The Effect of Option Listings: Evidence from American Depository Receipts

by

Spencer A. Case*

Hankamer School of Business

Baylor University

Janet D. Payne

McCoy College of Business

Texas State University

Abstract

This paper investigates the effect of option listings on underlying securities using the market for ADRs. Unlike the U.S. domestic equities, the short sale constraint differs widely for ADRs due to country specific legal restrictions on short selling and options availability. Our results show that U.S. options exchanges have no preference for listing ADRs that are subject to higher short sale constraints. We find no decrease in volatility or increase in short interest with option listing. The abnormal return upon option introduction isless negative when the ADR is less short sale constrained, consistent with option listings mitigating short sales constraints.

JEL Classification: G12, G14, G18

Keywords: options, ADRs, short sale constraint

* Please address correspondence to Spencer Case (), Hankamer School of Business, Baylor University, Waco, TX 76798. Ph:254-710-4145; Fax:254-710-1092.

The Effect of Option Listings: Evidence from American Depository Receipts

  1. Introduction

Danielson and Sorescu (2001) extend the theoretical model of Miller (1977) and Jarrow (1980) to examine the effect of removing the short sale constraint on prices of underlying securities. They conclude that removal of that constraint will result in predictable decreases in underlying security prices. They test their hypothesis in what they describe as a laboratory setting, where option introductions represent a reduction in short sale constraints. They find a puzzling result - prior to 1981, stock price reactions are positive, though after 1981, stock price reactions are negative. Since the legal and regulatory environment which causes the short sale constraint is virtually the same across all domestic firms, there is no clear control group forwhich options are introduced but where there is a substantially lower or nonexistent short sale constraint. In an ideal setting, a researcher would want to observe firms that had options introduced but weren’t short sales constrained, as well as firms that have options introduced that do have binding short sales constraints. A more robust test would take place in an environment where short sale constraints have identifiable cross-sectional variation. We examine the price effect of option introductions in an environment where the short sale constraint is likely to differ across firms, namely the market for American Depository Receipts (ADRs). Because we have cross-sectional variation in the sample (firms that have high levels versus firms that have relatively low levels of short sale constraint due to differences in the home market legal environment), we can examine the effect of introducing options on lesserconstrained firms versus more constrained firms. These lesser constrained firms in essence act as the control group that would otherwise be absentor problematic to identify in a single country setting. We find that, if neither put options are available for trading nor short sales are allowed in the home country, the ADR is no more likely to have an option introduced in the U.S. The (negative, on average) abnormal return associated with option introduction is positively associated with put option availability and unrelated to the availability of short sales in the home country (the country in which the underlying security originates).

American Depository Receipts are negotiable certificates traded on a U.S. exchange, issued by a U.S. bank, representing a specified number of shares in a foreign stock. We examine 190 option introductions on ADRs over the period of 1982 to 2006. We first examine the choice by the options exchanges to list options on ADRs, similar to the examination of options listings on domestic firms as seen in Mayhew and Mihov (2004). Several authors includingTrennepohl and Dukes (1979), Conrad (1989), Bansal, Pruitt, and Wei (1989), Skinner (1989),and Damodoran and Lim (1991) find that there is a decrease in the volatility of the underlying security after the options are introduced, with Mayhew and Mihov (2004) and Bollen (1998) being exceptions, they do not find a decrease in volatility. Likewise, we do not find this decrease in volatility among optioned ADRs. Mayhew and Mihov (2004) find that trading volume, volatility and market capitalization are determinants of option listing. We find thatrelative daily volume during the period before the listing and market capitalization in the U.S., arethe onlysignificant determinantsof selection ofADRs for option listing. Volatility as measured by standard deviation or beta does not appear to be a factor by which ADRs are selected for option listing.

Like previous studies on domestic securities, which have excluded ADRs, we find that the introduction of options on ADRs has a significant and negative influence on the price of the underlying security. In contrast, the non-optioned ADR matches have a positive price reaction to news of an option listed on their counterpart. When examining the optioned firms, the price reaction is positively related to changes in relative short interest and the availability of puts in the home country. The change in volume, positive price change and the standard deviation of returns prior to the listing all have a negative influence on returns around the option listing date. The price reaction is unrelated to previous capital-raising by the ADR firms in the U.S. market or the legality of short sales in the home country. We find that when home country investors have a fairly strict constraint of “short sale illegality,” the price response to option listing in the US market is insignificant. However, if the home country for the ADR already has put options available, the stock price response is far more positive than would otherwise be the case.

