THE JOURNAL OF FINANCE • VOL. LXIV, NO. 1 • FEBRUARY 2009
`
`It’s SHO Time!
`Short-Sale Price Tests and Market Quality
`
`KARL B. DIETHER, KUAN-HUI LEE, and INGRID M. WERNER∗
`
`ABSTRACT
`We examine the effects of the Securities and Exchange Commission (SEC)-mandated
`temporary suspension of short-sale price tests for a set of Pilot securities. While short-
`selling activity increases both for NYSE- and Nasdaq-listed Pilot stocks, returns and
`volatility at the daily level are unaffected. NYSE-listed Pilot stocks experience more
`symmetric trading patterns and a slight increase in spreads and intraday volatility
`after the suspension while there is a smaller effect on market quality for Nasdaq-
`listed Pilot stocks. The results suggest that the effect of the price tests on market
`quality can largely be attributed to distortions in order flow created by the price tests
`themselves.
`
`REGULATION SHO (REG SHO) PROVIDES a new regulatory framework governing
`short-selling of securities in U.S. equity markets. The rules were passed on
`September 7, 2004 and became effective on January 3, 2005.1 Reg SHO is in-
`tended to establish uniform locate and delivery requirements, create uniform
`marking requirements for sales of all equity securities, and establish a proce-
`dure to temporarily suspend the “provisions of Rule 10a-1 under the Securities
`Exchange Act of 1934 and any short-sale price test of any exchange or na-
`tional securities association for short sales of certain securities for certain time
`periods” in order to “evaluate the overall effectiveness and necessity of such
`restrictions.”2
`In this paper, we study the effect on market quality of the Securities and
`Exchange Commission’s (SEC) mandated temporary suspension of short-sale
`price tests for a set of designated pilot securities (Rule 202T—Pilot Program).3
`
`∗Diether and Werner are at Fisher College of Business, The Ohio State University. Lee is at
`Rutgers Business School at Newark and New Brunswick and at Korea University Business School,
`Seoul. We thank the New York Stock Exchange for financial support and seminar participants at
`The Ohio State University, the Wharton School, the New York Stock Exchange, and Brigham Young
`University for comments. We are also grateful for feedback received at the SEC Roundtable on the
`Regulation SHO Pilot. We thank Paul Irvine for comments and Yingdi Wang for research support.
`All errors are our own.
`1 Securities Exchange Act Release No. 50103 (July 28, 2004), 69 FR 48008 (August 6, 2004).
`2 Division of market regulation: Responses to frequently asked questions concerning regulation
`SHO (January 4, 2004).
`3 Securities Exchange Act Release No. 50104 (July 28, 2004), 69 FR 48032 (August 6, 2004).
`
`37
`
`Coalition for Affordable Drugs IV LLC - Exhibit 1032
`
`

`
`38
`
`The Journal of Finance R(cid:4)
`
`On May 2, 2005,4 roughly 1,000 U.S. stocks—so called Pilot stocks—began to
`trade without short-sale price tests (Uptick test for the NYSE and bid price
`test for Nasdaq). These stocks were selected by the SEC to represent a broad
`cross-section of the U.S. equity market. The Pilot stocks were drawn from the
`Russell 3000 index, comprising every third stock ranked by volume. We label
`the remaining Russell 3000 index securities Control stocks. The experiment
`was designed by the SEC to investigate whether Rule 10a-1, NYSE’s Uptick
`rule, and Nasdaq’s bid price test affect market quality, and to develop uniform
`price tests if such rules were deemed necessary going forward. The temporary
`suspension was originally set to expire on April 28, 2006, but was extended to
`August 6, 2007.5
`The extent to which specific price tests are likely to have an effect depends
`on what fraction of overall trading activity is represented by short sales. Re-
`cent empirical evidence shows that short-selling is much more common than
`most market observers previously imagined. For example, Boehmer, Jones, and
`Zhang (2008) find that short sales represent 13% of NYSE (SuperDOT) share
`volume during 2000 to 2004. Even more striking, Diether, Lee, and Werner
`(2008) find that short sales represent 31% of share volume for Nasdaq-listed
`stocks and 24% of share volume for NYSE-listed stocks in 2005. Hence, there
`is a potential for short-sale price tests to affect price levels, volatility, as well
`as high-frequency measures of market quality. Short-sale price tests may affect
`price levels if they create frictions that are strong enough to limit the extent
`to which prices reflect the views of pessimists, that is, the investors who think
`stocks are overvalued (e.g., Miller (1977) and Diamond and Verrecchia (1987)).
