AN EXAMINATION OF THE UK CREST DATABASE ON ... - …



Patterns in Stock Lending

James Clunie a,[1], Yi Wu b

a University of Edinburgh, 50 George Square, Edinburgh, EH8 9JY, United Kingdom

bBarrie & Hibbert Ltd., 50 Lothian Road, Edinburgh, EH3 9AN.

Abstract

There is a growing body of evidence that short-sellers are well informed, and that highly-shorted stocks subsequently perform poorly. This suggests a trading strategy for those with full access to information on short-sales. Such full information is not publicly available, however, in most major markets. In the United Kingdom, though, stock lending data is available with a five day lag. As short-sales ordinarily require a loan of stock, stock lending data can serve as a proxy for short-sales. We test if there is exploitable information in stock lending data. We find that when confounding factors can be eliminated, stock lending data can provide valuable information to traders. We also identify a number of stock return and borrowing patterns consistent with the notion that borrowers are subject to ‘short-squeezes’. This suggests the use of stock lending data for an alternative purpose – to identify opportunities for predatory trading.

JEL classification: G11; G12; G14

Keywords: Securities lending; Stock lending; Short-selling; Predatory trading

1. Introduction

The association between securities lending and short-selling is of particular interest to academics, market participants and regulators alike, because of its impact on price discovery and market liquidity, and because of its role in the process of arbitrage. A common reason for borrowing is to facilitate short-selling (i.e. the sale of a security that the seller does not own). For this reason, securities lending data can serve as a proxy for short-selling data. There are, however, other reasons for borrowing securities, which weakens the relationship between short-selling and stock lending. In this paper, we examine a new, commercial database providing daily shares on loan information for the United Kingdom’s largest 350 listed companies. Our aim is to gain insights into the reasons for borrowing stock, to identify the types of stock preferred by borrowers, and to test if there is informational value in the data, by analysing the relationship between stock lending and abnormal returns.

Securities lending is the market practice whereby securities are transferred temporarily from one party (the lender) to another (the borrower) for a fee. The borrower must return the securities to the lender either at the end of an agreed term, or on demand. In law, securities lending is an absolute transfer of title (or sale) against an undertaking to return equivalent securities. Most securities loans are collateralised with cash or other securities. The process is facilitated by intermediaries such as custodians, investment banks or stockbrokers. Lending of securities is primarily motivated by the fee income received from the loan. D’Avolio (2002) studies a US stock lending database from April 2000 to September 2001 and shows that the value-weighted fee obtained for lending US equities is 25 b.p. per annum, but that only 7% of loan supply is borrowed. Although generally at a thin rate, lending improves the asset’s total performance and can offset custodial fees and administrative expenses. Lenders can also be motivated by the desire to borrow short-term money, and can do this by arranging transactions such as repurchase agreements or cash-collateralised securities lending. Securities lenders include long-term investors such as pension funds, insurance companies and mutual funds, but also banks and broker-dealers.

In addition to facilitating short-selling, reasons for borrowing stock include market makers borrowing securities to fill customer buy orders, exchange specialists borrowing to maintain price stability, and stockbrokers borrowing to cover a short position after failed settlement. Securities borrowing can also be related to hedging by the counterparties to contracts for differences, spread bets and swaps. The temporary transfer of ownership can also motivate securities borrowing. This includes dividend capture strategies such as ‘scrip dividend arbitrage’ and ‘dividend withholding tax arbitrage’[2].

Short-selling is the sale of securities that the seller does not own, or that the seller owns but chooses not to deliver. The short-seller must borrow securities in order to fulfil delivery obligations to the purchaser. ‘Naked short-selling’ occurs when the short-seller does not borrow, and so does not deliver, stock to the purchaser. In the United States securities markets, for example, Regulation SHO 2004 requires short-sellers to locate stock for borrowing, prior to selling a stock short, with the intention of prohibiting naked short-selling. Short-selling is particularly associated with the activities of arbitrageurs and hedge funds. Although some funds exclusively sell short, seeking to benefit from a decline in the value of a security, most short-selling is part of a broader trading strategy, designed to exploit perceived pricing anomalies between two or more securities. Examples of such trading strategies include capital structure arbitrage (see Yu, 2006), merger arbitrage (see Mitchell and Pulvino, 2001) and pairs trading (see Jacobs and Levy, 1993). Not all short-sales are driven by expectations of a price change; some sales are meant to stabilise prices. For instance, underwriters often sell short to reduce volatility in the price of public offering and buyback programs.

In this paper, we explore the frequency and distribution of borrowing, and the types of stocks that are preferred for borrowing. There is a growing body of evidence that short-sellers are well informed, and that highly-shorted stocks subsequently perform poorly. This suggests an initial trading strategy based on identifying and then shorting stocks which are already subject to high levels of shorting. Regular, timely short-sale information is not publicly available in most major markets. However, as short-sales generally require a loan of stock, stock lending data can serve as a proxy for short-selling. We test if successful trading strategies can be based on publicly available stock lending data, by studying the relationship between stock lending and abnormal returns. The remainder of this paper is organised as follows. Section 2 contains a literature review. Section 3 describes the data set being analysed. Section 4 describes the methods of analysis employed and the results obtained. Section 5 concludes.

2. Literature Review

Various studies examine the relationship between short-interest in a security (the value of shares shorted relative to market capitalization , or relative to average daily turnover) and abnormal return. These studies attempt to test the theory that short-sellers are, on average, well informed traders. Figlewski (1981), Brent et al. (1990), Figlewski and Webb (1993) and Woolridge and Dickinson (1994) find no evidence of a strong relation between short-interest and abnormal return. By contrast, Senchack and Starks (1993) investigate the market reaction to monthly short-sale announcements from both the New York and the American Stock Exchanges. They examine the wealth effects of short-interest announcements, and the relation between wealth effects and the degree of unexpected increases in short-interest. Using monthly common-stock short-interest figures from 1980 to 1986, they identify companies showing ‘unusually large’ increases in short interest. They find evidence that some significant negative price reaction occurs in an extended period around the announcement of a substantial increase in short-interest.

By focusing on firms with large short-interest only, Asquith and Muelbroek (1996) argue that the power of such tests can be improved. They find a strong and consistent relation between short-interest and excess returns. Shares with high levels of short interests perform significantly worse than comparable shares without high levels of short interest.

Only limited, monthly information on short interest has been publicly available in the USA prior to 2005, and this has limited the scope of research into this topic. Aitken et al. (1998) analyse information provided by the Australian Stock Exchange (ASX), covering intra-day information on short positions in listed ASX equities. Short trades were reported to the market soon after execution. The authors investigate the immediate market reaction to short sales and find a significantly negative abnormal return following short-sales. Abnormal returns are calculated by comparing short-sales to matched non-short sale trades. Ackert and Athanassakos (2004) study Canadian monthly data from 1991-1994 and from 1998-1999. For different levels of short interest (defined in their study as the ratio of the number of shares shorted to trading volume) they study subsequent abnormal returns and find that after high levels of short interest, poor performance persists into the future. In a study of NYSE proprietary system order data from 2000 – 2004, Boehmer et al. (2006) measure 20 day abnormal returns following from high levels of short-selling. They find that “As a group, [] short sellers are extremely well informed” and that “institutional, non-program short sales are the most informative.” However, the authors caution that the full information upon which their study was based, is not publically available, and that trading strategies based on their findings could not be formed in practice.