The remainder of this paper is organized as follows. Section 2 contains a concise review of the prior literature as well as presenting our research hypotheses. Section 3 includes institutional details that are relevant to this study. Section 4 provides a description of our sample selection and empirical methods. Results are in Section 5, and Section 6 concludes.

  1. Background and Hypotheses

Several previous studies have examined which stocks are chosen for options listing. In one of the earliest studies of listing choice, Cowan, Carter, Dark and Singh, (1992) empirically investigate the listing of firms on the NYSE. Among those firms that are eligible for NYSE listing, they examine the determinants of those that are ultimately chosen to be listed by the NYSE. However, as noted by Mayhew and Mihov (2004), thechoice of option listing is fundamentally different. Whereas firms apply to be listed on exchanges like the NYSE, the choice of option listing is dependent upon the options exchange and not the firm. In fact, the firm itself and or the government of the home country could protest in opposition to the listing of options, as was the case with many Japanese firms. Mayhew and Mihov (2004) investigate the determinants of that selection for domestic firms. They use a control sample method and find that stocks with high trading volume, volatility and market capitalization are more likely to have options listed. They also find that these relations have changed somewhat over time, with a shift from volume toward volatilityafter institutional changes in 1980.

Early studies, beginning withBranch and Finnery (1981) and corroborated by Skinner (1989) , DeTemple and Jorian (1990) and Conrad (1989),find that the introduction of options was associated with a positive excess returns for option introductions in the U.S. In later work Sorescu (2000) finds that there was a dramatic change in the price effect of option introductions on the underlying security. In the time period before 1980, price changes were positive, and after 1980 they became negative. Major regulatory changes in 1980-1981 as well as the introduction of index options in 1982 were cited as potential reasons for the shift. Danielsen and Sorescu (2001) include relative short interest to provide additional evidence related to the possibility that the decreases in price were due to diminishing short sale constraints. Asquith and Meulbroek (1995) outline several of these short sales constraints. First, because the proceeds of the short sale are held in the brokerage account with no interest earned, there is an indirect transaction cost to the short sale. The investor must be very bearish on the stock to entice him to enter this position. Until more recently, SEC rule 10a-1 required that short sales not be executed at a price below the preceding transaction price. In addition, the short seller must find a willing lender for the shares. Finally, the lack of ability to find shares for repurchase (a short squeeze)can be very costly for the investor.

In addition to facts about stock price changes upon option listing, it has been previously shownby several authors that stock return variances decrease significantly following the listing of options on the underlying security. The earliest evidence of this decline was provided by Trennepohl & Dukes (1979), Conrad (1989), Bansal, Pruitt, and Wei (1989) and Skinner (1989), all of whom find significant declines in volatility ranging from less than 1% to more than 6%. More recent evidence by Damadoran and Lim (1991) finds a drop in volatility of approximately 20%. In more recent studies this decrease in volatility is drawn into question. St. Pierre (1998) uses an EGARCH model and finds that the conditional volatility of security returns is unaffected by option introduction. Bollen (1998), who uses a control sample matched on industry, does not find a significant decrease in volatility after option listing. Mayhew and Mihov (2004) compare optioned firms to a pooled sample of all non-optioned firms that were eligible to be optioned and do not find a change in volatility after option listing. The use of differing control sample methods might partially explain the differing results from past studies. According theoretical work by Cao (1999), the volatility effect of option listing has the potential to spill over to other stocks, particularly when there is a high correlation with the optioned stock. Thus, non-optioned stocks in the control sample could be affected by the option listings.

The general relation of the pricing of an ADR with the equivalent shares in the home country has been the subject of much interest. Rosenthal (1983)shows that abnormal profits could not be earned on ADRs from any price dependence. Likewise, Kato , Linn and Schallheim (1991) and Wahab and Khandwala (1992) find that after accounting for transaction costs, there were few profitable opportunities to arbitrage the ADRs between the U.S. and the home market for the underlying security. Still other authors, such as Mathur, Gleason and Singh (1998),Kim, Szakmary and Mathur (2000) and Kim, Chen and Chou (2002), investigate the dynamic linkages of price movements between the home country and the U.S. Generally, the findings are that the linkage is strong, and that most price shocks are corrected quickly in each country.