`They may also reduce volatility if they make it more difficult for short-sellers to
`engage in downward manipulation of stock prices. Finally, short-sale price tests
`may affect the mix of passive and active trading strategies employed by short-
`sellers, which in turn could affect market quality measures such as spreads,
`depth, and order imbalances.
`Restrictions on short-selling activity were introduced in the United States in
`the 1930s following the stock market crash of 1929. Jones (2003) shows that
`short-selling in NYSE-listed stocks was more difficult after the introduction
`of shorting restrictions. He also finds that these events were associated with
`positive abnormal returns, consistent with the notion that optimists have more
`influence on prices in the presence of short-sale restrictions. By contrast, short-
`ing restrictions had no effect on the volatility of returns. Based on the historical
`experience, we predict that stock prices will fall on the announcement of a sus-
`pension of short-sale price tests, or at least after the suspension itself. However,
`we find no evidence that NYSE or Nasdaq Pilot stocks experience significantly
`lower average returns compared to their respective Control samples around
`
`4 The Pilot Program was originally intended to commence on January 3, 2005, but in response to
`information received by the SEC from market participants, the Pilot was postponed until May 2,
`2005 (Securities and Exchange Act Release No. 50747 (November 29, 2004), 69 FR 70480 (December
`6, 2004)).
`5 Securities and Exchange Act Release No. 53684 (April 20, 2006).
`
`Coalition for Affordable Drugs IV LLC - Exhibit 1032
`
`

`
`It’s SHO Time!
`
`39
`
`the announcement of the Pilot Program (July 28, 2004) or on the date the Pilot
`Program became effective (May 2, 2005). Moreover, we do not detect any change
`in daily volatility measures for Pilot compared to Control stocks. Hence, it ap-
`pears that there is no significant effect of suspending short-sale price tests on
`price levels or volatility at the daily frequency.
`The effect of short-selling on high-frequency market quality measures de-
`pends on the specific form of the price test that each market uses. NYSE uses a
`tick test to determine whether a short sale is permitted (NYSE Rule 440B, also
`called the Uptick rule). Consequently, a short sale is only allowed on a plus tick
`or on a zero tick, where the most recent price change preceding the trade was
`a plus tick (called a zero-plus tick).6 The way the specialist adjusts orders to
`ensure compliance with the Uptick rule means that short-sellers effectively be-
`come liquidity providers. As a result, we expect to see a disproportionate amount
`of limit orders on the offer side of the market, inflating the depth at the ask
`quotes and a disproportionate amount of trades executing above the midquote,
`creating a buy order imbalance. As suggested by Jones (2003), these trading
`strategies may also produce narrower quoted (and possibly effective) spreads.
`Finally, we argue that this bias toward passive short-sale order strategies may
`dampen short-term volatility.
`Nasdaq uses a bid price test to determine whether a short sale is allowed
`(Nasdaq Rule 3350). Short sales in Nasdaq National Market Securities (NM)
`are not allowed at or below the (inside) bid when the current inside bid is at or
`below the previous inside bid. We argue that the bid price test permits short-
`sellers to use a more natural mix of marketable limit order and limit order
`strategies than the NYSE’s Uptick rule. Moreover, Archipelago and INET, who
`together are responsible for a significant fraction (over 40%) of share volume
`in Nasdaq-listed stocks, did not enforce the bid price test during the sample
`period. As a result, we predict that the effect of short-selling activity on market
`quality will be smaller for Nasdaq-listed stocks.
