Event Study - Event Induced Volatility for One Firm

[eslickstephen]

Normally, in a stock price event study, we assume that the daily variance in the estimation period is the same as that during the event window.

The event-induced volatility literature (eg, Boehmer, 1991; Brown, Warner 1985) suggests that when an event induces higher volatility, then the standard deviation used during the event window should be higher than that during the estimation period.

The problem is, though, that the test statistics they determine (that I'm aware of) use the event windows of multiple firms to determine the standard error for the aggregated return of multiple firms.

What I want is to estimate the standard deviation to be applied to the returns for a single firm's IPO (to determine if its returns are significant). I know volatility is higher just after IPOs compared to that during the subsequent months. If I know what the firm's volatility is during the subsequent months, is there a way to show what the assumed standard deviation should be just after the IPO? Is there any literature that I can follow?