c y c l e o f 7 . 7 9 % p . a . , t h e L o n g s t a ff m o d e l g i v e s a v a l u e o f k 0 2 = 4 . 6 8 % c o m p a r e d t o k 2 w h i c h w e e m p i r i c a l l y
obtained to be 4.46%. These values are very close indicating a reasonable level.
Finally, the cost of the short sale constraint has to be verified. The most direct measure of short
sale costs is the lending or loan fee paid by the lender of a security to the borrower of that security.5
This fee arises because to sell a security short, an investor must borrow shares from an investor who
owns them and is willing to lend them. The lending fee serves to equilibrate supply and demand in the
While quantity data in the shorting market are readily available, price data are not. The lending market is not centralized and lending fees do not need to be disclosed.
Cohen, Diether, and Malloy (2007) report examples from a sample of proprietary stock lending data from September 1999 to August 2003. They report statistics on the full sample and on two sub-samples, large stocks and small stocks.6 The mean lending fee from the full sample is 2.60% p.a. while the 0.75 percentile value is at 4.20%. For large stocks, the mean lending fee is 0.39% and the 0.75 percentile value is only at 0.16%. Small stocks on the other hand have much higher loan fees with a mean value at 3.94% and a 0.75 percentile value at 5.30%. The most extreme lending fees in this sample are 7.25% and 14.75% respectively.
D’Avolio (2002) describes the market for borrowing stock and finds that 91% of the stocks lent out in the investigated sample exhibit a lending fee below 1%. The remaining 9% of the stocks have a mean fee of 4.3%. Less than 1% of the sample, typically small stocks with little institutional ownership, exhibit very high fees ranging from 10% up to 79%.
Investigating boundary levels, extreme values that reflect the maximum willingness to pay for a short sale possibility are critical. Stocks with very high lending fees can be considered special cases. For example, Lamont and Thaler (2003) study 18 equity carve-outs from April 1996 to August 2000 in which the parent has stated its intention to spin off its remaining shares of an already listed subsidiary company. Most subsidiaries were overpriced compared to their parent companies and had a significantly larger short interest than the parent. In the case of Palm Inc, a temporarily heavily overpriced subsidiary of 3Com, short interest was as high as 147.6% at the peak. That is, more than all floating shares had been sold short. Borrowed shares can be sold short to an investor who then lends them again. Lamont
and Thaler (2001) conclude that in the case of Palm, arbitrageurs could not find enough shares to satisfy 5A series of recent papers analyzes direct measures of shorting costs, e.g. Brent, Morse, and Stice (1990), D’Avolio
, Figlewski and Webb (1993), Lamont and Stein (2004), Ofek, Richardson, and Whitelaw (2004), Jones and Lamont
, Reed (2002) and Geczy, Musto, and Reed (2002).
6Large stocks exhibit a market capitalization above the NYSE median while small stocks are below the median.