Short Selling Activities and Convertible Bond Arbitrage

Empirical Evidence from the New York Stock Exchange

verfasst von: Sebastian P. Werner

Verlag: Gabler

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Wirtschaft"

Einloggen, um Zugang zu erhalten

Über dieses Buch

The main cause of financial crisis may be found in the over-optimistic investing of b- ers that leads market prices away from fundamental values. However, in the aftermath of “excess” when stock markets tumble, it is usually the pessimists or short sellers who get publicly blamed. Despite the longstanding controversy on short selling activities, this market instrument remains a widely misunderstood concept by the public while it is an essential tool used by hedge funds for speculation and arbitrage. That is why it is important to investigate short selling for its different motivations and the resulting effect on stock returns, a subject whose empirical study is in its infancy. In his doctoral thesis, Sebastian examines convertible bond arbitrage, which is a typical hedge fund strategy that involves a long position in a convertible bond and a significant short position in the underlying stock. The short selling is employed as a hedge against movements in the stock price. With every change in the stock price, the hedge needs to be continuously readjusted, a practice which should lead companies with convertible bonds outstanding to have on average higher short selling activity than companies without convertible bonds. Furthermore, fundamental information should be processed differently in stocks with convertible bonds as stock price reactions based on the information are accompanied by the short selling of the convertible bond arbit- geurs.

Inhaltsverzeichnis

Frontmatter

1. Introduction

Zusammenfassung

In the official statement of the Securities and Exchange Act of 1934 a short sale is defined according to Rule 3b-3 as “any sale of a security which the seller does not own or any sale which is consummated by the delivery of a security borrowed by, or for the account of, the seller.” In simple terms, a short sale transaction is like buying a stock – but in reversed order: it means to sell first and to buy later. Thus, short sellers benefit as corporate stock prices fall and lose when stock prices gain. However, legal and institutional restrictions as well as the limited availability of shares for short sales make short selling a costly transaction. Hence, Diamond and Verrechia (1987) argue that only informed investors with very negative information will enter into these trades assuming a rational expectations framework. Motivated by this theoretical prediction, the financial economics literature has tested this view empirically by investigating the relationship between levels or changes in short interest, i.e. the stock of open short positions that has not yet been repurchased and closed out, and subsequent stock returns.

Sebastian P. Werner

2. Background and Empirical Predictions

Zusammenfassung

Of central interest to this dissertation is the distinctiveness within aggregate short sales and, in particular, the short selling activities of convertible bond arbitrageurs. In order to motivate the empirical investigations that follow, this chapter therefore provides the theoretical background on short selling and convertible bond arbitrage discussed in Section 2.1 and 2.2, which is summarized in Section 2.3 to develop the empirical predictions of this work.

Sebastian P. Werner

3. The Event Study Methodology

Zusammenfassung

The previous chapter outlined the empirical predictions related to the research objective of this dissertation. These predictions will be methodologically specified and empirically investigated in Chapter 5 and 6 and require the determination of firm-specific abnormal stock returns and abnormal short selling activity following events of large stock price changes and extreme short selling activity. The examinations rely on the standard event study methodology, which goes back to Fama, Fisher, Jensen, and Roll (1969). This approach takes the investigation further towards identifying convertible arbitragebased short selling from the aggregate data and distinguishing this activity from valuation- based short selling by testing for significant differences in the trading pattern, information content, and impact on stock returns. Section 3.1 through 3.5 introduce the basic steps of the methodology and thereby provide the methodological framework for the event study’s specific application within the course of the empirical investigation of this work.

