State ownership in Southeast Asian multinational enterprises : do governments hamper innovation performance and growth opportunities?

(1)

State ownership in Southeast Asian multinational enterprises:

Do governments hamper innovation performance and growth

opportunities?

University of Amsterdam – Amsterdam Business School Faculty of Economics and Business

MSc Business Administration Track: International Management Master Thesis Final – 22.01.2018

Markus Sommer

Student number: 11449624

E-Mail: m.sommer1991@gmail.com Supervisor: Dr. Mashiho Mihalache

(2)

Statement of Originality

This document is written by Markus Sommer who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Although much research on state ownership has been done, none focuses on Southeast Asian economies (ASEAN), but rather on developed markets and Brazil, Russia, India and China as emerging countries. We contribute to existing literature as ASEAN benefited from stronger growth rates compared to their emerging market competitors over the past decade.

In this study, we try to provide evidence for a negative relationship between state ownership in ASEAN multinational enterprises and their financial performance. The analysis tries to answer if inward industry foreign direct investments and patent granting activity, an indicator for innovation performance of a country, moderate the relationship in any way. Our sample consists of financial information compiled from 2011 to 2016 for 160 publicly traded companies from Indonesia, Malaysia, Philippines, Singapore, Thailand and Vietnam. Our results show that there is no support for the main relationship or the moderation of it. However, certain direct effects on financial performance in our sample are identified. A surprising finding of our study is the direct negative impact of innovation activity per country on financial performance of all firms in scope. This observation is not in line with existing theory, so we conclude that the deviation might occur because of our specific sample size or innovation activity of a country is not a proper reflection of the one companies engage in. Last, the analysis has managerial and political implications. A stricter focus on innovation in ASEAN companies is essential because of a direct relationship between patent activity and financial performance. In addition, governments should promote research and development funding, which drive innovation and in the end financial growth.

(5)

1. Introduction

Emerging market1multinational enterprises (EMNEs) range among the top performing companies worldwide, from Alibaba, the Chinese e-commerce giant, to Embraer, the Brazilian aircraft manufacturer. How do these enterprises become global and domestic players in diverse industries? Certainly, innovation is a key factor in explaining this development, however, EMNEs cannot simply try to be alike western companies, or easily adopt their technologies. They are founded in environments with a scarcity of resources, human capital, know-how and often in instable political environments. In this sense, EMNEs are forced to act and operate differently than their competitors from developed countries. The economic and management literature has documented key determinants of innovation and their influence on financial- as well as innovation-performance (Wharton 2015).

From business groups, ownership concentration, research and development (R&D) to foreign direct investments (FDI), much research results in different arguments about the driving forces of innovation performance. Business groups are prevalent in multiple emerging markets and play an important role in the business environments (Fisman & Khanna 2004). The chaebol in South Korea, keiretsu in Japan, grupos in Brazil and qiye jituan in China are just some examples of business group affiliations (Khanna & Yafeh 2009; Khanna & Rivkin 2001). These groups are tight together in formal and informal structures within EMNEs. From supplier, customer and intra-firm networks to family owned MNEs, these groups have a considerable influence on EMNE’s strategies and thus financial performance and the way the companies internationalize (Khanna 2000; Chittoor 2009). Finally, also ownership structures are main drivers, most often large block shareholders, who have substantial power in managerial management of EMNEs. State owned enterprises (SOEs) resemble such government majority shareholder concentration in emerging economies. A higher degree of state ownership might be advantageously in order to acquire equity capital and specific

(6)

know-how easily. However, there is a tipping point when too much government influence results in inflexibilities and loss of innovative edge for EMNEs (Belloc 2010; Chang, Chung & Mahmood 2006). Technological development and innovation capabilities of firms are determined by patents. Once EMNEs’ innovation activity slows down, it results in lower frequencies of patent granting and thus lower financial performance. Patents are an intermediate result of R&D expenditures, a determinant of innovation performance in EMNEs and explain share price movements (SPM) (Yu & Hong 2016). Which specific role patents play in the relationship between ownership concentration and financial performance will be analyzed in this paper.

The limitation of existing research is their mutual focus on only developed markets or emerging economies like Brazil, Russia, India and China (BRIC). We find that there is a research gap because only little scholars target Southeast Asian economies in their papers. In our research we group Southeast Asian economies as Association of Southeast Asian Nations (ASEAN) (ASEAN.org 2017). ASEAN have been growth engines of Asia over the past five decades with 6% annual average growth and will sustain the gross domestic product (GDP) growth for the next five years, outperforming the BRICs. Productivity drivers on a firm level in ASEAN companies have been identified as ownership concentration, foreign company collaboration and firm size, which can further be increased with improvements in innovation and know-how (OECD 2017).

For these reasons, the thesis tries to shed light on different influencing factors on innovation and specifically financial performance in Indonesia, Malaysia, Philippines, Singapore, Thailand and Vietnam. Other ASEAN markets are economically too small to be significant in an international market environment. Singapore as unique example of the ASEAN economy is included in the study as it is representative for positive state intervention in domestic MNEs. Its economy over the past decade has been driven by inward FDI and might be seen as an outlier within ASEAN, but is as important for the analyses as the other markets in scope

(7)

of our study. The significant increase of ASEAN MNEs and their innovative products, including services, leave room to analyze key determinants of innovation capabilities and performance. It can be assumed that the BRICs, non-ASEAN emerging-economies and transitioning markets are representative for the literature review in regard to innovation, R&D, performance and firm characteristics. For these reasons, when mentioning EMNEs, emerging economies and BRICs, the related theories and observations are assumed to be mostly related to the ASEAN context.

The theoretical framework used in this study will help answering the following two research questions: How does ASEAN MNE state ownership influence companies’ financial performance? Does domestic innovation performance and industry FDI moderate the previously mentioned relationship positively?

We hope to present with our research valuable insights into ASEAN MNEs’ ownership structures and innovation activities with possible implications for financial performance. Next to those, we contribute to existing business and economic literature by extending the scope of EMNEs to an emerging economies’ high growth market environment. The outlined developments show the unique opportunities ASEAN companies have. This paper develops many valuable insights for similar growing emerging economies, which need to adapt firm structures to reach higher growth rates and innovative capabilities. From a managerial perspective, managers can use the insights to structure their ownership concentration differently to find optimal corporate performance and promote innovation performance within their organization.

In the Literature Review, existing research and theories on emerging markets, ASEAN, innovation, business groups, ownership structures, financial performance and R&D aspects are outlined. Afterwards we present our Theoretical Framework, followed by Methodology and Data Analysis. We conclude our research paper with a Discussion about the results and a final Conclusion section.

