In this study we attempt to answer the question – does the start of pre-announcing of S&P 500 index changes in October 1989 have an effect on the trading pattern of added or deleted firms? We document that prior to October 1989 the excess returns of added or deleted firms follow a white noise process around the event, whereas after the start of pre-announcing the excess returns can be described as nonstationary. This indicates significant excess profits to be captured around the addition or deletion event after S&P started pre-announcing changes in October 1989 but not prior to that date.
Stoyu Ivanov. "Analysis of the Effects of Pre-Announcement of S&P 500 Index Changes." Faculty Publications (2013).