Legal environments and investment barriers vary by country and those variations are often important determinates of market efficiency or firm level market performance. Foerster and Karolyi (2000) investigate the long run performance for capital raising depository receipts. They find that differences in long-term performance after raising capital are related to the scope and magnitude of investment barriers in the home country. Bris, Goetzmann, and Zhu (2004) analyze forty-six equity markets around the world. They find that prices incorporate information faster in countries where short sales are allowed and practiced. Daouk and Charoenrook (2005) study short sale and put option trading regulations and practices from 111 countries. They find that when short-selling is legally permissible, liquidity is higher. They also find a stock price increase associated with countries relaxing their short sale constraints. ADRs are different than domestic securities, in that important firm level characteristicscan vary widely due to differing legal environments, investment barriers and other factors specific to the home country of each firm.

Our hypotheses are similar to those of Danielson and Sorescu (2001) and Mayhew and Mihov (2004), but ourconjecture is that the ADR market is a superior environment in which to test the hypotheses. This is due to the potentially large variation in short sale constraints that results from the differences in their underlying home market legal and regulatory environments. Specifically, we test:

i) Options on ADRs are chosen for listing based on criteria similar to that found by Mayhew and Mihov (2004) i.e. size, volume, and volatility being significant determinates.

ii) Options exchanges are more likely to introduce options on ADRs from countries that are more short sale constrained; where the more constrained ADR’s home country doesnot allow short selling or there are not puts available for trading.

Similar to Danielson and Sorescu (2001), we hypothesize that, upon option introduction:

iii) Stock price reaction will be negative.

iv) Short interest will increase.

v) Stock price reaction will be negatively related to beta.

vi) Stock price reaction will be negatively related to changes in short interest.

Given our unique ADR sample and experimental design, we also test:

vii) The lack of short sales or put options in the underlying ADR’s home market will result in a more negative stock price reaction to the option introduction. i.e. ADRs with stronger short sales constraints will have a more negative stock price reaction when options are introduced in the U.S..

  1. Institutional Details

As described in Mayhew and Mihov (2004), option exchanges are member-owned self-regulating organizations, and are subject to federal securities laws and oversight by the Securities and Exchange Commission. The SEC has played a large role in selection of stocks for option listing, particularly in the beginning of the options market, in 1973. In July of 1977, the SEC announced plans for an extensive review of options trading, and asked for a voluntary moratorium on expansion.[1] That review, released in February of 1979, resulted in rules changes by the exchanges. At that time, the SEC lifted the moratorium on expansion and option trading was considered a permanent institution.[2]

Over time, there have been changes in eligibility requirements for domestic stocks, as outlined in Table I of Mayhew and Mihov (2004). Those different eligibility regimeswere used as delineating factors in their study. In a study of option listings on ADRs, however, those regimes are less relevant because ADRs and other foreign securities are often treated differently than domestic securities. According to the SEC, American Depository Receipts are exempt from many rules relating to options listing on domestic securities, except for the requirement that they be listed on a national exchange.[3] The same is true of short sale constraints. For example, the uptick rule, largely in place for domestic securities during our sample period, was not a requirement of a short sale on an ADR, under the “International Arbitrage Exemption.”[4] ADR specific criteria for listing include: the existence of a monitoring agreement between the U.S. stock exchange and the home market stock exchange, a majority of the trading volume taking place on the U.S. exchange and a minimum track record for trading in the U.S. Despite the actual guidelines provided by the SEC, approval for individual ADRs has largely been done on an ad hoc basis.

  1. Sample Selection and Empirical Methods

We used the Center for Research in Security Prices (CRSP) to identify American Depository Receipts. Our sample included ADRs traded from 1982 to 2006. There were 828 unique ADRs trading at various points in time on the CRSP database, 799 of those from countries classified by Daouk and Charoenrook (2005). The individual option listing dates wereprovided by the Chicago Board Options Exchange (CBOE). Of the securities in our sample, we identified 190that had listed options on any of the option exchanges tracked by the CBOE. The country of origin for each ADR was identified using Standard and Poor’s Research Insight. We used Standard and Poor’s Daily Stock Reports to obtain short interest data and used CRSP for the returns, prices and volume data.

Our empirical methods follow previous authors in the area of domestic options listing. To test differences in volatility, volume and short interest, we use the firm’s beta, standard deviations of returns before and after the option listing, volume before and after the option listing and relative short interest before and after the option listing. We use t-tests to calculate the statistical significance of any differences.

The question of how ADRs are chosen for options listing, and whether those criteria differ from those found using purely domestic options is addressed using logistic regressions with the dependent variable being a dummy variable equaling 1 if the security had an option listed, and 0 if it did not.

(1)