On Russell Index Reconstitution

Abstract

This paper investigates whether abnormal returns exist due to transparent changes in domestic U.S. Russell equity indexes. Newly-listed (delisted) companies in the Russell 1000 and Russell 2000 indexes have significant positive (negative) abnormal returns in May and June but not in July. Newly-listed companies in these Russell indexes have significant positive abnormal returns on the reconstitution date, not on the announcement date. This may be due to the reticence of money managers to rebalance their portfolios prior to the reconstitution date in order to avoid tracking error. This study provides evidence that some firm attributes are changed when a company is added or deleted from the Russell indexes. In addition, changes in attributes predict the excess returns associated with index reconstitution. Finally, using intraday data from TAQ, we find that Russell inclusion and deletion is associated with some permanent changes in liquidity.