Theses supervised by Prof. Dr. Kamil Yılmaz
11 theses · Koç University
Interstate labor market connectedness in the United States
This paper investigates the labor market connectedness across the US states. We apply the Diebold-Yilmaz Connectedness Index (DYCI) methodology to weekly initial unemployment claims to obtain connectedness measures across 51 states between 1989 to 2021. The system-wide connectedness increases during the US recessions since 1990. We also show that changes in the monetary policy stance affects the regional unemployment levels and their connectedness across states. A significant section of the thesis is devoted to analyzing the state-level initial unemployment claims and their connectedness across states during the COVID-19 recession. We show that in addition to the number of new COVID-19 cases, state-level government policy stringency and community mobility patterns play a critical role in explaining the unemployment dynamics during the COVID-19 recession. The same factors defined in relative terms over the pairs of states as well as distance and the value of shipments across states also explain the transmission of unemployment shocks across states during the COVID-19 recession.
Return connectedness across commodity futures
We obtain the connectedness measures in commodity futures by applying the Diebold-Yilmaz connectedness index methodology. We use an extensive data set of 126 commodity futures from July 1997 to January 2019. In the full-sample analysis, our findings show that crude oil, heating oil, copper, soybean have the highest to-connectedness; these are the commodities contributing more to the variance of other ones. Also, the contribution of futures at different maturities to the connectedness at the commodity level varies significantly; we find no significant maturity effect for gold and silver. In the dynamic analysis, we see that connectedness reaches its peak during the Great Recession. To further characterize the dynamic-connectedness, we decompose the connectedness index into within and cross-commodity components. Cross commodity connectedness dominates the within commodity connectedness after 2004, the year after which a significant amount of investment started flowing into commodity index trading. Finally, by focusing on the dynamic behavior of commodity connectedness over time, we show that the global business cycle and the U.S. Dollar index explain the substantial share of the variation in the connectedness and demand for commodities Granger-cause the total return connectedness of commodity futures.
Şirket satın alma ve birleşmelerinin hedef şirketlerin hisse fiyatlarına etkisi:Türkiye örneği
The objective of this thesis is to analyze the impact of the merger and acquisition announcements of the Turkish companies from August 1999 to June 2012 emphasizing Turkish shareholders' wealth while focusing on the Istanbul Stock Exchange (ISE), and to show whether the Turkish target firms generated unexpected, i.e. abnormal, returns, during this period. The expected stock return values are estimated by using the Capital Asset Pricing Model (CAPM) while the performances of target firms are compared with ISE-All and ISE Sectoral Indexes. According to the performance of ISE-All, target firms stocks responded negatively after the announcement. Negative performances reached their peak at a week, while the highest losses occurred at a month. It is not possible; however, to make a general comment about the performance of the stocks compared to their sectoral indices. The impact of foreign acquisitions is sharper compared to the impact of domestic acquisitions, but the underperformance of target firms? stocks is higher when acquired by local firms. The most significant response in transactions occurred when a small percentage of shares were acquired. Although abnormal returns are not statistically significantly different from zero, at weekly and monthly event periods these stocks underperform their previous period performances. The target firms with small trading volume carry negative returns over a longer period, and their underperformance is higher. These findings show the existence of insider trading in the ISE index, leading to information being leaked on the market before any official announcement.Keywords: Mergers and Acquisitions, Shareholder Wealth Effects, Domestic and Foreign acquisitions, Istanbul Stock Exchange, Sectoral Indices, Event Study, Informed Trading
Dinamik faktör modelleri ve finansal bağlanmışlık: Dünyadaki ulusal bankacılık sistemleri üzerine bir uygulama
This paper analyzes the volatility connectedness of major banking systems around the world. Incorporating dynamic factor models into Diebold-Yılmaz connectedness framework (Diebold and Yilmaz, 2011), we calculate volatility connectedness measures for countries rather than individual banks. In each country, bank stock return volatilities are assumed to be driven by a common country factor and an idiosyncratic component unique to each bank. The common country factor is calculated as the first principal component of the bank stock return volatilities. Constructing a VAR model of the country volatility factors, we undertake the variance decomposition analysis of volatility shocks to obtain the Diebold-Yılmaz connectedness measures. We obtain both static and dynamic measures of connectedness. In the static analysis we show that the volatility connectedness of countries are closely linked to major banking system characteristics. First, banking systems located in the same region tend to be more connected with each other in volatility than with those outside the region. Second, country banking systems tend to generate net volatility connectedness towards others as their size and financial development levels rise. In the dynamic rolling window analysis, we obtain important results related to both systemic volatility and volatility transmitted by individual countries. In terms of systemic risk, we manage to capture the main stages of the crisis by our total connectedness index. We find that the US was the main generator of volatility to other countries from the onset of the US financial crisis until the end of 2008. However, once the crisis became global and was followed by the sovereign debt/banking crisis in the Eurozone periphery, the "to-connectedness" of the European banking system, and especially the ones in the southern periphery, increased substantially. Keywords: Financial connectedness, dynamic factor models, risk measurement,systemic risk, systemically important financial institutions, vector autoregression,variance decomposition
Finansal krizler ve Avrupa bankalarının bağlanmışlığı
