Common-wide information flow in CDS markets: The case of earnings surprises of industry leaders

ABSTRACT

This paper examines the way common-wide information is disseminated in credit default swap (CDS) markets. I find strong supporting evidence that earnings announcements of CDS references with greater liquidity and higher credit risk help in predicting the subsequent CDS spread movements of their industry peers. This is consistent with the notion that the earnings announcements of certain firms may contain valuable market- and industry-wide information, which may be utilised afterwards to appraise the financial situation and the creditworthiness of industry peers, thereby influencing their CDS rates. Another major finding is that this common-wide information seems to be impounded with a lag into the CDS rates of industry peers. That pattern appears to be more pronounced for negative than for positive earnings surprises. The lag in the response to earnings surprises is higher among firms of the financial, cyclical consumer and energy sectors, and almost non-existent for firms of the sectors of technology and utilities. Attention costs and market frictions seem to fuel the pattern of under-reaction of CDS spreads to these earnings announcements.

KEYWORDS:

• CDS (credit default swaps)
• information flow
• earnings surprise
• price discovery

JEL CODES:

• G12
• G13
• G14

Notes

1 A CDS is a fixed-income derivative instrument that permits a protection buyer to purchase insurance against a contingent credit event on an underlying reference entity by paying a premium (generally quarterly payments) to the protection seller, generally referred to as the CDS spread, over the life of the contract. The CDS spread is usually defined as a percentage of the notional amount of the contract (or in basis points). The failure of an entity to meet its debt obligations is considered a credit event. A credit event triggers a payment by the protection seller to the buyer equal to the difference between the notional principal and the value of the underlying reference obligation. (See Augustin et al. (2014) for a survey on CDS markets.) Contrary to stocks, CDS is typically traded in over-the-counter markets through bilateral agreements.

2 Additional references on this rich literature can be found in Griffin (2014), who provides an extensive survey of the role of financial statement information in the determination of credit spreads.

3 Other alternative spans were analysed without producing significant changes in the results.

4 The model framework is adapted from Hotchkiss and Ronen (2002) and Hou (2007).

5 The hypothesis of poolability assumes homogeneous slope coefficients for the various obligors.

6 The use of random effects rests on the hypothesis that firms' heterogeneity does not correlate with the explanatory variables. Nevertheless, using fixed effects leads to virtually identical results.

7 Winsorisation is a common procedure to remove the influence of outliers on the estimation that consists of replacing any data value above the (1 – x) percentile of the sample data by the (1 – x) percentile and any value below the x percentile by the x percentile.

8 The advantage of using the span [ in the SCAR computation is that it is easily accommodated in a setup similar to equation (3).