Theses supervised by Doç. Dr. Sema Dube
12 theses · Yeditepe University
Effect of macroeconomic news announcement on bond market
The primary aim of this dissertation is to explain the factors effecting the behavior of the bond markets, with a specific focus on the elements affecting the local currency-denominated government bond yields in emerging markets. This analysis involves distinguishing the impact of both domestic and international factors and explaining the connections between foreign ownership and worldwide risk appetite. The first research question of this dissertation examines how foreign participation and global risk appetite affect the yields of government bonds with long-term maturity issued in local currencies within emerging markets. We use a novel approach by utilizing the panel two threshold variable model. Our analysis of panel data shows that foreign participation in the domestic government bond market leads to a decrease long-term government yield, and there is a threshold level in foreign participation. We also look at the global risk appetite interaction with the level of foreign participation. The analysis incorporates 19 emerging market countries, contributing to an understanding of the factors driving bond yields in these economies. This study concludes at the Federal Reserve's interest rate hike in 2015. The timing is pivotal, as it marks a shift in global financial conditions, offering a natural boundary for our analysis of emerging market economies' bond yields in relation to foreign ownership and debt levels. The second research question explores the impact of macroeconomic news and risk appetite on local currency government bond yields in Türkiye, utilizing an event study with data from 2005 to 2023. We explore the influence of unexpected macroeconomic news, monetary policy announcements, and risk events on the average and conditional volatility of Turkish bond returns. The research extends beyond the reaction of bond prices to unexpected news by exploring the dynamic response of the bond market to variations in risk appetite. The study reveals significant impacts of inflation and other domestic indicators on bond returns, a pattern consistent with mature markets, while also highlighting the intricate interplay between domestic economic conditions and global shifts. By incorporating a Risk State variable derived from the VIX index and considering local geopolitical events, the research provides an understanding of the dynamic bond market responses to varying risk perceptions and news events. This study contributes to the literature by utilizing a novel model to explain the determinants of domestic bond yields. The focus on is emerging markets, which is a segment that has been relatively ignored in past research. Also, this study explores the global and local news impact of government bond markets at a country level specifically focusing on the Turkish market. This part also addresses a gap in the existing literature by examining the effects on various bond maturities, offering insights into the entire yield curve's sensitivities in Türkiye. This analysis is facilitated by utilizing a comprehensive dataset consisting of economic forecast survey data. Overall, this comprehensive approach provides insights for investors, policymakers, and academics interested in the complex dynamics of emerging market bonds and the various domestic and international factors shaping their yields. Keywords: Emerging Market, Sovereign Bond Yield, Local Currency Bond Markets, Foreign Ownership Threshold, Global Risk Aversion, News Impact
Lease-debt relationship in corporations
This study is about the relationship between lease propensity and debt in corporations. We investigated whether there is a substitutionary or complementary relationship between lease and debt. Two different samples are used such as a sample of Turkish public companies traded on Borsa Istanbul (BIST) and a sample of public companies from the BRICS countries. The BIST sample consists of 147 firms and 62 quarterly data between the years 2004Q3-2019Q4. As a result, there are 9114 firm-level data. The BRICS sample consists of annual data for 760 public firms from Brazil, Russia, India, China, and South Africa between 2009 and 2018. Totally, 7600 firm-level data. Panel tobit regression method which is proper for limited dependent variable model ran to investigate the relationship between lease and debt for both samples. In both samples, the relationship between lease and debt decisions in corporations was examined. While examining the relationship between leasing and debt, other financial and economic factors were also used as control variables for robustness. Size of the firms, the collateral amount, the liquidity, the tax rate, the profitability, retained earnings, the growth opportunities, the financial distress level, age of the company, exchange rate difference and interest rate of companies are the factors we control when examining the relationship between leasing and debt. In the second part of the study, it was checked whether there was a significant change on the lease propensity according to the sectors. In the third part of the study, it was investigated