(8)

2. Literature Review

2.1 Emerging Markets

Over the past decades, EMNEs have experienced tremendous growth, while many shifted their focus from supplying developed market MNEs to innovation activities for domestic and international markets. There are many definitions of emerging markets and it is arguable which ones are better or worse, however, we use the most widely accepted one: Emerging economies are characterized by rapid economic growth, governments actively promoting such activity with economic liberalization and the acceptation of a free market system. The latter two are the main growth engines of emerging economies. Furthermore emerging markets are undergoing critical changes across industries, political environments and are mostly shaped by weak legal systems (Arnold & Quelch 1998; Hoskisson, Eden, Lau & Wright 2000; Luo & Tung 2007). Such markets present the opportunity for companies to become international enterprises. EMNEs are international operating firms, actively engaging in outward FDI and originated from emerging economies. These enterprises are deploying value-adding activities and are trying to gain control in single or multiple countries (Luo & Tung 2007).

Growth of EMNEs is closely related to FDI, which over a period of time lead to outward FDI. The latter globally increased from US$50 billion in 1980 to over US$1.4 trillion in 2007 (+2700%). BRICs have mostly been involved in outward FDI, representing about 30% of all emerging economies’, totaling US$600 billion in 2014 alone (UNCTAD WIR 2016). More interesting for this research paper, however, are developing countries in ASEAN. Although most FDI are in East Asia and BRICs, ASEAN is gaining more importance in the global investment environment. Economic growth rates are not comparable to China’s, and GDP are mostly lower than other emerging markets, but a supportive institutional environment in terms of doing business are significantly better in ASEAN in comparison to China or Russia for example (Ponlapat 2014). This shows the unique opportunity for ASEAN

(9)

and foreign companies to grow in those supportive environments. In order to do so, some EMNEs have profited from inward internationalization on their home market. Global enterprises from developed countries engage in acquisitions or joint ventures with EMNEs, eventually transferring knowledge, technological-, logistical-, and human resource management-skills to them. The transfers enable EMNEs to innovate, overcome their latecomer disadvantage and find suitable outward FDI opportunities (Pan & Tse 2000). Outward FDI by EMNEs is seen as a strategy to transfer subsidiaries’ capabilities back to home markets, and being able to compete against highly competitive MNEs from developed economies on their own territory (Anderson, Sever & Sutherland 2013). Luo and Tong (2007) describe this specific behavior of EMNEs as the “springboard theory of internationalization”, in which companies use outward FDI as a springboard to overcome disadvantages in the global market.

2.2 ASEAN

A great amount of Asia’s economic growth can be attributed to previously mentioned ASEAN, with its founding members Indonesia, Malaysia, Philippines, Singapore and Thailand growing on average by 6% over the past decade (Worldbank 2013). The Asian Financial Crisis in 1997, triggered by heavy speculative trading in exchange rates, interrupted economic growth and created immense volatility in the financial markets all over Asia for the following two years. It shows that there is a need for a better financial system and inter-regional economic activity to create sustainable growth and security within ASEAN (Aziz & Sundarasen 2015). In the next five years, GDP growth rates among the top five ASEAN countries, Indonesia, Malaysia, Philippines, Thailand and Vietnam, are projected to be approximately 6%, with Cambodia, Laos PDR and Myanmar even exceeding 7%, annually. The latter three so called frontier markets are subject to analysis in future research. Singapore is a special case in the ASEAN context, with impressive growth rates of 15% in 2010 and

(10)

averaging about 8% annually from 1998 until 2009. It shows the great inflow of FDI and economic development Singapore experienced years before the five other economies excelled. FDI in Singapore were predominantly in financial and construction sectors. Today, the country resembles those immense investment rates being one of the major financial hubs in the Asia pacific region. Over the last five years, Singapore’s GDP growth declined to 5% annually. It is projected that the city-state will grow over the next five years by 5% annually. We decided to include the economy into our focus countries as it plays an important role in the development of ASEAN in the global business field and also in regard to state owned MNEs (ASEANstats 2017, Worldbank 2017).

China on the other hand realizes slowing annual growth with a 6% average over the next five years. The economic outlooks and history put a heavy weight on the importance of ASEAN in the global economy. This dependence is researched in industry analyses of small and medium enterprises (SMEs), being the backbone of the ASEAN economy. Between 88% and 99% are considered SMEs in ASEAN, and show that the competitiveness of the region is depending on the performance of its SMEs. In general, productivity slowed down in multiple emerging countries in Asia. Firm level analyses on Vietnamese manufacturing industry for example show that productivity is affected by firm size, ownership concentration and collaborations with foreign companies. In Indonesia for instance large firms are more productive than smaller ones. Productivity in ASEAN MNEs can be increased with innovation, adaption of new technologies or knowledge, but also with improvement of existing skills, technology and know-how capabilities. These are all firm-level productivity growth drivers identified in ASEAN (OECD 2017). The history and future of emerging markets, specifically ASEAN, show that they are heavily dependent on innovation and improvement of existing technologies to compete in the competitive global market environment.

(11)

2.3 Innovation

Several scholars have researched innovation factors in emerging markets. One of the main underlying theories is Kremer’s (1993) O-ring theory of economic development. Developing countries’ firms have different productivity levels because of their failure to acquire best practices. Innovation can be viewed as firms’ ability to acquire or create new technologies, adapt them and introduce novel marketable products. Companies in emerging markets are sometimes distant to technological innovations and thus adopt different ways of production processes, helping them to market products in a more efficient way (Ayyagari, Demirgüç-Kunt & Maksimovic 2011). “New-to-firm” innovation, the set up of new products or processes, novel marketing approaches or work-locations, are an advanced definition of innovation when considering the aspect of imitation in emerging economies (Organization for Economic Co-Operation and Development (OECD) 2005).

Nelson and Winter (1982) argue that as part of the evolutionary economics view, companies unconsciously create new knowledge leading to new business capabilities while continuing the innovation process. The new resources gained from this innovation process help firms to reach exceptional performance in highly competitive markets.

Lewin and Massini (2003) present determinants of firm’s innovation, which are derived from internal R&D and imitation practice of other successful firms. They conclude that there is a positive relationship between firm size and R&D activities. However, with growing firm size R&D productivity declines, resulting in lower innovation. Thus, smaller firms tend to be more innovative, as they are able to transform processes and products in fast and efficient ways, leading to superior performance. A different viewpoint on innovation present Ayyagari et al. (2011); innovative firms in emerging markets are younger but larger in size than its competitors, and mostly export oriented with private ownership. According to the resource-based view of the firm, company’s resources, financial capital, information, human resources and equipment, are sustaining drivers of innovation (Barney 1991). These firm specific

(12)

characteristics lead to the conclusion that large-scale enterprises (LSEs) should be more affine to innovation than their smaller competitors.