Dechow et al. (2001) examine the extent of short selling during the period 1976-1993, using public US data. The authors identify a strong relation between the trading strategies of short sellers and ratios of fundamentals to market prices, such as book to market ratios. They show that short-sellers target equities that have low fundamental to price ratios, and then unwind their positions as these ratios revert to the mean. They also show that short sellers refine their trading strategies in three ways: by avoiding equities where short-selling is expensive; by using information other than fundamental to price ratios that has predictive ability with respect to future returns; and by avoiding equities with low fundamental to price ratios where the low ratios are due to temporarily low fundamentals (as opposed to temporarily high prices). Their evidence suggests that “short sellers are sophisticated investors who play an important role in keeping the price of stocks in line with fundamentals.”

Jones and Lamont (2002) study the centralized stock loan market on the floor of the New York Stock Exchange (known as the ‘loan crowd’) from 1926-1933. They show that as stocks ‘enter the loan crowd’, they generally have high valuations and low subsequent returns. Size-adjusted returns are 1-2% lower for stocks that enter the loan crowd for the first time, and despite the high costs of borrowing and shorting these securities, it is profitable to short them.

Angel et al. (2003) study 3 months of short trades reported to NASDAQ during 2000. They assess the frequency of short selling for their sample of NASDAQ trades. They find that 2.36% of trades are short trades, and that 2.88% of shares traded were shorted, with the median less than the mean in both cases, suggesting that short sales tend to be concentrated in certain shares on a subset of days. Where the degree of short-selling is greater than average, significantly negative market-adjusted returns follow in the next three days. Short-selling is more common in actively traded companies and in shares with higher price volatility. The authors also find that short-selling is focused on shares that had exhibited greater than average prior price performance.

D’Avolio (2002) examines stock lending fees and shows that ‘growth’ and ‘low-momentum’ stocks are relatively more likely to be ‘special’ (i.e. have higher than average lending fees), leading to practical difficulties and costs in creating the long/ short factor portfolios found in the finance literature. Geczy et al. (2002) analyse a private database of US securities lending. They examine if investors can actually realize the returns of such long-short factor portfolios, including the book-to market strategy from DeBondt and Thaler (1987) and Fama and French (1993), and the price momentum strategy from Jegadeesh and Titman (1993). The authors find that the expected-return difference between unconstrained factor portfolios found in the literature and portfolios that investors could actually hold is significantly smaller than the unconstrained factor portfolios’ documented profitability. They argue that if short-selling problems explain the availability of factor portfolio returns to unskilled managers, then these short selling problems are not borrowing costs, but perhaps prohibitions on short-selling, or liquidity constraints, such as those cited in Shleifer and Vishny (1997).

The literature thus reveals a growing body of evidence that short-sellers are well informed, and that highly-shorted stocks subsequently perform poorly. As shorts are, on average, well-informed, the potential importance of information about short positions has become clear. However, an awareness of the borrowing costs, constraints and risks of shorting are also highlighted, and these diminish any trading opportunities that might arise.

3. Data

3.1 The Data Set

CREST, the organisation responsible for settlement of all trades on the London Stock Exchange, has been providing data on stock lending to interested parties since September 1st 2003. It provides daily information, published with a 5 day lag, on total shares on loan for each company in the FTSE 100 index and FTSE 250 (mid-cap) index. These companies are amongst the largest companies listed on the London Stock Exchange at any time. Data Explorers Ltd., a commercial organisation, re-packages the CREST data for sale to financial analysts and researchers. Their database provides: date, name of company, number and value of shares traded for each company for each day, daily end of day share price and market capitalisation for each company, number and value of shares on loan for each company for each day, and dividend record dates. We make use of the Data Explorers database for this study.

For FTSE 350 companies as of September 27th 2004, we obtain stock lending data from the inception of the database on September 1st 2003. Stocks that were newly listed during the period are included from their entry date to the index. Stocks that were deleted from the index during the previous year are not included in our dataset. Deletions are likely to have occurred due to take-overs, mergers, or companies performing poorly relative to their peers, such that they are no longer included in the FTSE 350 index. Two companies have dual class shares for the sample period. We retain the voting shares and delete the non-voting shares from the analysis. To calculate daily total returns for each share, we adjust returns for dividends on the ex-dividend dates[3].

3.2 Securities Lending as a Proxy for Short-Selling

Except when naked short-selling[4] occurs, short-sales will form a subset of shares on loan. However, several factors serve to weaken the relationship between shares on loan and shares sold short. These include borrowing stock to build up an inventory, so as to facilitate future short selling or to protect against the risk of loan recall. A visual inspection of the pattern of shares on loan over time reveals that for some companies, there is a clear cyclical pattern of increasing shares on loan around the time of the dividend record date. Some examples are given in Appendix 1. This pattern indicates dividend tax arbitrage, and is particularly pronounced for shares with high dividend yields. This lessens the usefulness of securities on loan data as a proxy for shares sold-short, but provides additional information on how market participants operate. We are able to reduce the impact of dividend tax arbitrage by excluding periods around the dividend record for each company. We can eliminate the dividend tax arbitrage effect by considering only those companies that pay no dividend, although this results in a smaller sample of data. For the period considered, 35 of the FTSE 350 companies paid no dividends.

Traders and investors can achieve positions that are economically equivalent to short-sales through the use of contracts for differences, swaps, single stock futures and options. To the extent that the counterparties to these trades hedge their positions via short-sales, the securities on loan database will reflect this activity.

There is some possibility that CREST data is incomplete given that unethical lenders may for some reasons avoid reporting certain lending positions. While this is possible, it is illegal and we assume that in a developed and well-regulated market such as the United Kingdom, the effects are minimal, isolated and beyond the scope of considerable influence on this study.

Although stock lending is an imperfect proxy for short-selling, information on stock lending can offer insights into the operation of markets that cannot be obtained through observation of short-selling alone. For example, an understanding of borrowing for dividend-arbitrage and for the capture of voting rights is only revealed through analysis of stock lending data. These topics are explored further in Sections 4.2 and 4.8.

4. Method and Results

4.1 Frequency and Distribution of Stock Lending

We assess the frequency and distribution of stock lending among UK FTSE 100 and FTSE 250 companies. To reduce the effect of dividend tax arbitrage or scrip dividend capture, we repeat the examination for FTSE 100 companies, but exclude the periods one calendar week before and after the record date of each dividend for each company [justification for this time period?]. Table 1 provides the average number of shares on loan at any time:

Table 1. Mean Daily ‘Percentage on Loan’ for FTSE 100 and FTSE 250 Index Constituent Companies, September 1st 2003 to September 27th 2004

Statistic Max. 95% 90% 75% 50% 25% 10% 5% Min. Mean

Value (Median) Value

A. Distribution including periods around dividend record dates

FTSE 100 15.08 9.09 7.10 4.91 3.24 2.32 1.51 1.13 0.76 3.90 FTSE 250 12.61 6.23 4.71 2.90 1.73 1.03 0.63 0.49 0.14 2.33

B. Distribution excluding periods around dividend record dates

FTSE 100 15.00 8.94 6.97 4.69 3.04 2.08 1.43 1.13 0.68 3.69

Table 1 shows a lower level of stock lending for FTSE 250 index constituent companies than for FTSE 100 companies at all points of the distribution. Comparing FTSE 100 distributions including and excluding periods around the dividend record dates (panel A compared to panel B above) indicates increased stock lending activity around divided record dates.

