This thesis investigates dynamics of determinants of government bond spreads in EMU and non-EMU countries, using non-linear Markov-switching method and Dynamic model averaging. Utilizing Dynamic model averaging we found evidence of three bond pricing regimes - pre-crisis, crisis and post Outright monetary transaction announcements. These three regimes are characteristic for all EMU countries (except Slovak Republic) and Czech Republic. Announcements of OMTs triggered post OMTs announcement regime also in Slovak republic. Third regime is not present in Poland, Hungary and United Kingdom. Moreover United Kingdom has only one regime and is dominated solely by market expectations. We found that there is heterogeneity in the determinants of bond spreads across all examined countries. Moreover we found that spreads are significantly related to market and economic sentiments. JEL Classification F12, F21, F23, H25, H71, H87 Keywords Bond yields, bond spreads, DMA Author's e-mail jan.popadak@fsv.cuni.cz Supervisor's e-mail jaromir.baxa@fsv.cuni.cz

This paper looks at the determinants of bond yields for a select group of Eurozone countries, during the European sovereign debt crisis. In addition to traditional determinants of spreads, which include credit risk, liquidity risk and international risk aversion, this paper looks at the role of credit rating agencies. The movements of countries’ yields during the debt crisis played an integral role in the resulting bailouts by the European Union. Using expected data published by the European Commission, the results of the model were in line with current literature, with the exception of the effect of budget deficits. One interpretation of the conflicting results is that during a debt crisis what market participants care about is growth, not austerity. Including the effect of credit ratings showed that credit ratings have high predictive power.

This dissertation explores three current issues relevant to the accounting and business communities by empirically examining the effect these issues have on the bond credit market. The first study examines the effect earnings predictability has on both the initial bond rating and the initial pricing of the issue. Earnings predictability is measured as (1) the annual earnings surprise (actual minus analyst forecast) and (2) the dispersion of initial analyst forecasts. The results indicate a negative association between a lack of earnings predictability and both bond ratings and initial bond price. The results are consistent with creditors interpreting greater earnings variability as a dimension of default risk incremental to the benchmark model. These results add to the existing literature by documenting a favorable benefit in the credit arena for firms that have predictable earnings. The second study investigates the effect perceived auditor independence has on the rating assigned to newly issued bonds. The magnitude of non-audit service fees is utilized as a proxy for auditor independence. The results of the study document a consistent negative relationship between the level of non-audit fees provided by the external auditor and the bond rating received by the client for new issues. Several non-audit fee measures are used in the study (raw measure, log scaled, asset scaled, unexpected) and each possess a significant negative association with a firm's bond rating. Importantly, no economic effect was discernable in a classification accuracy analysis. The third study examines what effect, if any, longer auditor tenure has on the client's bond rating. There is some contention that longer auditor tenure can lead to substandard audits either through the auditor's excessive desire to retain the client or through general auditor complacency. However, the issue of auditor tenure is far from one-sided. An alternative view asserts that longer auditor tenure increases client-specific knowledge and, thus, results in increased audit quality. Results indicate a positive association between auditor tenure and the client's bond rating on new issues suggesting that longer auditor tenure is perceived to be beneficial by bond rating analysts. This is consistent with financial statement users perceiving longer tenured auditors to have more client specific knowledge thus increasing auditor competency and a better audit. Overall, these results contribute to the existing knowledge-base in accounting by empirically demonstrating how several important issues of interest to the accounting profession are impounded into a firm's bond rating. This research provides a detailed look at how one important group of knowledgeable financial statement users, i.e. bond rating analysts, incorporate several issues that are relevant and important to the professional community. / Ph. D.

