Investigating the self-fulfilling-prophecy effect of going-concern audit opinions with a joint model of the opinion decision and bankruptcy

The notion that a firm may fail as a consequence (at least in part) of receiving a going-concern audit opinion has been called the self-fulfilling-prophecy effect. Prior evidence as to the existence of this effect has been mixed. This study incorporates several improvements over prior studies. First, a joint model of the opinion decision and the probability of bankruptcy is employed to specifically consider whether the opinion decision is an endogenous variable in the bankruptcy model. A heteroskedasticity consistent, two-stage method of moments estimation technique is used to estimate the joint model. Second, the investigation of the going-concern opinion's effect is limited to initial going-concern opinions. Finally, interactive variables are employed to test whether any effect of the opinion on bankruptcy varies according to firm-specific characteristics thought to be linked to external resource providers' access to information.;The self-fulfilling-prophecy effect is posited to be a byproduct of the information value of the opinion. The self-fulfilling-prophecy effect is posited to vary according to the proportion of a firm's creditors who are relatively more or less informed about the details of the firm's financial matters. Holders of "outside" debt are theorized to be less well informed than holders of "inside" debt; the interaction between receipt of an initial going-concern opinion and the proportion of firms' liabilities that are publicly held is expected to be positively related to bankruptcy.;The simple logit estimation of the bankruptcy model provides evidence of a positive "treatment effect" of going-concern opinions on the probability of bankruptcy. Despite the lack of any evidence that the two-stage method is required to control for endogeneity of the opinion, there is some decline in significance of the GC variables when the two-stage method is used. Contrary to expectations, the results do not provide evidence that the proportion of publicly held debt financing increases the opinion's effect on bankruptcy. However, the positive effect of going-concern opinions on the probability of bankruptcy increases with firm size.