Relative Efficiency of G8 Sovereign Credit Default Swaps and Bond Scrips: An Adaptive Market Hypothesis Perspective

Abstract

This article is aimed at examining the degree of efficiency exhibited by the sovereign credit default swaps (SCDS) and the sovereign bonds (SBs) pertaining to the G8 countries, namely, US, UK, Japan, Germany, Italy, France and Russia, from an Adaptive Market Hypothesis (AMH) perspective. At the outset, the authors employ rolling AR(1)–GARCH(1,1) filter so as to remove short-term dependency in the different time series considered for this study. The AR(1)–GARCH(1,1)-filtered rolling standardized residuals were then subjected to Mandelbrot’s classical R/S test so as to obtain rolling Hurst exponents. Subsequently, the evolving efficiencies of G8 SCDS and SB scrips as reflected by the scrip-wise rolling Hurst exponents are made available. In doing so, scrip-wise transient inefficiencies, as characterized by periods that exhibit strong evidence of long memory, were identified. The authors then offer a ranking of G8 SCDS and SB scrips, in the decreasing order of relative efficiency, based on proportion of scrip-wise rolling windows that exhibited strong evidence of long memory. The findings pertaining to this study are as follows. First, SBs pertaining to UK, Japan, US, France and Russia were found to be the most efficient, while SBs pertaining to Russia were found to be the least efficient. Second, when it comes to SCDS, Russian SCDS scrips were found to be relatively more efficient than the Russian SBs. Having said so, G8 SCDS scrips pertaining to the other countries were found to be inefficient vis-à-vis their respective SBs. Of all the G8 countries, US was the only country wherein no strong evidence of long memory was witnessed in SCDS and/or SB scrips. Finally, there is discernable difference in the degree of efficiency exhibited by G8 SCDS scrips vis-à-vis their underlying reference obligations (ROs).