Environmental remediation liabilities are generated primarily as a result of past actions by a firm. The most important of these liabilities for domestic U.S. firms are related to Superfund sites as designated by the Environmental Protection Agency (EPA). These liabilities are important for domestic firms because of their size, which is estimated to be approximately $300 billion (Congressional Budget Office 1994) and because of public concern for the environment. This study examines the relation among bond ratings, bond yields, and EPA-based estimates of contingent environmental remediation liabilities to test if the relationships hold as theory implies it would. Extant theory suggests that financial variables, such as environmental remediation liabilities, have incremental explanatory power beyond the information included in bond ratings for bond yield. The purpose of this study is to determine the importance of external estimates of a firm's contingent environmental liabilities for a firm's cost of debt. In addition, the manner in which a firm's contingent environmental liabilities are included in the costs of debt is examined in this study. The results of this study indicate that external estimates for environmental liabilities are associated with the bond ratings and bond yield for a data set of new bond issues collected from the period 1995 to 1997. Despite that firms are increasing their recognition of environmental liabilities, either due to regulatory pressure or other factors, the measures based on EPA data still have significant explanatory power. The results imply that firms are either still lagging in appropriate recognition or that the external measures proxy for amounts imputed by the capital markets for some probable unspecified future costs. The latter explanation is supported by additional evidence in this study that the largest monetary measure of the liability is the most significantly associated with bond ratings and bond yields. Further, the results indicate that the external estimates are incorporated in bond ratings as part of the firm's default risk and have no direct influence over bond yield beyond that included in the bond ratings. This implies that bond ratings are particularly important for any evaluation of investment in debt securities from firms that have contingent environmental liabilities. / Ph. D.

Thesis advisor: Jeffrey Pontiff / The first essay of this dissertation studies the effect of Exchange Traded Funds (ETFs) on the yields and liquidity of the underlying corporate bonds. I find that ETFs lower the yield, have an insignificant or negative impact on the liquidity, and decrease the retail volume of constituent bonds. Overall, these results support theoretical predications that basket securities entice liquidity traders to exit the underlying market. The second essay analyzes the role of ETFs in mutual fund families and is joint work with Harold Spilker. We study mutual fund and ETF twins - index funds from the same family that follow the same benchmark. Mutual fund twins are shown to have lower tax burdens, long-term capital gains yields, and unrealized capital gains. Conversely, ETF twins have higher long-term yields and unrealized capital gains, but are compensated with lower expense ratios. Fund families benefit because twin offerings generate higher flows than their non-twin peers. These results support previous research that mutual fund families use diversification and subsidization to benefit the overall family. The third essay provides academics with a detailed understanding of the history, structure, regulation, and prospects of ETFs. The essay documents that the growth of index investing can largely be attributed to ETFs. The information and nuances discussed provide a baseline for developing future research questions and data. / Thesis (PhD) — Boston College, 2015. / Submitted to: Boston College. Carroll School of Management. / Discipline: Finance.

This paper proposes that managerial incentive compensation affects the firm choice between public and bank debt. To motivate the case I analyze a simple model with complete and perfect information that implies a positive relation between managers’ incentive compensation and preference toward bank debt. Using firm-level data over the period 1992-2005, I empirically examine the relation between managerial incentives and financing decisions. Specifically, I examine whether managers whose compensation is tied to firm performance choose bank over public debt as a commitment mechanism to reduce the cost of debt. Consistent with a monitoring role of banks, I find that the probability of choosing bank over public debt is positively related to the level of incentive compensation. Further, I find that public lenders price the incentive alignment between manager and shareholders by increasing the cost of debt, while the overall cost of bank loan does not depend on the manager’s incentive compensation. Finally, I find that banks are more likely to include a collateral provision in the debt contract if the manager’s compensation is tied to firm performance.