`We find that the temporary suspension of short-sale price tests affects short-
`selling activity for both NYSE- and Nasdaq-listed securities. For NYSE-listed
`stocks, there is no significant change in short-sale share volume or short in-
`terest, but short-sale trade size decreases significantly and short-sale trade
`frequency increases significantly for Pilot relative to Control stocks. In other
`words, NYSE short-sellers split their orders more as they switch from passive
`to more active trading strategies. Moreover, short sales relative to share vol-
`ume on the NYSE increase significantly after the suspension of the price tests.
`For Nasdaq-listed stocks, both short-sale share volume and the short-sale fre-
`quency increase significantly for Pilot relative to Control stocks. However, there
`is no evidence of increased order splitting for Nasdaq Pilot stocks. In sum, short
`sales relative to share volume increase significantly for Pilot relative to Control
`
`6 Types of short sales that are exempt from short-sale rules include certain odd-lot short-sales,
`certain sales by registered specialists or market makers, certain sales necessary to equalize prices
`on a nonprimary market with the primary market, certain sales in special arbitrage accounts, and
`certain sales by underwriters (see SEC Rule 10a-1, section e (3), (4), (5), (6), (7), (8), (9), and (10)).
`
`Coalition for Affordable Drugs IV LLC - Exhibit 1032
`
`

`
`40
`
`The Journal of Finance R(cid:4)
`
`stocks. Thus, it appears that suspension of NYSE’s Uptick rule and Nasdaq’s
`bid price test makes it somewhat easier to execute short sales.
`We argue that NYSE’s Uptick rule causes short-sellers to engage in more
`passive trading strategies, which results in narrower spreads, lower volatility,
`higher ask depth, and a disproportionate amount of order flow executing above
`the midquote. Thus, we predict that the suspension of the Uptick rule will
`significantly reduce the quote and order flow asymmetries, and that it may
`result in wider spreads and higher short-term volatility. We also argue that the
`impact of short sales on Nasdaq is smaller for two reasons: The bid price test
`is not very restrictive, and Archipelago and INET permitted unfettered short
`sales in Nasdaq-listed stocks. Consequently, we predict that the suspension of
`the bid price test will have more limited impact on market quality for Nasdaq
`Pilots.
`We find strong evidence supporting our hypotheses. The relative bid depth
`increases significantly for NYSE-listed Pilot stocks but there is no significant
`change for Control stocks. The buy order imbalance declines significantly for
`Pilot stocks while it actually increases significantly for Control stocks. Relative
`to Control stocks, NYSE-listed Pilot stocks experience a slight but statistically
`significant increase in both quoted and effective spreads. Trade-to-trade re-
`turns are significantly more volatile for NYSE Pilot relative to Control stocks
`after May 2, 2005. However, there is no evidence of a disproportionate increase
`in downside volatility. Lower-frequency intraday volatility measures (at 5, 15,
`and 30 minutes) suggest that the increase in volatility for Pilot stocks rela-
`tive to Control stocks disappears as returns are measured over longer intervals
`(30 minutes). Finally, variance ratio tests suggest that short-term volatility
`increases relative to longer-term volatility for NYSE Pilot stocks compared to
`Control stocks.
`As predicted, we find smaller differences between the changes in market qual-
`ity measures such as quote asymmetries and spreads for Nasdaq-listed Pilot
`stocks relative to Control stocks. There is also no significant change in trade-
`to-trade volatility, or in midquote volatility at lower frequencies (5, 15, and 30
`minutes), for Nasdaq-listed Pilot stocks relative to Control stocks. These find-
`ings are confirmed by variance ratio tests, which show no significant changes
`for Nasdaq-listed Pilot stocks relative to Control stocks.
`Throughout the paper, we conduct cross-sectional tests to examine how the
`suspension of price tests affects less liquid stocks. While we do find evidence
`that the magnitude of the effects of the suspension of the Uptick rule are in
`some cases larger for less liquid NYSE-listed stocks, we attribute this mostly
`to the fact that the distortions in order flow are more severe for stocks with
`wider spreads and lower price. The evidence on how suspending the bid price
`test affects small, less liquid Nasdaq-listed stocks is more mixed.
