The Impact of External Shocks on Pricing Mechanism of Sovereign CDS and Bonds in Advanced Economies

The paper analyzes the pricing of sovereign bonds and CDS of 7 developed countries in terms of the impact of external shocks on volatility dynamics. The external shocks corresponded to the crisis events of 2019 - 2022. Among them, three external shocks of particular interest for the analysis stand out: 1) the beginning of the COVID-19 pandemic in March 2020 and the corresponding increase in volatility in financial markets; 2) the global energy crisis of late 2021 - 2022; 3) the crisis of the abrupt change of monetary policy by the Fed and the ECB in 2022. The analysis of the long-term impact of external shocks was carried out using the study of the mean reversion phenomenon, which characterizes the rate of return of volatility to its long-term average value. In the framework of the paper, we hypothesized that there is no long-term impact of external shocks on the dynamics of volatility of sovereign bonds and CDS of developed countries in the period from 2019 - 2022. In accordance with this hypothesis, we predicted an increase in the mean reversion rate due to the introduction of external shocks into the ARMA-(E)GARCH model. The primary data were five-year zero-coupon bond prices and five-year CDS premiums of seven developed countries. These included the US, Germany, Japan, Italy, Spain, France, Germany, Japan, Italy, Spain and the UK. The processed data were logarithmic series of weekly yields from January 2019 to December 2022. The key idea in the methodology of testing the hypothesis of the study was the comparison of volatility parameters of two model modifications and their mean reversion rates in the conditional variance equation. The first modification of the model was obtained in the second chapter and did not take into account external shocks in the returns and volatility of the analyzed series. The second modfication of the model was estimated in the third chapter by introducing dummy variables corresponding to the dates of structural breaks into the return and variance equations. This resulted in obtaining optimal models whose parameters could be compared with the corresponding models from the second chapter. Both models with/without accounting for external shocks exhibit characteristics of a mean reversion process in the level of volatility, with the rate of mean reversion increasing after accounting for external shocks. The leverage effect was not detected in the models with/without external shocks. Thus, it can be stated that there is no appreciable influence of investors' riskophobia on the volatility of the studied markets. The best CDS market for arbitrage deals to be considered is France CDS market, for which the speed of volatility return to its average value is less than one trading week. The best bond market in this aspect is the UK market, where the speed of mean reversion is estimated at 7 trading days. Bond and CDS markets of Spain can be considered the worst markets in terms of arbitrage opportunities, as they require the longest possible time to close a position at the moment when the past volatility shocks are fully absorbed and the arbitrage window closes - 118 and 83 trading days, respectively. The hypothesis of the study concerning the absence of long-term impact of crises on volatility dynamics was highely likely confirmed in the presented paper. This result provides strong evidence that the pricing of sovereign bonds and CDS does not follow the efficient market hypothesis. Thus, the prices and volatility of the studied assets return to their average values in the long run. Sovereign bonds and CDSs of the considered developed countries provide significant arbitrage opportunities for capital market players in case of successful modeling of long-term average volatility/returns and the speed of return to them. The findings may also be useful for regulatory institutions engaged in market pricing research to implement effective regulatory and monetary policies.