Michael G. Ferri

A Close Look at Short Selling on Nasdaq

We examine the frequency of short selling in stocks listed in the Nasdaq market. Using previously unavailable transaction data, we can report several findings: (1) overall, 1 of every 42 trades involves a short sale; (2) short selling is more common among stocks with high returns than stocks with weaker performance; (3) actively traded stocks experience more short sales than stocks of limited trading volume; (4) short selling varies directly with share price volatility; (5) short selling does not appear to be systematically different on various days of the week; and (6) days of high short selling precede days of unusually low returns. To examine the frequency of short selling, we used a unique data set that identifies all short-sale transactions reported to Nasdaq through its ACT trade-reporting system during the fall of 2000. Our study differs from most prior empirical studies in that our sample consists of day-by-day short selling of individual stocks by investors expecting price declines whereas most prior research relied on the once-a-month report of total short interest. To investigate the frequency of short selling, we used two complementary metrics. The first is the “percentage of short trades,” which is the ratio of short trades to the total number of trades in a stock within a day. This percentage addresses the likelihood that the seller in a transaction is shorting. The second is “percentage of shorted shares,” which is the ratio of the shares in the short trades to the total number of a companys shares traded in the day. This percentage concerns the probability that a traded share is being shorted. We found that, on average, the seller in 1 of every 42 trades is shorting and that 1 of every 35 traded shares is a shorted share. These results indicate that short sellers tend to transact in larger volume than the typical nonshorting investor. The second issue we examined is whether short selling varies with a stocks price performance in a manner that is consistent with either a momentum or contrarian investing style. After partitioning the sample into quintiles based on the within-sample percentage change in the stocks price, we found that both the percentage of short trades and the percentage of shorted shares were highest for the quintiles containing the stocks with the highest returns and lowest for those with the lowest returns. These results are most consistent with the proposition that short sellers tend to follow a contrarian strategy. Next, we addressed the link between short selling and trading volume. The liquidity offered by high-volume stocks potentially reduces the probability that the short seller will experience a “short squeeze” (when the shares that have been lent to the investor for the short sale are recalled). After partitioning the sample according to trading volume, we found that both the percentage of short trades and the percentage of shorted shares declined monotonically with trading volume. Therefore, as hypothesized, short sellers tend to be more interested in high-volume stocks. Fourth, we examined the relationship between short selling and stock price volatility. Specifically, we searched for any contemporaneous connection between the two. After partitioning the sample into quintiles according to the standard deviation of within-sample stock return, we found that short selling is highest for the most volatile stocks and lowest for low-volatility stocks. A number of prior studies documented a trend in which returns on certain days of the week tend to be significantly lower than returns on other days. Therefore, we considered whether any such day-of-the-week pattern is to be found in short selling. Somewhat surprisingly, our results indicate little day-to-day variation in short selling. Finally, we examined the potential short-term profitability of short selling by examining market- adjusted returns following days of significantly high short selling. We found statistically significant three-day excess returns of −1.23 percent following days of high short selling, implying that an investor who opens a short position on a day of high short selling in a stock and closes it three days later can obtain a positive net profit.

What does Nasdaq's high-yield bond market reveal about bondholder-stockholder conflicts?

We use data from Nasdaq’s FIPS system for reporting transactions in selected high-yield corporate bonds to investigate the relationship between the returns on a firm’s stock and the returns on its publicly traded, high-yield debt. Regression models and analysis of the behavior of the returns around events associated with agency conflict show that the returns follow complex patterns of similarity and divergence. Positive co-movement is the dominant form of the relationship, but opposite movement of the bond and stock returns around those events indicates agency conflicts between bondholders and stockholders.We use data from Nasdaq’s FIPS system for reporting transactions in selected high-yield corporate bonds to investigate the relationship between the returns on a firm’s stock and the returns on its publicly traded, high-yield debt. Regression models and analysis of the behavior of the returns around events associated with agency conflict show that the returns follow complex patterns of similarity and divergence. Positive co-movement is the dominant form of the relationship, but opposite movement of the bond and stock returns around those events indicates agency conflicts between bondholders and stockholders.

