Resumen de Four essays on the interaction between credit derivatives and fixed income markets

Sergio Mayordomo

• In this thesis, I study the interaction between Credit Derivatives and Fixed Income Markets, both corporate and sovereign, from different perspectives. In the case of corporate, I study arbitrage, price discovery and financial integration. In the case of sovereign, I focus on the European Monetary Union (EMU) sovereign bond market and analyze the potential arrival of a common risk free rate for the EMU and the advantages derived from it. First, we analyze long-run and statistical arbitrage opportunities in credit deriva- tives markets using strategies combining Credit Default Swaps (CDSs) and Asset Swaps (ASPs). We present a new statistical arbitrage test which has lower Type I error and selects arbitrage opportunities with lower downside risk than existing alternatives. This test allows us to study for arbitrage opportunities in the appropriate way by focusing our analysis on the cases in which long positions in CDSs and ASPs are needed. Using four di¤erent databases from 2005 to 2009, we find long-run and statistical arbitrage opportunities before the current crisis in 27% and 29% of the cases, respectively. During the crisis, they decrease to 9% and 17%, respectively. Specifically, CDS spreads are too low in comparison with asset swap spreads. This fact puts into question the e¢ ciency of this segment of the CDS market. After considering funding and trading costs, we find SA opportunities in 16% of the cases before the crisis but never during the crisis. Thus, once the crisis started, noticeable deviations from the parity relation appeared although the increase in funding costs makes the apparent arbitrage opportunities non profitable. Finally, we find that arbitrage opportunities are more frequent in low rated bonds. Second, we analyze whether liquidity affects the price discovery process in credit derivatives markets and we find that it does. Specifically, we focus on the credit derivatives markets in the context of the subprime crisis. We fill a gap in the price discovery literature in credit markets, because no analysis of the price discovery process between ASPs or bonds and CDSs has been carried out up to now. The goal of the price discovery model is to analyze the dynamics and interaction between CDS and ASP or bond spreads in an equilibrium non-arbitrage model. Thus, this paper contributes to the price discovery literature by considering another facet of the credit markets (ASPs) and by adapting Garbade and Silver (1983) model to this segment of credit markets. We first present an agent-based theoretical price discovery model for the ASP, bond and CDS markets that allows for simultaneous agent participation in different markets. We then empirically test this model and obtain that the CDS market leads the ASP market in the price discovery process before the crisis (the high liquid scenario). This result changes with the appearance of the subprime crisis. Thus, during the crisis (the low liquid scenario) the ASP leads price discovery. We obtain similar results when we relate the CDS and bond markets. The difference in the relative liquidity between CDS and ASP/bond markets helps to explain these results. We also extend our analysis to test the suggestion that ASP spreads are a more accurate measure of credit risk than bond spreads. Our results indicate that ASP spreads consistently lead bond spreads up to the point that according to Garbade and Silver (1983) terms, we find that the bond spread is a "pure satellite" of the ASP spread. Third, we analyze the internal (intra-market) and external (inter-market) integra- tion of three European corporate credit markets (corporate bonds, asset swap packages and corporate credit default swaps) in the context of the current financial crisis. For this aim, we use a DCC-GARCH model for each market's innovations which are obtained after subtracting from the credit spreads the effect of the fundamentals of the under- lying entities. Our measure of internal market integration is based on the average DCCs between an individual (firm-specific) credit spread innovation, and the corresponding average market credit spread innovations. Similarly, our measure of external market integration is based on the average DCCs between the two individual credit spreads innovations of the same frm but in different markets. The higher the innovations'correlation, the stronger is the market integration. We find that credit spread changes are largely driven by firm-specific innovations and to a lesser extent by changes in funda- mentals. Internal market integration increases during the crisis for CDSs but decreases for bonds and ASPs. External market integration decreases during the crisis between the CDS and the other two markets. Both facts suggest that the CDS market tends to follow its own way to a considerable extent in times of nancial distress. The degree of internal and external integration is signi cantly a¤ected by liquidity and global risk factors. Finally, we study the impact of a hypothetical common European Monetary Union (EMU) sovereign bond yielding a common European Monetary Union risk free rate. The possibility of a common European bond has attracted the interest of the financial press and is receiving increased attention from policy makers. However, there is no published quanti cation of a common risk free rate, nor a detailed comparison with other possible alternatives is available. This paper addresses both questions and presents a tentative estimate of this common risk free for the European Monetary Union countries from 2004 to 2009 using variables motivated by a theoretical portfolio selection model. First, we analyze the determinants of EMU sovereign yield spreads and find significant effects of the credit quality, macro, correlation, liquidity and interaction variables. Robustness tests with different data frequencies, benchmarks, liquidity and risk variables, cross section regressions, balanced panels and maturities confirm the initial results. Motivated by these results, we present a tentative estimate of the common risk free rate which will be free, at least to some extent, from the e¤ect of the risk factors (credit, liquidity, macro, correlation) that influence the yield of individual sovereign bonds. Finally, we find tentative evidence in favour of the hypothesis that a common bond and a common risk free rate in the EMU could produce substantial savings in borrowing costs for all the countries involved. Of course, there are many institutional design features that must be resolved (seniority, amount relative to total debt issues, guarantee fund, etc.) before such a common bond can be launched. But our paper provides a first insight into one central issue.--------------------------------------------------------------------------------------