The literature on financial connectedness has expended significantly since the last financial crisis. While there is a voluminous literature on financial linkages, mostly due to lack of empirical work these studies fail to provide conclusive policy recommendation. Complementary to this literature, my study focuses on all aspects of the linkages among the large banks of the EU member countries. My thesis identifies how financial shocks propagate during crisis times and which banks become main transmitters of these shocks. In empirical analysis, I use Diebold and Yilmaz connectedness measurement to calculate the reciprocal effects of each banks. The data set includes daily stock return volatilities for 45 European banks over the period 1998-2014. The resulting connectedness matrix serves as my main data in full sample and dynamic analysis. Keywords: Connectedness, Eurozone Crisis, European Banks, Diebold and Yilmaz Connectedness Measurement
Kısa vadeli faizlerin bağlanmışlığını anlamak
I analyze the connectedness of short-term interest rates. In particular, I use the Diebold and Yilmaz methodology to study the behavior of financial connectedness across the 3-month interbank offered rates for 33 countries over the 1991-2014 period. I find that short-term interest rates have become highly interconnected during the financial crises. More importantly, my results reveal that policy interest rate decisions by the central banks have a crucial influence on the connectedness of short-term interest rates. As a next step, I apply variance decomposition to the financial network framework to clearly observe the linkages between the short-term interest rates. The total connectedness index emphasizes both the minor and major crisis episodes around the world. The empirical results show that before the introduction of the single European currency, Euro, in 1999, the bulk of the shock transmission took place among the EU member countries. Once Euro was introduced, the connectedness of the US and the EU short-term interest rates happened to be the most significant. After establishing these general results, I study the directional connectedness across countries during the most important crisis episodes. Keywords: Financial connectedness, risk measurement, systemic risk, Libor rate, vector autoregression, variance decomposition
Küresel kamu kredi riski bağlanmışlığının tahmini
This paper applies the Diebold-Yilmaz connectedness index methodology on sovereign credit default swaps (SCDS) to estimate the network structure of the sovereign credit default risks. In particular, using the elastic net estimation method, we separately estimate networks of daily SCDS returns and return volatilities for 38 countries between 2009 and 2014. Our results reveal striking differences between the network structures of SCDS returns and return volatilities. In the SCDS spread networks, emerging market and developed countries stand apart in two big clusters; major emerging market countries being the main determinants of spreads in the network. In the case of the SCDS volatility networks, however, we observe regional clusters among emerging market countries along with the developed-country cluster.
Sigortacılık endüstrisinde oynaklık bağlanmışlığı ve sistemik risk
This thesis studies the systemic importance of the insurance industry in terms of volatility connectedness. We first analyze the volatility connectedness between the banking and insurance industry in the US and then we repeat the same analysis for the insurance companies from 26 countries. In both cases, we obtain both the static and dynamic total connectedness measures. Next, using pairwise directional connectedness measures, we analyze the revealed network structure. First, we display the financial networks, before and after some important systemic events, to understand the relative position of the insurance companies. Second, we use a community detection algorithm based on random walks to see how individual companies creates subgroups within the network. Additionally, we also report results showing how the contribution of the insurance industry to the systemic risk changes over time in the US. In our first result, the analysis of the US banking and insurance industry reveals that the total risk in the US financial system reaches to rather high levels once the insurance companies are included in the analysis. Second, we show that the insurance industry has become more connected globally after the global financial crisis. This is an important result because unlike the banking sector, insurance companies have no bilateral lending practices among each other. In the global insurance industry analysis, we show that the insurance companies are clustered on a geographical basis. In case of the analysis of US banks and insurers, we find that the so-called systemically important financial institutions tend to be clustered on the basis of their size and sector. On the other hand, unlike the large banks and insurers, small-sized banks and insurers tend to fall in the same community. Keywords: Financial connectedness, risk measurement, systemic risk, systemically important financial institutions, vector autoregression, variance decomposition, nonparametric estimation, lasso, adaptive elastic net, networks, communities
Varlıklar, komşu riskleri ve banka ağları
This thesis consists of two essays discussing shock propagation and counterparty risk in financial networks. The first essay explores the resilience of financial networks to systemic shocks under regulatory solvency constraints. We generalize the contagion under fire sale model of Cifuentes, Ferucci and Shin (2005) by allowing financial institutions to be connected through assets they hold in their portfolios. We simulate the model under different combinations of debt and asset networks and observe how shocks spread across markets. In the second essay, we provide a dynamic model of financial contagion to identify the source of systemic risk when banks can borrow from each other as well as from external creditors to invest in a risky portfolio. Our framework differs from earlier work as it describes how a bank's value function depends on counterparties' risky behavior. We analyze the implications of the model in the case of a ring network of banks where liabilities of a bank are held by a single counterparty. We show that the network effect is positive for the banks whose risky investment is less than the average of the rest. In other words, we show that the counterparties' risky behavior increases the probability of default of a bank. We also show that the uniform Value at Risk (VaR) constraint doesn't reflect the real probability of default when the network effect is considered. Therefore, we propose a policy function which assigns different VaR values for each bank in the network, and obtain that the target level imposed by the social planner is achieved.
