The Accumulative Dynamic Capability effects of Strategic Alliances on Firm Success:

University of Groningen Faculty of Economics and Business

International Business & Management 2014/2015 Nettelbosje 2 9747 AE Groningen

The Netherlands

The Accumulative Dynamic Capability

effects of Strategic Alliances on Firm Success:

A study of the Brazilian Automotive Industry

June, 2015

MSc Thesis (EBM719A20.2014-2015.2) IB&M

Pedro Celli – s2515601

p.g.pieroni.celli@rug.nl

Aweg 5C – 309 9718 CS Groningen +31610472009

The Accumulative Dynamic Capability effects of Strategic Alliances on Firm Success: A study of the Brazilian Automotive Industry

MSc Thesis by P. Celli – Master program of International Business & Management Faculty of Economics and Business, University of Groningen

June 2015

ABSTRACT

While attempting to answer the research question ‘What are the effects caused by the accumulative capabilities and knowledge sharing on alliance success when looking into the Brazilian automotive industry?’ the study was able to assess the implications for future businesses entering the Brazilian automotive market, as well as which strategic path to take in order to achieve success in terms of abnormal stock returns. The thesis at hand looks at the effects of Strategic Alliances in the Brazilian automotive industry. This is particularly interesting due to large automotive market in Brazil, as well as its large network of subsidiaries, suppliers and manufacturers; and soon to become the third largest automotive industry in the world.

By exploring the theories of strategic alliances, the study evaluated the different effects that are cause by comparing horizontal and vertical alliances, as well as variables such as alliance experience and effectiveness. From the empirical study it was possible to deduce that alliance effectiveness has a positive impact when looking into the Brazilian automotive market. The tests ran on alliance type, as well as experience, did not seem to result in significant implications, and require future studies in order to access their validity.

ACKNOWLEDGEMENTS

!

!

ACCUMULATIVE'DYNAMIC'CAPABILITIES' 11!

EVENT'STUDY'AND'STOCK'VALUATION' 13!

HORIZONTAL,'VERTICAL'AND'THE'BRAZILIAN'INDUSTRY' 14!

List of tables

Table 1 List of Variables Table 2 Correlation Matrix

Table 3 Breusch-Pagan test for heteroscedasticity Table 4 Colinearity statistics

INTRODUCTION

In the world of the automotive industry, strategic alliances have become a common direction for firms to improve their financial performance by means of knowledge sharing in order to produce better products (Burgers et al. 1993). In recent years we have seen an increase in firm collaboration by means of horizontal and vertical alliances in the automotive industry, providing companies with great economical growth and core competencies development. Ranging from all over the world, from European and American brands, to Japanese and other Asian ventures, examples like, Fiat-GM, Nissan-Renault, Toyota-BMW or Daimler-Chrysler, show that there is great potential in inter-firm collaboration and development. Once known as an industry of isolated proprietary knowledge, now companies are slowly lowering their guards and sharing resources and knowledge in order improve their competencies and gain competitive advantage against other brands (Dyer & Nobeoka, 2002).

modern approaches of strategic alliances, companies are able to remain competitive while generating new capabilities and ultimately improve its performance and growth. This becomes particularly interesting when assessing the situation of the automotive industry in Brazil. Although many vertical alliances are achieved by selecting local suppliers and manufacturers, it is virtually unheard of horizontal alliances with Brazilian subsidiaries and local brands, restricting the opportunity of knowledge sharing and increase of accumulative capabilities between the allied firms. The main component studied in this thesis is the consideration of experience versus effectiveness of alliances. Essentially the paper will investigate what remains at the core of improved alliance performance and success, looking into: a) the ability to participate in many alliances and therefore gain knowledge from the sheer number of interactions, or b) the effectiveness of the alliance as seen by the products and patents created in each individual alliance.

alliances and the different approaches taken by companies in the Brazilian automotive industry.

The study will discuss several aspects that influence the aforementioned factors. Firstly, it will provide theoretical explanation of what constitutes alliance capabilities and its success, explaining the nature of these relationships within the automotive industry by means of dynamic capabilities and stock valuation. Secondly, alliance experience versus effectiveness, as being the crucial source of competitive advantage in the industry. Furthermore it will discuss the distinction of vertical and horizontal alliances, and how it effects the relation of firms in Brazil. This becomes increasingly relevant as Brazil will soon become the third largest automotive market in the world after USA and China (Geromel, 2011). Lastly, it will look into the accumulative capacity of strategic alliances, meaning dissecting the influence of continuous partnerships that demonstrate an improvement not only in performance but also of knowledge sharing and creation. For the purpose of the study, we encapsulate strategic alliances into horizontal and vertical alliances as well as to research & development, production or technology agreements. The relevance of these terms will be explained in the following chapter.

The study will be guided by the main research question ‘What are the effects caused by the accumulative capabilities and knowledge sharing on alliance success when looking into the Brazilian automotive industry?’ This will help identify sub questions as to which factors of the strategic alliance are particularly relevant for a successful union, as well as, can the companies opt for a strategy that will guarantee a successful and effective alliance.

BACKGROUND & LITERATURE REVIEW

Previous literature assessed for this study suggests that strategic alliances have a beneficial effect for firms in the automotive industry (Burgers et al. 1993). Horizontal alliances between large firms and vertical alliances with suppliers and manufacturers do not look at the alliance as a collective result of innovation, growth, performance and capabilities; but instead the singular act of the agreement bound by the contract to achieve a certain goal. This creates a research gap in order to find the accumulative effects of such components on performance and success over time. The guiding proposition for the research is to identify what factors influence success and performance, and most importantly from a relational point of view, does this have a more significant impact when considering knowledge sharing and application. Certain external factors also have to be taken into consideration, such as the complexity of the projects and relations, as well as the role of structure and knowledge flow. This study looks into the Brazilian market as previous research has briefly identified the benefits of alliances of large firms with local suppliers in Brazil, but omit horizontal alliances, as well as the accumulative effects of the relationship. A partial reason for this occurrence is the fact that automotive horizontal alliances in Brazil are few and far between and mostly done with private or family owned companies.

strengthened the popularity of cooperation in the country. The synergy that is created by complementary technology and accumulative capabilities has prosperous effects on company performance as well as the overall relation built between firms (George et al. 2001). In order to further understand and begin to explain these effects it is necessary that each theory is explained and related to the topic.

Strategic Alliances

Accumulative Dynamic Capabilities

Current studies on strategic alliances support the notion that all alliances must follow a specific protocol in order to achieve the desired effect of the accumulative nature of alliance capabilities (Kale & Singh, 2007). Initially, the creation of a dedicated function within the firm to supervise and coordinate all alliance activities has shown significant benefits on alliance success (Kale et al., 2002). This is followed by continuous feedback between the parties involved, gaining experience by implementing and generating constructive criticisms on their processes and behavior. Lastly, this is taken into consideration to accumulate the knowledge and practically apply it back into the business, so that results can be viewed. This process, when multiplied by the multitude of alliances formed by the firm, ensures the ability to absorb capabilities and further assimilate, valuate and apply the new found capabilities (Cohen & Levinthal, 1990).

