Large AI and hyperscaler stocks have delivered unusually sharp post-earnings moves this season, overturning the typical pattern in which smaller companies produce the biggest reactions after results. ORATS said major names including Amazon, Microsoft, Google and Meta moved more than their historical averages as investors responded strongly to whether heavy AI spending appeared to be generating returns. The shift also showed up in options trading: for companies reporting in the first week of the second-quarter earnings season, which began in mid-July, long straddles (buying a call and a put to bet on volatility) generated average gains of 23%, versus an average 2% loss in the first week over the prior 12 quarters. By week four, the strategy posted an average 6% loss, close to the historical average loss of 5%, suggesting the season's biggest volatility was concentrated early rather than later among smaller companies.