In this regard, especially foreign competitors in the product market are associated with superior company innovation. Considering the largest publicly traded companies, foreign competition also influences SMEs, which have to innovate constantly in order to survive in the dynamic environment with highly concentrated competition (Knight & Cavusgil 2004). Existing literature reveals that ASEAN companies raised their knowledge and innovative capabilities to compete in an international environment. If the increase in innovation performance is significantly higher in SMEs or LSEs is open to discussion, but obviously there are determinant factors influencing innovation and performance of EMNEs.

2.4 Business Groups

Business groups are one of the many influencing factors of EMNEs. Scholars of international business (IB) find and confirm the positive influence of business group affiliation on innovation performance in transition economies (Chang et al. 2006; Choi, Lee & Williams 2009). For this research paper a business group is defined as “a group of companies, more than 30% of whose shares are owned by some individuals or by companies controlled by those individuals.” (Chang & Hong 2000, pp.437-438). Business groups in most cases are taking direct ownership in EMNEs and depending on the structure might help explaining determinants of innovation performance. SOEs are categorized by a government’s shareholder ownership of 30% or more in our research.

In their paper, Mahmood and Mitchell (2004) find that the mentioned relationship is further deepened in emerging economies, because business group constitute a trade off for firms. Either business groups are able to set ground for important infrastructure needed to enable innovation at times of weak market institutions, or these business groups are converging regarding technologies, thus limiting “possible adaptations and mutations possible in an

(13)

industry” (Mahmood et al. 2004, p.1361). Independent firms are main creators of ideas needed for innovation to foster. An increase of the group dominance will lead to a non-monotonic impact on innovation, because first, the infrastructure development leads to higher innovation but second, independent firms experience high entry barriers to these markets which means inhibiting innovation (Mahmood et al. 2004).

Another important driver of innovation performance are inter-firm relationships with foreign firms. These relationships and close ties to customers and suppliers enable EMNEs to gain new know-how and insights, which companies can leverage to increase innovation performance significantly (Prahalad & Ramaswamy 2000; Skaggs & Youndt 2004). Companies gain access to external knowledge mostly up and down the supply chain as companies share same knowledge, competencies and are on a frequent knowledge exchange basis (Lasagni 2012). A distinction between SMEs and LSEs is that SMEs have to rely on external sources of knowledge for innovation (Zhou & Li 2012). These differences show the importance of inter-firm relationships and supply chains of EMNEs. Lasagni (2012) also noted that innovation performance is significantly higher among SMEs that are proactively engaging in relationships with customers, suppliers and end users. So external knowledge, infrastructure investments and inter-firm relationships can be derived from different ownership structures, which might play a vital vole in innovation performance of EMNEs. The mentioned findings show that business groups in their different dimensions influence innovation performance in emerging economies.

2.5 Domestic- and Foreign-Ownership

In the context of emerging markets, with higher density of SOEs, and its weak market infrastructures, governments very often serve as replacements for those market failure. Governments are shaping EMNEs’ strategic decisions in internationalizing and controlling opportunities of EMNEs. So non-SOEs have to “pursue unconventional non-market strategies

(14)

to exploit government-related advantages and compensate for their weaknesses” (Hong, Wang & Kafouros 2015, pp. 45-46). In SOEs, the government as main shareholder can influence the EMNEs’ strategies directly to reach certain economic, political or social goals. A critical point is that state ownership and institutional pressures, which are not always congruent with a profit-maximizing scope of a company, lead to economic irrational decision making for EMNEs, pushing them towards not optimal decision making (Hong et al. 2015; Zhang, Zhou & Ebbers 2011).

There are three key reasons why governments in emerging economies are much more influential in steering EMNEs compared to developed market governments. (1) The degree of state ownership among emerging markets is high. (2) Emerging economies resemble younger markets compared to developed economies, thus state involvement is higher and more influential resulting in increases of coercive and regulatory pressures (Child & Rodrigues 2005; Hoskisson, Eden, Lau & Wright 2000). (3) There are limited resources, capabilities and experience of EMNEs, in those firms state-influence and state-ownership are then driving forces in the way they grow (Peng, Wang & Jian 2008; Hong et al. 2015).

Many scholars researched ownership concentration and structure of EMNEs as influencing factors on innovation performance. According to Financial Times (2017), ownership concentration is defined as the stock in a company owned by individual investors or large block shareholders. The latter need to have five percent equity capital in the company to qualify as a block shareholder. In most publicly traded firms the large block shareholders are institutional investors like pensions funds or mutual funds. In this context, it is noteworthy that the power of such investors can lead to aggressive actions on strategy, board members, or CEO (Financial Times 2017).

Concentrated ownership results in higher performance because of an efficient structure, solving occurring agency issues (Claessens & Djankov 1999). In addition, a high degree of concentrated ownership means the possibility of effective monitoring, leading to better

(15)

innovation activity (Belloc 2010). In China, for example, large block-shareholders, being part of firm’s ownership structures, have a positive influence on company’s innovation performance (Chang et al. 2006). SOEs are characterized by large block-shareholders, mostly the government as main influencer, shaping the nature of the SOE. Latter are prevalent in most emerging economies and fight with a trade off between performance and governmental involvement. On one hand, the state can have a positive effect on company’s performance in emerging- as well as developed-economies (Kole & Mulherin 1997; Sun, Tong & Tong 2002). In China, evidence shows that 100% government owned SOEs are not a preferable option, but no government ownership is also not beneficial. This results in an inverted U-shape relationship between ownership concentration and firm performance. So the optimal degree of ownership is somewhere in between (Sun et al. 2002). On the other hand, however, it is questionable if SOEs are able to innovate in an efficient way, as government’s policy decisions are not solely based on profit maximization ideals. Furthermore governments have a negative influence on firm’s performance because of a lack of managerial know-how (Child 1994; Child & Lu 1996; Choi, Lee & Williams 2009). Another aspect is that SOEs lack market experience, are inefficient and are associated with lacking competitive advantages. In order to create and sustain a competitive edge, SOEs need to upgrade processes and facilities constantly with innovation (Li, Liu & Ren 2007; Dougherty & Bowman 1995; Lu & Lazonick 2001). According to Roland and Sekkat (2000), government involvement in SOEs is sometimes constructive for the speed of diffusion of innovation. A comparison of non-independent SOEs in China to more non-independent SOEs with less interference of government entities shows a higher flexibility and daily decision making management for SOEs. A possible downside of the involvement by states might also be the resource constraints of re-investments into certain R&D business areas, which are essential to innovate and compete on a national and international scale (Boardman & Vining 1989; Li et al. 2007).

