= log {[RV(Nikkei 225) RV(Hang Seng) RV(KOSPI) RV(FT Straits Times) ]}.37 We assess the importance of RV information in other European and/or Asian equity markets by taking the aug- mented HAR model that adds local VIX data as predictors, defined in Eq.(21), and further adding fRV t ,EU and fRV t ,ASIA to it as regressors. That is, we form the predictive regression model: y t+ 1 = local volatility info  x tβ +xVIX t β VIX + xUS t β US    US volatility info    augmented HAR as inEq.(6) + forward-looking local volatility info    x VIX t ,FC βVIX FC + fRV t ,EU βf EU + fRV t ,ASIA βf ASIA    other RV info + ϵUS t + 1, (22) where fRV t ,EU and fRV t ,EU are the (1 × 3)-dimensional Eu- ropean and Asian RV HAR factor vectors defined above, and βf EU and βf ASIA are corresponding (3 × 1) parameter vectors that capture their influence on the international RV. Improvements in out-of-sample forecasts relative to our augmented HAR model are examined in the same set- ting as in Section5.2; that is, informally by the magni- tude of the 1R2 os ( h ), statistically using theClarkandWest (2007) MSFE-adjusted t-test, and visually from plots of the cumSFE sequence. These evaluation results are reported in Figure A.3 and Table A.9. Figure A.3 shows the incremen- tal improvement in the cumSFE from including other RV information, in the form of a European RV factor HAR and an Asian RV factor HAR, in addition to the predictor set of the augmented HAR with local VIX information defined in Eq.(21). The blue line in Figure A.3 again shows the cumSFE of the augmented HAR as a reference point, as was done before. The red line shows the cumSFE when only the Eu- ropean RV factor HAR vector fRV t ,EU is included in Eq.(21) in addition to local VIX information (legend entry +FC 37 For the European RV data, the first three principal components explain 89.1745\%, 4.6748\%, and 1.8267\%, respectively, of the variation in Y(EU ). Thus, using only the first principal component seems to be justified, as it explains nearly 90\% of the variation in the data. For the Asian RV data, these values are 68.9685\%, 15.4803\%, and 11.2654\%, respectively. These results are less clear as to whether one factor is enough to capture all of the important movements in Y(ASIA ). We address the issue that more than one factor may be driving Y(ASIA )by also performing forecast evaluations using a HAR structure on the first two factors, together with equally weighted and R2 weighted linear combinations. The latter was performed in order to keep the number of additional regressors added small, so as to minimise overfitting and the ensuing poor out-of-sample performance. However, the forecasts in all of these assessments were always worse than those based on the first PC only. VIX HAR +f(EU) HAR). The light green line in Figure A.3 shows the improvement when both the European RV factor HAR vector fRV t ,EU and the Asian RV factor HAR vector fRV t ,ASIA are added to Eq.(21)as predictors (legend entry +f(ASIA) HAR). From the reported results, we can summarise the effect of adding other RV information into the predictor set in the form of factors as follows. First, examining the time series evolution of the cumSFE sequence visually, one can see that there is no improvement, or at the very best only a very mild improvement, in the out-of-sample forecasts as a result of adding other RV information that is not already realised from the addition of the local VIX series assessed earlier. In fact, the results for some equity markets worsen (see for instance the cumSFE series for the Bovespa, the IPC Mexico, the S&P CNX Nifty, and the All Ordinaries index). Second, conditioning on the Asian RV factor HAR generally produces marginally worse out-of-sample forecasts. 38 This can be seen most clearly from the FT Straits Times and the IPC Mexico, and also somewhat more mildly from the DAX, the Euro STOXX 50, the Swiss Market Index and the KOSPI series. For these equity markets, the loss in precision from adding irrelevant predictors worsens the out-of-sample forecast performance. Third, it is evident from the multiple-horizon statistical comparison in Table A.9 that any gains in R2 os over the benchmark augmented HAR model and their significance levels are very similar to those obtained by only incorporating local VIX information, as was done in Eq.(21), see Table A.8. Moreover, again in line with the results in Section5.2, any forecast gains that are statistically significant at the one-step-ahead horizon disappear fairly quickly as hincreases, with the 22-day- ahead forecast even for the Euro STOXX 50 resulting in a rather small and statistically insignificant improvement. Overall, we conclude this last robustness check with the finding that, once we condition on local VIX information, as described in Section5.2, including additional RV data in the predictor set does not add any further information to improve the out-of-sample forecasts of RV in international equity markets. 6. Conclusion This study extends the work ofRapachet al.(2013) and investigates whether US-based equity market volatility information has predictive value for volatility forecasts in a large cross-section of international equity markets. We assess the role of the US by augmenting the benchmark HAR model ofCorsi(2009) with daily, weekly and monthly US RV and log VIX HAR components, and evaluating the in-sample and out-of-sample contributions of this information to realized volatility in international equity markets. We find the US to play a strong role as a 38 We have also include current time, i.e., t+ 1, factors for Asia when forming forecasts for the European and North and South American indices, but the difference between using lagged or current time information is immaterial, producing the same statistical conclusions in terms of significance levels. 1338 D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 source of relevant volatility information, being particularly important for the Australian and all of the European equity markets that we consider, with a sizeable part of this relevant volatility information coming from forward- looking (or implied) volatility. Using a large out-of-sample forecast evaluation period, we find that the volatility forecasts for all 17 equity markets improve substantially and are highly statistically significant when US volatility information is included in the predictor set. The daily out-of-sample R2 values range between 4.56\% (Hang Seng) and 14.48\% (All Ordinaries), and are above 10\% for nine of the 17 equity markets that we analyse. Moreover, our results show that the Australian and all of the European equity markets benefit the most from the inclusion of US-based volatility information, while the South and North American equity markets and some of the Asian markets benefit the least. An assessment of the forecast performance over time shows that this improvement in predictive performance is consistent over the entire out-of-sample period, and is not driven solely by a few individual events. Moreover, we show that the improvements remain significant for forecast horizons of up to 22 days ahead for 15 of the 17 equity markets, still yielding sizable out-of-sample R2 values of around 5\%–6\% for four of the 17 equity markets that we include. One interesting finding from our in-sample analysis is that the low frequency US volatility component has a negative effect. That is, the parameter estimates on the weekly log VIX HAR component are negative and highly significant for all 17 equity markets. The values that we obtain range from −1.03 to −0.37, with the majority being in the range −0.9 to −0.8. The monthly US RV HAR component is significantly negative for 12 of the 17 equity markets, with values largely ranging from −0.20 to −0.10. So far, there does not seem to have been any discussion in the literature as to why this negative effect occurs, and what economic forces lie behind it, particularly with regard to the weekly log VIX HAR component. In summary, our analysis confirms that the US plays a leading role as a source of equity market information. This role is important not only for international equity return forecasts, as documented byRapachet al.(2013), but also for forecasts of the volatilityin international equity markets. Acknowledgments We are grateful to Francis Diebold, Adrian Pagan, Francesco Ravazzolo, Valentyn Panchenko, Dave Rapach, Paul Söderlind, Angelo Ranaldo, Francesco Audrino, Matthias Fengler, Lorenzo Camponovo, Davide La Vecchia, Jeroen Rombouts, Kamil Yilmaz, Giampiero Gallo, Victor Todorov and Jonathan Wright, as well as seminar partic- ipants at the University of St. Gallen, the University of Pennsylvania, and the 9th International Conference on Computational and Financial Econometrics (CFE 2015) in London for helpful discussions and comments on earlier drafts of the paper. Katja Gisler gratefully acknowledges financial support from the Swiss National Science Founda- tion through grants 144033 and 161796. Appendix A. Supplementary data Supplementary material related to this article can be found online athttp://dx.doi.org/10.1016/j.ijforecast.2016. 05.001. References Andersen, Torben G., Bollerslev, Tim, & Diebold, Francis X.(2007). Roughing it up: Including jump components in the measurement, modeling and forecasting of return volatility. Review of Economics and Statistics ,89 (4), 701–720. Andersen, Torben G., Bollerslev, Tim, Diebold, Francis X., & Labys, Paul (2003). Modeling and forecasting realized volatility. Econometrica, 71 (2), 579–625. Andrews, Donald W. K., & Monahan, Christopher J.(1992). An improved heteroskedasticity and autocorrelation consistent covariance matrix estimator. Econometrica ,60 (4), 953–966. Baillie, Richard T.(1996). Long memory processes and fractional integration in econometrics. Journal of Financial Econometrics ,73 (1), 5–59. Baillie, Richard T., Bollerslev, Tim, & Mikkelsen, Hans O.(1996). Fraction- ally integrated generalized autoregressive conditional heteroskedas- ticity. Journal of Econometrics ,74 , 3–30. Baur, D., & Jung, R. C.(2006). Return and volatility linkages between the US and the German stock market. Journal of International Money and Finance ,25 , 598–613. Becker, K. K., Finnerty, J. E., & Friedman, J. E.(1995). Economic news and equity market linkages between the US and UK. Journal of Banking & Finance ,19 , 1191–1210. Bollerslev, Tim, Litvinova, Julia, & Tauchen, George(2006). Leverage and volatility feedback effect in high-frequency data. Journal of Financial Econometrics ,4 (3), 353–384. Bonato, Matteo, Caporin, Massimiliano, & Ranaldo, Angelo(2013). Risk spillovers in international equity portfolios. Journal of Empirical Finance ,24 , 121–137. Brand, Claus, Buncic, Daniel, & Turunen, Jarkko(2010). The impact of ECB monetary policy decisions and communication on the yield curve. Journal of the European Economic Association ,8 (6), 1266–1298. Brunnermeier, Markus K., Nagel, Stefan, & Pedersen, Lasse H.(2009). Carry trades and currency crashes. In Daron Acemoglu, Kenneth Rogoff, & Michael Woodford (Eds.), NBER macroeconomics annual 2008, Vol. 23 (pp. 313–347). University of Chicago Press. Buncic, Daniel, & Melecky, Martin(2013). Macroprudential stress testing of credit risk: A practical approach for policy makers. Journal of Financial Stability ,9 (3), 347–370. Buncic, Daniel, & Melecky, Martin(2014). Equilibrium credit: The reference point for macroprudential supervisors. Journal of Banking & Finance ,41 (April), 135–154. Buncic, Daniel, & Moretto, Carlo(2015). Forecasting copper prices with dynamic averaging and selection models. North American Journal of Economics and Finance ,33 (1), 1–38. Buncic, Daniel, & Piras, Gion Donat(2016). Heterogenous agents, the financial crisis and exchange rate predictability. Journal of International Money and Finance ,60 (February), 313–359. Campbell, John Y., & Thompson, Samuel B.(2008). Predicting excess stock returns out of sample: Can anything beat the historical average? Review of Financial Studies ,21 (4), 1509–1531. Chen, Xilong, & Ghysels, Eric(2011). News - good or bad - and its impact on volatility predictions over multiple horizons. Review of Financial Studies ,24 (1), 46–81. Chen, Ying, Härdle, Wolfgang K., & Pigorsch, Uta(2010). Localized realized volatility modelling. Journal of the American Statistical Association , 105 (492), 1376–1393. Chevillon, Guillaume(2007). Direct multi-step estimation and forecast- ing. Journal of Economic Surveys ,21 (4), 746–785. Chevillon, Guillaume, & Hendry, David F.(2005). Non-parametric direct multi-step estimation for forecasting economic processes. International Journal of Forecasting ,21 (2), 201–218. Chow, Victor, Jiang, Wanjun, & Li, Jingrui (2014). Does VIX truly measure return volatility? SSRN Working Paper No. 2489345. Available from: http://ssrn.com/abstract=2489345. Clark, Todd E., & West, Kenneth D.(2007). Approximately normal tests for equal predictive accuracy in nested models. Journal of Econometrics, 138 (1), 291–311. Clements, Michael P., & Hendry, David F.(1996). Multi-step estimation for forecasting. Oxford Bulletin of Economics and Statistics ,58 (4), 657–684. D. Buncic, K.I.M. Gisler / International Journal of Forecasting 32 (2016) 1317–1339 1339 Comte, Fabienne, & Renault, Eric(1996). Long-memory continuous time models. Journal of Econometrics ,73 (1), 101–149. Comte, Fabienne, & Renault, Eric(1998). Long memory in continuous- time stochastic volatility models. Mathematical Finance,8(4), 291–323. Corsi, Fulvio(2009). A simple approximate long-memory model of realized volatility. Journal of Financial Econometrics ,7 (2), 174–196. Corsi, Fulvio, Audrino, Francesco, & Renó, Roberto(2012). HAR modeling for realized volatility forecasting. In Luc Bauwens, Christian Hafner, & Sebastien Laurent (Eds.), Handbook of volatility models and their applications (pp. 363–382). John Wiley and Sons, Inc.. Corsi, Fulvio, Pirino, Davide, & Renó, Roberto(2010). Threshold bipower variation and the impact of jumps on volatility forecasting. Journal of Econometrics ,159 (2), 276–288. Corsi, Fulvio, & Renó, Roberto(2012). Discrete-time volatility forecasting with persistent leverage effect and the link with continuous-time volatility modeling. Journal of Business and Economic Statistics ,30 (3), 368–380. Diebold, Francis X.(2015). Comparing predictive accuracy, twenty years later: A personal perspective on the use and abuse of Diebold–Mariano tests. Journal of Business and Economic Statistics , 33 (1), 1–9. Diebold, Francis X., & Mariano, Roberto S.(1995). Comparing predictive accuracy. Journal of Business and Economic Statistics ,13 (1), 253–263. Diebold, Francis X., & Yilmaz, Kamil(2012). Better to give than to receive: Predictive directional measurement of volatility spillovers. International Journal of Forecasting ,28 , 57–66. Diebold, Francis X., & Yilmaz, Kamil(2014). On the network topology of variance decompositions: Measuring the connectedness of financial firms. Journal of Econometrics ,182 (1), 119–134. Diebold, Francis X., & Yilmaz, Kamil(2016). Trans-atlantic volatility connectedness among financial institutions. Journal of Financial Econometrics ,14 (1), 81–127. Dimpfl, Thomas, & Jung, Robert C.