Key to understanding the accumulative nature of alliance competences is the concept of dynamic capabilities. Defined as the ability to incorporate, construct and reconfigure internal and external resources while quickly adapting to the demands of the complex environment that is the automotive industry (Teece et al. 1997). When exploiting dynamic capabilities by means of path dependencies and current position in the market, firms are able to innovate and grow while reaching competitive advantage (Leonard-Barton, 1992). This concept surrounds two important aspects that are not generally involved when discussing firm capabilities and its effect on success. The dynamism of the capabilities is attributed to the response to the market, and rapid changes to answer the demand. This implies that the continuous accumulation of knowledge and resources by means of several alliances throughout time would amalgamate to a superior form of capabilities, and ultimately reaching greater success due to its implementation, creation and reconfiguration. Thus, dynamic capabilities become relevant as it intertwines with the notion of alliance experience or effectiveness; strength in numbers or superior interactions.

Event Study and Stock Valuation

2002). This increase in stock earnings can be attributed to the accumulation of dynamic capabilities that firms are able to amalgamate over their collaboration. Practically speaking, the abnormal returns show that the market perceives the alliance as a successful one, either by their recent developments and products, or simply as a new addition to the alliance portfolio, mirroring their success from past agreements. This in turn will cause a disruption in the market, where the demand for the stock will rise and consequently the stock price will too. This notion has been the usual paradigm for strategic alliances and abnormal returns, more specifically in the pharmaceutical and telecom industries (Kale et al., 2002). Nowhere in current and extensive literature has this been conclude with regard to the automotive industry and most significantly in such a crucial market as Brazil has become. This study intends to show that the use of abnormal returns in the Brazilian automotive industry serves the purpose of proving the alliances validity, with either a successful or unsuccessful outcome. This not only will shed light on the type of alliance with the greatest record, but it also will dissect in which instance is the alliance most prominent, when collecting skills and knowledge from a large portfolio, or curating meaningful alliances where capabilities and knowledge can blossom from its interaction. Furthermore, it will expand on the notion of accumulative dynamic capabilities in the automotive industry, an industry until recently focused mostly on mergers and acquisitions (Snavely, 2014).

Horizontal, Vertical and the Brazilian Industry

The Brazilian market has often dealt with welcoming large corporations into the country, interested in merging or acquiring smaller companies to fully obtain and control their resources, as a main strategy for growth (Ghisi et al., 2008). Due to new policies on tax reduction, in the early 90s this lead to seven of the largest car manufacturers in the world to set up production plants all over Brazil, doubling its production in less than ten years. Most recently, producing just fewer than 4 million units per year, the manufacturing segment of the industry is over saturated, and does not allow for much room for local companies to compete or grow (Correa et al., 1998). Ford, Fiat, GM, Renault, VW are a few of the names that have entered Brazil from the very beginning and have helped shape the industry in Brazil. These however have entered the market with the sole purpose of manufacturing, with no intent of knowledge growth or local innovation. Brazil is soon to become the third largest automobile manufacturer after USA and China, expected to reach 6.2 million units per year by 2025 (Muller, 2012). This is mainly due to global brands looking to enter the Brazilian market following the footsteps of its competitors. The well established Kia, Hyundai, Honda, Nissan and BMW already have their sights on setting up plants in the coming years. Emerging brands like the Chinese JAC and Chery, and Indian Tata and Mahindra will inevitably look into gaining market share in the South American market. Due to trade agreements in the Mercosur trade block, cars produced in Brazil are exempt from sales tax within South America, making it highly attractive for new comers (MercoPress, 2011). However, this tax openness will continue the process of local expansion and growth, but will hinder the possibility of local innovation and emergence, as smaller companies cannot compete with the incoming MNCs. Other forms of entrance into the country can perhaps benefit both the parent and the local companies.

success and compromise (Ford Motor Company, 2008). This provides a clear, if only scarce, example of the mutual benefits encountered when opting for a strategic alliance as opposed to a mergers and acquisitions. Not only can the company accumulate knowledge and share resources, but reduce risks of entering the market segment. In an industry where merging and acquiring is the norm, the dynamic and accumulative benefits of an alliance might just be the next step forward for the incoming card brands, and the local car producers in Brazil. This example as well as the theory behind it leads the research to question the effects of alliances on firm success, and produces the initial hypothesis for the study.

Experience vs. Effectiveness

portfolio. This is possible by the use of multiple sets of skills and knowledge provided by the network of alliances, which can offer cheaper resources, processes and technology from various sides of the value and production chain spectrums (Zollo et al., 2002). Thus, following on the previous theoretical basis, the empirical research will aim to confirm the following hypothesis:

HYPOTHESIS 2. Firms with a greater network of strategic alliances, i.e. experience, will relish increased performance and long-term success than firms with fewer strategic alliances in the Brazilian market.

2003). In this study we will focus on the accomplishment of goals as well as the capability accumulation due to the mutual development of products and patents. This implies that companies that stay within a tightly formed alliance and due to time and interaction, will eventually produce more and better products and patents, which will in turn increase performance and guarantee long term success and competitive edge. Compared directly against hypothesis 2 regarding experience, the idea of firm effectiveness generates the third hypothesis for the study:

HYPOTHESIS 3. Firms with greater product and patent development within strategic alliances, i.e. effectiveness, will relish increased performance and long-term success than firms with fewer developments in the Brazilian market.

RESEARCH DESIGN & METHODOLOGY

Research Design

Conceptual Model

Figure 1. Conceptual Model

The conceptual model illustrates the logical progression derived from the extensive literature review and theoretical background. It visually demonstrates the intellectual path that the research will cover from the beginning to its conclusion. As we can see the initial topic of strategic alliance, including its many different forms, lead the study towards investigating the potential benefits of accumulative dynamic capabilities in different alliances as well as the direct and indirect effects on success. This in turn would effect the stock valuation of the parent firm in the set time where the data was collected, generating Hypothesis 1, and eventually proving or disproving that this is indeed a measurement of success and increased performance. Another view covered in the literature review looked into the distinction between capabilities due to the general amount of alliances a firm has, versus the perceived effectiveness of these alliances, in this case seen as development and innovation. These two distinct measures are to create Hypothesis 2 and 3 that will further explain the effects of accumulative capabilities on stock valuation and ultimately the success of firms and alliances in the Brazilian automotive industry. The model will guide the research in its entirety to make sure all variables are considered in the empirical testing, as well as when devising the corresponding implications from the results.1

Method

The appropriate path followed by a deductive empirical study is once the theoretical background has been paved and hypotheses built upon it, a quantitative analysis is carried out by means of a statistical test to validate the previously mentioned propositions. This way, it is possible to predict the effects of strategic alliances on the long-term success and improved performance of companies entering the Brazilian automotive market. Furthermore, it dissects deeper into what aspects of the accumulative dynamic capabilities will influence this improvement, meaning the experience or effectiveness of alliances. As the research aims to explain the magnitude of effects from strategic alliances capabilities on their stock valuation, we will be looking at separate yet relatable variables. The specifics of the data samples, variables and analysis methods will be presented in the coming sections of the chapter.