(16)

large block shareholder, which as part of the ownership structure play a significant role for strategic and financial decisions. In this regard, one can distinguish between equity owned foreign ownership and relationship based ownership, both resulting in knowledge spillovers and resource acquisitions beyond just financial contributions (Child 1994; Douma et al. 2006). For this study, however, equity ownership is of greater importance, because first it is easier to compile data for and second it is a good quantitative measurement to analyze relationships and influences of innovation performance, FDI and ownership concentration on financial performance. Do foreign or domestic owners change the influence on certain firm characteristics?

In transition economies for example, domestically owned companies are less innovative than foreign owned ones (Falk 2008). Cheng et al. (2006) find three key reasons for this positive relationship between foreign ownership and innovation. First, foreign companies want to invest into the emerging market, which means investors have an ambition to equip domestic firms with innovation capabilities from their home country. Second, foreign owners encourage innovation in the domestic company, resulting in an increase of R&D expenditure. Third, the foreign owners “force” investments into technological developments, which eventually lead to higher innovation performance. Choi et al. (2011) and Douma et al. (2006) support the previously mentioned research on positive relationship between foreign ownership of domestic firms and innovation performance with their academic papers.

Taking the findings on ownership structures in MNEs together we can say that foreign and state ownership show significant influence on the financial performance of companies. Nevertheless, R&D and innovation activity are also influential attributes for financial stock moves of a company.

(17)

2.6 R&D Innovation & Financial Performance

R&D is not just a critical factor for a country’s technology and science development, but also for the economy. It leads to higher employment, technological development and higher productivity gains (Sikka 1998). A lot of research has been conducted on Chinese MNEs’ R&D activities. One can distinguish between inter-, intra-industry and in-house R&D. Inter-industry R&D spillovers are found to be more important than R&D spillovers within the same industry (Bernstein 1988, Steurs 1995). Inter-industry R&D spillover effects are of particular importance for Chinese LSEs, especially the ones operating in manufacturing business. In-house R&D and inter-industry R&D spillover result in a complementary relationship; first shaping the innovation environment, and second influencing factor productivity. In-house R&D is related to foreign technology transfer, and serves complimentary to labor productivity (Bin 2008). Inter-industry R&D spillovers in emerging countries like China are a better determinant for industry-level innovation than in developed countries, because of the heavy dependence on imitation rather than innovation in China (Bin 2008).

Specific industry sectors also portray unique characteristics in explaining innovation performance of EMNEs. A higher concentration of R&D performers can be found in capital-intensive industries and along firms with large volumes like pharmaceutical or equipment and machine manufacturing. Most R&D performers are concentrated among SOEs and the least performing ones among foreign and overseas enterprises (Jefferson, Huamao, Xiaojing & Xiaoyun 2006). Again, one has to say that the interesting results of existing literature are limited to BRICs and transitioning economies, excluding ASEAN, thus generalizations can only be made with caution. Moreover, many scholars are providing different solutions on explaining how R&D, technology and innovation are influencing each other. Aghion, Blundell, Griffith, Howitt and Prantl (2009) argue that innovation in sectors, which are further distant from the technology frontier, may be slowed down by a high degree of

(18)

competition, while the opposite will profit from it. On the other hand other academics like Vives (2004) argue that competition slows down R&D, which results in slower innovation processes.

Wang and Kafouros (2009) find that industrial R&D effects are significantly higher compared to those in developing markets. There is a shift over time in emerging economies from an imitation approach to innovation, R&D capabilities and technological efforts (Kafouros 2008). Another key finding is the higher significant impact of R&D on innovation performance than FDI. This means that emerging market firms still have to undergo R&D even if certain technologies and know-how are sourced from foreign companies, specifically the ones from developed economies. In addition, determinants of innovation experience resemble stronger effects in industries with high-technological future opportunities and industries involving less foreign presence (Wang et al. 2009). Nevertheless, FDI is an important driver of innovation performance in emerging economies as past empirical research shows (Liu & Buck 2007).

Ren, Eisingerich and Tsai (2014) analyze the search scope effect along supply chains of firms on R&D and innovation performance of EMNEs. They find that there is a negative influence of search scope on the relationship between R&D investments and innovation performance. So the question is if R&D expenditures play a positive or negative role in the corporate environment of EMNEs, or is there another indicator for measuring innovation performance in companies?

R&D investments are firms’ long-term obligations, which eventually result in patent granting. Latter are viewed as a measure of firm’s technological activity and thus innovation performance. The number of patents granted can be related to innovation performance of a country or firm (Huang, Sung, Wang & Chen 2013). In addition, patents can be seen as “intermediate outcome of R&D expenditure” (Yu & Hong 2016, p.198), and contributing to corporate performance (Hsu & Ziedonis 2013). Patents also give information about

(19)

technological developments and innovation of a firm. Exploitation and exploration are two effects on firm’s performance resulting from innovation activities (Wang & Li 2008). Patents are complimentary to R&D expenditures, but are more significant in explaining SPM than R&D investments (Yu & Hong 2016). For example, pyramidal ownership structures in the biotech industry show that there are different impacts of R&D investments, firm’s innovation activities, exploration and exploitation on SPM and thus corporate performance in different industries (Gavious, Hirsh & Kaufman 2015; Yu & Hong 2016). In the light of previous research, our study focuses on patent activity in the ASEAN region as substitute for R&D expenditures, because of better availability and indicator for innovation of a country.

The previously outlined literature contributes to the phenomenon of innovation performance among EMNEs in BRICs and transition economies. However, little attention has been paid to other emerging economies such as ASEAN. As outlined, the focus on ASEAN economies is specifically interesting because of the immense growth opportunities for companies within the next five years in those markets. Although a lot of research discusses determinants of EMNEs’ innovation performance, none so far analyzes the moderating effect of industry FDI and innovation performance on the relationship between state ownership in EMNEs and their financial performance.For this reason, our research analyzes the influence of state ownership on firm performance. In a second step we analyze how domestic innovation performance, measured as patents granted per year, and industry FDI, moderate the previously mentioned relationship. It is assumed that ownership concentration is positively influencing EMNE’s financial performance and innovation performance also affects the relationship positively. This assumption is based on the previously outlined literature, which found positive relationships in different studies involving the mentioned variables.

(20)

3. Theoretical Framework

The following sections are the key building blocks for the theoretical framework, used to analyze the relationship between ownership, financial performance, innovation performance and FDI. It is illustrated in Figure 1 at the end of this section.

