GC opinions also have several limitations. One is that, like restatements and AAERs, the egregious nature of GCs means
that they are not useful in capturing more subtle compromises in audit quality. GCs are also relatively rare and exclusively
issued to financially distressed clients. This reduces statistical power in tests using large samples of healthy firms. While
researchers often restrict their analysis to distressed firms to increase power, it reduces generalizability. Another limitation
is that GCs reflect a fairly narrow aspect of the auditor’s role, and do not fully capture the broad value of auditing. Finally,
while the literature interprets more GCs as greater auditor independence, more GCs may also indicate excessive auditor
conservatism, which arguably reduces audit quality. Auditors have incentives to issue more GCs than are appropriate
because they reduce the auditor’s liability in court (Kaplan and Williams, 2013).29 The risk of erroneously interpreting
excessive auditor conservatism as increased audit quality is a problem that affects all output-based audit quality proxies,
including restatements and DAC. However, clients have incentives to resist excessive auditor conservatism by dismissing
overly conservative auditors (DeFond and Subramanyam, 1998)