(2012). Financial market spillovers around the globe. Applied Financial Economics ,22 , 45–57. Engle, Robert F.(1990). Stock volatility and the crash of ’87: Discussion. The Review of Financial Studies ,3 (1), 103–106. Eun, C. S., & Shim, S.(1989). International transmission of stock market movements. The Journal of Financial and Quantitative Analysis ,24 (2), 241–256. Fengler, Matthias R., & Gisler, Katja I. M.(2015). A variance spillover anal- ysis without covariances: What do we miss? Journal of International Money and Finance ,51 , 174–195. Fengler, Matthias R., Mammen, Enno, & Vogt, Michael(2015). Specifica- tion and structural break tests for additive models with applications to realized variance data. Journal of Econometrics,188 (1), 196–218. Fernandes, Marcelo, Medeiros, Marcelo C., & Scharth, Marcel(2014). Modeling and predicting the CBOE market volatility index. Journal of Banking & Finance ,40 , 1–10. Goyal, Amit, & Welch, Ivo(2008). A comprehensive look at the empirical performance of equity premium prediction. Review of Financial Studies ,21 (4), 1455–1508. Grassi, Stefano, Nonejad, Nima, & de Magistritis, Paolo Santucci (2014). Forecasting with the standardized self-perturbed Kalman filter. CREATES Research Paper 2014-12. Aahrhus University. Hamao, Yasushi, Masulis, Ronald, & Ng, Victor(1990). Correlations in price changes and volatility across international stock markets. The Review of Financial Studies ,3 , 281–307. Heber, Gerd, Lunde, Asger, Shephard, Neil, & Sheppard, Kevin 2009. Oxford-man institute’s realized library. Oxford-Man Institute. Univer- sity of Oxford. Available from:http://realized.oxford-man.ox.ac.uk/. Jacod, Jean, & Shiryaev, Albert N.(1987). Limit theorems for stochastic processes . Springer. King, Mervyn A., & Wadhwani, Sushil(1990). Transmission of volatility between stock markets. The Review of Financial Studies ,3 , 5–33. Knaus, Simon D.(2014). Essays in computational statistics with applications to volatility forecasting and forecast combination . (Ph.D. thesis), University of St. Gallen. Lin, Wen-Ling, Engle, Robert F., & Ito, Takatoshi(1994). Do bulls and bears move across borders? International transmission of stock returns and volatility. The Review of Financial Studies ,7 (3), 507–538. Liu, Lily Y., Patton, Andrew J., & Sheppard, Kevin(2015). Does anything beat 5-min RV? A comparison of realized measures across multiple asset classes. Journal of Econometrics ,187 (1), 293–311. Marcellino, Massimiliano, Stock, James H., & Watson, Mark W.(2006). A comparison of direct and iterated multistep AR methods for forecasting macroeconomic time series. Journal of Econometrics, 135 (1–2), 499–526. McAleer, Michael, & Medeiros, Marcelo C.(2008). Realized volatility: a review. Econometric Reviews ,27 (1), 10–45. Müller, Ulrich A., Dacorogna, Michel M., Davé, Rakhal D., Olsen, Richard B., Pictet, Olivier V., & Ward, John R. (1993). Fractals and intrinsic time - a challenge to econometricians. In Invited presentation at the 39th international AEA conference on real time econometrics, 14–15 October 1993, Luxembourg . Neely, Christopher J., Rapach, David E., Tu, Jun, & Zhou, Guofu(2014). Forecasting the equity risk premium: The role of technical indicators. Management Science ,60 (7), 1772–1791. Newey, Whitney K., & West, Kenneth D.(1994). Automatic lag selection in covariance matrix estimation. Review of Economic Studies,61 (4), 631–653. Nikkinen, Jussi, Mohammed, Omran, Petri, Sahlström, & Äjiö, Janne (2006). Global stock market reactions to schedulded US macroeco- nomic news announcements. Global Finance Journal,17 , 92–104. Patton, Andrew J., & Sheppard, Kevin(2015). Good volatility, bad volatility: Signed jumps and the persistence of volatility. Review of Economics and Statistics ,97 (3), 683–697. Pesaran, M. Hashem, Pick, Andreas, & Timmermann, Allan(2011). Vari- able selection, estimation and inference for multi-period forecasting problems. Journal of Econometrics ,164 (1), 173–187. Prokopczuk, Marcel, Symeonidis, Lazaros, & Wese Simen, Chardin(2016). Do jumps matter for volatility forecasting? Evidence from energy markets. Journal of Futures Markets , (in press). Rapach, David E., Strauss, Jack K., & Zhou, Guofu(2013). International stock return predictability: What is the role of the United States? The Journal of Finance ,68 (4), 1633–1662. Savva, C. S., Osborn, D. R., & Gill, L.(2009). Spillovers and correlations between US and major European stock markets: the role of the Euro. Applied Financial Economics ,19 , 1595–1604. Daniel Buncic is Assistant Professor of Quantitative Economics at the Institute of Mathematics and Statistics in the School of Economics and Political Sciences at the University of St. Gallen. He holds a Ph.D. in Economics from the University of New South Wales in Sydney, Australia. He has published in various journals, including the Journal of the European Economic Association , theJournal of Banking and Finance and the Journal of Financial Stability . He is currently interested in forecasting of financial assets, particularly, commodities, exchange rates and the equity premium using various forecast averaging and aggregation methods. He has previously worked on issues related to stress testing and financial stability issues in general, as well as macro-econometric modeling and its use in policy analysis. Katja I.M. Gisler holds a Masters in Quantitative Economics and Finance from the University of St. Gallen and is currently a Ph.D. student in the School of Economics and Political Sciences at the University of St. Gallen. Her research interests are in areas of realized volatility modeling, volatility spillovers and financial econometrics in general. She has recently published a paper on volatility spillovers in the Journal of International Money and Finance . An annotated bibliography of at least five additional resources related to the research topic by Day 7 of Week 2.***No plagiarism*** It must be less than 30\% match on authenticity report.***PLEASE DO Knowledge and innovation in emerging market multinationals: The expansion paradox☆ Richard Lynch, Zhongqi Jin ⁎ Middlesex University Business School, The Burroughs, London NW4 4BT, UK abstract article info Article history: Received 1 February 2015 Received in revised form 1 August 2015 Accepted 1 September 2015 Available online 23 October 2015 Keywords: Global strategy Technical innovation Non-technical innovation Knowledge acquisition Emerging market multinationals Chinese car industry This article examines the innovation and knowledge strategies that allow emerging-market companies to be- come major international players. By adopting a qualitative approach, this study identiﬁes a signiﬁcant paradox between the desire of some leading Chinese car companies to expand internationally and the current relationship of such companies with leading global car companies, which signiﬁcantly inhibits Chinese international expan- sion. This study unpacks that paradox using innovation theory and the resource-based view, and develops a ma- trix of strategic options that can assist emerging market multinational companies to expand internationally. © 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 1. Introduction The internationalization strategies of companies from developing countries attracts increasing attention from researchers (Gammeltoft, Filatotchev, & Hobdari, 2012; Tsai, 2014). Most of the research, however, focuses on two aspects of global expansion for companies from such countries: The motivation and the internationalization processes (Cuervo-Cazurra, 2008). Research seldom addresses how innovation content and processes in potential multinational companies from emerging markets (EM MNEs) enable these companies to compete with their international rivals (usually from developed markets), the DM MNEs.Buckley and Hashai (2014)identify a signiﬁcant gap in re- search regarding the extent to which EM MNEs are able to compete ef- fectively with DM MNEs. To address this research gap, this study explores the effective strategic options for EM MNEs aiming to expand in mature world markets against the backdrop of the positions of the existing DM MNEs, which presents a particular challenge for new entrants (Banerjee, Prabhu, & Chandy, 2015). This research explores the role of technical and non-technical innovation at EM MNEs as part of this process. Speciﬁcally, this study seeks to develop an approach to solve a signiﬁcant strategic puzzle: Why do some EM MNEs with global ambitions have little success so far? How can thesecompanies solve this problem? This research adopts a qualitative ap- proach focusing on the car industry and the strategies of car companies from China that seek internationalization. This article contributes to the literature by presenting new insights into the contribution and processes of innovation at EM MNEs as they develop their international strategies, especially in the areas of knowl- edge and learning in the context of mature world markets. Second, the article provides a potential pathway to internationalization. Third, by looking at the international process of EM MNEs from both ambidex- trous innovation and resource-based perspectives, this study provides further insights into the processes necessary for the effective interna- tionalization strategies of EM MNEs. 2. Literature review This study draws on innovation theory and the resource-based view. In addition, the study also employs the concept of strategicﬁttoexplore the options available to EM MNEs to reduce the gap with their DM MNE rivals. The importance of research into innovation hardly needs any justiﬁca- tion (Damanpour, Walker, & Avellaneda, 2009). Globalization makes in- novation even more important (Berry, 2014). Two issues regarding the importance of innovation for EM MNEs deserve special attention. First, EM MNEs may not have the innovative capability to exploit disruptive new technologies in comparison to their DM MNE rivals (Fleury, Fleury, & Borini, 2013). Second, a signiﬁcant gap in innovation capability exists between EM MNEs and their established rivals, especially in high-tech Journal of Business Research 69 (2016) 1593–1597 ☆ The authors thank T. C. Melewar, Middlesex University, Arno Haslberger, Webster University Vienna, and Thomas Lawton, Open University for their careful reading and suggestions. ⁎Corresponding author. E-mail addresses:[email protected](R. Lynch),[email protected](Z. Jin). http://dx.doi.org/10.1016/j.jbusres.2015.10.023 0148-2963/© 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Contents lists available atScienceDirect Journal of Business Research industries. DM MNEs have far more experience in and orient more toward innovation (Christensen, Chang-Chieh, Kah-Hin, & Subramanian, 2010). From a resource-based perspective, EM MNEs may also be at a disadvantage in comparison with their developed-country competitors in management,ﬁnance, and technical knowledge (Fagerberg & Godinho, 2006; Kumar, Mudambi, & Gray, 2013). Therefore, to investi- gate process differentiation in terms of the resources, learning, knowl- edge development, and the acquisition of new knowledge between the EM MNEs and the DM MNEs is important (Zhong, Peng, & Liu, 2013).Zhu, Lynch, and Jin (2011)argue that the catch-up process re- quires EM MNEs to develop innovative strategies by focusing on over- coming their weaknesses and improving their competitiveness. This process involves the acquisition of new knowledge and the further development of existing knowledge within theﬁrm. Such knowledge may come from outside the company as well as inside (Leonard- Barton, 1999). This research therefore proposes that: P1.In a mature industry, innovation is an essential strategy component for EM MNEs that wish to close the gap with their global rivals through the strategic processes of learning, knowledge development, and the ac- quisition of new knowledge. This research focuses on two large categories of innovation: techni- cal (technological) innovation and non-technical innovation (see Damanpour, Szabat, & Evan, 1989for deﬁnitions). In a mature industry, EM MNEs are initially at a competitive disad- vantage with their established rivals.Li (2010)argues that MNE late- comers may employ an accelerated trajectory of cross-border learning in which co-exploitation and co-exploration take place during the inter- nationalization process. These changes come from non-technical inno- vations such as cross-border alliances and joint ventures, rather than from technical innovations (Li, 2010) and from organizational transfor- mation (Dixon, Meyer, & Day, 2010). However, the changes apply if, and only if, the implementation process runs smoothly and the EM MNE is capable of learning (Zhu et al., 2011). Hence, EM MNEs maintain clear resource-based beneﬁts in non-technical innovations if these compa- nies are capable of implementing such a learning process. P2.In a mature industry, EM MNEs are more likely to close the gap with their established rivals through the development and exploitation of non-technical innovative resources and capabilities rather than through the development and exploitation of technical innovations. This process involves learning, new knowledge, and investment in new resources that relate to innovation. Turning to technical innovation, the acquisition of aﬁrm from a de- veloped country by an EM MNE to produce new technical innovations may take longer and even yield negative output (Ahuja & Katila, 2001). In addition, in a mature industry, the established players rather than new EM MNEs are most likely to possess technological knowledge such as pools of patents. Due to the nature of technology diffusion in a mature industry, EM MNEs can build their technical innovative re- sources and capabilities by learning from their established rivals through approaches such as licensing, OEM, and joint ventures (Choung & Hwang, 2000). However, the implementation of these ap- proaches takes time and may require the building of intangible assets drawing on the technology transfer of intellectual capital, which is par- ticularly difﬁcult for the aspiring EM MNE (Tsai, 2014; Zhu et al., 2011). P3.In a mature industry, EM MNEs are more likely to close the gap with their established rivals through incremental technology advances rather than a major technology breakthrough. 