3.1 Ownership concentration and financial performance

As outlined in the previous section, business groups, ownership, R&D and FDI influence the performance of EMNEs. State ownership has certain positive effects on the performance. However, examples from China show that a majority of government controlled EMNEs experience negative performances. We know from the Literature Review that there is a U-shape relationship between financial growth and SOEs. In this regard, a certain degree of ownership is desirable, but when the state takes too much control, financial performance and innovation can be influenced negatively. This is partially because of little managerial knowhow of the state as controlling party (Sun et al. 2002; Choi et al. 2009). ASEAN firms can take advantage of this research and segregate themselves from the governments to become more independent SOEs with less state interference, or even none. This step creates more flexibility, reinvestment possibilities for innovation activity, which are needed to compete on a global scale and finally creating long-term competitive advantages (Boardman et al. 1989; Li et al. 2007).

A transition towards more independent SOEs and privatized EMNEs is supported by SOE reforms in ASEAN. These reforms, targeting ASEAN and pacific island states, try to move away from 100% SOEs to hybrid forms, a commercialization of SOEs with a full or partial privatization. A commercialization leads to higher efficiencies, decreased costs and improved outcomes regarding productivity and resource allocation for the economy. In addition, commercial governance structures result in higher transparency and accountability of

(21)

managers, creating an incentive structure to maximize value for stakeholders and shareholders. Furthermore, a privatization, fully or partially, is not a necessary result of commercialization, but ensures that SOEs do not fall back into their old structures. For these reasons, SOE reforms in ASEAN help increase efficiency and innovation capabilities (ADB 2008).

The relationship between ownership type and firm performance have been researched in many developed economies. There is an observed negative effect of ownership type on the financial performance when SOEs are engaging in other aims but profit maximization (Bozec, Breton & Côte 2002). Another issue with state intervention and firm performance is corruption. Many emerging economies deal with corruption, as Brazil for example. Petrobas, a large Brazilian SOE from the oil sector, had to list its shares in the 1990s on the New York Stock exchange. Afterwards the EMNE used to progress political investments, engage with political allies and directly control gasoline prices. These actions had negative effects on the firm’s profitability (Lazzarini & Musacchio 2015). This is also likely to be the case for SOEs in ASEAN economies where governments work closely with companies, trying to reach political aims with SOEs and do not base their decisions on profit- and shareholder-value-maximizing ideals. An example from Thailand shows that the five largest publicly listed companies are 100% owned by the state and in additional twelve EMNEs the state is a majority shareholder. Aligning with existing research, most of these EMNEs run at loss and experience non-sustainable business performances. In addition, the Thai government allocates 3-4% of its annual budget to the funding of SOEs (OECD 2015). The findings clearly demonstrate that SOEs are not used to increase financial performance but rather as a political steering wheel for power and influence.

In contrast, in Central European- and East Asian-economies large block ownership has positive effects on financial performance, because of better monitoring capabilities within the MNEs (Hu & Izumida 2008). If these large block shareholders are governments, foreign

(22)

investors or holding companies is not clear within the study. However, when large block shareholders are foreign investors or MNEs, foreign owned companies are more innovative than their domestic peers (Falk 2008). The foreign owned companies have an ambition to invest in emerging economies, thus the encouragement to invest heavily into innovation capabilities, know-how and increase their R&D expenditure, with a high probability resulting in higher performance (Cheng et al. 2006).

As outlined, there are positive and negative aspects to state ownership in EMNEs and its affiliated firm performance, however, the majority of cases from China and developed countries outline rather negative relationship between state ownership and financial performance. How SOEs perform in ASEAN has not been researched so far and is the main research question the framework analyzes. We cannot predict that ASEAN SOEs show the same impact on performance as BRICs or other economies. This can be the case because of different geographical and political environments of the ASEAN economies compared to BRICs. However, I predict that non-SOEs, also referred to as privately owned enterprises (POEs), experience higher financial returns than SOEs in ASEAN countries. This brings us to the first hypothesis of the theoretical framework:

Hypothesis 1: State ownership in ASEAN EMNEs has a negative effect on firm’s financial

performance compared to non-SOEs.

3.2 Moderating effect of industry FDI

FDI in general is one of the most significant drivers of globalization and economic growth. ASEAN economies are heavily dependent on international trade as domestic markets are mostly too small. In addition to attract FDI, macroeconomic performance as well as institutional factors are of importance. In most ASEAN countries there is general low governance effectiveness, poor regulatory quality, and a large presence of poor institutions,

(23)

not being able to create good governance at all. So a key determinant for higher FDI is the improvement of overall institutional environments among ASEAN economies (Buracom 2014).

Malaysia for example, experienced high FDI growth up to mid 2000s, followed by continuing decrease eventually reaching surpassing of outflowing FDI compared to inward FDI. The other high growth ASEAN nations including Indonesia, Singapore, Thailand and Vietnam surpassed Malaysia’s FDI inflows by the end of 2010 (Athukorala & Waglé 2011). Singapore had particularly high FDI inflows into its financial sector. These examples show that certain Southeast Asian economies are more attractive to FDI than others.

Empirical results show positive effects of inward FDI on the productivity of local firms, but a negative one on the productivity of foreign firms (Aitken & Harrison 1999; Haskel, Pereira & Slaughter 2007; Keller & Yeaple 2009). Nevertheless, an increase in productivity is not necessarily an indicator for innovation or yielding in innovation (García, Jin & Salomon 2013). If local firms are at disadvantage within their home market, they can use FDI and knowledge from foreign firms and reverse-engineer products or services in order to gain technological insights and remain competitive (Salomon 2006).

A higher degree of competition within certain market segments and industries results in shifts of local MNEs into less-profitable and less-innovative segments. So an increase in FDI may evolve into exits of local firms (Garcia et al. 2013). FDI also impacts the innovation output of domestic firms negatively (Cheung & Lin 2004). In addition, foreign owned firms might shift innovation activity back to their home country, away from the host country and its affiliates. These findings strengthen the importance of FDI in industries with indirect effects on innovation and firm performance. Garcia et al. (2013) find a negative relationship between industry- and firm-level FDI and the innovation activities of local companies. This means that FDI slows down domestic innovation activity. Nonetheless, inward FDI on industry- and firm-level helps increasing the efficiency of domestic EMNEs with a combination of

(24)

knowledge and competition spillovers. In the context of FDI into ASEAN it is difficult to predict the effects of industry FDI on the relationship between SOEs and financial performance. But as efficiency and productivity of local firms increases with higher FDI we hypothesize the following:

Hypothesis 2: All else equal, industry FDI positively moderates the relationship between SOE and financial performance in ASEAN.