`Our paper proceeds as follows. In Section I, we outline the short-sale rules for
`the NYSE and Nasdaq and develop our testable hypotheses. We describe the
`data in Section II. We test whether the level of short sales changed significantly
`between the period before and after the suspension of price tests in Section
`III. In Section IV, we examine the effect of suspension of price tests on daily
`
`Coalition for Affordable Drugs IV LLC - Exhibit 1032
`
`

`
`It’s SHO Time!
`
`41
`
`returns and volatility. Section V is devoted to testing for changes in market
`quality measures. In Section VI, we investigate whether intraday volatility
`increased for Pilot stocks. We discuss robustness tests in Section VII. Section
`VIII concludes. A short epilogue can be found in Section IX.
`
`I. Testable Hypotheses
`In this paper we first test whether the actual form of price tests used by
`NYSE or Nasdaq constrain short-selling activity. Angel (1997) and Alexander
`and Peterson (1999) conclude that the Uptick rule impedes short-selling activ-
`ity on the NYSE by examining the fill rate of short-sale orders.7 We provide an
`alternative test of the effect of the Uptick rule by examining the overall volume
`of short-selling after price tests are lifted. In addition, we test whether Nas-
`daq’s bid price test impedes short-selling. If the applicable short-sale rules are
`restricting market participants from executing their desired trades, we predict
`that short-selling activity for Pilot stocks will increase significantly after the
`suspension of the price tests. Of course, there could be a secular trend in short-
`selling activity, so we need to express our hypothesis relative to what happens
`to Control stocks, for which there is no change in rules. Hence, our first hypoth-
`esis is that the suspension of price tests will increase short-selling activity in
`Pilot stocks relative to Control stocks.
`Much of the empirical short-selling literature focuses on how short-sale con-
`straints affect stock prices and returns. Theoretical papers such as Miller
`(1977), Harrison and Kreps (1978), Morris (1996), Scheinkman and Xiong
`(2003), and Duffie, Gˆarleanu, and Pedersen (2002) develop models in which
`the presence of short-sale constraints and opinion divergence leads to over-
`pricing and abnormally low subsequent returns. Empirically, the previous lit-
`erature finds that stocks are sometimes expensive to short and that there is
`a link between short-selling and subsequent returns. Using data from April
`2000 through September 2001, D’Avolio (2002) reports that about 9% of the
`stocks in his sample have loan fees greater than 1% per year. Jones and Lam-
`ont (2002), Ofek, Richardson, and Whitelaw (2004), and Cohen, Diether, and
`Malloy (2007) find that subsequent average returns are low for stocks that are
`expensive to short. Figlewski and Webb (1993), Dechow et al. (2001), Desai et al.
`(2002), Boehmer, Jones, and Zhang (2008), Diether, Lee, and Werner (2008),
`and Asquith, Pathak, and Ritter (2005) find a negative relation between short-
`selling activity and subsequent returns. If short-sale price tests help push stock
`prices above fundamental value, then suspending them should result in neg-
`ative returns. Hence, our second hypothesis is that the suspension of price
`tests will be associated with lower returns for Pilot stocks relative to Control
`stocks.
`The price tests for listed stocks were developed by the SEC in the 1930s to
`prevent short sales from executing in declining markets (see, e.g., Jones (2003)).
`
`7 Ferri, Christophe, and Angel (2004) argue that the Nasdaq bid price test is not effective and
`does not curtail short-selling activity.
`
`Coalition for Affordable Drugs IV LLC - Exhibit 1032
`
`

`
`42
`
`The Journal of Finance R(cid:4)
`
`The Nasdaq bid price test is much newer, and was approved as a temporary
`rule by the SEC in September of 1994.8 If these price tests are effective, they
`should dampen downside volatility of returns. However, the literature has not
`generally found the rules to be effective. SEC (1963) finds that the Uptick rule
`is ineffective at relieving pressure from short-sellers. Ferri, Christophe, and
`Angel (2004) find no significant difference in the level of shorting activity during
`down markets for Nasdaq stocks that are subject to the bid price test (National
`Market stocks) and those (small cap stocks) that are not. Moreover, Jones (2003)
`finds that the introduction of the tick tests has no effect on volatility at the daily
`frequency. For completeness, we reexamine the question of the impact of price
`tests on volatility at the daily frequency. We hypothesize that if the price tests
`are effective in reducing volatility, we should find that Pilot stocks experience a
`significant increase in volatility relative to Control stocks after the price tests
`are suspended.