3. Research method This research adopts a qualitative approach (Rouse & Daellenbach, 2002 ) following a replication logic with multi-source data collection,asEisenhardt (1989)andWoodside and Baxter (2013)suggest. The study focuses on the global passenger car industry rather than the broader markets for automotive parts, buses, trucks, and other commer- cial vehicles. The target population of this research is the passenger car companies operating in China over the last 30 years. The unit of analysis includes both the DM MNEs that have set up joint ventures with Chinese companies and the Chinese EM MNE car companies that either expand internationally or plan to undertake this route over time. The study em- ploys both secondary data and primary data. For the secondary data stage, the data analysis method is content analysis in three steps. First, the study uses global car industry statistics to identify the leading world companies and Chinese car companies. Second, this study ana- lyzes the innovation input and output of these companies in terms of patent applications, R&D investment, and the development of new products and services (Muller, Välikangas, & Merlyn, 2005; Tseng & Wu, 2007). Third, the study analyzes the strategic intent regarding in- ternational and global expansion of the major Chinese companies from their published statements, independent stockbrokers, and con- sultants reports. For the primary data collection stage, this study adopts a two-stage process using in-depth interviews with a predetermined protocol. The protocol design process draws onYin (1994)and follows the OECD Oslo Manual on the collection and interpretation of innovation data (OECD, 2005)andFagerberg and Godinho (2006)on the innovation ac- tivities of MNEs. The selection of interviewees employs the following criteria: extensive work experience in the Chinese car industry, employ- ment either directly for a major Chinese car company or for a supplier, and sufﬁcient managerial seniority to provide informed judgments. Four Chinese middle managers participated in theﬁrst stages in- depth interviews in 2012 via personal contact. The second stage inter- view followed a year later with a similar protocol: one group interview andﬁve in-depth interviews.Table 1lists the proﬁle of the study sample and major points the interviews covered. 4. Research results 4.1. Strategic context: EM MNEs vs. DM MNEs on technical innovation Table 2lists the leading world companies in the car industry in order of turnover. The list also offers their expenditure on research and devel- opment in the years 2007–2010, and data from the leading Chinese companies. In general, the R&D expenditure among international lead- ing car companies (all of which are DM MNEs) remains remarkably steady over the four years that this study observes. Table 2shows that ten of the DM MNE car companies together spent over US$50.2 billion in the year 2007 alone on research and develop- ment. The average expenditure in relation to turnover is 4.5\%. Following the downturn in the world car industry, this ratio decreased to 4.2\% in 2009. Few, if any, of the Chinese car EM MNEs invest at such levels. The high level of technical R&D innovation at the EM MNEs results from two external causes: High fuel prices and governmental pressure toﬁnd alternative fuels because of environmental issues. 4.2. Strategic context: The development of the Chinese car industry since 1978 In 1978, the Chinese Government began its“open door”policy re- garding the entry of foreign companies into the Chinese domestic market. The four largest Chinese domestic car companies, FAW, Dongfeng, Guangzhou, and SAIC, then set up joint ventures with various European, Japanese, and U.S car companies in which the local Chinese companies were the majority shareholders. The result was that the for- eign joint ventures now dominate the Chinese market (seeTable 3). The prime focus of the four leading Chinese joint ventures is on the rapidly expanding Chinese domestic market. These leading companies have not attempted, so far, to expand internationally. However, some 1594R. Lynch, Z. Jin / Journal of Business Research 69 (2016) 1593–1597 smaller Chinese car companies engage in international activities: Seeking overseas expansion through exports, seeking overseas pro- duction, and seeking co-operation with companies in other develop- ing countries. Table 3also shows that much of the branding investment in the Chi- nese domestic market is in the name of the DM MNE car companies rather than in name of the local Chinese EM MNE companies. Thissituation makes it difﬁcult for the development of a distinctive national brand that could expand internationally. 4.3. Strategic options of EM MNEs: the evidence from primary data analysis All the executives agreed in the interviews that Chinese car compa- nies aim to move beyond the domestic market over time. Manager H, for Table 1 Qualitative interviews with middle and senior managers of Chinese car and car supply companies. Manager IDCompany Role of manager Comment on knowledge and learning Field work 2012 A Major Chinese joint venture car company with foreign MNESenior production manager In-depth knowledge of one company B Major Chinese joint venture company with foreign MNEMiddle rank marketing manager In-depth knowledge of one company C Tier 1 foreign MNE supplier to several Chinese car and Chinese car joint venture companiesSenior sales manager working across a region of China Able to give a wider perspective across Chinese car manufacturing and also some knowledge of MNEs control of its R&D D Tier 1 foreign MNE supplier to several Chinese car and Chinese car joint venture companiesSenior engineering consultant working with a number of Chinese car production companiesAble to give a wider perspective across Chinese car manufacturing and also some knowledge of MNEs control of its R&D Field work 2013 E Major Chinese joint venture car company with foreign MNEFour middle managers from the same company as A and B above in sales, marketing, human resources, productionInterviewed as a group F Major Chinese joint venture car company with foreign MNEResearch and development manager—another joint venture company different from A and B aboveDifferent strategic issues to the A and B company in 2012 on model development and j.v. relationships G Major Chinese joint venture car company with foreign MNESenior sales manager—another joint venture company different from A, B, and F aboveDifferent strategic issues to the A and B company in 2012 on model development and j.v. relationships H Major Chinese car company—not part of a joint ventureSenior production manager Company keen to develop global sales strategy I Tier 2 foreign MNE supplier to several Chinese car and Chinese car joint venture companiesMiddle rank sales manager Able to give a wider perspective across Chinese car manufacturing and also some knowledge of MNEs control of its R&D J Tier 2 foreign MNE supplier to several Chinese car and Chinese car joint venture companiesSenior director with responsibility across a region of ChinaAble to give a wider perspective across Chinese car manufacturing and also some knowledge of MNEs control of its R&D Table 2 R&D expenditure of selected car companies (ranked by decreasing sales in US$). Company Country Total turnover 2007 (US$ Bn)Total annual R&D expenditure 2007 (US$ Bn)Annual R&D expenditure as \% turnover 2007 2008 2009 2010 Toyota Cars Japan 187.4 7.6 4.1\% 3.5\% 4.0\% 3.9\% General Motors USA 181.1 8.1 4.4\% 5.4\% 2.9\% 5.1\% Volkswagen Germany 160.1 7.8 4.8\% 2.6\% 2.9\% 2.4\% Ford USA 154.4 7.5 4.8\% 5.6\% 4.5\% 3.9\% Daimler Mercedes Germany 133.3 4.6 3.4\% 4.5\% 5.1\% 4.9\% Nissan Japan 92.5 n/a n/a 5.4\% 5.1\% 4.9\% Fiat Italy 86.0 2.3 2.7\% n/a 3.0\% 2.8\% BMW Cars Germany 79.1 4.6 5.8\% 2.8\% 2.6\% 3.1\% Honda Motors Japan 76.0 4.1 5.4\% 4.4\% 5.8\% 5.6\% PSA France 73.6 3.0 4.0\% 4.4\% 4.7\% 4.2\% Renault France 55.5 3.6 6.5\% 4.9\% 5.5\% 4.9\% Mazda Japan 36.6 n/a n/a 3.3\% 3.8\% 3.9\% Hyundai South Korea 32.3 n/a n/a n/a n/a n/a Suzuki Japan 21.2 n/a n/a n/a n/a n/a Selected Chinese companies FAW China 24.6 0.38 1.5\% 1.5\%E 1.0\% 1.1\% SAIC China 20.7 0.41 2.0\% 2.0\%E 1.0\% 1.6\% Guangzhou Auto China 14.7 n/a n/a n/a n/a n/a ChangAn Auto China 7.9 0.05 0.6\% 1.0\%E 1.0\% 1.0\%E Dongfeng Auto China 7.8 0.01 1.3\% 1.3\%E 2.3\% 1\% Geely China 5.6 n/a n/a n/a n/a n/a JAC China 4.7 n/a n/a n/a n/a n/a Chery China 2.1 n/a n/a n/a 1.0\% n/a Note: By 2010, the total sales of both SAIC and Dongfeng Auto were greater than FAW Auto. Source: authors compilation from company accounts. The FAW R&D data comes from Xinhuanet (29 April 2009). R&D expenditure includes development for commercial vehicles as well as cars in some companies. However, the majority of the funds go to cars, according to the text commentary in the company annual reports. See comment on deﬁnition of R&D expenditure in text. E = estimated.1595 R. Lynch, Z. Jin / Journal of Business Research 69 (2016) 1593–1597 example, says that their company is“world-oriented and moving ag- gressively to becoming internationalized.”Manager A explains that “after China has entered into the WTO [World Trade Organization], local automobile companies have opportunities to compete with the global automobile giants in global markets.” Theﬁrst proposition argues that in a mature industry, innovation is an essential component strategy for EM MNEs that wish to close the gap with their global rivals with learning being an essential element. Manager A explains that little opportunity for innovation exists: [The Chinese car company] performs well on manufacture, but not so well on innovation. In fact, it only holds some techniques with less importance. Almost all the core techs are held in [the Chinese companys DM MNE joint venture] partners. However, the Chinese companies learn from their DM MNE joint venture partners through training, new equipment from the DM com- pany, and from the secondment of personnel to the DM companys de- veloped markets. Manager C says that he“will be leaving in September for at least three months at our head ofﬁce in a [European city].” The second proposition suggests that in a mature industry, EM MNEs are more likely to close the gap with their established rivals through the development and exploitation of non-technical innovative resources and capabilities. The evidence from the interviews and from the second- ary research shows that Chinese car companies somewhat succeed in non-technical innovations in the last few years. The Chinese managers explain that DM MNE partners perform the main technical activities and that foreign partners control those activi- ties very tightly. “We simply make to speciﬁcation. We do not know any details or have any sharing of technology”(Manager F). One Chinese manager complains that, for two years, the DM MNE partner did not share its new models with the Chinese car company. “[The DM MNE] gave the new models to a rival [Chinese] joint ven- ture company. As a consequence, our sales have shown little growth over the period”(Manager G). Despite this problem, the Chinese joint venture company receives training for its managers from the DM MNE. Furthermore, DM MNEs provide aﬂow of non-technical innovation in the form of help with building car service centers, supplier networks, and other support activities. Proposition 3 argues that in a mature industry, EM MNEs are more likely to close the gap with their established rivals through incremental technology advances rather than a major technology breakthrough. The evidence provides support for this proposition. The Chinese joint ventures (including some patents according to the websites of the companies) conduct some R&D activities. However, theinterviewees only mention these activities occasionally. The inter- viewees explain that the EM MNE joint venture uses the R&D center to make incremental technical adjustments to models from the parent DM MNE company“to satisfy the habits and tastes of Chinese cus- tomers”(Manager A). The company does not use the center to develop basic designs for the cars. Part of the reason for this was that R&D centers need plenty of experience and their own powerful tech- nical strength based on R&D talent with such innovation taking a long period to become mature…[This] lack of an R&D talent pool and being too dependent on the [parent companys] product plat- form was a signiﬁcant barrier for the long-term growth strategy of the Chinese subsidiary. (Manager A). 5. Discussion and conclusions This study provides support for all three propositions. The difﬁculty for Chinese car EM MNEs to move internationally relates to the levels of knowledge and innovation that the Chinese car companies currently possess. A constant theme, particularly during the qualitative inter- views, is an awareness of the reliance of the EM MNE on the knowledge and innovation databases of their DM MNEs partners. The knowledge and innovationﬂ ow in the Chinese car industry up to 2013 comes large- ly from the global DM MNEs to their Chinese joint venture partners, with the DM MNEs being reluctant to lose control of this strategic re- source. Hence, despite the strategic intent to expand internationally, the Chinese car companies are effectively locked in to the knowledge and innovation strategies that their joint venture DM MNE partners pro- vide. The Chinese car companies are aware of the problem but do not appear to have the solution. Therein rests the paradox—not only for Chinese car companies but also for all EM MNEs. This study calls this sit- uation the“locked-in paradox”of EM MNEs. This research suggests that the answer to this paradox lies in two main areas: the innovation and the knowledge base of the EM MNEs. EM MNEs are unlikely to develop radical innovations. However, many Chinese EM car companies are developing various forms of incremental technology innovation, partly by recruiting new R&D staff members and partly through contacts with other local research institutions, including university engineering departments. Chinese EM MNE car companies should be searching much more widely across national borders to ac- quire technology to beat the paradox. The research shows that a key in- gredient in responding to the considerable resources of the DM MNEs lies in the area of the knowledge base of the EM MNE—not just about Radical market reconfiguration strategy Acquisition, joint venture, and/or alliance strategy Search and acquire technology from external sources National technology search and R&D recruitment strategy Non-technical innovation Technical innovation Resources inside the EM MNE Resources external to the EM MNE Fig. 1.Effective international strategy options for EM MNEs in mature world markets. Table 3 China market shares of leading car brands (2011) Chinese Domestic Market Brand Local joint venture partner(s) China market share \% Volkswagen First Auto Works (Beijing), SAIC 16 Toyota FAW, Guangzhou Auto 12 GM SAIC, Liuzhou Wuling 12 Honda Guangzhou Auto, Dongfeng Auto 9 Nissan Dongfeng Auto 7 Chery 5 Hyundai Beijing Auto 5 BYD 4 Geely 4 Note: Source for Chinese share data: authors compilation based on DataMonitor,Market Watch Automotive(December 2011). Source for local Chinese partners: Company reports. 