An auditor's reputation lends credibility to the earnings report that
he audits. An unresolved issue is whether auditor size is correlated w
ith auditor quality, where a high-quality auditor is defined as one wh
o brings about more credible earnings reports. According to basic intu
ition and a modified Holthausen-Verrecchia (1988) model, investors' re
sponse to an earnings surprise will depend on the perceived credibilit
y of the earnings report. In this study, we examine whether the earnin
gs response coefficient (ERC) differs between Big Eight (B8) and non-B
ig Eight (NB8) audited firms. This provides a test of the joint hypoth
eses that auditor size is a proxy for auditor credibility and of the m
odified H-V model. Consistent with the joint hypotheses, we find that
the ERCs of Big Eight clients are statistically significantly higher t
han for non-Big Eight clients. The result obtains in both a matched sa
mple of firms paired according to industry membership, and a switch sa
mple of firms grouped according to shifts from and to B8 and NB8 audit
ors. Furthermore, the result is robust with respect to the inclusion o
f other explanatory factors for ERC that have been suggested by previo
us studies: growth and persistence, risk, firm size, and predisclosure
information environment.