3.3 Moderating effect of domestic innovation performance

Schumpeter (1934) already argued that “creative destruction” and “technological innovation” are the basis for firm’s competitive advantage. So MNEs’ success is not solely based on market capitalization or specific industry structures, but rather on the ability to create new technologies, which are key to influence competition and the external environment. In addition, innovation is a main driver of economic growth, thus countries that can create and sustain new technologies coupled with an innovative environment will grow faster than its competitors. So technological capabilities, innovation culture, innovation strategy and formal structures affect innovation performance positively. In this context, especially technological capabilities show a significant impact on innovation performance in Turkish or Chinese MNEs. The same observation holds for EMNEs trying to gain a competitive edge (Kamasak 2015).

Moreover, scholars identified the need for long-term investment decisions in R&D and innovation sectors in order to reach high returns. Such long-term investments are most often undertaken by large block shareholders represented by the state or foreign investors. In comparison to managers, who have short-term views on strategies as they are evaluated and reimbursed based on quarterly financial results, block shareholders have an interest in long-term profit maximization. These shareholders support R&D activities that are important

(25)

indicators for innovation activity of an EMNE (Choi, Lee & Williams 2011; Zhang, Li, Hitt & Cui 2007; Wright, Ferris, Sarin & Awasthi 1996).

SOEs in transitional economies engage in innovation because of a lack of resources, which is backed by financial capital and infrastructure by the government (Li, Meng, Wang & Zhou 2013). In this regard, state ownership allows companies to receive resources that are invested in R&D and innovation (Zhou, Gao & Zhao 2016). POEs, on the other hand, with limited resources, barriers and uncertainties in the market environment, face difficulties to undergo efficient innovation (Zeng, Xie & Tam 2010). For these reasons, POEs compared to SOEs emphasize their strategic horizon on short-term risk-free investments. This can be explained by a shortage of human- and capital-resources, limited and poor distribution channels, and weak marketing networks (Li & Xia 2008). Especially foreign ownership of domestic firms and innovation performance show a positive relationship (Choi et al. 2011; Douma et al. 2006). These insights show the importance of innovation activity in EMNEs and firm performances, creating a need to analyze the moderating role of innovation activity on the main relationship in ASEANs’ MNEs.

When patents, indicator for innovation performance in a country, result in positive financial performance of EMNEs, it is interesting to analyze how patents on a domestic level affect the relationship in Hypothesis 1. So the outlined effects of innovation activity on financial performance and SOEs, which have less resource constraints, are assumed to result in higher innovative activity. The following hypothesis combines this statement and tries to answer it in the analysis in the next section.

Hypothesis 3: All else equal, a higher domestic innovation performance positively moderates the relationship between SOE and its financial performance.

(26)

Figure 1: Theoretical Framework – ASEAN EMNEs – Own Illustration (2017)

The outlined theoretical framework in Figure 1 shows the relationship between state ownership and financial firm performance. Innovation performance and industry FDI are moderating the previously mentioned relationship. Although industry FDI and innovation activity can add value to the question if state ownership influences financial performance in ASEAN, both indicators have not been used as moderators in previous academic research. This is especially true for the ASEAN region, which has largely been neglected in international management research so far. It can be assumed that the generalization of the results will only hold for the ASEAN members, as government structures and regulatory environments differ significantly between emerging economies. In addition, EMNEs do not operate in one single industry. They have multiple industry pillars, which might result in different observations for different industry segments. So the data selection has to focus on the appropriate selection of SOEs and non-SOEs in order to mitigate the risk of skewed results. The Methodology section will outline the sample, data collection and variables, followed by the Data Analysis and Discussion parts.

SOE Ownership SOE financial

performance in US$

Number of patents / country per year H1 (-)

US$ FDI / industry per year

H3 (+) H2 (+)

(27)

4. Methodology

This study uses a quantitative research approach with an underlying collected dataset for statistical analysis. We use an experimental design study that tries to create causal relationships between the variables and increases internal validity. As the dataset is stretched over a six year time period from 2011 until 2016, a longitudinal design is used, trying to capture the relationship over time. In addition, all information collected is secondary data, accessible through statistical yearbooks, databases and company websites.

The next section discusses the sample in detail, followed by the data collection steps. Dependent-, independent-, control- and moderating-variables are presented in the last part of Methodology.

4.1 Sample and data collection

We focus in our data sample on the ASEAN economies of Indonesia, Malaysia, Philippines, Singapore, Thailand and Vietnam. As mentioned in the Literature Review, the countries are chosen because of strong similarities and economic growth over the past decade. Smaller economies like Cambodia or Myanmar are neglected because of a lack of financial information availability and greater economic differences to the countries of focus. The explanatory and longitudinal study helps to create causality between the variables and researches the relationship of SOEs, respectively non-SOEs, and financial performance over time.

All company data is collected from domestic stock exchange databases and financial service providers like Bloomberg. Time series, stock prices, are directly sourced from Compustat – Capital IQ – Security Daily. In regard to the moderating variables, the number of patents granted per year, are collected from the World Intellectual Property Organization (WIPO).

(28)

Inward industry FDI data is collected from domestic statistical yearbooks and statistics office databases when accessible.

Considering the domestic stock exchange indices of the target countries, the largest 25-30 EMNEs are collected based on their index weighting, measured as market capitalization in billion US-Dollars. Our final population consists of 160 companies. The EMNEs represent a diverse dataset in terms of sector allocation, with eleven different industry sectors from accommodation, mining to wholesale and retail companies (Appendix 1). Out of those 160 companies, 47 are SOEs and the other 113 are non-SOEs or controlled by foreign conglomerates. A larger number of SOEs in the data set is preferable, however, the certain countries like the Philippines do not have as many publicly traded SOEs as Singapore or Thailand for example. Nevertheless, the data set shows a diverse and large number of company observations.

Choosing the time period from 2011 until 2016 for our data observations we create a better predictive model and can reach economic significance (Britten-Jones, Neuberger & Nolte 2011). Moreover, by applying a six-year time frame we mitigate the risk of extreme outperforming or underperforming EMNEs, which can be the case for a single time period observation (de Jong & van Houten 2014). In addition, we choose this specific period because of two reasons. First, most historical company information is not easily accessible for a time frame of one decade or more. Second, a time period after the financial crisis of 2008 neutralizes the risk of large outliers and changes in financial performance because of the financial crash and global recession.

The control variables used in the analysis help eliminating alternative explanations of the relationship and causality between the dependent and independent variable. Firm size measured by number of employees, and firm age with the company’s founding year as reference point, are controlling the main relationship. Our third control variable is the annual

(29)

GDP growth rate per country, which serves as economic indicator. All data has been collected from the previously mentioned financial service databases.

4.2 Measures

Corporate performance of EMNEs with a focus on Indonesia, Malaysia, Philippines, Thailand, Singapore and Vietnam is measured as yearly excess returns. This is discussed in detail in the next sections. The moderators of the framework are innovation performance and industry FDI. Latter is measured in billion US Dollars per industry per year of each country. We analyze the degree of innovation performance as annual number of patents granted per country. Unfortunately, R&D expenditure has not been used as innovation performance measure due to time constraints and availability of financial information.