Data Source

The majority of the data for the study was obtained with Thomson Reuters Databases, this one in particular named SDC Platinum. The database specializes in maintaining and publishing global records of strategic alliances, merger & acquisitions and all its relevant financial data. Furthermore, the database has been used in several studies pertaining strategic alliances, performance of multinationals and global industries (Kale, 2002; Oxley, 2009; Singh, 2008; Gulati, 2009). Complementary databases available at Rijksuniversiteit Groningen were used to mine for corporate information that was not promptly available via SDC, among them are Orbis, Zephyr and Business Source Premier. For the complementary stock market information Yahoo Finance was used.

Sample

every and any company involved in automotive functions, production, manufacturing and development.2 _{With this standard it was possible to obtain results from over 20} multinational and local companies involved in approximately 300 strategic alliances over a period of 20 years (1995-2015). Most of the variables required for the study were immediately found from this search alone, however as it will explained in the variable section, certain values were gathered using different methods and databases.

Independent Variables Type of alliance

The initial point of departure from the investigation is identifying among the many types of alliances (e.g horizontal, vertical) performed in the Brazilian automotive market, which are prominent, provided by the SDC database values. Furthermore, to look at what are their effects on the residual success, i.e. abnormal returns, of the alliance that will later be explained as the stock valuation and abnormal returns on company shares.

Experience

Literally, the number of alliances formed, firm experience is a variable that causes much discussion in the field of strategic management and alliances (Kale, 2000, 2002; Anand & Khanna, 2000). Often determined as the quantity of previous strategic alliances engaged by a firm, meaning the larger the alliance portfolio, the greater the experience. As discussed in the previous chapter however, there are others who argue that there is a ‘Goldilocks’ ratio, and too many or too little alliances will also hinder the firm. That is why this variable is relevant for the research, as it will help identify not only the optimal amount of alliances for stock valuations, but if indeed it matters more than the effectiveness of alliances ties. By using the SDC database, this variable was simply gathered by selecting multinational parent companies that operate in Brazil and have engaged in local alliances in the past 20 years, including current and dismantled ones. Thus experience, as an explanatory variable, is measured by the total alliance portfolio of the automotive company in Brazil.

Effectiveness

Literally, the number of patents and products developed, the other argument pertaining the impact of strategic alliances, is the subject of effectiveness. Ignoring the time and quantity factor of alliances, effectiveness looks into embeddedness of resources, skills and capabilities. Meaning that a company does not infinitely share knowledge with infinite alliance partners, but rather invests its focus on correctly diffusing and implementing these accumulative dynamic capabilities. During several studies of strategic alliances in the high-tech industry (George et al., 2001; Belderbos et al., 2012; Teece et al, 1997), this was attributed to the research and development portion of the company. Meaning explicitly, the creation and development of new patents and new products was a measure of a firm’s success when entering an alliance. This would debunk the assumption that knowledge is gained by quantity, but rather by quality of interactions and knowledge diffusion. While SDC had a large database for patents and product development, it did not fulfill all companies at the moment of the search. This is soon be rectified by a second data pull from the SDC Platinum, or complementing the information with the very capable Orbis database, which also deals with patents, licensing and lists of new products.

Dependent Variable

Used in many studies (Oxley, 2009; Gulati, 2009; Kale, 2002) of strategic alliances and its inherent success effects, the academic standard often adopted by scholars is the dependent variable called abnormal stock-market returns. It measures the incremental value generation that can happen in the subsequent months of the alliance announcement in the public stock market. In order to measure this effect, it is required to use a widely accepted method for these studies, namely the Event Study Method. The variable is measured by using an asset-pricing model, where the daily data regarding the firm’s returns is gathered for a period of 180 days (6 months) before the announcement of the alliance. The current stock-market returns being calculated by the standard formula:

rit =αi +βirmt +εit

Once these returns are collected and calculated, the model calculates the daily returns incurred after the alliance period has been announced on the market. This is then later compared to a predetermined time window after the alliance has been made, in order to compare the fluctuation from what was predicted to be the valuation rate, according to the previous 6 months, and the actual valuation that happened after the announcement. As such we can calculate the abnormal returns that were gained, in essence, the additional value that was created by the alliance announcement. This information was gathered by using Yahoo Finance, a public financial database that provides publicly traded stock market information dating back to 1980. These figures are then value-weighted against the financial index S&P 500.

Control Variables

As the success of the alliance most likely does not solely depend on the independent variables previously looked at, some controls must be considered to strengthen the validity of the argument and account for adverse effects.

Firm Size - Due to a larger arsenal of already existing resources and capabilities, larger firms might have a larger probability of alliance success, hence it must be monitored to isolate the effects of the independent variables, measured by employee count.

Market fluctuation – A common measure in the investment world, the variable simply named Beta, measures the fluctuation of the market at a current point in time. This can be obtained by financial sites such as Yahoo, and serves as benchmark for comparing the performance of firms with regard to the situation of the market. Most relevant for this research, it also allows for calculating the difference of abnormal returns generated in comparison to the market movement. But it also serves as a control, as in times of crises or booms, the Beta will give an indication of the market health.

Logistic regression Analysis

Considering the variables chosen for the study, the most appropriate method for analysis in this case will be logistic regression method. The method was chosen as it is possible to separate the values into two categories. In this case the rate of abnormal return, being smaller or larger than five percent, which indicates the success or failure of the alliance in terms of experienced abnormal returns. This technique allows investigating the relationships and interconnectivity of categorical and continuous variables from a secondary statistical data set (Thomas, 2004). As stipulated by the following equation:

Y = ß0 + ß1x1 + ß2x2 + ß3x3 + ε

Y being the dependent variable of alliance success, or abnormal stock returns, which is effected by the following independent variables illustrated by X. ß represents the regression coefficient, whereas ß0 represents the slop of the regression line, and so a constant. The error term ε accounts for the residual differences that the test might incur. Combining the formula with the predetermined variables will create the following:

Ŷ (Abnormal Returns) = ß0 + ß1 (Horizontal Alliances) + ß2 (Experience) + ß3 (Effectiveness) + ε

DATA ANALYSIS & RESULTS

Variable Analysis

Compiled from three separate databases, the information gathered for the different variables were manually researched over a period of one week from Thomson Reuters’s SDC Platinum database, Yahoo Finance Historical Stock and Orbis Database for Alliances, Merges and Acquisitions. A complete list of the variables used for testing can be found below, where the dependent, independent and control variables are labeled, described and appropriately characterized as a measurement.

As the dependent variable, abnormal return, takes on a dichotomous nature measured above or below 5% return and the independent and control variables take on both categorical and continuous nature, the appropriate test required would be the logistic regression. This test not only tests for the effects of the macro and micro independent variables on the dependent one, but it also looks at the interaction effects of different explanatory variables in regard to the descriptive variable.

