Does a nearby split change this? (event study)

Prices already use split-adjusted values (see the correction note on the event-study page), so a split should not distort the return by itself. This checks something extra: whether excluding filings that happen to fall near a real split — for the noise a split day can bring (order-book resets, speculative interest) beyond the price adjustment itself — changes the result. Four exclusion windows, same convention as the monthly panel's split scenarios.