`To develop testable hypotheses regarding the effect of price tests on market
`quality measures, we rely on the specific form of the rules used by NYSE and
`Nasdaq. NYSE uses a tick test to determine whether a short sale is allowed, and
`short sales are only allowed on a plus tick or a zero-plus tick. A plus tick occurs
`when the last sale price is above the price of the previous trade. A zero-plus tick
`occurs when the last sale price is the same as the price of the previous trade, but
`the most recent different price is below the last sale. Whether a price is a plus
`tick, zero-plus tick, zero-minus tick, or minus tick is determined by comparing
`the price with the last sale on the NYSE.9 There are a few rare situations
`where the Rule permits short sales without regard to the tick test.10 Generally
`speaking, NYSE specialists, option market makers, and third-market market
`makers for NYSE-listed stocks are subject to the tick-test. However, there is an
`exemption (Rule 10a-1 (e)(5)) for a short sale by a market maker for his own
`account effected at a price equal to or above the last sale or effected at a price
`equal to that participant’s most recent offer if the offer, when quoted, was equal
`to or above the last sale.
`We hypothesize that the Uptick rule significantly affects how short-sellers’
`orders are presented to the market. Specifically, the NYSE specialist has to
`ensure that an order is compliant with the Uptick rule and he does so by re-
`lying on the display book software. Whatever order type (market or limit) the
`short-seller submits, the display book software adjusts the order so that it is
`compliant. Effectively, this means that short-sale orders are changed into non-
`marketable limit orders. Suppose that the last sale is at $28.05 on a plus tick
`and the quotes are $28.00 to $28.05. If the short-seller submits a market sell
`
`8 The bid test has been reapproved annually as a temporary rule ever since (SEC (2006)).
`9 Rule 10a-1 uses a tick test based on the consolidated transaction reporting system as a default,
`but permits each market center to use its own tape as the reference for short-sale compliance
`provided that this is a real-time reporting system.
`10 These include: odd lot transactions that change the market maker’s position by less than a unit
`of trading, domestic arbitrage, and international arbitrage. The SEC has also granted exemptive
`relief for certain eligible volume weighted average price (VWAP) transactions, exchange traded
`funds, and certain short sales executed at the closing price in after-hours crossing sessions.
`
`Coalition for Affordable Drugs IV LLC - Exhibit 1032
`
`

`
`It’s SHO Time!
`
`43
`
`order, the order cannot be executed at the bid ($28.00) because of the Uptick
`rule. Instead, the order will be treated by the display book software as an ef-
`fective limit order to sell at $28.05. Similarly, if the short-seller submits a limit
`order to sell at $28.03 (or anywhere below $28.05) the limit price will be changed
`to $28.05 before being displayed in the book. Note that the effective limit prices
`of short-sale orders have to be adjusted over time as prices move in such a way
`that they remain compliant. For example, if the following trade takes place at
`$28.00 then the short-sell limit order is cancelled and replaced by a new limit
`sell order at $28.03 (the original limit price). The new order will carry a time
`stamp of the price change that triggered the adjustment to the order.
`Similarly, short-sale orders in NYSE-listed stocks that are routed to ArcaEx
`are altered by the system so that they are compliant with the Uptick rule. To
`illustrate, suppose the market is at $20.00 to $20.03 and last sales are $20.00
`followed by $20.03. A short limit sell order arrives to ArcaEx priced at $20.02.