Preliminary Analysis

Table 2 portrays the correlation matrix, with relevant information regarding the variables means, standard deviations, and possible correlations found by the different coefficients. Furthermore it is possible to interpret the effect sizes comparative to the faults in the sample. Measured in terms of r, a coefficient above .50 will account for 25% of the variance encountered. r coefficients of .30 represent roughly 9% of variance, and .10 will generally account of 1% of the variance (Tabachnick, 2001). As seen on the table, most coefficients lay below.20, meaning a variance of roughly

1-Name Type Description Measurement

Abnormal Return (↑5%) DV 5% increase of AR with regard to market Dummy variable (= 0 below , = 1 above) Horizontal Alliance IV Type of alliance; Horizontal vs Vertical Dummy variable (= 0 vertical , = 1 horizontal) R&D Patents IV Total number of patent development in SA Continuous; 50000 max.

Strategic Alliances IV Total number of strategic alliances formed Continuous; 82 max.

Firm Size CV Total number of employees; SME <20000 , MNC > 20000 Dummy variable; (= 0 SME , = 1 MNC) Firm Power CV Total yearly revenue; Weak <2m , Strong > 2m Dummy variable; (= 0 SME , = 1 MNC) Market Volatility CV Fluctuation of market index Beta; Non-volatile<1 , Volatile >1 Dummy variable; (= 0 Non-V , = 1 V) Sources: SDC, Orbis, Yahoo Finance

5%. However the coefficients with higher significant (.01 or.05 alpha) are above r= .50 meaning a high variance of 25-30% on the significant variables.

Significance can be observed by the symbols indicating a 0.01 or 0.05 level of significance on the correlation. It is visible that firm type and size have connections, possibly meaning the larger the firm the higher chance of partaking in a horizontal alliance rather than a vertical one. We can also see the control variable power to be correlated with the abnormal return above 5%. This would indicate that firms that have a large yearly income, also will have abnormal returns when entering alliances. Experience and effectiveness are other variables that show a high degree of significance and correlation. This could mean that the independent variables work in unison and that the amount of alliances partaken is just as important as the quality of these alliances for reaching abnormal returns. Less relevant to the study, but still interesting we can see that size, power and experience are all interconnected and seem relevant to each other with a high level of significance and correlation.

Assumptions

In order to perform a logistic regression, it is assumed that the dependent and independent variable do not have a linear relationship. The dependent variable must be dichotomous, and explained by categorical and/or continuous independent variables (Bogdan, 1998). The explanatory variables are also assumed to be non-interval, non-linear, non-normally distributed and without equal variance within each group. The categorical variables must be mutually exclusive and exhaustive, meaning that there can only be one case per group and cases must be assigned a group.

Mean SD 1 2 3 4 5 6 7 1 AR (↑5%) 0,55 0,50 1,00 2 Type 0,87 0,34 -0,20 1,00 3 Eff 30,80 21,34 -0,12 -0,07 1,00 4 Exp 5069,00 10452,91 0,22 0,16 0,359** 1,00 5 Size 0,55 0,51 0,12 0,309* 0,26 0,639** 1,00 6 Power 0,62 0,49 0,34* 0,15 0,329* 0,716** 0,560** 1,00 7 Volatility 0,56 0,50 0,08 -0,01 0,12 0,35 0,01 0,02 1,00

** Correlation is significant at the 0.01 level (two-tailed). * Correlation is significant at the 0.05 level (two-tailed).

Homoscedacity

With a sample size of 55 firms, the assumption of homoscedasticity dictates that variables must have similar levels of variance, meaning that the predictors must have constant residuals as a result of the testing. If this is not the case, the variables face heteroscedasticity, limiting the validity of the results. The Breusch-Pagan aims to identify potential heteroscedasticity, which when found to be above 0.05 of significance level, the sample data is homoscedastic (Tabachnick, 2001). The table below portrays the results of the Breusch-Pagan test for this particular study. As one can see, the significance is greater than .05, meaning that the variables are homoscedastic, enabling it to assume that the residuals are constant and that variance levels are similar.

Multicolinearity

An additional concern in most studies that involve several predictors is the existence of multicolinearity. Here it is assumed that the variables have a strong correlation between predictors, making it unreliable when explaining the effects of the independent variables on the dependent ones (Bogdan,1998). In addition to the calculation of the variance inflation factor, the colinearity statistics table also shows the tolerance level of the variables. The VIF being significantly below a value of 3 and the tolerance levels surpassing the 0.1 threshold, it appears that the variables do not present any sign of multicolinearity.

Table 3 - Breusch-Pagan test for heteroscedasticity

Sample Size Number of predictors Breusch-Pagan test Significance

55 6 13,636 0,115 >.05 no heteroscedasticity

Dependent variable abnormal returns (+5%)

Predictors are all independent variables: Firm age, firm size, industry, ownership, bribery involvement, control of corruption, H0: homoscedasticity

Table 4 - Collinearity statistics

Tolerance VIF Horizontal Alliance ,881 1,135 R&D Patents ,386 2,591 Strategic Alliances ,833 1,200 Firm Size ,525 1,905 Firm Power ,430 2,326 Market Volatility ,913 1,096

Empirical Results

This section will discuss the descriptive statistics found in the studied sample, as well as the results from the logistic regression analysis. With these results it will be possible to observe the effects and impact of the explanatory variables on the dependent variable, as well as interaction effects between them.

Descriptive Statistics

into the results from the logistic regression analysis, showing any potential descriptive and interaction effects of the variables.

Logistic regression Results

In order to identify any influential effects of the control and independent variables on the abnormal returns experienced by companies partaking in a strategic alliance, the logistic regression analysis was selected for the study. By observing significance levels as well as chi-squares and goodness-of-fit tests, it is possible to ascertain any significant results and possible effects. In table 6, the results although not significant are in line with the descriptive statistics previously analyzed. All Hosmer and Lemeshow (H-L) goodness-of-fit results remain similar in Models 1,2 and 3, only to increase drastically with Model 4. The increase of the Nagelkerke R² is desirable as it demonstrates a better fit as each model progresses, it does exceed the 10% threshold, making it a high variance with regard to the dependent variable, abnormal returns. The complete table can be found below and it demonstrates 4 different tested models. Model 1 includes only the control variables with regard to the constant. Model 2 includes the micro independent variables, seen here as the type of alliance partaken, horizontal or vertical. Model 3 moves on to include the continuous independent and macro variables, experience and effectiveness of strategic alliances. Finally, Model 4 integrates the interaction effects experienced by crossing certain macro and micro variables to generate new results.

Model 1 demonstrates the effects of the control variables with relation to abnormal returns above 5% in strategic alliances. The controls were measured as firm size, represented by the amount of employee of the parent company; firm power, represented by the operating yearly revenue generated by the parent company; and Table 5 - Descriptive statistics of variables

N Minimum Maximum Mean Std. Dev. Variance

stock volatility, also knows a beta, a measurement of the stock movement with regard to the industry index S&P500. Firm power appears to be the only significant effect in Model 1, and incidentally a positive one (β = 1,720; p = 0.027) regarding the yearly operating revenue power of a company and their abnormal stock return when venturing into a strategic alliance. Firm size appears to have an negative effect on the abnormal returns of strategic alliance, although the metrics are insignificant (β = -,475; p = 0.525). The last control variable, stock volatility, or beta, also has a negative effect on abnormal returns, but once again a largely insignificant results according to the its significance level (β = -,012; p = 0.984).