4.3 Dependent Variable – Financial Performance

There are many different performance indicators measuring MNEs’ financial performance, however, excess stock returns as risk-adjusted measure for financial performance is favored by academic research in the field of finance and thus used for our models in this paper (Brown & Warner 1985). Measuring the EMNE’s annual return compared to a risk free investment, most often a US treasury bill, results in the excess return (Oswald & Tahera 1991). Therefore, excess returns can be used as performance measurement. We integrate the 3-months US treasury bill as risk free investment. The difference between the MNEs’ yearly stock performance and the risk free investment results in the following formula:

𝑟𝑥_!!" = ln 𝑃! − ln 𝑃!!! − ln 𝑅𝑓!

𝑃_! describes the aggregated company’s stock index price at time t ,(𝑅𝑓_!) represents the return rate of the 3-months US treasury bill, in this analysis used as risk free investment rate

(30)

for the calculation of the excess returns. In addition, a natural logarithm of the change in stock- and risk free investment-prices is applied. This transformation is necessary as most relationship in economics and business are not linear, so applying this transformation is accepted in econometric analysis (Wang, Buckley, Clegg & Kafouros 2007). Furthermore, natural logarithmic changes in stock return prices create a more stable and normalized model, with returns that can be compared easily while pure prices cannot.

4.4 Independent Variable - SOE

The independent variable of the framework is SOE, respectively POE or foreign owned EMNEs, which are grouped as either SOE or non-SOE in our analysis. We follow Chang and Hongs’ (2000) suggestion and decide that a minimum of 30% shareholder ownership in any given EMNE will label the company as SOE. Foreign conglomerates control some non-SOEs with shareholder ownership of more than 30%, but we focus on SOEs in this study and neglect this observation. The independent variable is coded as categorical variable, also known as dummy variable, with “1” for SOE and “0” for non-SOE.

4.5 Moderator

The first of the two moderators is innovation performance of ASEAN countries. As mentioned earlier, R&D investments resemble firm’s long-term obligations, which eventually result in patents. The technological activity and thus innovation performance can also be measured in number of patents granted (Huang et al. 2013). Yu and Hong (2016) find the significance of patents in explaining share price movements (SPM), the firm’s financial performance. So a higher patent granting activity per year should play a role in explaining the main relationship and excess returns. The collected patent data from WIPO covers the time frame from 2011 until 2016. In our population, the origin country of patent application serves as reference point for our data collection. So the data collected only covers the patents

(31)

applied, and granted for, in the country of focus. We only include domestic annual patent data, as detailed industry level data is not easily accessible through WIPO or statistics offices. The second moderator of the framework is industry inward FDI. As outlined in the theoretical framework section, inward FDI is a critical factor in firm growth and explaining SPM. All data for FDI inflows have been collected through statistical yearbooks and databases of domestic ASEAN statistics offices from 2011 until 2016. In order to have a congruent FDI population for the countries, the industries have been categorized according to the ASEAN statistics database. Next to this, we converted all investment data into US Dollar according to the respective US Dollar exchange rate of the years 2011 until 2016. Mentioned earlier, the industry segments are comprised of 11 categories according to the ASEAN statistics office (aseanstats 2017; see Appendix I).

4.6 Control Variables

We are controlling for three variables, firm age, firm size and GDP growth, when analyzing the main relationship and the moderating effects in our regressions.

Many scholars use size as controlling variable, and as size of the firm can influence the performance, it is important to use as controlling variable for the analysis. Similar to Yiu, Lau and Bruton’s (2007) approach, we take the average number of employees per firm from 2011 until 2016 and then apply a logarithm transformation. We apply the logarithm (log), as it allows us to easier interpret differences in a proportional way (Gelman & Hill 2007). In addition, the standardized logarithmic transformation of the variables mitigates risks of running into heteroscedasticity, extreme outliers and autocorrelation.

Firm age, with the founding year as reference point, is also an important indicator for firm performance. Furthermore, a common conception is that older companies have more experience, are considered to have first mover advantages and thus experience the benefits of a longer learning curve. On the contrary, there are arguments that older MNEs are not as

(32)

flexible as younger ones, which directly influences the competitive advantages of a firm (Douma et al. 2006). For these reasons, firm age plays a key role and needs to be included as controlling variable in the analyses.

The third control variable is the annual GDP growth per observed country. GDP growth is an indicator for economic growth and thus also for innovation performance in economies. Galindo and Mendez (2014) found this relationship with the help of an Schumpeterian model. Economic growth, innovation and entrepreneurship are in a circular effect to each other so all three dimensions have a positive effect on each other.

5. Data Analysis

The independent variable, ownership, is measured dichotomously with “1” for SOE and “0” for non-SOE, while the dependent variable, excess returns, is numerical and continuous. For this reason a simple linear regression model is used with two moderators to test for moderation of the main relationship. As the observation is in a panel data set up, STATA statistics software is used to analyze the regressions. The following equation shows the first regression model and also the main relationship of the analysis between ownership (𝑜𝑤𝑛𝑒𝑟𝑠ℎ𝑖𝑝_!!2_{) and excess returns (𝑟𝑥}

!!). Firm size (𝑠𝑖𝑧𝑒!!), firm age (𝑎𝑔𝑒!!) and GDP growth

rate (𝑔𝑑𝑝_!!_{) are the control variables for all regressions in the three models. The subscript t}

represents the time and is equal to the years 2011 to 2016, superscript i stands for the company, respectively country for GDP growth and patents, and in for FDI for the specific industry the company does business in.

H1: 𝑟𝑥_!! _{= 𝛼 + 𝛽}

! 𝑜𝑤𝑛𝑒𝑟𝑠ℎ𝑖𝑝!!+ 𝛽! 𝑠𝑖𝑧𝑒!! + 𝛽! 𝑎𝑔𝑒!! + 𝛽! 𝑔𝑑𝑝!! + 𝜀!

(33)

In order to analyze the moderating effects of industry FDI and patents, interaction terms are introduced to the equation. The following two equations incorporate the moderating effects with the interaction terms of FDI and patents.