`The Uptick rule implies that the order cannot fill right away. Instead, ArcaEx
`would quote the short limit order at $20.03, because, with a last sale sequence
`of $20.00 and $20.03, the lowest “tick” available for shorting is $20.03 (zero-plus
`tick). If the order didn’t execute at $20.03 and the next trade was at $20.00,
`ArcaEx would move the short limit order back to its original limit price, since the
`order could now be executed at $20.02 (plus tick). By contrast, INET’s trading
`system simply rejects short-sale orders in NYSE-listed stocks at prices that
`would execute at a “bad” tick.
`As a result of how short-sale orders are handled by the main venues trading
`NYSE-listed stocks, we expect to see a disproportionate amount of limit orders
`on the offer side of the market. We also expect that a disproportionate amount
`of trades will occur above the midquote, which results in buy orders exceeding
`sell orders on average (buy order imbalance) when trades are classified using
`the Lee and Ready (1991) algorithm. The Uptick rule is also likely to produce
`narrower quoted (and possibly effective) spreads, and dampened short-term
`volatility, as short-sellers shadow the last sale to remain compliant with the
`rule.11
`The suspension of the Uptick rule permits short-sellers in NYSE-listed Pilot
`stocks to use more aggressive marketable orders. As short-sellers switch from
`being (forced) liquidity providers to demanding liquidity, their incentives to
`engage in order-splitting to price discriminate among liquidity providers on
`the bid side of the market increases. In addition, short-sellers are more likely
`to be perceived by other traders in the market as aggressive sellers and this
`implies that they are more susceptible to order anticipators. As a result, we
`predict that short-sellers will engage more heavily in order-splitting to reduce
`the price impact of their trades after the suspension of the Uptick rule.
`The usage of more aggressive marketable orders also means that the asym-
`metric patterns that we expect to see for NYSE-listed stocks will attenuate.
`Specifically, we predict that the suspension of the Uptick rule will result in a
`
`11 Angel (1997) finds that the national best bid and offer for NYSE-listed stocks is typically
`lowered after the placement of a short-sell order.
`
`Coalition for Affordable Drugs IV LLC - Exhibit 1032
`
`

`
`44
`
`The Journal of Finance R(cid:4)
`
`reduction in the depth at the offer relative to the bid, a reduction in the buy
`order imbalance, a widening of quoted and effective spreads, and an increase
`in short-term volatility for NYSE-listed Pilot stocks relative to Control stocks.
`Nasdaq uses a bid price test to verify whether a short sale is permitted (Nas-
`daq Rule 3350), and short sales in Nasdaq National Market Securities (NM)
`are not allowed at or below the (inside) bid when the current inside bid is at or
`below the previous inside bid. To execute a “legal” short-sale in a NM security
`on a down bid, the short sale must be executed at a price at least $0.01 above
`the current inside bid. As of February 1, 2004, all market participants should
`use the Nasdaq (SuperMontage) inside bid to verify compliance with the bid
`price test. Registered market makers for a Nasdaq-listed stock, as well as mar-
`ket makers in options on this stock, are exempt from the bid price rule when
`trading the stock. To illustrate, suppose again that the last sale is $28.05 on a
`plus tick and that the quotes are $28.00 to $28.05. A Nasdaq short-seller can
`place a marketable limit sell order at $28.00 and still be in compliance with
`the bid price test as long as the most recent bid was $28.00 or below. Moreover,
`a short-seller can always place a limit sell order at $28.01 regardless of past
`bids and past trade prices. By contrast, as described above, on the NYSE the
`short-sale order would in this case enter the display as a limit sell order at
`$28.05, which is 4 cents higher. Of course, a Nasdaq short-seller may also want
`to be more passive and place the order at $28.05 (or even higher), but the point
`is that he is not forced to do so by the rules. He has more freedom to choose the
`optimal limit price for his order than his NYSE counterpart. The only situation
`for which the price tests are equally restrictive is if the last sale is at $28.00 on
`a down bid and a down tick. In this case, both Nasdaq and NYSE rules permit
`a limit sell order at $28.01.