Asquith and Moelbroek (1996) report that most firms have less than 0.5% of outstanding shares shorted, but that a few firms have large positions (greater than 5% of outstanding shares) shorted. Two reasons for the greater levels of shares on loan shown above are that short-interest will, to all intents and purposes, be a subset of shares on loan figures; and that there has been substantial growth in hedge fund assets since Asquith and Moelbroek’s study. For example, Hedge Fund Research estimates that total hedge fund assets have grown from under USD 200 million in 1995 to over USD 1 trillion by 2005. Assets available for use in long/ short equity strategies have out-grown equity market capitalisation during this period. During the period of study, the proportion of shares on loan varied daily. Weighting equally each company in the FTSE 350, we calculate the mean proportion of shares on loan for each day, as shown in Figure 1.

Figure 1: Equally-Weighted Mean Proportion of FTSE 350 Shares on Loan across the Study Period

[pic]

The mean proportion of shares on loan grew from approximately 2.5% at the start of the period to over 3.5% at the end of the period. This might reflect the market conditions prevailing during the time of the study: share prices were generally rising as markets continued to recover from the ‘bear market’ of 2000-2002, and these higher share prices might have attracted more short-sellers. Alternatively, it might reflect an increased tendency on the part of investors to practice short-selling, consistent with the observation of growing short interest by Dechow et al. (2001). The reductions in proportion of shares shorted that can be observed in December 2003, April 2004 and August 2004 are most likely associated with the passing of clusters of dividend record dates. By instead weighting by market capitalization, the daily mean proportion of shares on loan becomes generally higher, reflecting the greater proportion of shares on loan amongst the larger, FTSE 100 companies.

4.2 Dividend Capture Strategies

Having calculated the frequency of securities lending for FTSE 100 companies including and excluding a period around the dividend record date, we use the difference between the two figures for each company to assess the extent of dividend tax arbitrage and scrip dividend capture as a proportion of securities lending. We also wish to test if the extent of dividend capture strategies increases with the size of the dividend, and if it is related to the cost and ease of securities borrowing. The database does not provide data on the costs and ease of borrowing, but we can proxy cost and ease of large-scale borrowing through market capitalisation, in the belief that shares in larger companies will generally be more plentiful, and so cheaper and easier to borrow. We can also observe from the database that the value of borrowing in those companies with the greatest market capitalization is large in money terms, thus facilitating large-scale dividend capture. Thus we form equation 1 below:

Yi = αDivYldi + βMktCapi + εi 1)

Where:

Yi is the mean daily proportion of shares borrowed in company i over the full period, minus the mean daily proportion of shares borrowed in company i excluding one calendar week around the dividend record date.

DivYldi is dividend yield for company i, calculated as the mean of the annualized dividend yield reported on Datastream for each day in the sample period.

MktCapi is the market capitalisation of company i, calculated as the mean over the full sample period of the daily market capitalisation.

εi is an error term

The expected signs of the coefficients α and β are both positive. Four FTSE 100 companies paying no dividends were excluded from this analysis of dividend capture. Table 2 summarises the results.

Table 2. Regression Results for Dividend Capture Ratio

Number of Observations 96

Adjusted R2 0.345

F Statistic of Regression 25.98

Significance of F 1.08E-09

Coefficients α β γ

Coefficient Estimate 0.0007 1.09E-14 -0.0005

t-statistic 7.12 1.19 -1.12

Significance of t-statistic 2.28E-10 0.238 0.265

We find very highly statistically significant evidence of a positive relationship between dividend yield and tax arbitrage activity, but do not find statistically significant evidence of a relationship between market capitalization and dividend tax arbitrage.

4.3 Trading Activity and Securities Lending

We investigate whether more actively traded shares are subject to a greater degree of borrowing than other shares. Short-sellers are concerned with the cost of trading, and more liquid stocks are likely to have lower trading costs. They are also concerned at the risk of a ‘short squeeze’ – being forced (through loan recall or a collateral-call) to cover their short sale at a time of an increasing share price. Forced buying can have meaningful market impact, and this can be exploited by ‘predatory traders’ (see Brunermeier and Pederson, 2004). By focusing on shares with greater liquidity, short-sellers can reduce their risk of exposure to predatory trading.

Miller (1977) theorises that when short-selling is constrained and investors’ expectations are heterogeneous, the most optimistic investors will own all the shares in a company, and their estimates of the fair-value of a stock will be above that of the average investor. Stocks for which investor opinions exhibit the greatest differences will tend to be the most over-valued. Trading activity can be a proxy for differences in opinion amongst market participants, and thus heavily traded stocks might be most over-valued in the market when short-selling is constrained. This could partly explain why short-sellers target stocks with high levels of trading activity.

In their study of the relationship between liquidity and shorting, Angel et al. (2003) use the total number of shares traded in a three month interval during 2000 as their measure of liquidity. However, there is a risk that their measure of liquidity is more a proxy for market capitalisation. To remove this potential bias, we obtain a measure of the proportion of a company’s market capitalisation traded each day, on average. For each company we multiply end of day share price by the number of shares traded during the day, and divide by end of day market capitalisation. We then determine the mean of these daily figures for each company, utilising data from the full period under review. We rank the FTSE 100 companies into quintiles of liquidity. We calculate the mean and median of the daily average percentage of shares on loan for each quintile, using the full sample and then the sample excluding the 11 day dividend capture period. Results are shown in Table 3 below.

Table 3. Percentage of Shares on Loan for FTSE 100 Companies by Quintile based on Trading Activity

Statistic 20 Most Stocks Stocks Stocks Stocks

Actively Traded ranked 21-40 ranked 41-60 ranked 61-80 ranked 81-102

A. Including periods around dividend record dates

Mean 5.81 3.83 3.47 3.86 2.67

Median 4.68 3.45 3.17 3.22 2.03

B. Excluding periods around dividend record dates

Mean 5.63 3.72 3.22 3.53 2.49

Median 4.39 3.29 3.04 2.90 1.84

Unlike Angel et al. (2003), we do not find a monotonically declining figure for daily percentage of short trades as trading activity declined. Nevertheless, the percentage of shares on loan for the least actively traded category is less than half that for the most actively traded category in each case. Furthermore, the most actively traded category shows the greatest percentage of shares on loan in each case, and the least actively traded category shows the lowest percentage of shares on loan in each case. This provides support for the view that securities borrowers (and short-sellers) are more interested in highly liquid stocks.

Table 4 shows the relationship between the proportion of shares on loan and trading activity for the FTSE 250 stocks.

Table 4. Percentage of Shares on Loan for FTSE 250 Companies by Quintile based on Trading Activity

Statistic 50 Most Stocks Stocks Stocks Stocks ranked

Actively Traded ranked 51-100 ranked 101-150 ranked 151-200 201-251

Mean 4.25 2.81 1.96 1.53 1.12

Median 3.80 2.55 1.52 1.28 0.83

Table 4 shows monotonically declining daily percentages of FTSE 250 shares on loan as trading activity declines. Table 5 compares liquidity for FTSE 100 companies to that for FTSE 250 companies. It shows that for FTSE 250 companies, mean liquidity is lower than for FTSE 100 companies, but liquidity levels are more dispersed: the most actively traded FTSE 250 companies see a greater proportion of market capitalisation traded than do the most actively traded FTSE 100 companies, but the least actively traded FTSE 250 companies see a much smaller proportion of market capitalisation traded than do the least actively traded FTSE 100 companies.

