Failures to Deliver, Short Sale Constraints, and Stock Overvaluation

• Author: Thomas J. Boulton,Don M. Autore,Marcus V. Braga‐Alves
• Published date: 01 May 2015
• Date: 01 May 2015
• DOI: http://doi.org/10.1111/fire.12063

The Financial Review 50 (2015) 143–172

Failures to Deliver, Short Sale Constraints,

and Stock Overvaluation

Don M. Autore

Florida State University

Thomas J. Boulton∗

Miami University

Marcus V. Braga-Alves

University of Akron

Abstract

Studying a large sample of publicly available data on failures to deliver, we find that

stocks reaching threshold levels of failures become significantly overvalued.Where short sale

constraints are especially binding, we report extreme overpricing and subsequent reversals.

These findings support the overvaluation hypothesis, although the mispricing is likely to be

difficult to arbitrage because of extreme shorting costs. In addition, threshold stocks with

low short interest become more overvalued than threshold stocks with high short interest.

This suggests that the level of short interest reflects supply-side effects when the examination

conditions on the difficulty of borrowing shares.

∗Corresponding author: 3013 Farmer School of Business, Miami University,Oxford, OH 45056; Phone:

(513) 529-1563; Fax: (513) 529-6992; Email: boultotj@miamioh.edu.

The authors thank Robert Van Ness (editor), two anonymous reviewers, Matteo Arena, Robert Batallio,

Leslie Boni, Brandon Carl, Jerry Dwyer, Larry Harris, Zsuzsa R´

eka Husz´

ar, Terry Nixon, Sarah Peck,

Adam Reed, Andrei Shleifer, Sorin Sorescu, James Upson, Steve Wyatt, Chad Zutter, and seminar par-

ticipants at the American Finance Association Meetings (Chicago), Federal Reserve Bank of Atlanta,

Financial Management Association Meetings (New York), Miami University, Midwest Finance Associ-

ation Meetings (Chicago), and the University of Kansas for valuable comments. Financial support for

this project was provided by the Florida State University College of Business. Any remaining errors or

omissions remain the responsibility of the authors.

C2015 The Eastern Finance Association 143

144 D. M. Autore et al./The Financial Review 50 (2015) 143–172

Keywords: failures to deliver, short interest, institutional ownership, short sale constraints,

stock returns

JEL Classifications: D02, G14, G28

1. Introduction

Short sale constraints can result in overpriced securities in the presence of in-

vestor heterogeneity (Harrison and Kreps, 1978; Duffie, Garleanu and Pedersen,

2002) or less than fully rational investors (Miller, 1977; Scheinkman and Xiong,

2003). Studies often test the overvaluation hypothesis by identifying a sample of

short sale constrained stocks and examining subsequent stock returns. For exam-

ple, stocks with high short interest or low institutional ownership tend to experience

negative future returns.1Boehme, Danielsen and Sorescu (2006) find that under-

performance is especially severe for stocks with both binding short sale constraints

and wide dispersion of opinion. Chang, Cheng and Yu (2007) report that stocks that

become eligible for shorting exhibit poor future stock performance.

There is less attention paid to the predicted abnormal increases in stock prices

when constraints are tightening. Jones and Lamont (2002) is one exception. They

study loan crowd entrants during the 1920s and 1930s, a period in which NYSE

stocks entered a centralized loan list when shorting demand could not be met through

traditional channels. They find that market-to-book ratios increase dramatically lead-

ing up to a stock’s entrance onto the loan list, peak at the point of entrance, and

subsequently fall as the overvaluation subsides.

The absence of additional studies that take a dynamic view of short sale con-

straints is arguably due to the fact that modern U.S. data provide limited opportunity

to examine large samples of well-defined, firm-specific events that capture the tight-

ening and subsequent easing of short sale constraints. A notable exception is the

threshold lists of excessive failures to deliver released daily by the three major ex-

changes. Since 2005, securities with aggregate open failures to deliver equal to, or

greater than, 10,000 shares and 0.5% of the total shares outstanding for five consec-

utive settlement days appear on their exchange’s daily threshold list. The exchanges’

threshold files enable us to identify both the addition date and the subsequent removal

date for each threshold security. Evans, Geczy, Musto and Reed (2009, p. 1975) sug-

gest that the advent of threshold events “has the potential to alter the cost of short

exposure, so its impact is an important new empirical question.” In this study, we

conduct an empirical investigation of threshold events.

1See, for example, Asquith and Meulbroek (1995), Desai, Ramesh, Thiagarajan and Balachandran (2002),

and Diether, Lee and Werner(2009) for evidence on short interest and Asquith, Pathak and Ritter (2005)

and Nagel (2005) regarding institutional ownership.