Sebastian P. Werner

4. Data, Full Sample and Variable Construction

Zusammenfassung

The empirical investigation focuses on a sample of New York Stock Exchange (NYSE) listed stocks for a two-year period from January 2005 to December 2006. The data is obtained from multiple sources, which includes information on short sale volume, convertible bonds, accounting items, institutional ownership, analyst estimates, and option availability. Most of the data is used as input parameters to determine firm, stock and trading activity variables that account for the different motives of short selling, short sale constraints, and general control issues as investigated by the previous literature and discussed in Section 2.1 and 2.2. In the following, Section 4.1 identifies the main data sources, while Section 4.2 describes the principal matching process for the construction of the full data sample. Section 4.3 defines and explains the calculation of different variables from the different data sources and their matching process to the full data sample.

Sebastian P. Werner

5. Difference in Abnormal Short Selling Activity Following Events of Large Positive Stock Price Changes

Zusammenfassung

Based on the theoretical, methodological, and empirical foundations set in the previous chapters, in this chapter, I begin with the first part of the empirical investigation. As the chapter’s title suggests, methodologically it measures abnormal short selling activity following events of large positive stock price changes, which is investigated for significant differences between firms with and without convertible bonds. With regard to content, this analysis is primarily intended to test all predictions linked to Proposition 1. The objective is therefore to examine whether arbitrage-based short selling activity, i.e. the aggregate short selling in firms with convertible bonds, shows a different reaction to large positive stock return events and thus a significantly different trading pattern as compared to valuation-based short selling activity, i.e. the aggregate short selling in firms without convertible bond. The event study of Section 5.1 provides evidence for this, whereas Section 5.2 supplements these results by investigating the absolute and relative magnitude of abnormal relative short sales within a cross-sectional framework of short selling determinants.

Sebastian P. Werner

6. Difference in Information Content of Extreme Short Selling Activity Events and the Impact on Stock Returns

Zusammenfassung

This chapter continues with the second part of the empirical investigation. It measures and investigates the information content of extreme short selling activity events and their impact on stock returns of firms with and without convertible bonds. This analysis is primarily intended to test the remaining predictions linked to Proposition 2. The objective is therefore to examine whether extreme arbitrage-based short selling activity, i.e. the extreme aggregate short selling in firms with convertible bonds, has a significantly lower information content and thus a weaker negative impact on stock returns as compared to extreme valuation-based short selling activity, i.e. the extreme aggregate short selling in firms without convertible bonds. To do so, the event study of Section 6.1 measures abnormal buy-and-hold returns following the events. The results imply a lower information content of short sales for event observations with convertible bonds outstanding and indicate a temporary downward price pressure as a reflection of uninformed short selling activity over the 20 day post-event window. This is further analyzed in Section 6.2, which investigates the direct impact of events’ extreme short selling activity on the stock returns and expands the testing for price pressure.

Sebastian P. Werner

7. Overall Conclusion

Zusammenfassung

The literature and prior empirical evidence analyze short sellers as informed traders who predict negative stock returns as they assess a stock price to be overvalued. However, there exist other types of short sellers who act purely uninformed as they short stock as part of an arbitrage or hedging activity. This motive for short selling has gained in importance due to the explosive growth in hedge funds and market neutral strategies over the past decade. As a result, the recent literature provides evidence that the increased presence of these arbitrage- and hedging-related short sales has contributed to a weakened negative relationship between short interest and future returns, which suggests that the level of short interest has lost its precision as a measure for the negative sentiment of short sellers. Under the continuing growth of arbitrage and hedging strategies, it is therefore important to distinguish between valuation- and arbitrage-based short selling activities. Motivated by this insight and based on the theoretical background of short selling and convertible bond arbitrage, this dissertation investigates aggregate daily short sales for the trading pattern of arbitrage-based short selling activities of a particular type of trader, i.e. the hedging activities of convertible bond arbitrageurs, compared to valuation-based short selling activities.

Sebastian P. Werner

Backmatter

Titel: Short Selling Activities and Convertible Bond Arbitrage
verfasst von: Sebastian P. Werner
Verlag: Gabler
Electronic ISBN: 978-3-8349-6003-0
Print ISBN: 978-3-8349-1886-4
DOI: https://doi.org/10.1007/978-3-8349-6003-0