The independent variables effects are included in Model 2 and 3, divided as the micro and macro variables, or categorical and continuous. Model 2 includes the effects of the choice in alliance, whether horizontal or vertical, and if that has an impact on the abnormal return outcome when a company partakes in a strategic alliance. Although the preliminary findings looked promising, the logistic regression test shows that the type of alliance, horizontal over vertical, has a negative effect, but this can be dismissed as it shows a largely insignificant result (β = -,366; p = 0.692). Model 3 on the other hand, expands the findings by including the macro variables, firstly it test the alliance experience, regarded as the number of participating alliances in the network, showing a positive result, but again insignificant (β = ,008; p = 0.726). Alliance effectiveness (or the amount of patents and products developed) on the other hand has a slight positive result with regard to the abnormal returns of strategic alliances, and also slightly significant, as it resides below .1 (β = ,010; p = 0.083). The inclusion of the macro variables still achieved a positive result when looking at the firm power variable. Here once again the variable has a positive impact on abnormal returns and a significant one (β = 1,865; p = 0.050).

horizontal alliance (β = ,001; p = 0.028). The other interaction variable that includes the alliance experience shows a negative effect but with no significance (β = -,514; p = 0.304). Once again the variable firm power appears to have a positive and significant effect on the dependent variable, as it remains so thru all four models tested in the losgistic regression (β = 1,802; p = 0.06).

Although the logistic regression test did not generate many significant results, it provided the study with new insight into certain effects and characteristics of the variables. In the coming chapter, the potential explanation for these effects as well as the conclusions that can be drawn from these results will be discussed. Also in the following chapter the study will attempt to explain the possible reasons for the non-significant results, and potential limitations.

! ! !

Table 6 Logistic regression results

Model 1 Model 2 Model 3 Model 4

β Exp(B) Sig. β Exp(B) Sig. β Exp(B) Sig. β Exp(B) Sig. Constant -,604 ,547 ,276 -,327 ,721 ,711 -,210 ,810 ,819 -3,915 ,020 ,316 Firm Size -,475 ,622 ,525 -,396 ,673 ,608 -,285 ,752 ,735 ,014 1,014 ,987 Firm Power 1,720 5,584 ,027** 1,716 5,563 ,027** 1,865 6,457 ,050** 1,802 6,063 ,06* Stock volatility -,012 ,988 ,984 -,008 ,992 ,990 ,126 1,134 ,843 ,194 1,214 ,763 Type of alliance -,366 ,694 ,692 -,675 ,509 ,470 2,938 18,869 ,454 Alliance experience ,008 1,008 ,726 ,518 1,678 ,300 Alliance efficiency ,000 1,000 0,083* -,001 ,999 ,254 TA x Aexp -,514 ,598 ,304 TA x Aeff ,001 1,001 ,028** Observations 55 55 55 55 Nagelkerke R² 0,152 0,156 0,236 0,3127 2LL 69,112 68,953 65,108 61,138 Chi² 6,679 6,837 10,683 14,653 H-L Goodness-of-fit 5,207 4,147 6,179 11,262 * p < 0.10; ** p < 0.05; *** p < 0.001 Dependent variable Abnormal Return (↑5%)

DISCUSSION

Hypothesis 1 - Firms in horizontal alliances will relish increased performance and long-term success compared to firms in vertical alliances in the Brazilian automotive industry.

The results presented in table 5 and 6 helps the research by giving an answer to the previously stipulated hypotheses. With hardly any significance in most cases, the results still paint an image of what effects impact the rate of abnormal stock return when companies are venturing into a strategic alliance. Hypothesis 1 proposed that horizontal alliances would have a positive effect on a firm’s alliance success, as seen by the abnormal increase in stock value after the event. Although this is observed in the APPENDIX A by quickly scanning the numbers gathered from the 55 observable results, the same cannot be said when performing the logistic regression. In fact although not significant, the results suggest that companies participating in a horizontal strategic alliance will a negative effect regarding abnormal returns. When the interaction variables are added, the type of alliance variable demonstrates a positive effect is observed, but again not significant. This would suggest that Hypothesis 1 is rejected. The only time where type of alliance becomes a relevant and significant value is when interacted with alliance effectiveness. This suggests that when horizontal alliances focus on the effectiveness of their relation, meaning the creation of patents and products, they will have a positive and significant effect on abnormal stock returns. This does not validate Hypothesis 1, but illustrates the impact of alliances type and effectiveness with regard to the positive effects on firm success, aka returns.

Hypothesis 2 - Firms with a greater network of strategic alliances, i.e. experience, will relish increased performance and long-term success than firms with fewer strategic alliances in the Brazilian market.

the variables, it is similarly insignificant. From analyzing the logistic regression results, it can be concluded that Hypothesis 2 is rejected, and had no foundation or support to justify its statement. This suggests that strategic alliance experience plays no role in the effect of strategic alliance abnormal return.

Hypothesis 3 - Firms with greater product and patent development within strategic alliances, i.e. effectiveness, will relish increased performance and long-term success than firms with fewer developments in the Brazilian market.

LIMITATIONS AND FUTURE RESEARCH

Besides the enlightening results presented by this quantitative study, there are several limitations that must be considered, and elaborated on for future research. Most limitations arise in the areas of reliability, generalizability and accuracy. Firstly, the use of secondary data from a large database inherently presents limitations with regard to the reliability of data, as there is no crosschecking of information. Although many sources were used for the different variables tested, for future studies it would be valuable to seek the confirmation of the values that were used with other sources and/or databases.

The use of secondary data also led to another limitation that might have largely affected the outcome of the research, meaning the usable sample size. Even though the database produced over 150 results regarding strategic alliances in the Brazilian automotive industry, using abnormal returns as the dependent variable meant that public stock information had to be gathered. Since several of the companies that participated in the alliances were privately or family owned, there was no financial information regarding their stocks to be found, as they were not publicly traded. This reduced the sample to a modest 55 usable variables, with complete information. In order to evolve from this limitation, a future study might want to seek alliances, which are strictly traded publicly, or perhaps seek another measure for firm success, other than abnormal stock returns.

Staying within the generalizability issue, the specific choice of horizontal and vertical strategic alliances might prove to be a limiting factor in the research. Essentially this means that other forms of company collaboration; e.g. M&As, Hybrids, JVs, R&D agreements, would have different effects, and were ignored in this research. Ideally a larger and complete study would address all different types of alliances to ensure complete and unbiased results.

The hypotheses and data are not the only methodological shortcomings, as the testing method itself has its biases. The use of logistic regression assumes that the dependent variable is dichotomous (and categorical) and that the independent variables are continuous, reducing the accuracy of the data. This study makes use of dummy variables to achieve such characteristics, transforming the abnormal returns into categorical variables by selecting an increase of 5% as the threshold. Furthermore, the different independent variables are both categorical and continuous, making it less reliable and accurate when performing a logistic regression. Also, the inclusion of the interaction effects could have been developed in further detail as in its initial form it does not present much insight into the topic, neither does it contribute largely to proving the hypothesis.