1596R. Lynch, Z. Jin / Journal of Business Research 69 (2016) 1593–1597 the knowledge of technology but about markets, opportunities, and the implementation process. The EM MNEs need to distinguish be- tween what they already possess internally and what they can ac- quire externally. Combining both technical and non-technical innovation with the ac- tual and potential competitive resources of the EM MNE, this study de- velops a matrix (seeFig. 1) to show the effective international strategy expansion options for EM MNEs. By illustrating the role of technical and non-technical innovation to- gether with resource-based knowledge in the process of catching-up for EM MNEs in a mature industry, this research contributes to the litera- ture from the following perspectives: First, the study illustrates the challenge that EM MNEs face when they try to catch up with their established rivals in a mature industry in which latecomers do not possess the technological knowledge, which constitutes the locked-in paradox. Theﬁndings are consistent with existing literature regarding the difﬁculty of such catch-up process (e.g.Nolan et al., 2008), regarding contingency factors of catching up (Gammeltoft et al., 2012), and regarding institutional barriers of innova- tion from emerging economies (McCarthy, Puffer, Graham, & Satinsky, 2014). Second, the expansion matrix this study proposes provides a poten- tial pathway for those EM MNEs aiming to make inroads into the inter- national market. The expansion matrix involves a comprehensive roadmap that combines technical and non-technical innovations to- gether with learning- and knowledge-acquisition processes. The expan- sion matrix complementsMathews (2006)andContractors (2007) iterative learning models for EM MNEs. Innovation and knowledge resource development are not the only routes for EM MNEs to expand internationally. By looking at the interna- tional process of EM MNEs from both ambidextrous-innovation and resource-based perspectives, this study provides further insights into which processes are necessary for the effective internationalization of EM MNEs. Thus, the study provides an extension forZhu et al.’s (2011)catch-up framework andZhong et al.’s(2013)ﬁndings regard- ing knowledge integration of EM MNEs. This research focuses on only one industry and one country. Future research should replicate this study with other mature industries such as the food industry and EM MNEs in other countries and continents in order to validate theﬁndings and conclusions of this study. References Ahuja, G., & Katila, R. (2001).Technological acquisitions and the innovation performance of acquiringﬁrms: a longitudinal study.Strategic Management Journal,22(3), 197–220. Banerjee, S., Prabhu, J.C., & Chandy, R.K. (2015).Indirect learning: how emerging-market ﬁrms grow in developed markets.Journal of Marketing,79(1), 10–28. Berry, H. (2014).Global integration and innovation: multicountry knowledge generation within MNCs.Strategic Management Journal,35(6), 869–890. Buckley, P.J., & Hashai, N. (2014).The role of technological catch up and domestic market growth in the genesis of emerging country based multinationals.Research Policy, 43(2), 423–437.Choung, J., & Hwang, H. (2000).Transition of latecomerﬁrms from technology users to technology generators: Korean semiconductorﬁrms.World Development,28(5), 969–983. Christensen, C., Chang-Chieh, H., Kah-Hin, C., & Subramanian, A. (2010).Managing innova- tion in emerging economies. (editorial)IEEE Transactions on Engineering Management, 57(1), 4–8. Contractor, F. (2007).Is international business good for companies? The evolutionary or multi-stage theory of internationalization vs. the transaction cost perspective. Management International Review,47(3), 453–475. Cuervo-Cazurra, A. (2008).The multinationalization of developing country MNEs: the case of multilatinas.Journal of International Management,14(2), 138–154. Damanpour, F., Szabat, K., & Evan, W. (1989).The relationship between types of innova- tion and organizational performance.Journal of Management Studies,26(6), 587–601. Damanpour, F., Walker, R., & Avellaneda, C. (2009).Combinative effects of innovation types and organizational performance: a longitudinal study of service organizations. Journal of Management Studies,46(4), 650–675. Dixon, S., Meyer, K., & Day, M. (2010).Stages of organizational transformation in transi- tion economies: a dynamic capabilities approach.Journal of Management Studies , 47(3), 416–436. Eisenhardt, K.M. (1989).Building theories from case study research.Academy of Management Review,14(4), 532–550. Fagerberg, J., & Godinho, M.M. (2006).Innovation and catching-up. In J. Fagerbberg, D.C. Mowery, & R.R. Nelson (Eds.),The Oxford handbook of innovation(pp. 514–542). Ox- ford: Oxford University Press. Fleury, A., Fleury, M.L., & Borini, F.M. (2013).The Brazilian multinationals approaches to innovation.Journal of International Management,19(3), 260–275. Gammeltoft, P., Filatotchev, I., & Hobdari, B. (2012).Emerging multinational companies and strategicﬁt: A contingency framework and future research agenda.European Management Journal,30(3), 175–188. Kumar, V., Mudambi, R., & Gray, S. (2013).Internationalization, innovation and institu- tions: The 3 Is underpinning the competitiveness of emerging marketﬁrms.Journal of International Management,19(3), 203–206. Leonard-Barton, D. (1999).Wellsprings of knowledge.Boston, MA: Harvard Business School Press. Li, P. (2010).Toward a learning-based view of internationalization: The accelerated trajecto- ries of cross-border learning for latecomers.Journal of International Management,16(1), 43–59. Mathews, J. (2006).Dragon multinationals: new players in 21 stcentury globalization.Asia Paciﬁc Journal of Management,23,5–27. McCarthy, D.J., Puffer, S.M., Graham, L.R., & Satinsky, D.M. (2014).Emerging innovation in emerging economies: can institutional reforms help Russia break through its histor- ical barriers?Thunderbird International Business Review,56(3), 243–260. Muller, A., Välikangas, L., & Merlyn, P. (2005).Metrics for innovation: Guidelines for de- veloping a customized suite innovation metrics.Strategy & Leadership,33(1), 37–45. Nolan, P., Jin, Z., & Liu, C. (2008).The global business revolution, the cascade effect, and the challenge forﬁrms from developing countries.Cambridge Journal of Economics, 32(1), 29–47. OECD (2005).Oslo manual: Guidelines for collecting and interpreting innovation data(3rd Edition ). Luxembourg: OECD, Statistical Ofﬁce of the European Communities. Rouse, M.J., & Daellenbach, U.S. (2002).More thinking on research methods for the resource-based perspective.Strategic Management Journal,23(10), 963–967. Tsai, H. (2014).Moderators on international diversiﬁcation of advanced emerging market ﬁrms.Journal of Business Research,67(6), 1243–1248. Tseng, C., & Wu, L. (2007).Innovation quality in the automobile industry: Measurement in- dicators and performance implications.International Journal of Technology Management, 37(1/2), 162–177. Woodside, A.G., & Baxter, R. (2013).Achieving accuracy, generalization-to-contexts, and complexity in theories of business-to-business decision processes.Industrial Marketing Management,42(3), 382–393. Yin, R.K. (1994).Case study research: Design and methods.Thousand Oaks, CA: Sage Publications. Zhong, W., Peng, J., & Liu, C. (2013).Internationalization performance of Chinese multina- tional companies in the developed markets.Journal of Business Research,66(12), 2479–2484. Zhu, Y., Lynch, R., & Jin, Z. (2011). Playing the game of catching-up: Global strategy build- ing in a Chinese company.Asia PaciﬁcBusinessReview,17(4), 511–533.1597 R. Lynch, Z. Jin / Journal of Business Research 69 (2016) 1593–1597 An annotated bibliography of at least five additional resources related to the research topic by Day 7 of Week 2.***No plagiarism*** It must be less than 30\% match on authenticity report.***PLEASE DO Early globalizations: The integration of Asia in the world economy, 1800–1938 David Chilosi a,⁎, Giovanni Federico b aEconomic History Department, London School of Economics, Houghton Street, London WC2A 2AE, UK bDepartment of Economics, University of Pisa, Via Ridolfi 10, 56124 Pisa, Italy Received 14 April 2014 Available online 23 April 2015 Abstract This paper contributes to the debate on globalization and the great divergence with a comprehensive analysis of the integration of Asia in the world market from 1800 to the eve of World War II. We examine the patterns of convergence in prices for a wide range of commodities between Europe and the main Asian countries (India, Indonesia, Japan and China) and we compare them with convergence between Europe and the East Coast of the United States, hitherto the yardstick for the 19th century. Most price convergence occurred before 1870, mainly as a consequence of the abolition of the European trading monopolies with Asia, and, to a lesser extent, the repeal of duties on Atlantic trade. After 1870, price differentials continued to decline thanks to falling freights and to better communication after the lay-out of telegraph cables. There was only little disintegration in the inter-war years. © 2015 Elsevier Inc. All rights reserved. JEL classification:F14; F15; F63; O40; N74; N75 Keywords:Globalization; Market integration; International trade; Economic growth; Asia; Nineteenth century 1. Introduction Standard economic theory holds that trade and market integration foster economic growth. Indeed, the era of the so-called first globalization, before World War I, coincid- ed with a period of unprecedented economic growth in Europe and in its Western Offshoots (Maddison Project, 2013). Yet, at the same time, the Asian countries (with the partial exception of Japan) fell increasingly behind the advanced European ones (Broadberry, 2013), in spite of rapidly growing exports (Federico and Tena, 2013). Some scholars have tackled the paradox posed by this“greatdivergence”(Pomeranz, 2000) by pointing out that exports of primary products did benefit the Asian economies, but their effect was too small to foster economy-wide growth (Feuerwerker, 1980, 1983; Booth, 1988; Tomlinson, 1993; van der Eng, 1996; Roy, 2000; Brandt et al., 2013). Others blame the colonial powers for forcing the Asian economies to export primary products, thus damaging their growth potential (Dutt, 1969; Parthasarathi, 2011). ForWilliamson (2008, 2011, 2012, 2013), too, specialization in primary products damaged the long-term prospects for industrial- ization in the periphery.In his view, however, this specialization wasthe unintended consequence of market ⁎ Corresponding author. E-mail address:[email protected](D. Chilosi). http://dx.doi.org/10.1016/j.eeh.2015.04.001 0014-4983/© 2015 Elsevier Inc. All rights reserved. Available online at www.sciencedirect.com ScienceDirect Explorations in Economic History 57 (2015) 1–18 www.elsevier.com/locate/eeh integration, which improved the terms of trade before the 1870s. In the same vein,Allen (2011)argues that peripheral countries could have escaped this“curse of primary products”(Sachs and Warner, 2001)onlyby adopting a coherent industrialization policy, which was conspicuously lacking in all Asian countries but Japan. Testing these competing views about exports and economic growth entails a huge and very challenging research agenda. This paper contributes to this agenda by exploring price convergence, an essential component of integration, between Europe (United Kingdom, the Netherlands, or France) and the four main Asian countries, China, British India, the Dutch East Indies (henceforth Indonesia) and Japan from the beginning of the 19th century to the eve of World War II. In so doing, we fill in two key gaps in the literature on the integration of Asia in the world economy: we analyze the period 1800–1870 and quantify the impact of the abolition of Western trading monopolies. Previous work has shown that price gaps were high before 1800 (ORourke and Williamson, 2002; Rönnbäck, 2009), narrowed after 1870 (ORourke and Williamson, 2002; Hynes et al., 2012), and widened during the Great Depression (Hynes et al., 2012). However, no empirical research, to date, has dealt with the period 1800– 1870. Yet, these years featured massive processes of integration both within Europe (Federico, 2011, 2012)and in the Atlantic economy (Jacks, 2005; Uebele, 2011; Sharp and Weisdorf, 2013), raising the question of how does Asia compare? The same years also saw the abolition of the monopolies by the Western companies trading with Asia. Their demise must have boosted integration, but we lack measures of the actual size of this effect, relative to those of the decline in duties and advancements in transport and communication technology. We present our dataset inSection 2and discuss the patterns of price convergence inSection 3. The key period of integration across routes was the early, rather than the late nineteenth century, while price differentials remained roughly stable in interwar years.Section 4deals with the main barriers to trade, focusing on the trading companies and on their abolition, whileSection 5 estimates the contribution of different causes (institution- al change, fall in transport costs, trade liberalization and so on) to price convergence with a panel regression. In spite of the similarities in trends across oceans, the processes of integration had roots in different institutions: while in the Atlantic economy the repeal of duties was a key determinant, much of the price convergence between Asia and Europe was due to the demise of the British East India Company (EIC) and, to a lesser extent, the Dutch trading monopoly (Nederlandsche Handel-Maatschappij or NHM).Section 6concludes.2. The data-base The quantitative analysis of integration faces a trade-off between the quality of the data and their representativeness of changes in the overall market. In particular, asFederico (2012)argues, in order to produce reliable results, the price series should meet three conditions: i) The price ratios should refer to pairs of markets which were actually trading. Otherwise, price differentials can be lower than costs and move (quasi-) randomly within the band of commodity points. If markets trade and are efficient à laFama (1970), in equilibrium price gaps must be equal to transaction costs, inclusive of monopoly mark-ups. ii) Each price series should refer to a specific quality rather than to the market average and each pair of series should refer to the same quality. Otherwise, price gaps might reflect quality differentials, and any change in quality in a market might introduce spurious trends. iii) The commodities should be representative of the actual trade flows. Extending inferences from one product only (e.g. cereals), is tantamount to assume that that movements in transport costs, barriers to trade and market efficiency are similar across all traded goods. Unfortunately, we cannot examine integration on the import side because the data on prices of manufactures are very scattered and refer to different qualities. The data for primary products are much more abundant and thus we have been able to collect series of prices for the same commodities in 26 pairs of markets (Table 1). 