Tüketim eylem halinde: Avrupa Birliği'nde enflasyon yayılmaları
Economists and policymakers focus on aggregate inflation measures, such as headline and core inflation. In this thesis, we argue that the inflation spillovers across consumer good sub-categories are as important as their direct contributions to the aggregate inflation measures. We identify and characterize the spillovers among 12 major consumption groups by applying Diebold-Yilmaz connectedness index methodology. We utilize the disaggregated Harmonized Index of Consumer Prices for the European Union member countries, 1996-2017. We estimate the within-country inflation connectedness across 12 groups statically using the data for the full-sample, and dynamically with the rolling-window estimation of 60 months. We observe resembling full-sample networks among 12 groups, which indicates similar inflationary connectedness due to similarities in consumption patterns. The countries experienced different inflationary behavior mostly due to national policy shocks. The rolling-window estimation enables us to trace the connectedness throughout the sample, and we are able to pinpoint the national policy implementation dates. We also estimate the within-group inflation connectedness across EU members. International connectedness increases for Alcohol&Tobacco, Recreational&Cultural Activities and Restaurants&Hotels, positively responding to policy harmonization and developments in openness to trade. We argue that harmonization efforts in policy-making will further increase the inflation connectedness across EU. Keywords: Consumer price inflation, Inflation connectedness, Spillovers, Vector autoregression, Forecast error variance decomposition, Network relations, European Union.
Üretim ağlarında denemeler, verimlilik ve enflasyon bağlantılılık
This thesis consists of three chapters on the firm- and industry-level production networks and their role in firm-level productivity and performance over time, as well as the amplification of inflation shocks across sectors. The first chapter analyzes the interaction between firm heterogeneity, production networks, and productivity. A detailed analysis of 2006-2017 data shows that input-output linkages across Turkish manufacturing firms provide an amplification mechanism for shocks. After finding strong evidence that supports the asymmetry of the firm-level production networks, I build a simple model that captures the interaction between productivity and firm performance in production networks. Empirical results demonstrate the close relationship between the sophistication of a firm's production network, its productivity, and its entry decision to export markets. This study provides evidence on how firms become more productive if they are part of a sophisticated production network while proposing a hypothesis of learning-by-networking with other firms in their production network. It is shown that productivity gains correlated with the firm's position in the supply chain, industry class, and the diversity of its export destinations. The second chapter applies the Diebold-Yilmaz Connectedness methodology to producer price inflation (1947-2018) and industrial production growth series (1976-2018) for 17 U.S. manufacturing sectors to analyze how supply and demand shocks propagated in the manufacturing industry. The empirical results show that supply (demand) shocks are transmitted downstream (upstream) through the input-output network in the form of price/cost increases (production/input use decreases). Going into further detail, the paper shows that the aggregate and granular input-output network measures Granger-cause the system-wide and pairwise producer price inflation connectedness measures, respectively. The Granger causality from the input-output network to the inflation connectedness is stronger during periods of major supply-side shocks, such as the global oil and metal price hikes. Similarly, it is shown that the input-output network also Granger cause the industrial production connectedness during times of major aggregate demand shocks, such as the Volcker disinflation of 1981-84 and the Great Recession of 2008. The third chapter investigates the relationship between input-output networks and the transmission of inflation shocks across manufacturing industries in South Korea, an economy that is more open to external shocks than the United States. Using the dynamic inflation connectedness measures for 1971-2020, we show that production networks are responsible for the amplification of inflation shocks during times of supply shocks, such as the oil price shocks of 1973-74 and 1979-80. On the contrary, production networks are weakly associated with inflation transmission across sectors if the shocks originate from the demand-side such as the East Asian Financial Crisis of 1997.