CONCLUSION

In the recent years of the automotive world, strategic alliances have been used as a tool for collaboration, increased productivity and development of new and better products. The horizontal alliances formed between established companies or the vertical partnerships between suppliers and manufacturers have shown great promise in the Brazilian automotive industry. Previous research on the topic has established that strategic alliances often have a positive effect on company performance, knowledge sharing and competency development. However these studies identify alliances as singular partnerships and are deemed successful by the achievement of a common goal.

The final Hypothesis looked to address similar questions as 1 and 2, and provided deeper insight to the effects of having an effective alliance versus an experienced one. The results show that there is a significant relation between the increase of cumulative abnormal returns of stock value due to the effective alliance, in other words, the ability to develop and produce new products and patents. From this, the study can infer that it is more valuable to invest on durable relations with the allied parties, and seek to develop and innovate by means of new technology and products.

From the empirical results we can derive that as predicted, the ideal scenario for an automotive company looking to enter or grow in the Brazilian market, should seek the path of horizontal alliances. This implication is extremely relevant as Brazil is to become the third largest automotive industry in the world, and with many companies already situated there, all new entrants must find a way to gain some kind of specific advantage. Therefore not only has the study shed light on the gaps between the previous strategic alliances studies, but also from a superficial point of view, shows the possibilities and advantages of choosing one strategy over another. On the other hand, there are several limitations that must be considered; therefore the validity of the study remains in dispute. With the aid of future research and corrections in the areas that were discussed in the previous chapter, this topic can gain further insight and strength for more relevance and impact. It is important to remember however that this topic although not new, has many unexplored areas, especially when considering the Brazilian environment, and all must be taken into account. For academic purposes, the study satisfies the pursuit of the question to what are the effects of strategic alliances, but omits the other factors that might play a part in this specific scenario. For managerial purposes, companies can look at strategic alliances as a source for more significant relations and development by means accumulative capabilities, knowledge sharing and competency development.

REFERENCES

Anand, B., T. Khanna. 2000. Do companies learn to create value? Strategic Management

Journal. 21(3) 295–316.

Ariño, A. 2003. Measures of strategic alliance performance: An analysis of construct validity.

Journal of international Business studies, 34(1), 66-79.

Baum, J.B. & Ingram, P. 2002. Inter-organizational learning and network organization: toward a behavioral theory of the inter-firm. In M. Augier & J.G. March (Eds.), The economics of choice, change and organization. Cheltenham Review: Edward Elgar.

Belderbos, R., Carree, M., & Lokshin B. 2006. Complementarity in R&D cooperation strategies. Review of Industrial Organization, 28: 401-426.

Belderbos, R., Gilsing, V., & Lokshin, B. 2012. Persistence of, and interrelation between, horizontal and vertical technology alliances. Journal of Management, 38(6), 1812-1834. Bogdan, R. C., & Biklen, S. K. (1998). Qualitative research in education. An introduction to theory and methods. Allyn & Bacon, A Viacom Company, 160 Gould St., Needham Heights, MA 02194; Internet: www. abacon. com.

Burgers, W. P., Hill, C. W., & Kim, W. C. 1993. A theory of global strategic alliances: the case of the global auto industry. Strategic Management Journal, 14(6), 419- 432.

Campbell, J. Y., Lo, A. W. C., & MacKinlay, A. C. 1997. The econometrics of financial markets (Vol. 2). Princeton, NJ: Princeton University press.

Coleman, J.S., 1988. Organizations and Institutions: Sociological and Economic Approaches to the Analysis of Social Structure. The American Journal of Sociology, 94: 95-120.

Corrêa, H., & Gisela Martes de Miranda, N. 1998. Supply network management in the Brazilian automotive industry. Integrated Manufacturing Systems, 9(5), 261-271.

Correia, H.L., Miranda, N.G.M. 2001. Supply network management in the Brazilian automotive industry. MCB University Press. 261–271

Dadashian, F., Shakibfar, S., Fazel Zarandi M. H. 2007. Strategic alliance for core competencies improvement in textile industries. International Journal of Management

Science and Engineering Management Vol. 2 No. 2, pp. 98-107

Draulans, J., & Volberda, H. W. 2003. Building alliance capability: Management techniques for superior alliance performance. Long Range Planning, 36(2), 151-166.

Dyer, J.H. & Nobeoka, K. 2000. Creating and managing a high performance knowledge sharing network: the Toyota case. Strategic Management Journal 21(3): 345-367

Dyer, J.H., Singh, H. 1998. The relational view: Cooperative strategy and sources of interorganizational competitive advantage. Academy of Management Review 23(4): 660-679 Emden, Z., Yaprak, A., & Cavusgil, S. T. 2005. Learning from experience in international alliances: antecedents and firm performance implications. Journal of Business Research, 58(7), 883-892.

George, G., Zahra, S. A., Wheatley, K. K., & Khan, R. 2001. The effects of alliance portfolio characteristics and absorptive capacity on performance: a study of biotechnology firms. The

Journal of High Technology Management Research, 12(2), 205-226.

Geromel, R. 2011. Automakers Announce $19 Bi Investments In Brazil; Country Will Be The World's Third-largest Auto Producer. Forbes Magazine

Ghisi, F. A., da Silveira, J. A., Kristensen, T., Hingley, M., & Lindgreen, A. 2008. Horizontal alliances amongst small retailers in Brazil. British Food Journal, 110(4/5), 514-538.

Gulati R. 1998. Alliances and networks. Strategic Management Journal, Special Issue 19(4): 293–318.

Gulati, N. Nohria. 1998. The dynamics of learning alliances: Competition, cooperation and relative scope. Strategic Management J. 19(3) 193–210.

Gulati, R., Lavie, D., & Singh, H. 2009. The nature of partnering experience and the gains from alliances. Strategic Management Journal, 30(11), 1213-1233.

Hagedoorn, J., & Schakenraad, J. 1994. The effect of strategic technology alliances on company performance. Strategic management journal, 15(4), 291-309.

Heimeriks, K. H., Duysters, G., & Vanhaverbeke, W. 2005. Developing alliance capabilities: an empirical study. In Academy of Management Conference, Honolulu (Vol. 5, No. 10.08, p. 2005).

Hoang, H., & Rothaermel, F. T. 2005. The effect of general and partner-specific alliance experience on joint R&D project performance. Academy of Management Journal, 48(2), 332-345.

Kale, P., & Singh, H. 2007. Building firm capabilities through learning: The role of the alliance learning process in alliance capability and firm-level alliance success. Strategic

Management Journal, 28(10), 981-1000.

Kale, P., & Singh, H. 2007. Building firm capabilities through learning: The role of the alliance learning process in alliance capability and firm-level alliance success. Strategic

Management Journal, 28(10), 981-1000.

Kale, P., Dyer, J.H., & Singh, H. 2002. Alliance capability, stock market response, and long‐ term alliance success: the role of the alliance function. Strategic Management Journal, 23(8), 747-767.

Kale, P., Singh, H., & Perlmutter, H. 2000. Learning and protection of proprietary assets in strategic alliances: Building relational capital. University of Michigan Journal.

Lambe, C., Spekman, R. and Hunt, S. 2002, “Alliances competence, resources, and alliance success: conceptualization, measurement, and initial test”, Journal of the Academy of

Marketing Science, Vol. 30 No. 2, pp. 141-158.