Table 5. Mean Daily ‘Percentage of Market Capitalisation Traded’ for FTSE 100 and FTSE 250 Index Constituent Companies, September 1st 2003 to September 27th 2004

Statistic Max. 95% 90% 75% 50% 25% 10% 5% Min. Mean

Value (Median) Value

Distribution for:

FTSE 100 1.49 1.09 1.00 0.82 0.65 0.54 0.40 0.31 0.22 0.69

FTSE 250 1.75 1.20 0.87 0.62 0.41 0.26 0.15 0.11 0.03 0.49

For FTSE 250 companies, the greater dispersion in liquidity reveals a clearer positive relationship between trading activity and shares on loan.

4.4 Volatility and Securities Lending

McDonald and Baron (1973) suggest that “the greater the variability of returns of the security to be sold short, the greater are the advantages for hedging.” This leads them to expect relatively greater short interest in stocks with more variable returns. They also suggest that speculators, gambling on a general decline in the stock market, are likely to prefer short positions in stocks with greater volatility and variability, as these would produce greater returns in a declining market. The authors regress short interest on stock betas, and find that security risk is significant (at the 5% level) in explaining the mean short interest for an individual stock. This supports their hypothesis that riskier stocks have relatively higher short positions. The authors also hypothesised that more volatile stocks have more variable short positions, as measured by the variance of short interest divided by the mean short position over a five year period. Their regression analysis found a statistically significant relationship between risk in a stock and variability in its short positions. Angel et al. (2003) find a strong indication that short-selling activity is highest for the most volatile equities, and that it declines monotonically with decreasing volatility. We calculate the standard deviation of daily total returns for each company in the FTSE 100 and FTSE 250 index over the full sample period. We then rank the securities into quintiles based on standard deviation of daily returns, and compare to the mean percentage of shares on loan during the period. We also study the relationship between volatility of returns for FTSE 100 companies and mean percentage on loan, excluding percentage of securities on loan for the 11 day period around the dividend record date, so as to reduce the impact of dividend tax arbitrage. Table 6 summarises the results for the FTSE 100 companies and Table 7 summarises the results for FTSE 250 companies.

Table 6. Percentage of Shares on Loan for FTSE 100 Companies by Quintile based on Volatility

Statistic 20 Least Shares Shares Shares 20 Most

Volatile ranked 21-40 ranked 41-60 ranked 61-80 Volatile

A. Including periods around dividend record dates

Mean 4.66 3.60 2.88 4.08 4.49

Median 4.41 3.33 2.43 2.89 3.87

S.D. 2.31 1.81 1.37 3.29 2.99

B. Excluding 11-day periods around dividend record dates

Mean 4.34 3.32 2.73 3.88 4.38

Median 4.12 2.97 2.31 2.49 3.56

S.D. 2.36 1.79 1.37 3.31 3.01

Table 7. Percentage of Shares on Loan for FTSE 250 Companies by Quintile based on Volatility

Statistic 50 Most Shares Shares Shares Shares ranked

Volatile ranked 51-101 ranked 101-150 ranked 151-200 201-251

Mean 3.41 2.41 2.22 1.84 1.78

Median 2.93 1.93 1.82 1.44 1.24

For FTSE 250 companies, securities lending declines monotonically with volatility, as expected. However, for FTSE 100 companies, the fifth quintile (lowest volatility) shows a surprisingly strong (in fact the highest) percentage of securities on loan. The same pattern is seen whether an 11 day period around the dividend record date is included or excluded, suggesting that dividend capture does not explain this pattern. Closer examination reveals that certain industry sectors (especially utilities, banks and brewers) dominate the list of securities with low volatility but a high percentage of securities on loan. These sectors tend to have above-average yields, with earnings that have greater than average interest-rate sensitivity. UK short-term interest rates were rising, and the yield curve was flattening, around the period of the study. This might explain why short-sellers have targeted these interest-rate sensitive sectors, despite their low volatility.

[Cross-sectional Tests section here?]

4.5 Weekend Effect

Chen and Singal (2003) argue that many short-sellers close their positions on Fridays, ahead of the weekend trading halt, and re-establish their short positions on Mondays, contributing to the so-called Weekend Effect (or pattern of superior returns on Fridays relative to Mondays). They find a significantly stronger ‘weekend effect’ for stocks with greater short-interest relative to shares outstanding. However, Angel et al. (2003) do not find any ‘day of the week’ effect associated with short-selling for their sample. We investigate our database for patterns that might allow profitable trading strategies.

A limitation associated with our database is that when short-sale positions are closed, the borrowed securities might be retained so as to re-establish the short position in the near future, rather then returned to the borrower. Also, securities can be borrowed in anticipation of short selling. As a result, the loan data thus might not fully reveal the closure and re-establishment of short positions.

We calculate the mean number of shares on loan for each day of the week for the full dataset. We do the same for stocks showing the highest proportion of shares on loan, by taking the highest decile of stocks by proportion of shares on loan. Three of these 35 stocks were introduced to the dataset part way through the sample period – we eliminate these and calculate the daily means for the 32 remaining stocks. We also calculate daily means for the 35 non-dividend paying stocks only, so as to eliminate any noise due to dividend tax arbitrage. Further, we take the highest quintile from these 35 non-dividend paying stocks. Two of these 7 stocks were introduced to the dataset part way through the sample period – we eliminate these and calculate the daily means for the 5 remaining stocks. Results are shown in Table 8 below:

Table 8. Daily Proportion of Shares on Loan by Day of the Week

Statistic Monday Tuesday Wednesday Thursday Friday

A. Full Dataset (352 companies)

Mean 2.78 2.78 2.78 2.80 2.81

B. Top Decile by Proportion of Shares on Loan (32 companies)

Mean 7.87 7.86 7.85 7.89 7.90

C. Non-Dividend Paying Stocks Only (35 companies)

Mean 3.28 3.27 3.27 3.27 3.28

D. Top Quintile by Proportion of Shares on Loan amongst Non-Dividend Paying Stocks (5 companies)

Mean 7.54 7.55 7.54 7.53 7.54

None of the panels show large variation by day of the week. Based on Chen and Singal (2003), and our understanding of how dividend tax arbitrage might influence borrowing for dividend paying stocks, we would expect Panel D to be most likely to show any Weekend Effect. However, we find no such pattern. This need not imply that no such pattern exists amongst short-sellers. Instead, it merely shows that there is no supporting evidence to emerge from this stock lending data.

4.6 Abnormal Returns Associated with High and/or Rising Levels of Short Interest

Aitken et al. (1998) highlight the immediate negative impact of information about short sales. In measuring the subsequent performance of shares with different levels of short interest, Asquith and Muelbroek (1996) focus on shares with high levels of short interest. Senchack and Starks (1993) focus on shares with unexpected increases in short positions. We investigate the securities lending data to identify companies with a high proportion of shares on loan, and also those showing unexpected increases in shares on loan.