This study is among the first to provide insight into the assessment of the Swedish central bank’s (Riksbank) three unconventional monetary policies (UMPs) and their influence on Swedish commercial banks. The three UMPs include forward guidance (FG), quantitative easing (QE) and negative interest rate policy (repo rate). Riksbank introduced the UMPs in order to revive inflation and support Sweden’s economic recovery. The banks’ ability to certainly forecast their operations is highly dependent on the communication availed by the Riksbank on its expected future monetary policies through FG. QE is paramount because this is when commercial banks sell government bonds to the Riksbank. Repo rate determines interest rates set by banks. Four indicators (uncertainty, government bond yields, bank interest rates, borrowing and lending) were used in this study to investigate the perception of the commercial banks on the three UMPs. There are limited studies on Swedish banks’ perception of the UMPs which leaves a research gap in this area.Previous studies indicate that dominant banks in terms of asset shares and deposits are more sensitive to monetary policy shocks. The four dominant commercial banks studied include: Nordea, Handelsbanken, Swedbank and Skandinaviska Enskilda Banken. This thesis considers the evidence of the results from previous empirical studies. Empirical material for this study was collected through semi-structured interviews from respondents by the Riksbank and the four commercial banks. A deductive approach was used to interpret the information collected.Our results presents various perceptions of the dominant commercial banks on the three UMPs in relation to the four indicators. Some commercial banks perceived the increased transparency and clarity during the increased FG to have reduced their uncertainty. Other banks perceived that FG had increased their uncertainty. They questioned the credibility of the FG since they could not predict Riksbank’s monetary policies with the FG availed. In regards to the perception of QE on uncertainty, an increased signalling channel during QE implementation had resulted in a decline of their uncertainty since they were experiencing a surplus of liquidity in the banking sector. However, they stated other factors that increased market volatility during QE. The increased market volatility during QE increased their uncertainty. The four commercial banks agreed that the demand for government bonds increased while the yields of the government bonds declined. They perceived these changes to have been influenced by QE. The commercial banks’ lending, deposit and interbank interest rates have declined systematically correlating the trend of the declining repo rate. The four banks experienced a decline in their average net interest income, an improved flow of credit through higher lending volumes and stable lending margins to households and firms. Commercial banks perceived these changes to have resulted from the declining market interest rates because of the negative repo rate.Riksbank can use this study to assess the effectiveness of its UMPs on commercial banks based on the perception of the employees from these respective banks. This study discusses implications of the findings for commercial banks and the Riksbank, as well as academics in the realm of implementations and influences of UMPs.

The thesis analyses co-movement of 10Y sovereign bond yields of 11 EU mem- bers (Greece, Spain, Portugal, Italy, France, Germany, Netherlands, Great Britain, Belgium, Sweden and Denmark) divided into the three groups (the Core of the Eurozone, the Periphery of the Eurozone, the states outside the Eurozone). In the center of attention are changes of co-movement in the crisis period, especially near the two significant dates - the fall of Lehman Brothers (15.9.2008) and the day, when increase of Greek public deficit was announced (20.10.2009). Main contribution of the thesis is usage of alternative methodol- ogy - wavelet transformation. It allows to research how co-movement changes across scales (frequencies) and through time. Wavelet coherence is used as well as wavelet bivariate and multiple correlation. The thesis brings three main findings: (1) co-movement significantly decreased in the crisis period, but the results differ in the groups, (2) co-movement significantly differs across scales, but its heterogeneity decreased in the crisis period, (3) near to the examined dates sharp and significant decrease of wavelet correlation was observable across lower scales in some states. JEL Classification C32, C49, C58, H63 Keywords Co-movement, Wavelet Transformation, Sovereign Debt Crisis, Sovereign Bond Yields,...