1 All cities in our sample were major trading centers in their own countries, and trade statistics report bilateral trade of that specific good for about 93\% of the observations. Missing data are mostly scattered, which suggests failure to record rather than absence of trade. There is a small chance that absence of trade could be an issue in only about 2\% of the cases. Quality is homogeneous across markets (Yes in the relevant column) in the overwhelming majority of pairs, 22 out of 29 pairs. In two of the other cases, one series only can be considered as qualitatively homogeneous (Yes in the Column“within market”), but the quality surely differs between series (No in the Column“across markets”), 1For a more detailed discussion of our sources seeChilosi and Federico (2013, Appendix B). We have considered a larger sample, but we have decided to drop some series (e.g. tobacco from Indonesia) which did not meet the minimum quality threshold. 2D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 while in the remaining ones the quality differs between markets and changes in time in each market (No in all three columns). In short, in the worst case, homogeneity might be a problem at most for one fourth of cases. Last but not least, these products accounted for about half or more of exports from India and Indonesia nearly the whole period, from United States before the Civil War and from China until 1900 (Table 2). The coverage is poor only for Japan. Summing up, the data-base, though falling short of being perfect, can be considered as fit for our purposes.3. The process of integration: A statistical analysis A visual inspection of some representative series of price ratios (Fig. 1) reveals that, on the one hand, as expected, the price ratios were fairly high at the beginning of the 19th century and fell in the long run, with a spike during World War I. On the other hand, the differentials differedhugelyamongcommodities at the same time (as seen in the vertical axis) and they often suddenly collapsed rather than steadily declined. Table 1 The data-base. Period Market Homogeneous quality? Within markets Across Origin Destination Origin Destination Markets Atlantic Cotton 1801–1938 New York Liverpool Yes Yes Yes Wheat 1800–1937 New York London No No No Far East Silk 1834–1877 Canton London Yes Yes Yes Silk 1874–1913 China London Yes Yes Yes Silk 1874–1914 China Lyon Yes Yes Yes Silk 1894–1914 Yokohama Lyon No No No Silk 1894–1938 Yokohama New York No No No Tea 1811–1831 Canton England No No Yes a Tea 1820–1877 Canton London Yes Yes Yes India Cotton 1796–1845 Calcutta London Yes Yes Yes Cotton 1867–1938 Bombay London Yes Yes Yes Indigo 1822–1931 Calcutta London Yes Yes Yes Jute 1844–1938 Calcutta London Yes Yes Yes Linseed 1846–1938 Calcutta London Yes Yes Yes Rapeseed 1871–1921 Calcutta London Yes Yes Yes Rice 1870–1938 Rangoon London Yes Yes Yes Saltpetre 1796–1853 Calcutta London Yes Yes Yes Silk 1796–1856 Calcutta London Yes Yes Yes Silk 1857–1877 Calcutta Lyon Yes Yes Yes Sugar 1796–1856 Calcutta London Yes Yes Yes Tea 1893–1931 Calcutta London Yes Yes Yes Wheat 1861–1931 Calcutta London Yes No No Indonesia Coffee 1833–1913 Batavia Rotterdam No Yes No Pepper 1828–1938 Batavia Amsterdam No No No Rice 1848–1913 Batavia Amsterdam Yes Yes Yes Rubber 1913–1938 Batavia London Yes Yes Yes Sugar 1822–1938 Batavia London No No No Tea 1893–1938 Batavia Amsterdam Yes Yes Yes Tin 1863–1913 Batavia Amsterdam Yes No No Sources: seeChilosi and Federico (2013: Appendix B). aSince all theseﬁgures refer to tea imported and sold at auctions by the East India Company, we expect the quality mix to be similar in the same year, but not necessarily across years.3 D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 More precise information on convergence can be obtained by estimating the equation (Razzaque et al., 2007): ΔLn RP i¼αþβTIMEþψln RP i t 1 þφlnΔLn RP i t 1 þuð1Þ whereRP iis the relative price of theith good between two markets andTIMEis a linear trend. We compute the long-run rate of change ast= (β/ψ). The error correction model coefficientψ(ranging between 1 and 0) tests whether and estimates how rapidly price ratios return to this trend after a shock, while the lagged change in relative prices is added to address possible serial correlation. We first run Eq.(1)as a fixed effects panel for the whole sample and separately by country, considering jointly the (comparatively few) observa- tions for the two Far Eastern countries, China and Japan. The results (Table 3) confirm the conventional wisdom: at the beginning of the series (which differed by product and route), European prices were on average double the Asian ones and by the end this difference had been cut by a half. All rates are statistically significant and very similar across routes. The half-lives of shocks are comparatively quite high. One might sum up that the market was becoming increasingly integrated, but overall it had still a long way before becoming really efficient. However, such a conclusion might be a tad hasty: it assumes stationary efficiency and linear price conver- gence with equal rates by product. We explore difference in timing of integration by running separate panel regressions for five periods: the twilight of mercantilism (1796–1815), the early globalization (1815–1870), the heyday of globalization (1870– 1913) and the war and interwar (1914–1938). We alsorun for these periods a pooled group estimator, which allows rate of change to differ across products and then averages them (Table 4). As posited by the conventional wisdom, the results show no integration before 1815, although they rely on a small and thus potentially non-representative number of observations. The data for the two subsequent periods are undoubtedly representative and they yield a clear conclusion: overall, convergence was twice faster in the“early globalization”than during its (alleged)“heyday”. The difference between the two periods is particularly wide for the Far East, while convergence between Indonesia and Europe was only 45\% faster in 1815–1870 than in 1870–1913. Integra- tion of Indonesia during the“early globalization”is also the only case where the two estimators yield different results. The negative coefficient(s) for the war and interwar period reflects the return to normal levels of price differentials, after the sharp war-time increase. Indeed, dropping the first 5 years, the rates become positive for the total and for all areas but the Far East. Yet the rates of inter-war disintegration remain rela- tively small and are not statistically significant. 2In other words, the inter-war disintegration of the world trading network affected little the Asian and American exports of primary products analyzed here. The increase in the speed of reaction between the second and third periods suggests an increase in market efficiency. However, this is modest and the reaction remains rather slow: one may surmise that in modern markets, most shocks were arbitraged away within the year and that, consequently, our yearly series capture only very large shocks, which needed more time to be absorbed. This conjecture should be tested with higher frequency data (Brunt and Cannon, 2014). One might suspect the inter-temporal comparisons to be biased by composition effects as the coverage by product differs between periods. The absence of some bulky product from the panel before 1870 implies a negative bias in the pace of integration during the “twilight of mercantilism”, while the sample for the “heyday of globalization”includes high-value goods whose price differentials were already very small by 1870 and could not fall further. We address this concern by running fixed-effect panel regression of the natural logarithm of price ratio as a function of 5 year time dummies (Bateman, 2011). 2The overall rate of change (in percentage) is 0.50 (not signiﬁcant), while rates by area are 0.39 for the Far East (signiﬁcant at 10\%), 0.63 for India, 0.53 for Dutch East Indies and 0.51 for the United States (not signiﬁcant). Table 2 Shares of covered products on total exports (in percentage). China India Indonesia Japan United States 1810 26.4 84.0 a 34.8 1830 45.0 b 67.1 50.8 1850 40.5 76.3 46.5 1870 95.4 55.5 73.3 28.8 64.9 1890 58.7 57.1 59.4 24.0 33.5 1900 33.3 46.5 56.4 20.6 22.2 1913 29.0 53.3 43.4 26.3 26.7 1929 20.3 46.0 57.2 29.4 17.1 1938 9.5 42.2 43.1 9.0 10.0 Sources: seeChilosi and Federico (2013: Appendix B)andFederico and Tena (2013). a1823.b1828. 4D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 The results (Fig. 2) confirm that if anything the linear trends understate the difference between 1815–1870 and 1870–1913. The pattern is almost identical also if we omit non-homogenous goods and very similar (although statistically less precise) if we run separateregressions for the four trade routes. On top of this, Fig. 2shows that price convergence concentrated in fairly short periods of time, most notably between 1815 and 1825, in the 1850s, and in the early 1870s and the 1880s. Thus, we look for structural breaks in the series Price ratios: selected commodities 0.5 1.0 1.5 2.0 2.5 3.0 3.5 1796 1821 1846 1871 1896 1921 Cotton New York Liverpool 1 2 3 4 5 1796 1821 1846 1871 1896 1921 Rice Rangoon London 1.0 1.5 2.0 2.5 3.0 3.5 4.0 1796 1821 1846 1871 1896 1921 Pepper Batavia Am s terdam 0 2 4 6 8 10 1796 1821 1846 1871 1896 1921 Saltpetre Calcutta London 0.5 1.0 1.5 2.0 2.5 3.0 1796 1821 1846 1871 1896 1921 Sugar Batavia London 0.4 0.8 1.2 1.6 2.0 2.4 1796 1821 1846 1871 1896 1921 Te a C a n to n L o n d o n Notes:the price ratios are computed with the yearly means of import and export prices. Sources: see Chilosi and Federico (2013: Appendix B). Fig. 1. Price ratios: selected commodities.5 D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 of price ratios with aBai and Perron (1998, 2003)test. We find at least one break in 23 series out of 29 and two or more breaks in six of them, for a total of 31 breaks. About a third of these breaks feature also a shift in level exceeding a third of the differential before the break. Table 5sums up the results. Column 1 reminds the number of series by period and country (fromTable 1). Columns 2 to 4 report the number of trends, distinguishing between periods of convergence (negative and significantcoefficient of the time variable), divergence (positive and significant trend) and periods without a significant trend in price ratios. The number of trends (sum of Columns 2 to 4) exceeds the number of pair of markets (Column 1) whenever there is more than one trend within a given period. Then we count the number of breaks which entail also an upward (Column 5) or downward (Column 6) jump exceeding 10\% of the last year of the previous trend. We estimate the total change in each period/country (Column 7) and the contribution by shocks (Column 8), by averages of product specific figures. Last but not least, we compute an overall trend, comparable with the linear trends ofTable 4, as an average the product-specific yearly rates of change (Column 9). As a whole, our analysis confirms that convergence was consistently and significantly faster during the “twilight of mercantilism”than during the“heyday of globalization”, although the difference for Indonesia is very small. The analysis adds three important results. First, the timing of the breaks only weakly supports the traditional periodization: slightly less than half of the breaks (14 out of 31) fall within an interval of 6 years around 1815, 1870 or 1913 and almost as many (10) are Table 3 Long-run convergence: panel estimation. N Initial ratioHalf-life (months)Rate (in percentage)Cumulated change (in percentage) All 1725 1.940 24 0.443*** 46.72 Atlantic 257 1.618 20 0.401*** 42.46 Far East 191 2.024 18 0.520*** 52.22 India 805 2.085 26 0.501*** 50.89 Indonesia 472 1.746 18 0.402*** 37.25 Significant at *10\%, **5\%, ***10\%. Notes: N = sample size. Fixed-effects estimation of Eq.(1)is used in all cases. Sources: seeChilosi and Federico (2013: Appendix B). Table 4 Trends by period: panel regression. N Initial ratio Half-life (months) Rate (in percentage) Cumulated change (in percentage)Pooled group estimator Twilight of mercantilism (1796–1815) All 102 2.274 10 0.968 20.20 1.920 Atlantic 27 1.215 9 3.249 62.81 4.855 India 72 3.099 9 0.285 17.00 0.195 Early globalization (1815–1870) All 565 1.973 16 0.896*** 38.92 0.759** Atlantic 109 1.632 10 0.678*** 31.12 0.668*** Far East 48 2.245 12 1.501** 55.55 0.436 India 263 2.125 17 1.243*** 49.51 1.506*** Indonesia 145 1.892 12 0.504** 21.48 0.235 Heyday of globalization (1870–1913) All 765 1.291 8 0.418*** 16.44 0.379*** Atlantic 88 1.129 12 0.288*** 11.65 0.230 Far East 114 1.266 12 0.334** 13.37 0.235* India 330 1.326 7 0.517*** 19.95 0.461* Indonesia 233 1.311 8 0.347*** 13.85 0.359*** War and interwar (1914–1938) All 313 1.441 10 1.128*** 23.71 1.116*** Atlantic 37 1.084 7 0.160 3.78 0.260 Far East 27 1.157 0 0.250* 5.82 India 150 1.559 10 1.478** 28.86 1.425*** Indonesia 99 1.483 9 1.181** 24.69 1.045** Significant at * 10\%; ** 5\%; *** 10\%. Notes: N = sample size. N = sample size. Fixed-effects estimation of Eq.(1)is used in all cases. Sources: seeChilosi and Federico (2013: Appendix B). 6D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 farther than 10 years from these dates. Second, breaks mattered a lot. In the“twilight of mercantilism”phase, on average, they accounted for one third of changes, as compared to a quarter during the“heyday”and a fifth during the last period. Third, the timing of the breaks often coincided with important institutional changes, detailed in the next section. Almost all the major shifts, 10 out of 12, clustered around 1815 or 1913—i.e. at the end of the EIC monopoly in Indian trade (1813) or at the outbreak of World War I. One exception is the tea exported from Canton to London, whose price ratio saw a sudden fall in 1835, 2 years after the demise of the EIC monopoly on that route; the other one is the sugar exported from India, whose London price spiked up in 1839, after the ban on the import of West Indian sugar cultivated by slaves. Moreover, the three negative shifts around 1870 are all from Indonesia, where, as will be detailed in the next section, the monopoly on foreign trade was abolished from 1868, considerably later than in India and China. 