Short Selling and the Weekend Effect in Nasdaq Stock Returns

We examine daily short selling of Nasdaq stocks to explore whether speculative short selling causes a significant portion of the weekend effect in returns. We identify a weekend effect in speculative short selling whereby it constitutes a larger percentage of trading volume on Mondays versus Fridays. We find an opposite effect in dealer short selling, consistent with market makers adding liquidity and stability. Our main finding is that speculative short selling does not explain an economically meaningful portion of the weekend effect in returns, even among the firms most that are most actively shorted. This finding contradicts some prior studies.

Evidence that the Stock Market Overreacts and Adjusts

CHUNG-KI MIN is assistant professor of business administration at the School of Business Administration at George Mason University in F& (VA 22030). n this study, we examine whether changes in the general level of equity prices conform to the “overreaction theory,” which has attracted considI erable interest since publication of de Bondt and Thaler’s [1985] research. That research uncovered evidence that a stock whose price changes by an unusual-1y large amount in one direction tends to have a substantial return with the opposite sign in later periods. The de Bondt-Thaler theory consists of two propositions: Investors often respond to new information with an exaggerated shift in price that reflects excessive enthusiasm or fear; and, after assessing the new data more carefully, investors correct or adjust the exaggeration by moving the stock‘s price in the opposite direction. (For convenience, the theory that stock prices overreact and later adjust is called simply “the overreaction theory.”) The overreaction theory fascinates investors, because it promises opportunities to make excess profits from simple rules based only on the direction and size of sudden, abnormal movements in share prices. Academic analysts, too, are interested in the possibility of price reversals because such a pattern directly conflicts with the efficient market hypothesis. Some studies of individual equities, including de Bondt and Thaler [1985] and Fama and French [1988], have found evidence of price reversals following large changes, but other research, such as Cox and Peterson [1994], disputes that view. Likewise, the relatively few studies of the overall market do not reach a consensus

Day-of-the-Week Patterns in Volume and Prices of Nasdaq High-Yield Bonds

The authors investigate newly available dated from the Nasdaqs Fixed income Pricing System on transaction high–yield corporate bonds. Their tests of daily patterns for sixty issues traded for all or part of the interval between October 1994 and July 1997 show that: 1) the level of trading is consistently higher on Tuesday through Thursday than on either Monday or Friday, and 2) bond prices tend to significantly rise on Tuesday and decline on Friday. They interpret these results as supporting this rule of timing transactions: Investors seeking to sell high–yield corporate bonds should try to do so near the close on Tuesday or on either Wednesday or Thursday, while those who want to buy them should do so near the close on Friday or on Monday.

Should Owners of Nasdaq Stocks Fear Short-Selling?

It is interesting to look at the daily association between market-adjusted returns of Nasdaq stocks and the percentages of trading volume attributable to dealers and to their speculator customers. An unusually detailed and informative set of Nasdaq trading records reveals significantly negative average market-adjusted returns when speculative short-selling exceeds 10%. At the more common (lower) levels of speculative short-selling, average market-adjusted returns tend to be near zero. Even when there is considerable speculative short-selling, the associated negative market-adjusted return is, on average, only 4 to 6 basis points for each percentage point of short-selling. For many stocks, days of high speculative short-selling are not typically days of unusually low market-adjusted returns. Finally, high levels of short-selling by Nasdaq dealers are more common for stocks earning relatively higher market-adjusted returns.