whether factors such as size of the firms, the collateral amount, the liquidity, the tax rate, the profitability, retained earnings, the growth opportunities, the financial distress level, age of the company, exchange rate difference and interest rate of companies have an effect on the leasing propensity when the firms are at the same debt level. In the fourth part of the study, it was investigated whether there is a significant change according to the sectors in which the companies with the same debt level are included. Finally, the above-mentioned analyses were repeated using the subsamples of small & medium enterprises (SMEs) and large enterprises (LEs) for both BIST and BRICS firms. In the sample of Borsa Istanbul, the relationship between lease and debt is complement. Strong and significant relationships were found between lease and collateral amount, past and current profitability, financial distress, firm age, changes in interest rates, firm size, and USD-TL exchange rate change. According to industry research, we found that in the consumer staples, consumer discretionary, information technology, real estate sectors, they resort to borrowing rather than lease. For companies with the same debt level, a significant relationship was found between the lease and firm age, the amount of collateral, the interest rate, and the effective tax rate. Moreover, the effect of the change in VAT rates between 2007 and 2018, which is a factor that was studied for the Borsa Istanbul sample and not for the Brics sample, was found to be significant and negative. For the SMEs listed on the BIST, no significant result could be reached when the relationship between the lease and the debt is examined. When the relationship between leasing and debt for LEs listed on BIST is examined, it is concluded that there is a complementary relationship. In the empirical study, we conducted with the sample of Brics countries listed firms, the relationship between lease and debt was substitute. Strong and significant results were also obtained between the other financial and economic factors that we used in order to obtain more robust results and the lease propensity. Of these, significant and strong results emerged between profitability, firm age, exchange rate and leasing. At the same debt level, significant and strong results were obtained between leasing and exchange rate difference, interest rate, and firm age. We conducted the research among SMEs and LEs to check which financial factors are changing the leasing trends of SMEs and LEs. For SMEs, no significant relationship was found between lease and debt. The results of the relationship between lease and debt for LEs differ according to debt definitions, and strong results have been achieved. In addition, for the BRICS sample, the relationship between Gross Domestic Product (GDP) as a measure of development level and lease propensity is examined. We see that as the annual gross domestic product of the countries increases, they tend to lease more. Furthermore, lease rates of companies with the same debt level increase significantly with the increase in the annual gross domestic product GDP values of the countries. Firms with the same level of debt from countries with higher levels of development tend to use more leases. To this author's knowledge, we expand prior research on debt-lease relationships by considering a sample of all BRICS countries as well as Turkish public firms, and including interaction terms in the analyses. The findings of this study would be a guide for creditors and leasing companies in terms of identifying potential customers, and for investors in terms of understanding corporate decisions. Keywords: Lease, complementary, substitutiary, BRICS, BIST, Tobit, limited dependent variable model, random effects tobit model
The effect of cryptocurrencies on diverse portfolios
We examine whether cryptocurrencies provide any benefits if they are included in the well diversified portfolios. Their performance in a diversified portfolio is compared to those of standard currency and commodities. Optimal portfolios are created based on maximum Sharpe Ratio as well as on Global Minimum variance for two-asset, for three-asset and for four-asset portfolios, and the diversification effects obtained in each case are compared with the other assets, namely USD/TRY as standard currency and gold as commodity. For the four asset case, which is the most comprehensive case, where all types of currencies along with gold are added to the market portfolio, we repeat our analyses for the period of 04/05/21 - 04/28/23, which is the "main sample" period, for he "post-Covid-19 era," which includes the sub-sample period of 04/29/22 - 04/28/23, and for the "post-Covid-19 pre-Feb 6" period, which covers 04/29/22 – 02/03/23 excluding the period of the Feb. 6, 2023 great earthquake disaster in Turkey from the post-Covid-19 period. For each optimal portfolio, various risk and return metrics are estimated and compared. Our results show that cryptocurrencies can be good instruments for portfolio diversification in general. For our sample period, Etherium seems to perform better in diversified portfolios compared to Bitcoin.