3By contrast, the Great Depression was definitely not a shock in the market for primary products.4. The causes of convergence: Transport costs and barriers to trade The conventional wisdom attributes the fall in the transaction costs of world trade before World War I mostly to the combined effects of liberalization of trade and declines in transport costs; improvements in communication and the gold standard are also routinely stressed as contributory factors. Similarly, the disinte- gration after the war is seen as the product of the protectionist backlash. We measure transport costs with the freight factor (the ratio of freights to price at the origin, assuming other costs to have been proportional to freights), and explore barriers to trade by looking at the nominal duty on each product—i.e. the sum of import duties in consuming countries and export taxes from India and Indonesia divided by the price at the origin.Fig. 3plots the transport costs and the duties for four representative commodities. Levels are not directly comparable because the scales on the vertical axis differ, but trends are by and large similar across countries and products. As expect- ed, duties and freights did fall in the first part of the 19th century, while, consistent with the results ofTable 5, the rise in protection during the Great Depression is very limited. From 1930 to 1938, total duties were on average equivalent to 26.8\% of the price in the exporting country, exceeding 50\% in 22 observations out of 95 (most of them on sugar). A fortiori, the impact Long-run convergence: five year averages - 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1796 1821 1846 1871 1896 1921 Notes: The dependent variable is the natural logarithm of the ratio between import and export prices; the only independent variables are five years dummies. Fixed-effects estimation is used. Sources: see Chilosi and Federico (2013: Appendix B). Fig. 2. Long-run convergence: 5 year averages. 3The remaining shift near 1870 is positive and signals a slowing down in the pace of integration of linseed between Calcutta and London. In fact, only in one case out of four does a structural break signal that the pace of integration increased after the 1870s (tin from Batavia).7 D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 of protection was limited over the whole period 1796– 1938: total duties exceeded 5\% of prices in slightly more than a quarter of cases and 50\% only in about an eighth. On the same line, one should note that most of the fall in freights was over by 1875 and that the change was often quite sudden. The 75\% collapse in the freight factor for saltpetre in 1816–1817 is an extreme case, but there are several other instances of major jumps, including a 70\% fall in freights for tea from China to England in 1835. Such sudden changes cannot be accounted for by innovations in shipping organization or technology, the two competing explanations for the fall in costs of sea transport (North, 1968; Harley, 1988; Shah and Williamson, 2004), which by their nature are likely to spread gradually. Thus the evidence points to a major role for the dismantling of the institutional barriers to Asian trade.In Japan and China, these barriers were erected also by national governments to be dismantled by the force of Western powers: Japan was forced to open to trade in 1854, while the restrictions to Western merchants in China to trade in Canton only were lifted after the two opium wars in 1842 and 1858 (Dermigny, 1964; Fairbank, 1978; Wakeman, 1978). The British East India Company (EIC) enjoyed a monopoly on exports of Chinese tea to the United Kingdom and of imports of Indian opium into China until 1833, when it ceased all its mercantile activities. The monopoly of trade between its Indian possessions and the United Kingdom had been already abolished in July 1813. The Company had traditionally paid very high freights, because it used large and heavily manned and armed ships (Indiamen), very expensive to build and operate, which it rented from the owners. These Table 5 The four phases of globalization. Number of trends Number of breaks Total change in the periodContribution of shocksImplicit rate of change Pairs of marketsConvergence Divergence Trendless Upward shockDownward shock(in percentage) (in percentage) (in percentage) Twilight of mercantilism (1796–1815) Atlantic 2 2 1.10 0.00 0.10 India 4 4 3.50 0.00 0.31 All 6 6 2.70 0.00 0.24 Early globalization (1815–1870) Atlantic 2 1 2 2 27.60 45.50 0.45 Far East 3 1 1 2 1 28.50 25.80 0.79 India 8 4 5 1 5 46.10 41.80 1.90 Indonesia 5 3 1 3 2 10.60 16.20 0.47 All 18 9 2 12 1 10 32.50 33.39 1.20 Heyday globalization (1870–1913) Atlantic 2 2 8.40 21.40 0.21 Far East 4 3 1 1 1 2.80 5.70 0.06 India 8 6 1 2 1 2 19.30 22.70 0.54 Indonesia 6 3 3 2 14.30 41.30 0.40 All 20 12 2 8 1 5 13.40 24.76 0.37 War and interwar (1914–1938) Atlantic 2 2 1.80 0.00 0.08 Far East 1 1 2.90 0.00 0.12 India 8 4 2 2 3 1.80 17.50 0.28 Indonesia 4 1 1 2 2 1 4.90 43.50 0.16 All 15 6 3 6 5 1 1.84 20.91 0.17 Notes: The dependent variable in each regression is the natural logarithm of the ratio between import and export prices; the only independent variable is the time trend. TheBai and Perron (1998, 2003)tests detect structural breaks in the constant and the slope. The total change and the implicit rate of change for each product in each period (Columns 7 and 9) are based on the formulaDelta = (rp T rp 0)/rp 0, whereDeltais the total change and the rpsdenote the fitted values at the end and beginning of the period. As at times trends and shifts offset one another, to examine the extent to which cumulated changes are explained by them, we decompose the sum of the absolute values of the predicted changes (Chilosi, 2014), using the following formula:Delta_abs = |(rp n1 rp n0)/rp n0| + |(rp n2 rp n1)*rp n1/(rp n1*rp n0)| + |(rp n3 rp n2)*rp n2/(rp n2*rp n0)| +…whererp nichanges whenever there is a shift and again with the new trend. In practice, there are at most two trends in each period (i.e.i= 1, 2 or 3). The figures reported in Column 8 are based on the proportions ofDelta_absexplained by the shifts. Sources:Chilosi and Federico (2013: Table A1). 8D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 Barriers to trade for selected commodities a) Transport costs (freight factors) .00 .04 .08 .12 .16 .20 1800 1820 1840 1860 1880 1900 1920 .00 .01 .02 .03 .04 .05 1800 1820 1840 1860 1880 1900 1920 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 1800 1820 1840 1860 1880 1900 1920 .00 .05 .10 .15 .20 .25 .30 .35 1800 1820 1840 1860 1880 1900 1920 Cotton New York Liver poolIndigo Calcutta London Saltpetre Calcutta LondonSugar Batavia London b) Import and export duties Notes: the freight factors are equal to the nominal freights divided by the price at the origin. Similarly, the duties are equal to the sum of import and export duties divided by the price at the origin. Sources: see Chilosi and Federico (2013: Appendix B). .00 .02 .04 .06 .08 .10 .12 .14 .16 1800 1820 1840 1860 1880 1900 1920 .00 .02 .04 .06 .08 .10 1800 1820 1840 1860 1880 1900 1920 0.0 0.2 0.4 0.6 0.8 1.0 1800 1820 1840 1860 1880 1900 1920 0 1 2 3 4 5 1800 1820 1840 1860 1880 1900 1920 Cotton New York LondonIndigo Calcutta London Saltpetre Calcutta London Sugar Batavia London Fig. 3. Barriers to trade for selected commodities.9 D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 were a small group of Londoners, many of whom were also shareholders of the Company. Although the owners were barred from participation in the Court of Directors since 1710, they infiltrated it with the ships husbands and captains, thus forming a cohesive oligopoly within the monopoly. According to the conventional wisdom, the “shipping interest”, as it was known at the time, allowed the inflating of costs and thus the circumvention of the Companys official charter, which prevented it from earning extra-profits from trade (Wakeman, 1978; Mui and Mui, 1984:63–64;Bowen, 2006: 252–256;Webster, 2009:32–40). 4Solar (2013)has recently suggested that the Indiamen were necessary to stave off the military threat of competing Western trading companies. In his view freights collapsed in the 1810s thanks to the combined effect of technical progress (mostly copper sheathing) and to the successes of the Royal Navy, which made it possible to use smaller and cheaper ships. Anyway, since 1813 trade with India was free and in all likelihood the market was competitive. The expectation of great profits from liberalized trade caused a glut in the market for transportation for India (Webster, 2009: 72) and indeed our series of freight collapses in 1817. The Dutch trade with its colonies in the Indian Ocean had been a monopoly of theVereenigde Oost-Indische Compagniefrom the early 17th century to its bankruptcy in the 1790s. Trade remained free until 1825, when a monopoly on commerce with the Netherlands was granted to a new trading company, theNederlandsche Handel- Maatschappij, or NHM (Furnivall, 1976; Horlings, 1995: 142). Similarly to the EIC, the company rented ships from Dutch owners, at“exceptionally high”rates (Korthals Altes, 1994: 161;Horlings, 1995: 145). In its first years of activity, exports were reduced by a rebellion of natives, the so called Java War. At the end of the war, the Dutch government, desperate for revenues, established a system of compulsory delivery of coffee, sugar and indigo for exports, known as Cultivation System (de Klerck, 1938; Dobbin, 1983; Fasseur, 1992; Houben, 2002; van Zanden and Marks, 2012). Peasants were paid much less than the world market price—about a half of the Batavia price for coffee (Fasseur, 1992: 37). The goods were transported to Amsterdam, and there sold at auction: the profits, net of a fee for the NHM, accrued to the Dutch government. The amount was very substantial: at its peak, in the 1850s, itaccounted for over half the state revenues and for about 3.8\% of Dutch GDP excluding any hidden subsidy to Dutchshippingandindustry(Smits et al., 2000; van Zanden and van Riel, 2004). However, the Cultivation System was increasingly unpopular at home and was slowly phased out. In 1850, the NHM liberalized the bidding contracts for renting ships and its monopoly was abolished altogether in 1868 (NHM, 1924:23;Furnivall, 1976:168;Korthals Altes, 1994). The Dutch trade with the colonies remained free until 1918, when the sugar producers set up a private association (VJSP) to allocate the scarce available shipping (van der Eng, 1996:215– 216;Knight, 2010). The organization continued to manage sugarexportsaftertheendofthewar,andin1932itwas substituted by a governmental organization (NIVAS) to manage sugar production quotas under the international agreement (van der Eng, 1996:224–226). The Great Depression featured also the first intervention in American agriculture: the Agricultural Adjustment Act (AAA), part of the New Deal policies, established a loan facility for cotton farmers, which in practice set a minimum price of cotton since 1934 (Federico and Sharp, 2013). 5. The causes of integration: An econometric analysis As implied by the preceding discussion, the ratio of prices for theith good (RP it) at timetcan be explained by the barriers to trade (B), the efficiency of markets (E m) and the transport costs (T c) Log RP it¼FB;E m;Tc ð2Þ In our regression the setBincludes the total duties (LOG_DUTY) and dummies for monopolies—adummy for theEIC(1796 to 1816) and separate dummies for the NHMunder the full monopoly regime (NHM1), from 1824 to 1850, and for the partially liberalized one (NHM2) from 1851 to 1868. 5We also add dummies for the AAA support to American cotton prices (since 1933) and for the two marketing boards for Javanese sugar, the private VJSP (1918–1931) and the public NIVAS (after 1932). We expect duties and monopolies to increase price ratios, 4The EIC charter stipulated that the price of company wares in London“should not exceed the prime cost, the freight and charges of importation, the lawful interest of capital from the time of arrival of such tea in Britain, and the common premium in insurance (Mui and Mui, 1984: ix). Indeed, the ofﬁcial proﬁts of the Company were quite low (Wakeman, 1978: 167). 5The variableLOG_DUTYis computed aslog(1 + t), wheretis the ratio of (usually speciﬁc) duties to the price in the producing country. It is thus zero ift=0. The dummyEICis equal to 1 until 1816. We do not add a speciﬁc dummy for the end of the Napoleonic Wars because it would be collinear with theEICdummy and it might bias the estimate of its coefﬁcient. Anyway, price gaps with Asia in Britain were much less affected than in continental Europe by the Napoleonic Wars (O’Rourke, 2006). 10D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 while the interventions on the commodity market may augment or reduce price gaps according to the details of the policies. Arguably the most important contribution to increas- ing market efficiency was the connection by telegraph,which cut the time to transmit information from weeks to few minutes (Hoag, 2006). The cable between the United Kingdom and the United States was operational since 1866, and it was followed 4 years later by the line to India and around 1875 by the line between Europe Freight factors and adjusted freight factors: selected commodities i) Saltpetre Calcutta London 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 1796 1801 1806 1811 1816 1821 1826 1831 1836 1841 Freight factorAdjusted freight factor ii) Sugar Batavia London .00 .04 .08 .12 .16 .20 .24 1823 1833 1843 1853 1863 1873 1883 Freight factorAdjusted freight factor Notes: The freight factor is the nominal freight rate divided by the price at the origin. The adjusted freight series are constructed by firstly running the regression LOG_FREIGHT it=c +α i+β1EIC it+β2NHM1 it+β3NHM2 it+εitand secondly using the equation LOG_ADJFREIGHT it=LOG_FREIGHT it-(β1EIC it+β2NHM1 it+β3NHM2 it). Using alternative specifications, like including a time trend, yielded qualitatively identical results and only small quantitative differences. Sources: see the text and Chilosi and Federico (2013: Appendix B). Fig. 4. Freight factors and adjusted freight factors: selected commodities. i) Saltpetre Calcutta London. ii) Sugar Batavia London.11 D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 and Indonesia (Headrick, 1988: 101). We expect the TELEGRAPHdummy to be negative. We also add dummies for a number of political events which may have disrupted the orderly working of markets, such as the Java War in 1825–1827 (JAVA WAR), the Indian Mutiny in 1857–1859 (MUTINY), the American Civil War (CIVIL WAR), the anti-slavery campaign, which boycotted Caribbean sugar, in 1840–1845 (SLAVE) and World War I (WWI). By definition, the actual freights, the numerator of our measure of transport costs (LOG_FREIGHT), include any rent paid to privileged Dutch or British ship-owners under the NHM and EIC monopoly system. Consequently the EICandNHMdummies measure only any additional effect of the trading monopoly—e.g. from a manipulation of the market. We try to capture the total effect of monopoly, including the rents to ship-owners by using an alternative series of freight factor (LOG_ADJFREIGHT) net of the rents to shippers. We obtain this series (Fig. 4) by scaling down the actual series with coefficients from a fixed-effect panel regression, which explains freights with dummies for the