Noting that the top quartile in the distribution of proportion of shares on loan is approximately 5%, we take this as our definition of a ‘high proportion’ of shares on loan. We identify all days on which 5% or more of a company’s shares are on loan, and measure abnormal returns around those days. We break the observations into deciles based on the proportion of shares on loan, to study differences that might arise with different proportions of shares on loan. Dechow et al. (2001) calculate abnormal returns by adjusting for the equal-weighted return of all stocks in the markets they study. However, we choose instead to make a risk-adjustment, as we believe that an awareness of risk is important for leveraged investors such as hedge funds. Shleifer and Vishny (1997) describe how short-term arbitrage losses can lead to client redemptions. Abreu and Brunnermeier (2004) show how timing is important to arbitrageurs. Consequently, arbitrageurs tend to focus on short-term risk factors, and we deduce that sensitivity (beta) to stock market returns dominates the non-specific risk factors facing short-sellers. Accordingly, we calculate the abnormal return for each stock on each day by using the Market Model to calculate expected return, and subtracting this from the actual return. Noting that the stock lending data is released publicly with a five day lag, we include five days as one of the time periods to be assessed. We study same day performance, one day performance and 5 day performance – this latter to take account of the 5 day lag in reporting the stock lending data. We also calculate 5 day cumulative returns and one month (22 day) cumulative returns. Table 9 below summarises the results.

Table 9. Abnormal Returns of FTSE 100 Stocks, by Decile of Proportion of Shares on Loan

Decile Lowest 2 3 4 5 6 7 8 9 Highest

A. Cumulative Abnormal Returns One Month Before the Observation Date (6888 observations)

Mean 0.26 4.30 1.02 1.60 0.59 0.83 -0.41 0.59 1.55 0.78

Median 0.56 1.62 1.07 1.12 0.51 0.45 0.22 0.85 1.41 0.39

S.D. 6.44 20.08 7.31 8.76 7.58 6.07 5.35 6.48 10.92 5.36

B. Cumulative Abnormal Returns 5 Day Before the Observation Date (7263 observations)

Mean 0.07 0.09 0.18 0.38 -0.13 0.20 0.03 -0.23 0.97 0.16

Median 0.11 -0.04 0.12 0.07 -0.31 0.21 -0.15 -0.25 0.21 0.09

S.D. 2.65 3.26 2.78 5.90 3.28 3.14 2.86 2.91 10.45 2.83

C. Cumulative Abnormal Returns One Day Before the Observation Date (7338 observations)

Mean 0.02 0.02 0.00 0.22 -0.08 0.10 0.05 -0.10 0.04 -0.01

Median 0.00 0.00 -0.02 0.01 -0.11 0.03 -0.01 -0.13 0.03 -0.07

S.D. 1.34 1.31 1.48 5.45 1.34 1.44 1.25 1.30 1.18 1.27

D. Abnormal Returns On Day of Observation (7357 observations)

Mean 0.03 0.05 -0.01 0.20 -0.08 0.04 0.06 -0.11 0.02 -0.03

Median 0.04 0.04 -0.03 -0.02 -0.11 0.01 0.02 -0.15 -0.01 -0.08

S.D. 1.17 1.21 1.47 5.47 1.28 1.37 1.30 1.32 1.14 1.25

E. Abnormal Returns One Day After the Observation Date

Mean 0.00 0.08 -0.04 0.17 -0.05 0.02 0.00 -0.07 0.03 -0.02

Median 0.01 0.06 -0.05 -0.03 -0.11 -0.04 0.02 -0.12 0.00 -0.06

S.D. 1.18 1.19 1.47 5.47 1.26 1.39 1.31 1.31 1.12 1.24

F. Abnormal Returns 5 Days After the Observation Date

Mean -0.01 0.05 -0.02 0.02 0.16 -0.02 -0.02 -0.09 -0.01 0.00

Median -0.04 0.04 -0.05 -0.04 -0.07 -0.07 -0.01 -0.10 -0.04 -0.06

S.D. 1.19 1.16 1.48 1.27 5.52 1.42 1.29 1.28 1.12 1.22

G. Cumulative Abnormal Returns 5 Days After the Observation Date

Mean -0.34 -0.57 0.28 2.27 0.77 -0.55 0.00 -0.68 -0.94 -0.49

Median -0.07 -0.51 0.20 -0.20 0.07 -0.34 0.17 -0.47 -0.74 -0.56

S.D. 5.54 5.79 5.42 17.73 12.01 4.97 4.63 4.35 4.16 5.08

H. Cumulative Abnormal Returns One month After the Observation Date

Mean -0.35 -0.51 0.34 2.59 0.93 -0.25 0.12 -0.56 -1.01 -0.24

Median -0.03 -0.32 0.36 0.04 0.43 -0.29 0.30 -0.35 -0.71 -0.50

S.D. 5.73 6.01 5.58 18.49 12.47 4.99 4.62 4.42 4.43 5.06

These results show a suggestion of positive abnormal returns in the month prior to each observation date, indicating that the highest levels of borrowing are found in stocks that have performed well prior to borrowing. For the panels that consider more than one day of cumulative performance and thus involve overlaps, Newey-West t-statistics can be used to test the significance of any differences between the different deciles. Angel et al. (2003) find a similar pattern in their study of short-selling on NASDAQ in 2000. The post-borrowing results show no apparent pattern. These results compare with other studies (e.g. Jones and Lamont, 2002, Angel et al., 2003, Ackert and Athaanassakos, 2004) that find evidence of negative abnormal returns following from high levels of short-selling.

We also study the results excluding a 10 day window around the dividend record date, and find a similar pattern to that in Table 9 above. As a robustness check, we repeat the above analysis using market relative performance, and then nominal returns, and again find an absence of consistent patterns or statistically significant evidence of negative abnormal returns either before or after high levels of borrowing (the results are not illustrated but are available from the authors on request).

In Table 10 we identify situations where there is an increase in a company’s shares on loan relative to the previous day – this provides a proxy for an unexpected increase in short-selling – and find no apparent pattern.

Table 10. Next Day Abnormal Returns of FTSE 100 Stocks, by Quintile of Percentage Change in Shares on Loan

Quintile Lowest 2 3 4 Highest

A. Increasing Shares on Loan, Including Dividend Periods: Next Day Abnormal Returns (12956 observations)

Mean 0.00 0.01 0.04 -0.04 0.04

Median -0.05 -0.01 -0.01 -0.07 0.01

S.D. 1.26 1.25 1.38 1.26 1.30

B. Increasing Shares on Loan, Excluding Dividend Periods: Next Day Abnormal Returns (10668 observations)

Mean 0.00 0.01 0.06 -0.03 -0.03

Median -0.04 0.01 -0.01 -0.04 -0.05

S.D. 1.27 1.22 1.32 1.27 1.25

C. Decreasing Shares on Loan, Including Dividend Periods: Next Day Abnormal Returns (12972 observations)

Mean -0.05 0.05 -0.02 0.02 -0.01

Median -0.06 -0.04 -0.07 -0.04 -0.05

S.D. 1.22 1.29 1.44 1.29 1.51

D. Decreasing Shares on Loan, Excluding Dividend Periods: Next Day Abnormal Returns (11116 observations)

Mean -0.05 0.04 -0.02 0.03 0.01

Median -0.06 -0.04 -0.06 0.01 -0.04

S.D. 1.19 1.29 1.43 1.27 1.60

In Table 11 we analyse the non-dividend paying stocks only, so as to entirely eliminate the effect of dividend tax arbitrage from the study. We find lower subsequent abnormal returns for the most-shorted decile relative to the least-shorted decile for all panels. This includes panel E, which captures the daily release of 5-day-lagged stock lending data. This suggests that dividend tax arbitrage was obfuscating the underlying picture in the previous studies, despite our attempt to minimise this effect by excluding periods around the dividend record date. [INSERT Newey-West t-statistics].