Examination of the effect of esg ratings on dividend payouts and future earnings growth relationship: The case of Turkey
This study aims to examine how Environmental, Social, and Governance (ESG) ratings influence the relationship between dividends and earnings growth in Turkey. By addressing this research question, the study seeks to enhance our understanding of how ESG ratings influence the correlation between dividends and future earnings growth, while considering potential effects of other financial variables. The final sample covers a period of 10 years (2012-2021) with 26 cross-sections. The findings of this study provide insights into the relationship between dividends, earnings growth, and ESG ratings in the Turkish context. Importantly, the study demonstrates that ESG ratings have a positive and significant impact on future earnings growth. Firms with higher ESG ratings in Turkey are more likely to experience higher future earnings growth, indicating that integrating environmental and social considerations, along with strong governance practices, can contribute to improved long-term financial performance and growth prospects. Furthermore, the study finds a negative and significant interaction between ESG ratings and dividend payout, suggesting that the combined effect of higher ESG ratings and increased dividend payouts may dampen future earnings growth. In conclusion, this study contributes to our understanding of the influence of ESG ratings on the relationship between dividends and earnings growth in Turkey. The findings highlight the importance of considering ESG factors in corporate decision-making and strategic planning to enhance future earnings growth potential. By integrating sustainable business practices, maintaining high ESG ratings, improving profitability, and investing in future assets, Turkish firms can foster long-term financial success and resilience.
Essays on the relationship between investor sentiment and stock returns
This dissertation investigates the relationship between investor sentiment, sovereign credit ratings, and sectoral index returns within G20 countries that are members of the OECD, for the period from June 2018 to April 2024. Employing a panel data analysis framework, the study explores how consumer and business confidence affect stock market performance across five major index categories: prime, energy, finance, industry, and technology. It extends previous research by integrating sovereign credit ratings alongside sentiment indices to evaluate their combined impact on the stock exchanges of G20 nations within OECD countries. Regarding sovereign credit ratings, four types of dummy variables were constructed: (i) standalone ratings, (ii) rating upgrades (iii) rating downgrades, and (iv) all rating-related changes, including outlook revisions and watchlist adjustments Although most of the existing literature typically focuses on either sovereign ratings or consumer sentiment, this research integrates both consumer and business sentiment measures alongside sovereign credit ratings to evaluate their individual, interaction, and triple interaction effects on equity returns. The analysis includes macroeconomic control variables such as CPI, exchange rates, and industrial production and incorporates the effects of COVID-19 pandemic and the Mortgage Crisis. Empirical results show that while sovereign credit ratings do not display a consistent direct impact on sectoral returns, investor sentiment, particularly business confidence plays an important and positive effect in shaping market outcomes. Consumer confidence is also found significant in the majority of cases. Moreover, sovereign credit ratings have notable interaction effects mostly observed in the energy and prime sectors through carrying sentiment effect into stock prices. Triple interaction terms involving rating, consumer and business confidence indices demonstrate a statistically significant negative effect on the energy sector. This finding indicates that the joint deterioration in sovereign credit quality, consumer sentiment, and business confidence may increase downward pressure on returns in this sector. Additionally, the study finds that among macroeconomic variables, CPI influence all sector returns whereas exchange rates influence sector-specific returns, while industrial production do not demonstrate statistical significance in general. Among the two crisis periods covered, the COVID-19 crisis dummy reveals a significant effect solely on the energy sector, highlighting its sector-specific sensitivity during the pandemic period while the mortgage crisis does not show any statistically significant effect on the performance of the analyzed indices. This thesis contributes to the literature by demonstrating a comprehensive cross-country, cross-sectoral analysis that gathers sentiment and sovereign credit risk perspectives, providing new insights into how financial markets process macroeconomic information and behavioural signals. The findings carry implications for investors, policymakers, and credit rating agencies in understanding how structural and psychological factors interact to shape equity market patterns in advanced economies.
Market sentiment and its impact on the earnings yield- dividend payout relationship: Evidence from emerging markets
In this study, we examine the effect of market sentiment on the relationship between dividend payout ratio and earnings yield using software companies operating in developing countries. Emerging markets have high volatility and different risk dynamics compared to developed markets. The study aims to understand how market sentiment impacts the dynamics between dividend payouts and earnings yields as well as other financial metrics controlling for firm return and risk metrics. We employ a panel data set consisting of 49 cross-sections between 2013 and 2022 and covering companies from BRICS countries, Brazil, China, India and South Africa, that have common stocks traded on major stock exchanges. Our data includes dividend-paying public companies in the software sector. Our results document that there is a negative relation between dividend payout ratio and future earnings yield and market sentiment decreases the strength of this negative relationship. According to our analyses the impact of market sentiment on the relationship between dividend yield and future asset growth is only marginal for the software companies in BRICS countries during our sample period. Our analyses also indicate that the presence of Covid- 19 crisis have no significant effect our results.