H2: 𝑟𝑥_!! _{= 𝛼 + 𝛽} ! 𝑜𝑤𝑛𝑒𝑟𝑠ℎ𝑖𝑝!! + 𝛽! 𝑠𝑖𝑧𝑒!!+ 𝛽! 𝑎𝑔𝑒!!+ 𝛽! 𝑔𝑑𝑝!! + 𝑓𝑑𝑖!! + 𝛽_! 𝑜𝑤𝑛𝑒r𝑠ℎ𝑖𝑝_!! _{∗ 𝑓𝑑𝑖} !! + 𝜀! H3: 𝑟𝑥_!! _{= 𝛼 + 𝛽} ! 𝑜𝑤𝑛𝑒𝑟sℎ𝑖𝑝!! + 𝛽! 𝑠𝑖𝑧𝑒!!+ 𝛽! 𝑎𝑔𝑒!!+ 𝛽! 𝑔𝑑𝑝!! + 𝑝𝑎𝑡𝑒𝑛𝑡!! + 𝛽! 𝑜𝑤𝑛𝑒𝑟𝑠ℎ𝑖𝑝!!∗ 𝑝𝑎𝑡𝑒𝑛𝑡!!+ 𝜀!

Industry FDI (𝑓𝑑𝑖_!!_{) and patents granted per country (𝑝𝑎𝑡𝑒𝑛𝑡}

!!) are used as moderators in the

analysis. By creating interaction terms with the independent variable, the regressions analyze if there are significant effects and thus moderating the relationship positively or negatively, as mentioned in the Hypothesis 2 and 3.

All variables are analyzed in preliminary checks for normal distribution, frequency and correlation. Normal distribution is assessed with skewness parameters, histograms and outlier correction. The control variables firm size and firm age show moderate negative skewness in the descriptive variable analysis. Therefore, they are transformed accordingly, to fit the underlying assumptions of a linear regression model with normal distributed variables. Next to the log-transformation of the control variables we also log-transformed the amount of FDI and number of patents. As we already used a transformation on excess returns and thus is normal distributed, we do not have to make further changes to the variable. After the transformation the observations lie within a skewness range of -1 to 1, but most often for the years 2011 to 2016 within the boundaries of -0.5 to 0.5. In addition to skewness, histograms

(34)

show bell shaped distributions with insignificant deviations from the mean. These observations support the normal distributions of the variables used for the following analysis. In a second step, frequency checks are applied to examine extreme observations in the variables used for the regression. With the use of z-scores of the continuous variables, only excess returns demonstrate some extreme outliers with values larger than |3|. These observation points are viewed isolated in a boxplot. After isolated observation, the outliers are replaced with the respective mean score of each annual excess return. Analyzing each year as a separate group and comparing it to the other observations, reports that each group is normally distributed with skewness between -1 and 1. It is a normal observation that means of the same variable vary between the different groups, sorted by year. However, as the total population is viewed in the descriptive statistics the outliers are replaced with the mean of the overall analysis. In theory, a mean substitution might create an artificial deflation of variation, but the sample size is not large enough with a total of roughly 885 observations, so deleting the outliers would lead to a shift in means. Moreover, neglecting the outliers would result in loss of important company observations, so we keep them in favor of the analysis. Table 1 represents the descriptive statistics of dependent-, moderating-, and control-variables after transforming them to fit our assumptions mentioned before.

Table 1: Descriptive Statistics Independent Variable and Moderating Variables VARIABLE MEAN STD.

DEVIATION MIN MAX OBSERVATION EXCESS RETURN 0.274 0.820 -2.629 10.184 897 FDI 3.148 1.100 0.039 5.708 959 PATENTS 3.701 1.787 3.229 4.036 960 FIRM AGE 1.276 0.135 1.000 1.640 952 FIRM SIZE 0.363 0.129 0.000 0.710 954 GDP RATE 1.011 0.008 1.000 1.030 960

(35)

5.1 Correlation Testing

The descriptive analyses, including correlation matrices, frequency checks and outlier correction are analyzed and done in SPSS. In order to create a best fitted regression model with non-biased outputs, the dependent-, independent-, control-, and moderating-variables should show little correlation. The Pearson correlation coefficients are presented in the correlation matrix below, including the number of observations per group, mean and standard deviation.

Table 2: Correlation Matrix including Mean and Standard Deviation

The correlation matrix shows only high significant Pearson values for SOE with GDP rate (0.149**), FDI (0.219**) and patents (0.240**). Similar is true for GDP rate with FDI and patents (0.164**). Last, we observe that firm age and firm size are significant higher correlated (0.269**). These positive correlation results show that there is some linear relationship between the observed variables. In our case FDI is positive significant correlated with all variables besides excess returns. In this case, patents and FDI correlation (0.267**), means that a higher FDI activity would also lead to higher patent activity in the country. Similar can be inferred from company size and company age, while an older firm might also

Variables Mean Std. Deviation N 1. 2. 3. 4. 5. 6. 7. 1. Excess Return 0.136 0.682 895 1 -0.033 -0.002 0.004 -0.047 -0.027 -0.008 2. SOE 0.290 0.456 960 1 -0.084** 0.011 0.149** 0.219** 0.240** 3. Firm Age 1.276 0.135 952 1 0.269** -0.083* 0.112** -.090** 4. Firm Size 0.363 0.129 954 1 -0.029 0.182** -0.026 5 GDP Rate 1.011 0.008 960 1 0.164** 0.164** 6. FDI 3.148 1.100 959 1 0.267** 7. Patents 3.701 0.179 960 1

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).

(36)

have higher employee numbers than younger firms. Countries’ GDP growth rates are positively correlated with FDI and patents. This is in line with theory as higher GDP rates are indicators for an increase in economic activity, which leads to higher innovation performance and activity (Galindo and Mendez 2014).

Nevertheless, the correlation coefficients do not show significant high correlations, only between -0.090 and 0.269. In this context, it is difficult to reach any degree of causality, and the relationships have to be further analyzed with the help of the regression models. However, for our following regression analyses it is desirable to work with low correlated variables in order to mitigate risks of skewed results.

5.2 Regression Analysis & Results

Ordinary least square regressions with a panel data set up in SPSS can result in skewed data analysis, because the program views all observations in as a single group, thus creating non-heterogeneity. So we use STATA for all our regression models.

Before running the regressions, preliminary tests are conducted to detect which linear panel data regression model is used. The Hausman test is used to analyze if random or fixed effects are used. However, because the independent variable is a dummy variable with 1 and 0, a random effects model has to be applied. Categorical variables are omitted because they do not provide estimates for time invariant covariates when using a fixed effect model. The random effect model used in this regression analyzes the marginal effect of the variables used after accounting for effects of ownership (SOE or non-SOE). In addition, the Breusch-Pagan test is run to find out if a random effects model is favored over a pooled regression one. The test returned a significant Chi2 (p<0.05) so that we choose the random effects model. Last, we know that STATA reports omitted variables if there is reason for multicollinearity. As none were reported we can say that there is no multicollinearity within our model and between the variables (Rodríguez 2017, Princeton).

State ownership in Southeast Asian multinational enterprises : do governments hamper innovation performance and growth opportunities?