Investor expectations about callable warrants

I n the ea.rly 1980s, corporations began to issue ered from Value Line Convertibles. The sample includes stock purchase warrants with call features. The typical warrant includes a provision permitting the firm to accelerate the warrant’s expiration date and force its exercise when the stock price exceeds a specified level. While there are certain restrictions related to specified deferment periods and maxiinum decreases in a warrant’s life, the call feature gives a company substantial control over a warrant’s time to expiration. Thus far, firms have used the provisions unpredictably. Some corporations have retired warrants early, others have appeared reluctant to do so. Such reluctance has a precedent: As Ingersoll shows in a study of callable convertible bonds (1977), many firms delay calling long past the time when it has become both profitable and permitted. This study examines investor expectations on the likelihood of firms to call warrants. To do this we derive the implicit expected lives (or ”maturities”) of callable warrants from their market prices. The technique we use to obtain the market estimates of remaining life follows procedures used in previous studies to derive the implicit stock return variances from the option premiums.’ We also analyze whether implicit lives are consistently below corresponding nominal maturities and are related to variables that influence the firm in its 84 3 2 o 3 L every outstanding warrant followed by Value Line from January through May of 1986 that meets four criteria: The warrant is callable, either immediately or at a future date; the warrant has a fixed exercise price and a limited, nominal (or maximurn) life longer than one year, as of January; the firm’s mean dividend yield is not more than 2% over the five months; and the warrant was not called before July 1986. The number of such warrants is 34.’ We obtained five end-of-mon th prices (‘January through May) for these warrants and their stocks from Value Line, Standard & Poor’s Daily Stock Price Record, and the Wall Street Journal (which ’was the source of risk-free rates of various maturities). Closing or bid prices were available for most warrants, but in a few cases the only recorded price was Value Line’s average of dealers’ ask prices. In all, the data contain 170 observations on warrant prices. Value Line also provided information on each warrant’s “relative size” (the ratio of new warrant shares to outstanding common shares) and an estimate of each stock‘s return volatility. The estimate is the standard deviation of the natural log of weekly percentage price changes over the previous five years. Table 1 gives summary statistics for the sample’s important aspects. decision regarding call. Findings that implicit lives are both relatively low and sensitive to determinants of call would indicate that investors believe firms will readily force (early exercise. Results showing no significant difference between ex ante and nominal maturities woulcl suggest that investors expect firms to be reluctant to call. DERIVATION OF IMPLICIT MATURITIES

Evidence on the Market's Underestimation of Foreign Earnings Persistence

Can a hedging strategy consistently generate profits by exploiting a drift in returns following the announcement of large changes in the foreign earnings of U.S. multinational firms? Results of this study indicate such a strategy produces positive profits on average, but the returns vary considerably from year to year and are negative in two of eight years examined. A market-neutral hedge portfolio based on announced changes in foreign earnings represents not a pure arbitrage opportunity, but rather a potentially risky style of investment.

Short Selling and the Cross-Section of CorporateBond Returns

This article studies the effect of short selling in the equity market on corporate bond returns. We show that firms with heavily shorted shares or large short-trade sizes experience significantly negative future bond returns. Further tests indicate that the relationship between short-trade size and subsequent bond returns is more consistent with stealth trading of short sellers. The impact of short selling on bond returns is robust to various controls for risk, liquidity, and other pricing factors. In examining the information source of short selling, we find that firms associated with heavy shorting activities or large short-trade sizes are likely to subsequently experience negative earnings surprises, higher credit risk, and reduced dividends. The evidence provides little support for the overvaluation argument. Overall, the results are consistent with the proposition that short trades in the equity market exert important valuation consequences in the corporate bond market.

An Empirical Examination of the Determinants of Bond Yield Spreads

The yield on corporate bonds of good credit quality (issued by either industrial firms or public utilities) invariably is greater than the yield on U.S. government bonds of similar term and taxability. Furthermore, the yield on industrial bonds differs systematically from that of utility bonds of equal credit rating, term, and tax status. Each of these yield spreads is volatile. For example, the industrial-utility spread rose from a few basis points in 1973 to more than one full percentage point in 1975. The primary purpose of this paper is to examine the reasons for the existence and fluctuation of these yield differentials.