Trading on Coincidences

Abstract

This paper develops a model showing why traders might use coincidences to identify promising investment opportunities that are worth investigating further. The model predicts that, if both National Semiconductor and Sequans Communications realize top-10 returns (i.e., the semiconductor industry displays a coincidence), then the rest of the industry will have positive abnormal returns in the following month. The data confirm this prediction. Industries with a coincidence in the previous quarter have abnormal returns of 2.02% even after excluding the coincident firms. Industries with one top-10 return and one 11:20 return do not display this pattern. The results are robust to including controls for firm-specific momentum, industry-specific momentum, and within industry cross-autocorrelation. The findings also generalize to looking at geography-based top-10 coincidences.

Keywords: Market Scale, Cognitive Control, Behavioral Finance, Coincidences

JEL Classification: D83, D84, G02, G12, G14