Intersectoral linkages among non-performing loans
Bank loans are known as the most basic element for the advancement of markets and sectors and the turning of economic wheels. The existence of bank loans in different sectors or business sectors is of high importance, and the non-payment of these loans creates a great problem for the economic functioning. This unpaid loan is defined as a non-performing loan. The existence of unpaid loans not only affects the economic functioning, but also emerges as a situation in which market reliability is damaged on a sectoral basis. Loans of banks or financial institutions play a critical role in order to keep the economic wheels turning in free market conditions in Turkey. For this reason, it is possible to easily say that loans serve as a great support at some point. However, in countries with sensitive economies such as Turkey, non-performing loans cause the functioning of the markets to deteriorate and economic crises to arise. In this sensitive market environment, the risks of the sectors in benefiting and repaying bank loans are related. Methodologically examining this relationship in the light of statistical data is the healthiest method. In this work we examine whether there is an intersectoral relationship among NPL ratios of bank in Turkey, and if there is, whether this relationship depend on the bank type. We found that there are strong correlations among the NPL ratios of various sector and bank type seems to matter in intersectoral relationships during our sample period 2007-2022.
Return and volatility spillovers among trading partners: Evindence from Turkish market
In this thesis, the dynamic structures of returns, volatility spillovers, and conditional correlations between Turkish stock and bond indices and the relevant indices of countries with which Turkey has both high and low trade relationships in terms of exports, imports, foreign direct investment inflows, and foreign direct investment outflows during the period from January 1, 2015, to December 31, 2022, are examined. The specific indices used for the study, from each selected country include the primary stock indices representing the overall market, four sectoral indices representing manufacturing, finance, technology, and telecommunications, and a 10-year bond index. The countries in the High-Trade group are Germany, UK, Italy, Netherlands, USA, and China, while the Low-Trade group includes Australia, Sweden, and Canada. Another significant aspect of this work is the examination of how the global Covid-19 pandemic has affected the variability of returns, volatility spillovers, and conditional correlations for these indices among Turkey and its trading partners. The exact date of the Covid-19 pandemic is determined by relevant structural break tests. Finally, the effects of the returns of four different exogenous variables, namely the Volatility Index (VIX), Brent Crude Oil (BRENT), Currency Exchange (USD/TRY), and Credit Default Swap (CDS), on conditional correlations are separately examined. Structural break tests (ICSS) were applied to determine the exact date of the Covid-19 pandemic. Vector Autoregressive-Asymmetric Dynamic Conditional Correlation (VAR- ADCC-GARCH) method was used to examine returns and volatility spillovers. The Generalized Autoregressive Conditional Heteroskedasticity with External Variables (GARCHX) method was applied to analyze the effect of external variables on conditional correlations. The contributions of the study are five-fold: First, the existence of connections between Turkey's financial indices and those of countries with high or low trade volumes. Second, determining the actual break dates in each Turkish financial market instead of relying on the Covid dates announced by the Turkish Ministry of Health, and subsequently identifying the effects of Covid. Third, considering sectoral indices for portfolio diversification purposes for market makers and decision-makers. Fourth, integrating 10- year bond indices, which serve as a significant indicator of countries' macroeconomic data, into the system alongside stock indices. Lastly, measuring the impact of exogenous variables on conditional correlations between countries. The results of this study may help identifying contagion effects of spikes in returns and volatility in financial markets, and, therefore, providing investors with insights to diversify their portfolios and hedge risks effectively.
A cross-market comparison of the month of the year anomaly for bitcoin
According to the Efficient Markets Theory, asset prices reflect all information regarding the assets, and there should be no consistent patterns in returns based on calendar time or any other market differences. We examine if there are calendar time market anomalies in the crypto currency markets, and whether these anomalies, if any exists, vary across different markets based on the location. Specifically, we focus on Bitcoin, the largest crypto currency market, and search for the evidence regarding the month of the year effect anomaly in the Bitcoin markets during the period of 2015-2018. To see if market characteristics change based on the location, we compare the presence and the characteristics of the month of the year anomaly among three different Bitcoin exchanges: Bitfinex, Bitstamp and Okcoin. We find a strong month of the year effect anomaly in the Bitcoin markets during the period of our study. Although the general results are similar, we find some differences with respect to the month of the year effect anomaly among the three Bitcoin exchanges considered.