O objetivo desse trabalho é analisar a influência que fundamentos internos e fatores externos exercem sobre a taxa de juros nominal dos títulos soberanos de longo prazo nos países emergentes. A base de dados engloba 13 países ao longo de 33 trimestres entre 2006 e 2014. A metodologia utilizada aborda as técnicas econométricas mais recentes para tratar dados macroeconômicos. O modelo estimado é o Pooled Mean Group desenvolvido por Pesaran, Shin e Smith (1999). O modelo consegue separar as variáveis que afetam o nível de equilíbrio da taxa de juros das que exercem efeito apenas no curto prazo. Na melhor especificação apresentada, os resultados indicam que as variáveis que causam flutuações de curto prazo na taxa de juros dos países emergentes são a taxa de juros livre de risco (+0,3), a taxa de curto prazo (+0,4) e o déficit do governo americano (+0,1). Já as variáveis que afetam o nível de equilíbrio são a taxa de juros livre de risco (+0,5), a aversão ao risco dos investidores (+0,1), o saldo em conta corrente (-0,3), a inflação (+0,1), a abertura econômica (-0,04) e a dívida bruta (efeito positivo não linear). Por fim, ao analisar exclusivamente o caso brasileiro, nosso modelo indica que se o país tivesse mantido os fundamentos internos em linha com a média dos outros emergentes desde o final de 2006, o Brasil iria chegar ao final de 2014 com uma taxa de juros nominal de longo prazo de 6,1%, 6 p.p. abaixo do valor efetivamente observado que foi de 12,1%. / The aim of this study is to analyze the influence that internal fundamentals and external factors have on long-term sovereign bond yields in emerging markets. The database covers 13 countries over 33 quarters between 2006 and 2014. The methodology addresses the latest techniques to deal with macro panels, such as panel unit root and panel cointegration tests. The estimated model is the Pooled Mean Group developed by Pesaran, Shin e Smith (1999). This model can separate the variables that affect the interest rate equilibrium from the ones that only have an effect in the short-run. In the best specification presented, the results indicates that in the short-run the variables that affects the interest rates in emerging markets are the risk free interest rate (+0.3), the short term interest rate (+0.4) and the American government deficit (+0.1). The variables that affects the equilibrium level are the risk free interest rate (+0.5), risk aversion (+0.1), current account balance (-0.3), inflation (+0.1), trade openness (-0.04) and gross debt (non-linear positive effect). Finally, by examining only the Brazilian case, our model indicates that if the country had maintained the internal fundamentals in line with the average of other emerging markets since the end of 2006, Brazil would reach the end of 2014 with a long-term interest rate of 6.1%, 6 p.p. below the actual value of 12.1%.

O objetivo desse trabalho é analisar a influência que fundamentos internos e fatores externos exercem sobre a taxa de juros nominal dos títulos soberanos de longo prazo nos países emergentes. A base de dados engloba 13 países ao longo de 33 trimestres entre 2006 e 2014. A metodologia utilizada aborda as técnicas econométricas mais recentes para tratar dados macroeconômicos. O modelo estimado é o Pooled Mean Group desenvolvido por Pesaran, Shin e Smith (1999). O modelo consegue separar as variáveis que afetam o nível de equilíbrio da taxa de juros das que exercem efeito apenas no curto prazo. Na melhor especificação apresentada, os resultados indicam que as variáveis que causam flutuações de curto prazo na taxa de juros dos países emergentes são a taxa de juros livre de risco (+0,3), a taxa de curto prazo (+0,4) e o déficit do governo americano (+0,1). Já as variáveis que afetam o nível de equilíbrio são a taxa de juros livre de risco (+0,5), a aversão ao risco dos investidores (+0,1), o saldo em conta corrente (-0,3), a inflação (+0,1), a abertura econômica (-0,04) e a dívida bruta (efeito positivo não linear). Por fim, ao analisar exclusivamente o caso brasileiro, nosso modelo indica que se o país tivesse mantido os fundamentos internos em linha com a média dos outros emergentes desde o final de 2006, o Brasil iria chegar ao final de 2014 com uma taxa de juros nominal de longo prazo de 6,1%, 6 p.p. abaixo do valor efetivamente observado que foi de 12,1%. / The aim of this study is to analyze the influence that internal fundamentals and external factors have on long-term sovereign bond yields in emerging markets. The database covers 13 countries over 33 quarters between 2006 and 2014. The methodology addresses the latest techniques to deal with macro panels, such as panel unit root and panel cointegration tests. The estimated model is the Pooled Mean Group developed by Pesaran, Shin e Smith (1999). This model can separate the variables that affect the interest rate equilibrium from the ones that only have an effect in the short-run. In the best specification presented, the results indicates that in the short-run the variables that affects the interest rates in emerging markets are the risk free interest rate (+0.3), the short term interest rate (+0.4) and the American government deficit (+0.1). The variables that affects the equilibrium level are the risk free interest rate (+0.5), risk aversion (+0.1), current account balance (-0.3), inflation (+0.1), trade openness (-0.04) and gross debt (non-linear positive effect). Finally, by examining only the Brazilian case, our model indicates that if the country had maintained the internal fundamentals in line with the average of other emerging markets since the end of 2006, Brazil would reach the end of 2014 with a long-term interest rate of 6.1%, 6 p.p. below the actual value of 12.1%.