Do CDS maturities matter in the evaluation of the information content of regulatory banking stress tests? Evidence from European and US stress tests

This paper questions the relevance of using only the 5-year maturity CDS spreads to examine the CDS market response to the disclosure of a regulatory stress test results. Since the stress testing exercises are performed on short-term forward-looking stressed scenarios (1 to 3 years), we assume that short-term CDS maturities (from 6-month to 3-year) should better reflect the CDS market response compared to the 5-year maturity. Based on ten regulatory stress tests carried out in Europe and in the US in the time period from 2009 to 2017, we analyze the CDS market response by investigating its reaction through all the different CDS maturities. Our results show that after the results' disclosure, the CDS market reacts by correcting the CDS spreads of tested banks (upward or downward correction), at the level of all maturities. More precisely, we evidence that for a given stress test, the nature of the correction (upward or downward) is the same for all CDS maturities while the extent of the correction differs between shortterm maturities (from 6-month to 3-year) and the 5-year maturity or more. Indeed, we find that the extent is higher on short-term maturities and in most cases, the lower the maturity of the CDS, the higher the extent of the correction (i.e. the stronger the market reaction). We therefore argue that the only use of the 5-year maturity is not suitable. Short-term CDS maturities matter since they better reflect the CDS market response. Also, the use of these short-term maturities show that the information content of the different stress tests is more diverse than what is highlighted in the existing literature.