Lavie, D. 2007. The competitive advantage of interconnected firms: an extension of the resource-based view. Academy of Management Review, Vol. 31, No. 3, 638–658.

Makino, S., Lau, C., Yeh, R. 2002. Asset-Exploitation Versus Asset-Seeking: Implications for Location Choice of Foreign Direct Investment from Newly Industrialized Economies.

Journal Of International Business Studies, 33(3): 403-421.

Muller, J. 2012. Why The World's Automakers Love Brazil. Forbes Magazine

Oxley, J. E. 1999. Institutional environment and the mechanisms of governance: the impact of intellectual property protection on the structure of inter-firm alliances. Journal of Economic

Behavior & Organization, 38(3), 283-309.

Oxley, J. E., Sampson, R. C., & Silverman, B. S. 2009. Arms race or détente? How interfirm alliance announcements change the stock market valuation of rivals. Management Science, 55(8), 1321-1337.

Park, N. K. 2004. A guide to using event study methods in multi‐country settings. Strategic

Management Journal, 25(7), 655-668.

Rindfleisch, A. 2000. Organizational trust and interfirm cooperation: an examination of horizontal versus vertical alliances. Marketing Letters, 11(1), 81-95.

Singh, H. 2008. Building firm capabilities through learning: the role of the alliance learning process in alliance capability and firm-level alliance success. Development and Learning in Organizations: An International Journal, 22(3).

Snavely, B. 2014. Pace of auto industry mergers hits seven-year high. Detroit Free Press. Gannet Company

Teece, D. J., Pisano, G., & Shuen, A. 1997. Dynamic capabilities and strategic management.

Strategic Management Journal, Vol. 18, No. 7, pp. 509-533

Thomas, A.B. 2004. Research Skills for Management Studies. Routledge. ISBN: 0-415- 26898-2.

Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics.

Voss, K. E., Gammoh, B. S., & Fang, X. 2012. How does a Brand Ally Affect Consumer Evaluations of a Focal Brand?. Psychology & Marketing, 29(12), 929-940.

APPENDIX A - Data

Parent Companies SA Name SA Date SIC Code Horizontal SAR&D # of SAs Beta S&P500 Expected RActual RAbnormal REmploy. Revenue

A Raymond et Cie

Bollhoff Industrial Ltda A Raymond SARL(A Raymond)Bollhoff Industrial Ltda 04/17/96 34293429Yes 28 4 1,260 0,963 1,213 1,077 -0,136 1500 504.200 $ AG Simpson Automotive Co Ltd Usiminas Brasinca AG Simpson Automotive Co Ltd Usiminas Brasinca 09/16/97 3465 3312 6799 Yes 4 11 1,730 1,060 1,834 0,789 -1,045 3000 766.000 $ Alusuisse Lonza Group Ltd

Votorantim Participacoes SA

Alusuisse Automotive Cia Brasileira de Aluminio

05/13/98 3334 3365

Yes

980 24 1,263 1,396 1,763 1,839 0,076 23115 5.031.548 $ Amsted Industries Inc

Iochpe-Maxion SA

Amsted Industries Inc Maxion Foundry & Railroad

03/01/00 3462 3325

Yes

86 19 0,290 0,933 0,271 0,445 0,174 7500 3.000.000 $ ArvinMeritor Inc Randon SA

Implementos ArvinMeritor IncRandon SA Implementos 08/15/02 37114731Yes 3222 12 0,650 0,918 0,597 0,620 0,023 9050 3.766.000 $ Blackstone Group LP Hubner

Industria Mecanica Ltda

Amer Axle & Mnfg Hldg Inc Hubner Industria Mecanica Ltda

10/18/99 3714 3714

Yes

1135 12 1,220 0,972 1,186 1,274 0,088 2190 7.721.200 $ China Automotive Systems Inc

Undisclosed JV Partner Hengsheng Undisclosed JV Partner 10/24/11 3714 6799 Yes 2 27 2,350 1,004 2,359 2,485 0,126 7903 4.930.993 $ Chongqing Huiyang Hldg Co Ltd

Undisclosed JV Partner Lifan Industry (Group) Co LtdUndisclosed JV Partner 07/04/11 37116799Yes 3 37 2,210 1,012 2,237 2,263 0,026 50000 13.573.369 $ Chrysler Corp BMW AG Chrysler Corp BMW AG 04/14/98 3711 3711 Yes 6353 69 1,263 1,209 1,527 1,620 0,093 232165 116.902.047 $ CIE Automotive SA SAIC Autometal SA

Donghua Auto Industrial Co Ltd

09/18/12 3714 3714 Yes 67 54 0,900 1,015 0,914 1,100 0,187 50279 34.538.180 $ Continental AG Sansuy SA Benecke-Kaliko AG Sansuy SA 06/27/98 3714 3021 Yes 625 63 0,950 1,265 1,202 1,277 0,075 67401 43.724.364 $ Cooper Industries Inc

TRW Inc Lucas Electrical SystemsTRW Inc 06/04/98 36473714Yes 2975 32 0,650 1,396 1,050 1,390 0,340 66900 17.539.000 $ DaimlerChrysler AG Marcopolo SA DaimlerChrysler AG Marcopolo SA 09/26/00 3711 3711 Yes 29529 82 1,690 0,913 1,543 1,202 -0,341 279972 159.736.709 $ Dana Corp Varga Participacoes Dana Corp Freios Varga SA 08/26/96 3714 3714 Yes 11,167 43 1,142 1,396 1,594 1,601 0,007 71000 19.913.800 $ Delco Remy International Inc

Irmaos Zen SA Delco Remy International IncIrmaos Zen SA 07/10/98 37143621Yes 548 10 1,200 1,410 1,692 1,714 0,022 1183 649.790 $ Delphi Automotive Systems Cor Delphi Automotive Systems

Corp BGM Franchising International 03/23/01 ₃₇₁₄ 3714 Yes 2546 39 0,870 1,079 0,939 1,455 0,516 146600 18.060.000 $ Delphi Automotive Systems

Corp

BGM Franchising International

Delphi Automotive Systems Corp BGM Franchising International 03/23/01 ₃₇₁₄ 3714 Yes 2546 39 0,870 1,079 0,939 1,455 0,516 146600 18.060.000 $ Donnelly Corp Artur Eberhardt SA Donnelly Corp Industrias Arteb SA 08/18/97 3231 6799 Yes 1784 16 2,170 1,022 2,218 2,045 -0,172 6000 847.927 $ Eletronica Selenium SA Calearo Eletronica Selenium SA Calearo 04/17/98 3651 3663 Yes 25 5 0,780 1,396 1,089 0,986 -0,103 383 67.088 $ Fiat SpA

Fiat SpA Fiat SpAIveco SpA 02/06/01 37113713Yes 15113 74 0,650 1,079 0,701 0,901 0,200 251000 119.821.721 $ Fried Krupp AG Hoesch-Krupp

Fried Krupp AG Hoesch-Krupp Wilhelm Karmann GmbH

Krupp Metalurgica Campo Limpo

Pre-Star(Krupp Hoesch Auto) Wilhelm Karmann GmbH 07/15/96 ₃₄₆₅ 3465 3711 Yes 517 34 1,197 0,963 1,153 1,163 0,010 57913 11.080.288 $