Table 11. Abnormal Returns of FTSE 350 Non-Dividend Paying Stocks, by Quintile of Proportion of Shares on Loan for Shares on Loan in Excess of 5%

Quintile Lowest 2 3 4 Highest

A. Cumulative Abnormal Returns One Month Before the Observation Date (1453 observations)

Mean -0.39 -1.16 3.67 0.64 -2.48

Median 1.12 -0.71 2.73 0.53 -2.32

S.D. 15.17 10.97 11.30 10.44 13.81

B. Cumulative Abnormal Returns 5 Day Before the Observation Date (1519 observations)

Mean 0.59 0.05 0.02 0.19 -0.64

Median 1.02 0.14 -0.44 -0.22 -0.52

S.D. 5.62 5.83 5.39 5.51 7.35

C. Abnormal Returns One Day Before the Observation Date (1531 observations)

Mean 0.08 0.13 -0.10 0.03 -0.11

Median 0.04 0.06 -0.13 -0.09 -0.11

S.D. 2.35 2.61 2.35 2.39 3.08

D. Abnormal Returns on Day of Observation (1534 observations)

Mean -0.01 0.03 0.19 -0.13 -0.08

Median 0.03 -0.06 -0.02 -0.15 -0.09

S.D. 2.24 2.65 2.67 2.14 2.99

E. Abnormal Returns One Day After the Observation Date (1529 observations)

Mean 0.09 0.08 0.16 -0.14 -0.08

Median 0.15 -0.08 -0.03 -0.16 -0.04

S.D. 2.42 2.68 2.61 2.34 2.78

F. Abnormal Returns 5 Days After the Observation Date (1498 observations)

Mean 0.17 -0.15 0.41 -0.19 -0.10

Median 0.11 -0.17 0.17 -0.21 0.02

S.D. 2.70 2.47 2.76 2.30 2.70

G. Cumulative Abnormal Returns 5 Days After the Observation Date (1498 observations)

Mean 0.83 -0.47 1.28 -0.64 -0.38

Median 0.72 -0.69 0.55 -0.61 -0.24

S.D. 5.68 5.52 6.25 5.55 6.23

H. Cumulative Abnormal Returns One month After the Observation Date (1386 observations)

Mean 3.74 0.58 3.67 -3.81 -1.70

Median 1.70 0.76 3.03 -1.35 -2.59

S.D. 12.14 11.06 11.31 11.59 10.08

These results should be of interest to lenders of securities, and to intermediaries who promote securities lending. Potential lenders know that they can earn fees by lending, but are often unsure of the impact that lending might have on the returns of the securities that they lend. Whereas the typical lending fee for FTSE 100 stocks over the time period covered is 6-200b.p. per annum (Faulkner, 2004), the analysis in Table 9 shows no evidence of negative abnormal returns in the one month period following high levels of borrowing. This suggests that lending securities to earn fees is, in aggregate, a ‘safe’ activity. This result is somewhat unexpected, in light of the body of empirical evidence suggesting that negative abnormal returns follow high levels of short-selling. The reason for this is likely to lie in the diverse reasons for borrowing: for short-selling, dividend tax arbitrage, voting purposes, financing purposes etc. Once dividend tax arbitrage is eliminated, as in Table 11, we find the expected result.

[Cross-sectional Tests section here?]

4.7 Fundamental Ratios and Securities on Loan

Analysing data from 1976-1993, Dechow et al. (2001) identify a strong relation between the trading strategies of short-sellers and ratios of fundamentals to market prices. They show that short-sellers target equities that have low fundamental to price ratios, and then unwind their positions as these ratios revert to the mean. We wish to test if this relationship still holds. From discussions with investors, we suspect that hedge fund strategies have grown more diverse in nature since the Dechow et al. study.

We obtain published book values for each company in the FTSE 100 index around the period of our sample data. Noting that book values are published some time after the financial year end of a company (typically within about 3 months, but up to six months on occasions in the UK), we run three tests: first, assuming perfect knowledge of book values from the financial year end (as if an analyst was able to estimate book value accurately); second, assuming that book values are available from 3 months after the financial year end; and third, assuming that book values are available from 6 months after the financial year end. For each of these three tests, we divide each day’s (closing?) share price by the appropriate measure of book value per share, to obtain a ratio of price to fundamentals that is commonly used by investment analysts. For each test, we take all companies with positive book values, and assign those companies to quintiles by ratio of price to fundamentals. Within each quintile, we calculate the mean, median and standard deviation of the proportion of shares on loan for all companies in that quintile. We create a separate category for all companies that have negative book values, and again calculate mean, median and standard deviation. The results are shown in Table 12 below:

Table 12. Daily Proportion of Shares on Loan, by Quintile of Ratio of Price to Fundamentals

Price to Lowest 2 3 4 Highest -ve

Book book

Value values

Quintile

A. Assuming No Delay in Reporting Book Value (26,364 observations with positive book values; 608 observations with negative book values)

Mean 4.70 3.77 3.81 4.29 3.24 4.19

Median 3.58 2.48 2.36 2.77 2.25 3.88

S.D. 3.70 3.37 3.34 4.00 2.74 2.45

B. Assuming 3 Month Delay in Reporting Book Value (26,434 observations with positive book values; 590observations with negative book values)

Mean 4.80 3.61 3.85 4.26 3.28 4.11

Median 3.63 2.36 2.46 2.78 2.28 3.89

S.D. 3.79 3.10 3.37 4.00 2.83 2.55

C. Assuming 6 Month Delay in Reporting Book Value (26,365 observations with positive book values; 717 observations with negative book values)

Mean 4.81 3.70 3.83 4.16 3.24 4.39 Median 3.62 2.49 2.39 2.53 2.28 3.81 S.D. 3.80 3.11 3.39 4.10 2.60 2.91

These results are not consistent with the pattern identified by Dechow et al. (2001). Companies with the highest ratios of share price to (per share) book value do not experience greater proportions of securities on loan. This suggests either a greater diversity of hedge fund strategies since the Dechow et al. study, or obfuscating factors in the stock lending database.

