The Upstairs Market for Large-Block Transactions: Analysis and Measurement of Price Effects

Abstract

This paper analyzes the effect on equity prices of large- block transactions negotiated 'upstairs.' We develop a model of the upstairs market which yields testable hypotheses. We investigate these hypotheses with unique data for 5,625 block trades in 1985-1992. Unlike previous studies, all the trades in our data are negotiated upstairs and are identified as either buyer- or seller-initiated. This information is critical because many block trades examined in the previous literature occur downstairs. We find that price movements up to four weeks prior to the trade date are significantly related to trade size, consistent with information leakage as the block is "shopped" upstairs. This suggests that permanent price impacts measured relative to the price on the day preceding the block trade can severely underestimate the information contained in the block trade. The temporary price impact has a concave response to order size, which is consistent with our prediction of more intensive search in the upstairs market as trade size increases. Since our sample consists of mostly small market capitalization stocks, our estimated price impacts are substantially larger than found in previous studies. We also find that the price responses for buyer-and seller-initiated trades are asymmetric.

JEL Classification: G12