NHM and the EIC. By construction, the series move in parallel toLOG_FREIGHT,exceptinthe final year of the monopoly and thus the substitution affects the coefficient of transport costs only marginally. We add the lagged value of the dependent variable to reduce auto-correlation and to take into account the possible delay in adjustment to shocks, but its omission does not affect qualitatively the results. The coefficients are thus short term elasticities; long run elasticities can be computed asβ k/(1 γ) whereβ kis the coefficient of thekth variable andγis the coefficient of the lagged dependent variable. The descriptive statistics for all variables and the pairwise coefficient of correlation between them do not add much new information (Chilosi and Federico, 2013). The correlation between explicative variables is in general very low and thus there is no risk of multi-collinearity. On the other hand, most variables are clearly non-stationary (cf.Chilosi and Federico, 2013: Table A2, for a formal testing) and thus results might be spurious. We therefore test ex-post the stationarity of the residuals with aLevin et al. (2002)test for panel regressions. Our analysis omits from the regression the Far Eastern markets, because the series are very short and their quality is comparatively poor. This leaves a total of 22 cross- sections and 1568 observation. As said before, we cannot categorically exclude that quality may be an issue for wheat (both from the United States and from India), coffee sugar, pepper and tin (all from Indonesia). Thus, in Column 3 ofTable 6we drop these products, reducing the coverage to 16 products and 1020 observations. Last butnot least, Indian trade, with 838 observations, is overrepresented in the sample. We address this issue in two different ways. First, we run the regression with a restricted panel, featuring two products per route (wheat and cotton for the United States, sugar and indigo for India, and pepper and sugar for Indonesia), and we weight each of them with their share of the value of imports on that route. Thus, in this specification (Table 6, Column 6), each route has the same weight on the results. Second, we simply run separate regressions for the United States (Table 6, Column 7), India (Column 8) and Indonesia (Column 9), which supply additional insights on the different causes of integration. All regressions use fixed-effects specification with panel corrected standard errors to address cross-sectional heteroskedasticity (reflecting different levels of transac- tion costs) and contemporaneous correlation (which may arise from common shocks). Although we report results of an OLS estimate (Table 6, Column 1) for comparative purposes, we prefer to use instrumental variables because prices in the exporting country appear in the denominator of the dependent variable and of two explicative ones (duties and freights). Furthermore, freights might be endogenous in the short-run as well, if the supply of shipping on those routes were inelastic (Jacks and Pendakur, 2010). Specifically, we instrumentLOG_DUTY with the ratio of yearly duties to the average price throughout the whole period—so that within variations depend only on the exogenous changes in specific duties. Likewise we use average prices as denominator of the instrument forLOG_FREIGHT,while the numerator is the trend component of a Hodrick–Prescott decomposition of the series of nominal freights. The overall performance of the model is good. The residuals are stationary; the combined variables are highly significant (F-test) and explain about four fifths of the total variance. Almost all the signs agree with expectations. In all cases, theWooldridges (1995)and the regression-based tests find strong evidence of that the OLS estimates sufficiently differ from the instru- mental variables ones to recommend the use of the latter. All first stage tests find that the instruments are highly correlated with the variables, so that the expected small-sample bias is low. Turning to the results: i) As expected, the coefficients for freights are positive, highly significant and robust to the substitution of monopoly-adjusted series (LOG_ ADJFREIGHT). A 10\% increase in freights aug- mented price gaps by about 0.6–0.9\% in the short run and by about 1–1.5\% in the long run. Moreover, as expected, the impact of costs is much bigger for 12D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 bulky goods than for than for light products; over three times, in fact (Column 5). 6 ii) The variableLOG_DUTYis positive and significant in the baseline specification and its coefficient is similar to the coefficient of freights. Indeed, the averages of the two variables are almost identical: 18.1\% for the freight factor and 18.2\% (17\% on imports and 1.2\% on exports) for duties. Duties were high only on wheat, under the British Corn Laws before 1846, and sugar. Infact, the coefficient of LOG_DUTYis high and significant for the Atlantic trade (Column 7), low and significant only at the 13\% level for Indonesia (Column 9) and very low and not significant for India (Column 8). iii) As expected, the monopoly of the British East India Company widened the price differentials, by a quarter in the baseline specification withLOG_ FREIGHT(Column 2) and by more than a third (corresponding to a 70\% increase in the long run) if we add the effect of monopoly on freights (Columns 4 and 7). In contrast, the NHM apparently fostered convergence both before (NHM1)andafter1850 (NHM2). This effect however disappears if we use the monopoly-adjusted series of freights (Columns 4 and 8). TheNHMdummies become positive and significant, although the coefficients are less than a quarter of theEICone. In short, the NHM affected negatively long-range integration only because it chargedhighfreights,whileitmayhaveeven increased the efficiency of the market, by improving the transmission of information and by reducing the risks. 7In contrast, the monopoly of the EIC harmed trade even discounting the effect on freights. The overall effect of state intervention after World War I seems modest. Neither the Dutch (private) market- ing board (VJSP) in the 1920s nor the support to American agriculture after the Great Depression (AAA) affected significantly integration. Only the Dutch public marketing board (the NIVAS) had a positive impact, reducing price gaps between Indonesia and Europe by 10–15\%.v) The variableTELEGRAPHis negative as expect- ed and significant in all specifications. 8We have tried to interact it with a time trend to capture the effect of technical progress and increases in transmission (or changes in policy to set rates), but the results are poor. Our results confirm the earlier work byLew and Cater (2006)on the positive effect of telegraph on world trade. vi) The political shocks in producing countries seem not to have affected international price differen- tials, although of course they may have had important consequences in producing areas. In contrast the campaign against slave-produced sugar in the United Kingdom (SLAVE) had a massive effect on the sugar market. The dummies forWWIare positive and significant, as expected, for India and Indonesia but not for the Atlantic. We interpret them as the effect of disruption in the market on top of war-related increase in transport costs, which should already be captured by the variableLOG_FREIGHT. How much did each variable contribute to long run convergence? To answer, inTable 7we report the share of the total change accounted for by each variable. As the last row ofTable 7shows, the model performs very well: the divergence between the cumulated effect of the variables and the actual change is less than 5\% in four cases and exceeds 20\% only in the Atlantic trade. Most of the total convergence is explained by changes in transportation costs and by the abolition of the EIC, which looms large in the long run analysis, partly because the majority of observations in 1815 refer to India. 9 According to the baseline specification (Column 1), the 6We run an OLS regression withLOG_FREIGHTas dependent variable, explained by a linear trend and by route and product dummies. The product dummies yield a ranking of commodities from the lightest to the heaviest (silk, indigo, tin, cotton, tea, coffee, pepper, rubber, sugar, wheat, jute, saltpetre, linseed, rapeseed, and rice). We deﬁne‘light’theﬁrst nine products (“light”) and‘bulky’the others: this distinction closely mirrors the conventional one between“grain and seeds”and“lighter goods”. 7An ofﬁcial history of the Company offers an alternative explanation of the negative signs in the baseline speciﬁcations: it claims that the NHM sold at a loss to help Dutch middlemen to be competitive on the European market (NHM, 1924:18–19). 8We have tested separately three additional measures of efﬁciency: time trends, total traded quantity and a dummy forﬁxed exchange rates between countries (i.e. the gold standard). All variables are expected negative. More trade should increase theﬂow of informa- tion,ﬁxed exchange rates should reduce the risks of trading and time trends should capture all other improvements. The trends are signiﬁcant only in Indonesia, but the variable is unexpectedly positive—i.e.ceteris paribusmarkets have been disintegrating. The quantity variable is negative, but it worsens the overall results of the regression (available upon request), probably because of endogeneity issues. The dummy forﬁxed exchange rate is incorrectly signed but not signiﬁcant. Evidently, more reﬁned measures of exchange rate volatility than afforded by the available data are needed to adequately capture the effect of exchange rate risk. The time trend is positive and signiﬁcant for Indonesia—i.e. efﬁciency would have decreased, ceteris paribus. We speculate that the coefﬁcient reﬂects the increasing exposure to price shocks originating in other markets. 9But even if we give each route equal weight (i.e. using the sixth speciﬁcation fromTable 6), the EIC emerges as by far the most important factor, accounting for over 45\% of the overall decline.13 D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 Table 6 The causes of integration. (1) (2) (3) (4) (5) (6) (7) (8) (9) PCSE PCSE/IV PCSE/IV PCSE/IV PCSE/IV PCSE/IV PCSE/IV PCSE/IV PCSE/IV C0.517 0.365 0.327 0.366 0.329 0.224 0.312 0.386 0.416 (17.18)*** (8.73)*** (5.53)*** (8.73)*** (7.92)*** (4.48)*** (4.93)*** (5.51)*** (6.84)*** LOG_DUTY0.084 0.051 0.048 0.051 0.033 0.039 0.13 0.021 0.057 (3.64)*** (2.03)** (0.87) (2.04)** (1.28) (1.31) (2.75)*** (0.32) (1.54) EIC0.173 0.236 0.253 0.345 0.287 0.347 0.343 (5.05)*** (6.30)*** (5.66)*** (9.98)*** (7.38)*** (7.04)*** (9.12)*** NHM1 0.105 0.044 0.021 0.054 0.021 0.078 0.137 ( 4.48)*** ( 1.68)* ( 0.22) (2.26)** (0.74) (2.37)** (4.83)*** NHM2 0.07 0.046 0.013 0.023 0.022 0.049 0.087 ( 3.42)*** ( 2.26)** (0.34) (1.04) ( 1.05) (1.50) (3.43)*** AAA0.038 0.032 0.032 0.032 0.02 0.018 0.012 (0.79) (0.69) (0.67) (0.69) (0.46) (0.25) (0.28) VJSP 0.079 0.035 0.035 0.011 0.011 0.031 ( 2.08)** ( 0.90) ( 0.90) ( 0.28) ( 0.24) ( 0.78) NIVAS 0.185 0.125 0.126 0.105 0.105 0.153 ( 3.61)*** ( 2.35)** ( 2.35)** ( 1.93)* ( 1.60) ( 2.83)*** LOG_FREIGHT0.123 0.067 0.053 0.085 (13.27)*** (4.64)*** (2.48)** (3.65)*** LOG_ADJFREIGHT0.067 0.034 0.068 0.073 (4.64)*** (2.01)** (2.64)*** (4.44)*** LOG_FREIGHT*LIGHT0.033 (2.14)** LOG_FREIGHT*BULKY0.108 (6.73)*** TELEGRAPH 0.043 0.082 0.095 0.081 0.083 0.076 0.017 0.117 0.046 ( 2.57)** ( 4.31)*** ( 3.63)*** ( 4.31)*** ( 4.45)*** ( 3.75)*** ( 0.74) ( 3.32)*** ( 2.01)** 14D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 JAVA_WAR0.064 0.063 0.063 0.072 0.071 0.066 (0.83) (0.80) (0.80) (0.89) (0.86) (0.82) MUTINY 0.071 0.066 0.068 0.066 0.085 0.060 0.094 ( 0.82) ( 0.75) ( 0.76) ( 0.75) ( 0.99) ( 0.47) ( 1.06) SLAVE0.239 0.195 0.285 0.195 0.189 0.139 0.293 0.148 (5.62)*** (4.41)*** (4.91)*** (4.41)*** (4.20)*** (2.70)*** (4.98)*** (2.38)** CIVIL_WAR 0.009 0.055 0.08 0.055 0.085 0.082 0.006 ( 0.26) ( 1.47) ( 1.35) ( 1.46) ( 2.25)** ( 1.61) ( 0.15) WWI*ATLANTIC 0.067 0.012 0.033 0.012 0.048 0.044 0.034 ( 1.23) (0.21) (0.55) (0.21) (0.92) (0.62) ( 0.58) WWI*INDIA0.087 0.151 0.167 0.151 0.144 0.086 0.152 (2.08)** (3.40)*** (3.40)*** (3.40)*** (3.32)*** (1.14) (2.93)*** WWI*INDONESIA0.171 0.212 0.181 0.212 0.242 0.278 0.227 (5.67)*** (6.59)*** (4.34)*** (6.58)*** (7.09)*** (4.85)*** (7.27)*** LOG_RATIO10.445 0.497 0.513 0.496 0.463 0.452 0.503 0.51 0.338 (17.77)*** (18.00)*** (15.80)*** (18.00)*** (16.52)*** (12.11)*** (8.70)*** (14.90)*** (7.86)*** N1534 1534 1020 1534 1534 619 231 821 482 Adjusted R-square0.79 0.78 0.79 0.78 0.79 0.64 0.76 0.79 0.76 F153.67*** 129.36*** 120.24*** 129.37*** 129.61*** 49.38*** 77.07*** 133.66*** 76.88*** LLC t-stat 33.12*** 37.82*** 32.65*** 37.82*** 37.65*** 22.88*** 16.25*** 28.45*** 20.65*** Exogeneity chi-square test24.82*** 22.79*** 24.85*** 30.43*** 10.97*** 7.85** 20.20*** 7.70** Exogeneity F test12.90*** 12.90*** 12.83*** 11.31*** 3.81** 7.89*** 12.23*** 3.95** Sheas partial R-square 1st inst.0.89 0.86 0.89 0.88 0.75 0.96 0.94 0.74 Sheas partial R-square 2nd inst.0.46 0.42 0.46 0.46 0.32 0.41 0.45 0.44 Sheas partial R-square 3rd inst.0.6 Significant at *10\%, **5\%, *** 1\%. Notes: PCSE = panel corrected standard errors; IV = instrumental variables estimation;N= sample size;inst.= instrument. The dependent variable in each regression is the natural logarithm of the ratio between import and export prices. Fixed-effects estimation with panel corrected standard errors is used in all cases; instrumental variable estimation for specifications (2) to (9); weighted least squares estimation for specification (6). Specification (3) includes only homogeneous goods; specification (6) includes only two goods per route;specifications (7) to (9) include only goods exported from the U.S., India and Indonesia, respectively. Sources: seeChilosi and Federico (2013: Appendix B). 15 D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 fall in freights mattered almost as much, but this conclusion is decidedly changed if we use the coefficient fromLOG_ADJFREIGHTS(Column 4). The NHM does not appear among the variables in the aggregate analysis, because the series for Indonesia start later, but as Column 7 shows, it played a very important role in the convergence of prices between the colony and Europe, too. The telegraph did help a lot as well, accounting for between a sixth and a quarter of long-run price convergence. In contrast, the cut in duties contributed substantially to the integration only in the Atlantic economy (Column 5). The separate estimates by period (Columns 2 and 3) highlight the sharp differences in the causes of integration. The “early globalization”was mostly determined by the trade liberalization and the abolition of monopolies, while further convergence during the“heyday of globalization” was achieved thanks to the lay-out of the telegraph lines and to fall in sea-borne transport costs. On the whole, political decisions mattered more in the long run because, as shown inSection 4, most of total convergence pre-dated 1870. 