We repeat this for non-dividend paying stocks, to eliminate the effect of dividend tax arbitrage, and show the results in Table 13. The quintile with the highest ratio of price to book value shows the highest level of shorting, but otherwise the pattern is only weakly supportive of the Dechow et al. study

Table 13. Daily Proportion of Shares on Loan for Non-Dividend Paying Stocks, by Quintile of Ratio of Price to Fundamentals

Price to Lowest 2 3 4 Highest -ve

Book book

Value values

Quintile

A. Assuming No Delay in Reporting Book Value (6,350 observations with positive book values; 10 observations with negative book values)

Mean 3.69 3.42 2.41 2.17 5.04 3.44

Median 3.29 2.68 1.89 1.69 5.58 3.80

S.D. 2.49 2.74 1.84 1.49 3.26 1.51

B. Assuming 3 Month Delay in Reporting Book Value (6,350 observations with positive book values; 10 observations with negative book values)

Mean 3.60 3.49 2.48 2.06 5.09 3.44

Median 3.17 2.74 1.94 1.62 5.58 3.80

S.D. 2.46 2.75 1.89 1.43 3.22 1.51

C. Assuming 6 Month Delay in Reporting Book Value (6,330 observations with positive book values; 10 observations with negative book values)

Mean 3.38 3.67 2.54 2.03 5.13 3.44

Median 2.85 2.94 1.98 1.61 5.58 3.80

S.D. 2.34 2.83 1.96 1.44 3.19 1.51

4.8 Short Squeezes

One of the unique risks involved in short-selling is the risk of a short squeeze. In a short squeeze, one or more market participants buys shares in a company, pushing the share price upwards. At the same time, shares on loan in that company are recalled. The short-seller must thus cover his/ her short position at a time of a rising share price. This short covering will have a market impact that drives the share price higher, allowing the original buyers to then exit from their long positions at the now higher market price. Such ‘predatory trading’ has been illustrated via equilibrium pricing models, such as that developed by Brunnermeier and Pedersen (2005). Allen and Gale (1992) describe trade-based securities market manipulation, which they refer to as ‘pump and dump’ trading. Khwaja and Mian (2005) study a 32 month period of data for the Karachi Stock Exchange (KSE) data and find evidence of trade-based ‘pump and dump’ price manipulation by brokers. We seek to identify patterns that are consistent with such ‘pump and dump’ activity. Wen he manipulator’s trades cannot be observed, Zhou and Mei (2003) argue that “excessive trading volume and price movements without news on fundamentals” can assist in distinguishing manipulative trading from other forms of trading. Aggarwal and Wu (2003) study US SEC actions in stock manipulation cases and suggest that manipulation is associated with less liquid stocks and increased trading volume and volatilty. Prices trend throughout the manipulation period and reverse in the post-manipulation period.

As predatory trading relies on the market impact of the forced trader, we focus on FTSE 250 stocks, where trading activity has been shown (Table 5) to be on average more limited than for FTSE 100 stocks. We test if any stocks appear to have been subject to predatory trading, or short squeezes. Mahoney (1999) suggested that a large abnormal return in the absence of a news announcement, followed by a large negative return of similar magnitude (as investors learn that the trading was not information-based) is indicative of manipulation. We seek cases where a share price has risen over some time period, followed by a reduction in securities on loan (as a proxy for short-covering), followed by a reversion (upwards) in the stock price. We choose to identify situations where the price of a stock has risen in absolute terms by more than 5% over three consecutive days. This is not equivalent to the abnormal return suggested by Mahoney (1999). However, for an unhedged short-seller, it is absolute price rises (rather than relative or abnormal price rises) that cause losses. Rather than a sharp rise in price on a single day, which could be dismissed as a ‘clumsy trade’, we focus on three consecutive days of rising share prices, suggesting a building price momentum in the stock. Where prices have risen according to our defined criteria, we identify situations where shares on loan have also dropped by over 10% on the next day. We next strip out situations where this might be due to the end of a dividend arbitrage strategy, by comparing the timing of a drop in securities on loan to the dividend record date. We are left with occasions where securities on loan falls in response to a series of price rises. This runs counter to the general pattern of an increase in securities on loan following share price appreciation, as found by Angel et al. (2003) and in Section 4.6 of this study. Finally, we identify situations where all the above criteria hold, and where the share price falls in the five day period subsequent to the drop in securities on loan. We find 50 examples over the period from September 1st 2003 to September 27th, 2004 that fit the above pattern.

We now use the Perfect Analysis database to study regulatory news service announcements around the time of these patterns. We pay particular attention to earnings announcements and take-over bids. We find 9 examples where the patterns coincide with material, price sensitive regulatory news-flow, such as a corporate results statement or the announcement of new contracts. We find 6 examples where there are no regulatory news announcements within 7 calendar days (on either side) of the apparent short squeeze. In 35 cases, there were regulatory announcements related to share transactions but not company performance, including disclosure of declarable stakes in the company, share buybacks or disclosure of directors trading in shares of the company.

For the six examples with no associated news flow, we calculate summary statistics, to understand better the anatomy of the pattern and the characteristics of the underlying stocks. These statistics are shown in Table 12 below:

Table 12. Summary Statistics for Six Apparent Short Squeezes

Mean Median Standard Deviation

‘Pump’ Phase % Price Increase 7.1 6.1 2.1

‘Squeeze’ Day % Fall in Shares on Loan -20.8 -13.9 17.3

‘Dump’ Phase % Price Fall -3.5 -3.3 2.1

Squeeze Day Turnover Ratio1 2.20 .93 3.22

11 Day Window Turnover Ratio2 1.39 1.02 1.05

Market Capitalization Ratio3 0.50 .49 .11

Trading Activity Ratio4 .76 .66 .48

1Value of shares traded on the short squeeze day/ Mean daily value of shares traded in that company during the sample study period; 2 Mean daily value of shares traded during the 11 day period centred on the short squeeze day/ Mean daily value of shares traded in that company during the sample study period; 3 Market capitalization of the company at the short squeeze date/ Mean market capitalization of all companies in the FTSE 250 section of the database at that date; 4 Mean daily proportion of shares traded in the short squeeze company/ Mean daily proportion of shares traded in all companies in the FTSE 250 section of the database.

We observe from Table 12 that the apparent short squeezes are associated with companies with below average market capitalization; and that there is a suggestion that the stocks concerned have lower liquidity than their peers. There is also a suggestion that daily turnover around the time of the short squeezes rises above the mean daily turnover seen in the stocks concerned. Predatory traders can target their efforts on situations where similar patterns in stock lending data emerge. What is not known is the full distribution of returns that such traders might receive, in light of the possibility of failed attempts at squeezing short-sellers out of the market.

5. Summary and Conclusion

This paper undertakes an initial investigation into a new, commercial database of stock lending. We reveal the distribution of the proportion of shares on loan on any day and show that securities lending is generally greater in securities with higher volatility. We analyse securities lending around dividend record dates, and show that dividend capture strategies are, as expected, positively associated with high dividend yields, but not associated with market capitalization. We show support for the belief that poor stock liquidity acts as a constraint to short-selling and, consequently, to stock borrowing.

Whereas Dechow et al. (2001) find that short-sellers target stocks with high ratios of price to fundamentals, we find little support for the relationship between high levels of shorting and a high ratio of price to book value. This could be due to obfuscation in the dataset due to non-short-selling motivated borrowing, or from the employment of a broader range of short-selling strategies since the Dechow et al. study was conducted. For non-dividend paying stocks, we find weakly supportive results, suggesting that both effects apply.

We show that there is no evidence to support the notion that stocks that experience high or rising levels of borrowing suffer from abnormal returns in subsequent time periods. This has important practical applications for lenders of stock, and for marketers of stock lending programmes. However, if dividend tax arbitrage effects are entirely eliminated, as with the subset of non-dividend paying stocks, we find that negative abnormal returns follow from high levels of borrowing. This suggests that stock lending data can be used to identify trading strategies, but only for situations where the dividend tax arbitrage effect is not present. We identify a number of patterns in the data that might signify ‘short squeezes’ in operation.

