2016 05 19 new limits to arbitrage 883x432 1

New limits to arbitrage

15 June 2016

The article at a glance

The structure of institutional stock ownership – not just the per cent of total institutional ownership – is important in understanding how limits …

The structure of institutional stock ownership – not just the per cent of total institutional ownership – is important in understanding how limits to arbitrage prevent investors from adjusting prices to their efficient values, says study co-authored at Cambridge Judge Business School.2016-05-19-New limits to arbitrage-883x432

“Stocks with lower, more concentrated, short-term, and less passive ownership exhibit lower lending supply, higher costs of shorting, and higher arbitrage risk,” concludes the study of nearly 5,000 US stocks over a four-year period (2006-2010) – forthcoming in the Review of Financial Studies.

In simple terms: suppose that someone told you that a glass is 50 per cent full. While this is relevant information, it also matters whether the glass contains water, milk or wine, its temperature and expiry date. The article examines a wide list of institutional ownership characteristics, and shows that stocks with lower, more concentrated, short-term, and less passive institutional ownership exhibit lower lending supply, higher costs of shorting, and higher arbitrage risk.

Overall, it means that arbitrageurs will face higher costs to trade and more uncertainty when trying to exploit market inefficiencies, which in turn results in both less efficient prices and delays in incorporating pessimistic investors’ opinions.

“These results suggest a link between limits to arbitrage and ownership structure, which to the best of our knowledge has not been explored previously,” says the study.

The study – entitled “Ownership structure, limits to arbitrage, and stock returns: evidence from equity lending market” – was co-authored by Dr Pedro Saffi of Cambridge Judge Business School, Dr Melissa Porras Prado of Nova School of Business and Economics in Lisbon, and Dr Jason Sturgess of DePaul University in Chicago.

“We show that institutional ownership is not a sufficient statistic to proxy for lending supply as is often assumed,” the study concludes. “Instead, both institutional ownership levels and the structure of institutional ownership should be considered.” For example, short-sales constraints are different if a single institutional investor owns 80 per cent of the firm relative to when these 80 per cent are dispersed through several shareholders. This matters for investors trying to understand what are the costs and risks to short a stock.

The study also examines data from Schedule 13D filings required by the US Securities and Exchange Commission (SEC) whenever a shareholding crosses a five per cent threshold. The SEC requires investors to disclose the purpose of the transaction, which allows authors to identify if an investor will actively try to change or influence control, that is, engage in activism.

The presence of such activist investors, the study found, reduces supply of loaned shares for short-selling by two per cent, increases the loan fee by 17 basis points (from 0.56 per cent to 0.73 per cent), and increases the residual arbitrage risk by about 12 per cent (from 3.26 per cent to 3.64 per cent, or 38 basis points).

Dr Pedro Saffi
Dr Pedro Saffi

Q&A with Pedro Saffi

Dr Pedro Saffi, University Lecturer in Finance at Cambridge Judge Business School, answers some queries regarding the study:

What are your study’s basic findings?

Differences in ownership composition can explain the presence of limits to arbitrage through their impact on short-sales constraints.

Stocks in the top quartile of ownership concentration earn an additional average abnormal return following an increase in the demand to short of minus two per cent per month relative to an increase for stocks with dispersed ownership. This is consistent with short-sale constraints limiting negative information from being reflected in prices quickly. We also find a similar result when we examine reactions to earnings announcements.

How many companies did you study, and over what time period?

The study uses 4,487 US firms from 2006 through to 2010, combining data from several sources.

Why are institutional investors with large stakes more concerned about loaning out shares to short-sellers?

Short-sellers want prices to go down. They borrow shares and sell them today, hoping to purchase them in the future and return them to the lender at a lower price, keeping the difference as profit. Thus, a lender with a larger stake that lends out his shares would be giving ammunition for other investors to bet against its own holdings. The larger his stake, the less likely he is to lend shares.

The study looks at limits to arbitrage – in simple terms what is this?

Anything that reduces the ability of investors to correct inefficiencies in market prices, such as trading costs, regulatory constraints and risks of trading. One particular limit relates to short-selling stocks, such as high loan fees or the inability to locate shares.

Why have previous studies on short-selling and arbitrage focused on overall per cent rather than structure of institutional ownership?

This is a very opaque market and all transactions occur over the counter. Finding good data isn’t easy and it hasn’t been until very recently that researchers were able to examine it in detail.

This article was published on

15 June 2016.