6. Conclusion Our results are relevant for two literatures, that on market integration and that on trade and growth in Asia, which, although clearly related, have so far remained largely distinct. Our work contributes to filling in gaps in the literature on global market integration, strength- ening the emerging consensus view in the literature on Europe and the Atlantic trade. The process started early in the 19th century and it was determined to a large extent by institutional changes. Within Europe and betweenEurope and North America, barriers were raised essen- tially by protectionist trade policies, while commerce with Asia was hampered by the monopoly of Western trading companies. These barriers were progressively abolished and, at least for the sample of products/routes we are considering, were only partially re-instated during the Great Depression. Once trade was free from institutional constraints, further convergence was mainly achieved by cutting transportation costs. The cost of sea-borne trade, however, was fairly low already at the beginning of the 19th century and thus the scope for further convergence was limited. 10These findings are consistent with the view that market integration was at the root of the terms of trade boom experienced by Asian countries in the decades before 1870. 11 Future research should systematically assess this impact and examine the implications of market integration for trade and welfare. Appendix A. Supplementary data Supplementary data to this article can be found online athttp://dx.doi.org/10.1016/j.eeh.2015.04.001. 10In contrast, transport costs from the producing areas to the ports were surely high and indeed there is a strong evidence of growing integration in the second half of the 19th century in the domestic market in the United States (Federico and Sharp, 2013), in India (Hurd, 1975; Studer, 2008; Andrabi and Kuehlwein, 2010) and also in the Dutch East Indies (Marks, 2010; van Zanden and Marks, 2012: 25–26). Indeed, in a recent paper,Donaldson (forthcoming)estimates that on average trade created by railways increased Indian GDP by as much as a sixth from 1870 to 1930. 11This is particularly so as new estimates of terms of trade show that India took part in this boom, too (Chilosi and Federico, 2013: 8). Table 7 The causes of integration: decomposition analysis (in percentage). (1) (2) (3) (4) (5) (6) (7) Sample All All All All Atlantic India Indonesia Years 1816–1913 1816–1870 1871–1913 1816–1913 1816–1913 1815–1913 1849–1913 EIC39.27 49.57 57.25 69.14 NHM143.36 LOG_DUTY4.65 5.56 1.14 4.65 42.11 1.72 2.98 LOG_FREIGHT34.74 42.3 59.48 82.49 LOG_ADJFREIGHT16.79 8.16 24.99 TELEGRAPH20.34 15.8 37.3 20.31 13.08 23.47 14.6 Total 98.99 113.24 97.92 98.99 137.68 102.5 85.92 Notes: All the decompositions report the share of the percentage change in the dependent variable accounted for by each independent variable and are based on the results of the panel analysis presented inTable 6. Columns (1) to (3) are based on the second specification, Column (4) is based on the fourth specification, and Columns (5) to (8) are based on the seventh, eighth and ninth specifications, respectively. The exact starting date differs somewhat across samples in order to maximise the number of covered goods. We omit explicative variables which did not affect the dependent variables at the beginning and/or at the end of the period (such as the World War dummies). The cumulated change in the dependent variable is obtained as the average of the differences between the values fitted by the panel at time0and that of timeTdivided by the former. Sources: see the text andChilosi and Federico (2013: Appendix B). 16D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 References Allen, R.C., 2011.Global Economic History. A Very Short Introduction. Oxford University Press, Oxford. Andrabi, T., Kuehlwein, M., 2010.Railways and price convergence in British India. Journal of Economic History 70, 351–377 . Bai, J., Perron, P., 1998.Estimating and testing linear models with multiple structural changes. Econometrica 66 (1), 47–78 . Bai, J., Perron, P., 2003.Computation and analysis of multiple structural change models. Journal of Applied Econometrics 18, 1–22 . Bateman, V.N., 2011.The evolution of markets in early modern Europe: a study of wheat prices, 1350-1800. The Economic History Review 64, 447–471 . Booth, A., 1988.Agricultural Development in Indonesia. Allen and Unwin, Sidney. Bowen, H.V., 2006.The Business of Empire. The East Company and Imperial Britain, 1756-1833. Cambridge University Press, Cambridge. Brandt, L., Ma, D., Rawski, T., 2013.From divergence to convergence: re-evaluating the history behind Chinas economic boom. LSE Working Paper 175/13. Broadberry, S., 2013.Accounting for the Great Divergence. LSE Economic History Working Paper 184/2013. Brunt, L., Cannon, E., 2014.Integration in the English wheat market 1770-1820. Explorations in Economic History 52, 111–130 . Chilosi, D., 2014.Risky institutions: political regimes and the cost of public borrowing in early modern Italy. Journal of Economic History 74, 887–915 . Chilosi, D., Federico, G., 2013.Asian globalizations: market integration, trade and economic growth, 1800–1938. Economic History Working Papers No. 183/2013. London School of Economics. De Klerck, E.S., 1938.History of the Netherlands East Indies vol. 2 (Rotterdam: Brusse). Dermigny, L., 1964.La Chine et lOccident. Le commerce a Canton au XVIIIe Siècle 1719-1833. Tome III. SEVPEN, Paris. Dobbin, C., 1983.Islamic Revivalism in a Changing Peasant Economy. Central Sumatra, 1784-1847. Curzon Press, London and Malmö. Donaldson, D., 2014.Railroads and the Raj: estimating the Economic Impact of Transportation Infrastructure. American Economic Review (forthcoming) . Dutt, R., 1969.first ed. The Economic History of India vols. 1-2. A. Kelley, New York (1902-1904). Fairbank, J., 1978.The creation of the treaty system. In: Twitchett, D., Fairbank, J. (Eds.), The Cambridge History of China vol. 10, part I. Cambridge University Press, Cambridge, pp. 213–264. Fama, E., 1970.Efficient capital markets: a review of theory and empirical work. Journal of Finance 25, 383–417 . Fasseur, C., 1992.The Politics of Colonial Exploitation. Java, the Dutch and the Cultivation System. South-East Asia Program Cornell University, Ithaca. Federico, G., 2011.When did the European market integrate? European Review of Economic History 15, 93–126 . Federico, G., 2012.How much do we know about market integration in Europe? The Economic History Review 65 (2), 470–497 . Federico, G., Sharp, P., 2013.The Cost of railroad regulation: the disintegration of American agricultural markets in the interwar period. The Economic History Review 66, 1017–1038 . Federico, G., Tena, A., 2013.World trade 1800–1938. paper presented to the Seventh World Cliometric Conference (Hawaii). Feuerwerker, A., 1980.Economic trends in the late Ching Empire, 1870- 1911. In: Fairbank, J., Liu, K. (Eds.), The Cambridge History of China vol. 11, part II. Cambridge University Press, Cambridge, pp. 1–69.Feuerwerker, A., 1983.Economic trends, 1912–1949. In: Fairbank, J. (Ed.), The Cambridge History of China vol. 12, part I. Cambridge University Press, Cambridge, pp. 28–127. Furnivall, J.S., 1976.Netherlands India. A Study of a Plural Economy. B. M. Israël., Amsterdam. Harley, K., 1988.Ocean freight rates and productivity, 1740–1913: the primacy of mechanical invention reaffirmed. Journal of Economic History 48, 851–876 . Headrick, D., 1988.The Tentacles of Progress. Technology Transfer in the Age of Imperialism, 1850-1940. Oxford University Press, Oxford. Hoag, Christopher, 2006.The Atlantic cable and capital market information flows. Journal of Economic History 66, 342–353 . Horlings, E., 1995.The Economic Development of the Dutch Service Sector, 1800-1850. NEHA, Amsterdam. Houben, V.J.H., 2002.Java in the 19th century: consolidation of a territorial state. In: Dick, H., Houben, V.J.H., Lindblad, J.T., Wie, T.K. (Eds.), The Emergence of a National Economy. An Economic History of Indonesia, 1800-2000. Allen & Unwin and University of Hawaii Press, Crows Nest and Honolulu, pp. 56–81. Hurd, J., 1975.Railways and the expansion of markets in India, 1861- 1921. Explorations in Economic History 12, 263–288 . Hynes, W., Jacks, D.S., ORourke, K.H., 2012.Commodity market disintegration in the interwar period. European Review of Economic History 16, 119–143 . Jacks, D.S., 2005.Intra–and international commodity market integration in the Atlantic economy. Explorations in Economic History 42, 381–413 . Jacks, D.S., Pendakur, K., 2010.Global trade and the maritime transport revolution. The Review of Economics and Statistics 92, 745–755 . Knight, R., 2010.Exogenous colonialism: Java sugar between Nippon and Taikoo before and during the interwar depression, c. 1920–1940. Modern Asian Studies 44, 477–515 . Korthals Altes, W.L., 1994.Changing Economy in Indonesia vol. 15 Prices (non Rice) 1814-1940. Royal Tropical Institute, The Hague. Levin, A., Lin, C.F., Chu, C., 2002.Unit root tests in panel data: asymptotic and finite-sample properties. Journal of Econometrics 108, 1–24 . Lew, B., Cater, B., 2006.The telegraph, co-ordination of tramp shipping and growth in world trade 1870-1910. European Review of Economic History 10, 147–173 . Maddison Project, 2013. . At,www.ggdc.net/maddison/maddison- project/data.htm(Accessed on January 25th 2013). Marks, D., 2010.Unity or diversity? On the integration and efficiency of rice markets in Indonesia, c.1920-2006. Explorations in Economic History 47, 310–324 . Mui, H., Mui, L.H., 1984.The Management of Monopoly. A Study of the English East India Companys Conduct of its Tea Trade, 1784- 1833. University of British Columbia Press, Vancouver. Nederlandsche Handel-Maatschappij, 1924.A Brief History of the Netherlands Trading Society, 1824-1924. Mouton & Company, The Hague. North, D.C., 1968.Sources of productivity change in ocean shipping, 1600-1850. Journal of Political Economy 76, 953–970 . ORourke, K.H., 2006.The worldwide economic impact of the French Revolutionary and Napoleonic Wars, 1793-1815. Journal of Global History 1, 123–149 . ORourke, K.H., Williamson, J.G., 2002.When did globalization begin? European Review of Economic History 6, 23–50 . Parthasarathi, P., 2011.Why Europe Grew Rich and Asia Did Not. Global Economic Divergence, 1600-1850. Cambridge University Press, Cambridge.17 D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18 Pomeranz, K., 2000.The Great Divergence. Europe, China and the Making of the Modern World Economy. Princeton University Press, Princeton. Razzaque, M., Osafa-Kwaako, P., Grynberg, R., 2007.Long-run trend in the relative price: empirical estimation for individual commodities. In: Grynberg, R., Newton, S. (Eds.), Commodity Prices and Development. Oxford University Press, Oxford, pp. 35–67 (2007). Rönnbäck, K., 2009.Integration of global commodity markets in the early modern era. European Review of Economic History 13, 95–120 . Roy, T., 2000.The Economic History of India, 1857–1947. Oxford University Press, Oxford. Sachs, J., Warner, A.M., 2001.Natural resources and economic development: the curse of natural resources. European Economic Review 45, 827–838 . Shah, M., Williamson, J.G., 2004.Freight rates and productivity gains in British tramp shipping 1869-1950. Explorations in Economic History 41, 174–203 . Sharp, P., Weisdorf, J., 2013.Globalization revisited: market integration and the wheat trade between North America and Britain from the eighteenth century. Explorations in Economic History 50, 88–98 . Smits, J.P., Horlings, E., van Zanden, J.L., 2000.The Measurement of Gross National Product and its Components, 1800-1913(Groningen Growth and Development Centre Monograph series no 5 2000. data available at,http://nationalaccounts.niwi.knaw.nl/(accessed in January 2009). Solar, P.M., 2013.Opening to the East. Shipping between Europe and Asia, 1780-1830. Journal of Economic History 73, 625–661 . Studer, R., 2008.India and the great divergence: assessing the efficiency of grain markets in 18th and 19th century India. Journal of Economic History 68, 393–437 . Tomlinson, B.R., 1993.The Economy of Modern India, 1860–1970. Cambridge University Press, Cambridge.Uebele, M., 2011.National and international market integration in the 19th century: evidence from comovement. Explorations in Economic History 48, 226–242 . Van der Eng, P., 1996.Agricultural Growth in Indonesia. Macmillan, London and Basingstoke. Van Zanden, J.L., Marks, D., 2012.An Economic History of Indonesia, 1800–2010. Routledge, London and New York. Van Zanden, J.L., van Riel, A., 2004.The Strictures of Inheritance. The Dutch Economy in the Nineteenth Century. Princeton University Press, Princeton. Wakeman, F., 1978.The Canton trade and the Opium war. In: Twitchett, D., Fairbank, J. (Eds.), The Cambridge History of China vol. 10, part I. Cambridge University Press, Cambridge, pp. 163–212. Webster, A., 2009.The Twilight of the East India Company. Boydell Press, Woodbridge. Williamson, J.G., 2008.Globalization and the great divergence: terms of trade boom, volatility and the poor periphery, 1782-1913. European Review of Economic History 12 (3), 355–391 . Williamson, J.G., 2011.Trade and Poverty. When the Third World Fell Behind. MIT Press, Cambridge, Mass. Williamson, J.G., 2012.Commodity prices over two centuries: trends, volatility, and impact. Annual Review of Resource Economics 4, 185–206 . Williamson, J.G., 2013.The commodity export, growth, and distribution connection in Southeast Asia 1500–1940. paper presented to the Southeast Asian, Economic Growth Conference (Bangkok). Wooldridge, J.M., 1995.Score diagnostics for linear models estimated by two stage least squares. In: Maddala, G.S., Phillips, P.C.B., Srinivasan, T.N. (Eds.), Advances in Econometrics and Quantita- tive Economics: Essays in Honor of Professor C. R. Rao. Blackwell, Oxford, pp. 66–87. 18D. Chilosi, G. Federico / Explorations in Economic History 57 (2015) 1–18

68490

AN ANNOTATED BIBLIOGRAPHY OF AT LEAST FIVE ADDITIONAL RESOURCES RELATED TO THE RESEARCH TOPIC BY DAY 7 OF WEEK 2.***NO PLAGIARISM*** IT MUST BE LESS THAN 30\% MATCH ON AUTHENTICITY REPORT.***PLEASE DO

An annotated bibliography of at least five additional resources related to the research topic by Day 7 of Week 2.***No plagiarism*** It must be less than 30\% match on authenticity report.***PLEASE DO - Business & Finance

CATEGORIES

AN ANNOTATED BIBLIOGRAPHY OF AT LEAST FIVE ADDITIONAL RESOURCES RELATED TO THE RESEARCH TOPIC BY DAY 7 OF WEEK 2.NO PLAGIARISM IT MUST BE LESS THAN 30\% MATCH ON AUTHENTICITY REPORT.***PLEASE DO

An annotated bibliography of at least five additional resources related to the research topic by Day 7 of Week 2.No plagiarism It must be less than 30\% match on authenticity report.***PLEASE DO - Business & Finance