Future research is likely to cover several issues. A number of trading strategies, such as capital structure arbitrage, are undertaken across different asset classes. Research that links these asset classes together might provide further insights into the processes, distribution of returns and risks associated with arbitrage. The strategy of shorting stocks which are not subject to dividend tax arbitrage but which do exhibit high levels of borrowing can be further tested by incorporating borrowing costs and availability, as well as other trading frictions, to determine its expected profitability in practice.

References

Abraham A., and Ikenberry, D.L., 1994. The Individual Investor and the Weekend Effect. Journal of Financial and Quantitative Analysis 29, 2, 263-277.

Abreu, D., and Brunnermeier, M., 2002. Synchronisation Risk and Delayed Arbitrage. Journal of Finance 66, 341-360.

Ackert, L.F., and Athanassakos, G., 2004. Short Interest and Common Stock Returns: Evidence from the Canadian Market. Journal of Banking and Finance, forthcoming.

Aitken, M.J., Frino, A., McCorry, M.S. and Swan, P.L., 1998. Short Sales are Almost Instantaneously Bad News: Evidence form the Australian Stock Exchange. Journal of Finance 53, 6, 2205-2223.

Angel, J.J., Christophe, S.E., and Ferri, M.G., 2003. A Close Look at Short Selling on NASDAQ. Financial Analysts Journal 59, 6, 66-74.

Asquith, P., and Moelbroek, L., (1996), An Empirical Investigation of Short Interest. Working paper, Harvard University.

Boehmer, Ekkehart., Jones, Charles and Zhang, Xiaoyan, 2005. Which Shorts are Informed? .

Brent, A., Morse, D. and Stice, E.K., 1990. Short Interest: Explanations and Tests. Journal of Financial and Quantitative Analysis, 25, 273-289.

Bris, A., Goetzman, W.N. and Zhu, N., 2004. Efficiency and the Bear: Short Sales and Markets around the World. Working paper, Yale International Center for Finance.

Brunnermeier, Markus K., and Pedersen, Lasse Heje, 2005. Predatory Trading. Journal of Finance 60, (4), 1825-1863.

Chang, E.C., Pinegar, J.M., and Ravichandran R., 1993. International Evidence on the Robustness of the Day-of-the-Week Effect. Journal of Financial and Quantitative Analysis 28, 4, 497-513.

Chen, Honghui and Singal, Vijay, 2003. Role of Speculative Short Sales in Price Formation: the Case of the Weekend Effect. Journal of Finance 58, 685-705.

Connolly, R.A., 1989. An Examination of the Robustness of the Weekend Effect. Journal of Financial and Quantitative Analysis 24, 2, 133-169.

Connolly, R.A., 1991. A Posterior Odds Analysis of the Weekend Effect. Journal of Econometrics 49, 1-2, 51-104.

D’Avolio, G., 2002. The Market for Borrowing Stock. Journal of Financial Economics 66, 271-306.

DeBondt, W.F.M., and Thaler, R.H., 1987. Further Evidence on Investor Over-Reaction and Stock Market Seasonality. Journal of Finance 42, 557-581.

Dechow, P.M., Hutton, A.P., Meulbroek, L., & Sloan, R.G., 2001. Short Sellers, Fundamental Analysis, and Stock Returns. Journal of Financial Economics 61, 77-106.

DuBois, M. and Louvet, P. 1996. The Day-of-the-Week Effect: The International Evidence. Journal of Banking and Finance 20, 9, 1463-84.

Fabozzi, F.J., 2004. Short Selling: Strategies, Risks and Rewards. Wiley Finance, Hoboken, New Jersey, USA.

Fama, E., and French, K., 1993. Common Risk Factors in the Returns on Stocks and Bonds. Journal of Financial Economics 33, 3-56.

Faulkner, M.C., ‘An Introduction to Securities Lending’, Spitalfields Advisors, 2004.

Figlewski, S., 1981. The Informational Effects of Restrictions on Short Sales: Some Empirical Evidence. Journal of Financial and Quantitative Analysis 16, 463-476.

Figlewski, S., and Webb, G.P., 1993. Options, Short Sales, and Market Completeness. Journal of Finance 48, 761-777.

Geczy, C.C., Musto, D.K., and Reed, A.V., 2002. Stocks are Special Too: an Analysis of the Equity Lending Market. Journal of Financial Economics 66, 241-269.

Gibbons, M.R., and Hess, P., 1981. Day of the Week Effects and Asset Returns. Journal of Business 54, 4, 579-596.

Hong, H. and Stein, J.C., 2003. Differences of Opinion, Short-Sales Constraints and Market Crashes. The Review of Financial Studies 16, 2, 487-525.

Jegadeesh, N., and Titman, S., 1993. Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency. Journal of Finance 48, 93-130.

Jones, C. and Lamont, O., 2004. Short Sale Constraints and Stock Returns. Journal of Financial Economics 66, 207-240.

Khwaja, A.I., and Mian A. 2006. Unchecked Intermediaries: Price Manipulation in an Emerging Market. Journal of Financial Economics, forthcoming.

McDonald John G., and Baron, Donald C., 1973. Risk and Return on Short Positions in Common Stocks. Journal of Finance 28, 1, 97-107.

Mahoney, P.G., 1999. The Stock Pools and the Securities Exchange Act. Journal of Financial Economics 51, 343-369.

Miller, Edward M., 1977. Risk, Uncertainty and Divergence of Opinion. The Journal of Finance, 32, 4, 1151-1168.

Mitchell, M. and Pulvino, T., 2001. Characteristics of Risk and Return in Risk Arbitrage. Journal of Finance 56, 6, 2135-2175.

Senchack, A.J., and Starks, L.T., 1993. Short-Sale Restrictions and Market Reaction to Short-Interest. The Journal of Financial and Quantitative Analysis 28, 2, 177-194.

Shleifer, A. and Vishny, R., 1997. The Limits of Arbitrage. The Journal of Finance 52, 35-55.

Woolridge, J.R. and Dickinson, A., 1994. Short Selling and Common Stock Prices. Financial Analysts Journal 50, 20-28.

Yu, F., 2006. How profitable is capital structure arbitrage? Working Paper, University of California, Irvine.

Appendix 1 - Illustrations of dividend tax arbitrage

Proportion of shares on loan for a selected company with above average dividend yield:

[pic]

Royal Bank of Scotland PLC dividend record dates: 12th march 2004 and 13th August 2004.

Proportion of shares on loan for a selected company with below average dividend yield:

[pic]

Sage PLC dividend record dates: 13th February, 2004 and 21st May 2004.

Proportion of shares on loan for a selected company paying no dividend:

[pic]

-----------------------

[1] Corresponding author. Tel: +44 131 6503794; fax: +44 131 6683053. E-mail: James.Clunie@ed.ac.uk

[2] In this latter case, for example, the holder of securities is subject to withholding tax on interest or dividends, but the borrower would be free of withholding tax. Some of the benefits the borrower obtains from receiving the dividend free of withholding tax are shared with the lender.

[3] The dividends are actually paid on a later date, generally but not always within 30 days of the dividend record date. As such, there will be a slight ‘time value of money’ error in our calculation methodology.

[4] Naked short selling is unlikely to form more than a tiny subset of total short interest. Stockbrokers would eventually stop trading with a ‘repeat offender’, as it is time consuming to deal with the failed trades.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download