Fried Krupp AG Hoesch-Krupp Fried Krupp AG Hoesch-Krupp Wilhelm Karmann GmbH

Krupp Metalurgica Campo Limpo

Pre-Star(Krupp Hoesch Auto) Wilhelm Karmann GmbH 07/15/96 ₃₄₆₅ 3465 3711 Yes 517 36 1,197 0,963 1,153 1,163 0,010 57193 11.080.288 $ GB Auto SAE Marcopolo SA GB Auto SAE Marcopolo SA 08/17/09 5012 3711 Yes 37 19 0,480 1,049 0,504 0,601 0,097 22000 1.121.853 $ Georg Fischer AG Fundicao Tupy Georg Fischer AG Fundicao Tupy 08/11/00 3498 3544 No 4557 22 1,213 0,913 1,107 1,162 0,055 14140 3.878.273 $ Goodyear Tire & Rubber Co

General Motors Corp Goodyear Tire & Rubber CoGeneral Motors (GM) 01/28/98 30113711No 32518 49 1,540 1,281 1,973 1,774 -0,199 67000 18.138.000 $ Haldex AB

Master Sistemas Automotivos Haldex AB

Master Sistemas Automotivos

12/13/11 3714 3714 No

1230 15 0,650 1,077 0,700 0,789 0,089 2135 610.231 $ HyPower Fuel Inc

Trivest Acquisition Corp

HyPower Fuel Inc Trivest Acquisition Corp

10/12/06 3621 5099 No

36 3 0,120 1,071 0,129 0,139 0,010 75 112.500 $

Imo Industries Inc

Univel Industria E Comercio Roltra-Morse(Imo Industries)Univel Industria E Comercio 11/20/95 35683714Yes 180 13 1,430 1,050 1,502 1,805 0,304 3500 625.000 $ Inespo

Lear Seating Corp

Inespo Lear Seating Corp

06/28/95 3086 3714 Yes 9 3 1,112 1,050 1,168 1,243 0,075 2000 461.000 $ Kia Group Kia Group

Asia Motors Co Inc(Kia Group) Asia Motors do Brasil

06/25/97 3713 3711

Yes

97 67 0,873 1,022 0,892 1,002 0,110 33250 42.842.763 $ Kia Group

Undisc ian Investors Asia Motors Co Inc(Kia Group)Undisc ian Investors 04/07/97 37136799Yes 97 67 0,873 1,022 0,892 1,002 0,110 33250 42.842.763 $ Kikuchi Co Ltd Honda Motor Co Ltd Kikuchi Co Ltd Honda Motor Co Ltd 09/24/96 3714 3711 Yes 359 20 1,800 0,964 1,735 1,821 0,086 73050 1.764.871 $ Lupatech SA Neterwala Group Lupatech SA Neterwala Group 01/28/08 3491 3313 Yes 38 5 0,321 0,783 0,251 0,253 0,002 2900 50.268 $ Mahle-Stiftung GmbH Hirschvogel Umformtechnik GmbH

Mahle Metal Leve SA Hirschvogel Umformtechnik GmbH 05/21/08 ₃₅₉₂ 3462 Yes 6805 29 1,170 0,783 0,916 1,192 0,276 50496 9.915.217 $ Mando Corp Kayaba Industry Co Ltd Mando Corp Kayaba Industry Co Ltd 04/25/11 3714 3714 Yes 10678 21 0,650 1,209 0,786 1,036 0,250 3000 1.565.939 $ Marcopolo SA CONSULTREND ENTERPRISES Ltd Marcopolo SA GAZ 08/30/07 3711 3711 Yes 37 21 0,421 0,961 0,405 0,407 0,002 22000 1.121.853 $ Marcopolo SA

Kamaz Marcopolo SAKamaz 08/17/11 37113537Yes 37 21 0,421 0,961 0,405 0,407 0,002 22000 1.121.853 $

Mazda Motor Corp Sumitomo Corp

Mazda Motor Corp Sumitomo Corp 06/17/13 3711 5051 Yes 43460 42 1,200 1,063 1,276 1,277 0,001 40829 26.176.355 $ Motors Liquidation Co

Ituran Location & Control Ltd

General Motor 04/09/12 3711

3679 No

11 8 0,580 0,981 0,569 0,597 0,028 1392 170.167 $ NSK Ltd

Dow Jones & Co Inc NSK LtdDelovoj Standart 10/23/97 35622711Yes 22436 45 1,010 1,022 1,032 1,409 0,377 32948 23.431.309 $ PACCAR Inc

Porsche AG

Paccar Financial Corp Volkswagen Brasil SA(Volkswag)

11/07/95 ₆₁₄₁ 3711 Yes 1438 42 1,490 1,050 1,565 1,645 0,081 23300 18.997.000 $ Porsche AG Dana Corp Volkswagen AG Sistemas Modulares Ltda

05/11/98 3711 3714 No 10011 49 1,170 1,396 1,633 1,752 0,119 16165 20.342.508 $ Porsche AG Peccar Volkswagen AG Peccar 04/11/95 3711 3537 Yes 10011 49 1,170 1,396 1,633 1,752 0,119 16165 20.342.508 $ Porsche AG

Senna Group SA Audi AGSenna Import 10/08/99 37115511Yes 10011 49 1,170 1,396 1,633 1,752 0,119 16165 20.342.508 $ Riba Motos Industria Comercio

Vmoto Ltd

Riba Motos Industria Comercio Vmoto Ltd

04/24/13 3751 3751

Yes

21 15 0,090 1,200 0,108 0,112 0,004 2121 1.251.597 $ Rosoboronexport Republic of Rosoboronexport

Republic of 04/12/11 3795 999A Yes 16 6 1,112 1,111 1,235 1,243 0,008 1088 477.113 $ SAIC Porsche AG SAIC

Volkswagen Brasil SA(Volkswag) 04/30/97 3711 3711 Yes 1956 67 0,010 1,027 0,010 0,011 0,001 707700 92.155.141 $ SNS Co Ltd

Anhui Jianghuai Auto Grp Hldg SNS Co Ltd

Anhui Jianghuai Auto Grp Co

11/19/11 5012 3711

Yes

121 17 0,492 0,965 0,475 0,502 0,027 7900 1.879.765 $ Solvay SA

Holding Burelle Solvay Automotive IncPlastic Omnium(Burelle Hldg) 10/07/98 28213089Yes 3376 34 1,007 1,392 1,402 1,497 0,095 29400 18.102.066 $ Strattec Security Corp

Ifer Estamparia Witte-Strattec LLC Ifer Estamparia 11/30/00 3714 3465 No 272 7 1,800 0,917 1,651 1,723 0,072 3276 348.419 $ Sumitomo Electric Industries

Caixa Geral de Depositos SA

Declaration of Honor

“I do solemnly declare that I prepared this paper/thesis independently and that the thoughts taken directly or indirectly from other sources are referenced accordingly. The work has not been submitted to any other examination authority and has also not yet been published.”

Groningen, 12/06/2015