Interview with Armagan Gezici on Investment under Financial Liberalization in Turkey

July 2008

Armağan Gezici completed her Ph.D. dissertation, “Investment under financial liberalization: Channels of liquidity and uncertainty” in 2007. She is currently an Assistant Professor of Economics at Keene State College in New Hampshire. In this interview, current Ph.D. student Ann Werboff questions Armağan on her thesis topic.

What was the path that led you to pursuing your Ph.D. at the University of Massachusetts?

I got interested in economics during my undergraduate studies in mechanical engineering in Turkey. Growing up in a country that went through a military coup followed by structural adjustment and liberalization programs, I had developed an awareness of the connections between social problems and economic policies. Eventually I switched to economics. One my professors at Middle East Technical University in Ankara suggested UMass among a few other Ph.D. programs in the U.S. as a place to further study issues of policy and development. I was quite impressed by the wide range of theoretical approaches pursued by the faculty here.

How did you get interested in the issues of financial liberalization and investment? How did you choose your particular thesis topic?

Following the implementation of trade and financial liberalization programs in Turkey in the 1980s, the 1990s turned out to be the decade of financial crises and instability for the Turkish economy. The economy was exposed to the whims of international capital flows, trapped in boom-bust cycles of growth with high real interest rates and overvalued currency. In my research, I focused on the impact of liberalization on investment, because these liberalization programs were expected to boost productive private investment and promote growth in the long run.

My interest in the performance of productive industries and their investment approaches goes back to an internship I had with large manufacturing companies in Turkey during my engineering studies. Three summers spent on the shop floors of steel, machinery and textiles factories gave me the opportunity to observe the particular problems these companies face as well as their role in the development process. Before coming to UMass, I did research for my master’s thesis in Turkey, analyzing the connections between trade liberalization, technological change and manufacturing employment. So when I came to UMass, I had this background and interest in the manufacturing industry’s performance.

How did this range of experience and interests come together in your dissertation?

I explored the effects of financial liberalization on the investment decisions of manufacturing firms in Turkey, focusing mainly on two channels: credit availability and volatility. While the former channel is analyzed in the literature on financing constraints, the latter is left out of the studies investigating investment behavior under liberalization. Financial liberalization is expected to eliminate imperfections in the capital markets and provide firms with better access to credit. Under this scenario, the positive role of liquidity in the determination of investment is expected to be reduced with liberalization. I argue that such change in the importance of liquidity may not take place because of increased volatility in the economy. In addition to the lack of support for better credit access argument, I also argue that the negative impact of volatility on investment is further deepened under liberalization.

In my dissertation, I developed a firm-level investment model which contains output, liquidity and volatility as the determinants of investment. Econometric tests and survey results provided evidence for a negative relationship between firm-level investment and volatility variables and a positive relationship with liquidity and sales variables in the benchmark model. The negative impact of volatility on investment is worsened under financial liberalization due to nonlinearities in the investment-uncertainty relationship as suggested by the Post Keynesian theory. There is no evidence for a declining importance of liquidity. I also identified different elasticities across the different firm categories of size, export orientation and maturity. Overall, my results suggest that financial reform policies did not lead to the expected benefits for the investment of real sector firms while actually producing increased uncertainty that impedes investment further.

That firm-level data set was a unique element of your research. Would you describe the set and how you compiled the data?

The data set used in this study is a firm-level unbalanced panel data set constructed from the balance sheets and income statements of 165 manufacturing firms in Turkey for the period 1985-2003. The compilation of the whole data set took approximately 5 months. Given that my interest was in firm-level investment performance under financial liberalization, which was initiated in the mid-1980s, I could not use the widely used Worldscope database, since the Worldscope provides observations on only 23 firms in Turkey, beginning from the early 1990’s. I collected firm-level information from the publications of the Capital Markets Board of Turkey (CMB) for the years 1985-1988 and the Istanbul Stock Exchange (ISE) for 1989-2003. Because of the dynamic structure of investment models, I included only those firms with at least five years of continuous coverage after the initial year. I included only manufacturing firms, because physical capital accumulation was the focus of my study and because financial and insurance firms are subject to different accounting standards. Moreover, only firms with less than 50% government ownership were included in the sample because “State Owned Enterprises” have been subject to different managerial procedures than privately owned ones.

You utilized both field surveys and econometric analysis in your dissertation. How did each of these inform your research?

Doing field work is quite unusual for the type of research question that I was trying to answer. The main reason for including field work was the lack of firm-level investment literature in Turkey. I needed to develop a benchmark model of investment of Turkish manufacturing firms and there were not many studies that I could use as a guide for the model specification, mainly due to the lack of a data set. So instead of selecting among a mélange of possible empirical specifications to get econometrically significant results, I decided to interview key decision makers in firms about issues related to investment and uncertainty. The field research is qualitative in nature and complementary to the econometric analysis I adopted. Since I used a novel firm-level data set in my econometric analysis, the results of this qualitative field research were crucial for explaining and interpreting the findings of my quantitative analysis. I can certainly say that the fieldwork provided me with a more realistic framework that helped ground my work. I learned a lot about those manufacturing firms and how they perform in a changing business environment. It also helped me choose among competing theories of investment. Econometric analysis on the other hand, allowed to me to connect the firm performance to macro changes and trace connections by introducing specific measures of liberalization into an investment equation.

You utilize a Keynesian/Post Keynesian approach. Did you enter into your research having already decided that this was the theory you wanted to pursue or did your theoretical analysis come about as a result of your findings?

I was quite aware of, but certainly not invested in, Post Keynesian theory when I began my dissertation work. I chose Post Keynesian theory over the others possibilities simply because the field work supported a Keynesian/Post Keynesian framework.

Do you have any plans for future research related to your dissertation?

Theoretically, I want to further explore the issue of nonlinearities in the negative relationship between investment and volatility and look into the relevance of risk aversive behavior to explain this phenomenon. An inverted U shape, suggested in the literature for this relationship, might not be sufficient or relevant to the experiences of developing countries. Currently I am working on measuring this nonlinearity in an econometric framework. With this in mind I also want to extend my framework into other developing countries struck by crises, such as Mexico and Argentina. Another dimension of my future research will involve exploring the differences across various types of firms in their responses to macroeconomic conditions. Debt composition and export orientation are among the most important and relevant categories in the experience of developing countries.