A novel approach for optimal portfolio allocation: Feasible market factor estimation
This study gives the answer to the factor selection problem existing within the factor based covariance estimation framework. First, we illustrate the fact that the sample covariance estimation of the large scaled covariance matrices deteriorates the performance of the optimal asset allocation compared to naive diversification. Employing factors instead of assets into covariance estimation plays heroic role to lower estimation error and keep optimal asset allocation strategies alive for both asset management industry and the academic universe. However, choosing the optimal number of factors to reduce the estimation error to the rock bottom while not losing required information to estimate covariance profile of the asset universe, is still either sample size or time period dependent. To eliminate factor selection problem, we proposes a novel factor-based covariance estimation method, "Feasible Market Factor (FMF)" estimation, which generates a single artificial factor representing the sum of the individual variances of the observed factor universe. Under various constraint setting we compare the performance of the proposed approach to PCA, single and multi-factor and sample covariance estimation methods in terms of portfolio's performance ratios. Empirical results illustrate that FMF outperforms other factor based and sample covariance estimation methods in terms of Sharpe and Sortino ratios especially for the portfolios minimizing either variance or semi-variance. To also improve performance of the optimal portfolios maximizing return with given volatility, we implement the Autoregressive Markov Regime Switching model within FMF framework and enhance the performance of Maximum Sharpe Ratio (MSR) portfolio compared to CAPM.
Essays in financial economics, timing ability of mutual fund managers
This dissertation explores three different issues regarding mutual funds, consisting of three empirical studies. The first study investigates the evidence for market timing in the Turkish Real Estate Investment Trusts from 2004 to 2014. To this end, twenty-five Real Estate Investment Trusts in Borsa Istanbul are analyzed in three sub-sample periods around the Global Financial Crisis in 2007: Pre Crisis (2004-2006); Crisis (2007-2010) and Post-Crisis (2011-2014). As our findings indicate, there is no evidence that the Turkish Real Estate Investment Trusts have employed effective market timing. The second study examines the market timing abilities of Turkish variable-fund managers during 2011-2016 in a panel data framework, using interaction variables to assess the impact of global emerging-markets portfolio returns, country openness, technology advancement, economic growth, and derivative market size on these. For emerging markets, factors that impact the growth and development of the mutual fund industry may also be relevant. Our results show strong evidence for market timing ability, which increases with economic growth, country openness, global emerging-market portfolio returns, to a smaller extent for derivative market size, and decreases with technology advancement. The third study examines the liquidity timing ability of Turkish variable fund managers during 2011-2018, and how this ability is affected by the environmental factors such as the technological advancement level, the presence of derivatives market, the growth in the overall economy, the level of market openness, and the performance of portfolios for foreign exchange, gold, bond, real estate, and emerging markets. We use interaction variables within a panel data framework. We find strong evidence of liquidity timing ability of mutual fund managers even after controlling for environmental factors. The nature of the interactions for most of the control factors with liquidity timing ability is strongly significant and differs based on the factor.
Does market perception influence the relationship between dividends&earnings yield? Evidence from developed economies
In this study, we examine whether market sentiment alters the relationship between dividend payout and earnings yield in software and computer services firms across G7 economies during the 2013–2022 period. Our approach combines classical dividend theories with recent insights from behavioral finance, and applies a panel regression framework using the OECD Composite Leading Indicator (CLI) as a measure of market perception. We test whether sentiment strengthens or weakens the link between dividends and forward earnings yield, and whether this potential interaction changes when firms' growth characteristics are taken into account. Our findings show that the dividend payout ratio does not provide significant explanatory power for forward earnings yield, while sentiment maintains a consistently positive association with valuation indicators. Although the sentiment–dividend interaction remains insignificant for earnings yield, it becomes statistically relevant when growth-oriented valuation measures are used as dependent variables. This result suggests that dividends matter more through their contribution to perceived growth prospects than through direct earnings-based valuation channels in sentiment-driven market conditions. By focusing on a single technology-intensive industry in developed markets, we contribute to the literature by explaining why the importance of dividends may vary across sectors, time periods, and sentiment measures. The study also provides evidence on how investor perception influences payout interpretation and highlights the role of market sentiment in shaping valuation dynamics in technologically dynamic firms. Keywords: Dividend policy, earnings yield, investor sentiment, behavioral finance, OECD CLI, market perception, G7 software companies.