Rules and Regression Discontinuities in Asset Markets
46 Pages Posted: 29 Aug 2011 Last revised: 13 Aug 2013
Date Written: August 27, 2012
We show that institutional portfolio rules induce regression discontinuity experiments in stock markets using the Russell indices. Stocks are ranked each May 31st based on their market capitalization. Those ranked above the 1000 cut-off are in the Russell 1000. Those below the 1000 cut-off and above the 3000 cut-off are in the Russell 2000. Indices are value-weighted. Small stocks just below the 1000 cut-off are heavily weighted in Russell 2000 and receive more forced buying than large ones just above the cut-off that have little weight in Russell 1000. Just-included stocks have discontinuously higher returns in June compared to just-excluded ones, followed by return reversal in July. Just-included stocks discontinuously co-vary more with the market subsequent to index inclusion. There is also a discontinuity around the 3000 cut-off.
NOTE: This is an old draft. The new draft "Regression Discontinuity and the Price Effects of Stock Market Indexing" can be found at: http://ssrn.com/abstract=2288184
Keywords: portfolio rules, regression discontinuity, index inclusion
JEL Classification: G11, G12, G14, G2
Suggested Citation: Suggested Citation