A growing literature documents that various strategies of rotating across equity styles generate significant returns. However, the conventional risk adjustment regression is problematic in evaluating the gains from style rotation. I propose a weight-based multifactor risk adjustment approach as a solution. When interpreted as a performance attribution procedure, this approach extends Sharpe's (1992) classic method by emphasizing factor loading rotation. I use a logit-based timing strategy to show that the conventional procedure produces misleading results and the new method leads to the opposite conclusion.