`Note that neither ArcaEx12 nor INET13 enforced price tests for these stocks
`during our sample period. Granted, NASD member firms may have felt bound
`by NASD Rule 3350 and could have verified that their orders were compliant
`through third-party systems. These two venues were responsible for a signif-
`icant fraction of trading volume in Nasdaq-listed stocks during our sample
`period. For example, during May 2005 (the first month of the Pilot), ArcaEx
`executed 18.2% and INET executed 24.8% of Nasdaq share volume.14
`Thus, even though we expect to see some effect of the bid price test on Nasdaq
`depth, order flow, spreads, and volatility, it is likely to be smaller than in the
`case of the NYSE. It follows that the suspension of the bid price test should
`
`12 See http://www.tradearca.com/faqs gen.asp.
`13 INET’s ECN printed their trades on the National Securities Exchange (NSX) (previously
`the Cincinnati Stock Exchange) during our sample period. Since INET quotes and trade-reports
`through NSX, and NSX does not have a short-sale rule for Nasdaq-listed stocks, INET be-
`lieves that the NASD short-sale rule may not be applicable to orders sent to INET in Nasdaq-
`listed securities. (See: www.inetats.com/subcribers/emailarchive/2003/20030625.asp.) However, as
`of January 23, 2006, INET enforces the bid price test for Nasdaq-listed stocks because it has
`migrated its trade reporting from the National Stock Exchange (NSX) to Nasdaq’s ACT (see:
`www.isld.com/subscribers/emailarchive/2006/20060118 2.asp).
`14 See http://www.nasdaq.com.
`
`Coalition for Affordable Drugs IV LLC - Exhibit 1032
`
`

`
`It’s SHO Time!
`
`45
`
`have a smaller effect on order-splitting, the depth at the offer relative to the
`bid, the buy imbalance, quoted and effective spreads, and short-term volatility.
`Again, note that all these predictions need to be evaluated relative to Control
`stocks.
`Regulators have expressed concerns that small-capitalization, low-liquid and
`low-priced stocks will be most adversely affected by the suspension of the price
`tests. Less liquid stocks may be more vulnerable to the effects of more active
`short-selling since the order book is likely to be thinner. Low-price stocks may
`also be relatively more affected by more active short-selling since a penny tick
`may be a more significant impediment to shorting for these stocks. At the same
`time, it is important to recognize that less liquid stocks tend to have wider
`spreads, and often also lower prices. To the extent that short-sale volume is
`still significant for less liquid stocks, we expect the distortions introduced by
`the price tests to be larger.
`To see why, consider two NYSE-listed stocks L(iquid) and I(lliquid). Stock L
`has a spread of 1 cent and Stock I has a spread of 5 cents. Start by assuming
`that they both have the same bid price, $30.00, and that the last sale was at the
`bid on a down tick. An aggressive short-seller in stock L (that is someone who
`would have submitted a market sell order absent the Uptick rule) will have to
`place a limit order at $30.01 to be compliant. Since the quoted spread for stock
`L is 1 cent, this short-sale order will not affect the quoted spread. Similarly,
`an aggressive short-seller in stock I will place a limit order at $30.01 to be
`compliant. However, in this case the short-sale order narrows the quoted spread
`from 5 cents to 1 cent or from 16.67 to 3.33 basis points. Thus, if aggressive
`short-sellers are equally active in both stocks, that is, they represent the same
`fraction of order flow, the Uptick rule would have a larger effect on stock I’s
`quoted spread than on stock’s L’s quoted spread. Now change the assumptions
`so that the bid price for stock I is $15, while the price remains $30 for stock L.
`Then the basis point spread for stock I would go from 33.33 to 6.67. In other
`words, the Uptick rule affects stocks with wider spreads disproportionately, and
`the effect is exacerbated if stocks also have a lower price.15 Hence, we should
`expect a larger increase in spreads and intraday volatility for less liquid stocks
`after price tests are suspended simply because the price tests cause larger
`distortions in order flow for these securities.
`
`II. Data and Methodology
`Our study covers the period from February 1, 2005 through July 31, 2005.
`The initial sample includes all Pilot stocks as defined by the Securities Ex-
`change Act Release No. 50104 (July 28, 2004), 69 FR 48032 (August 6, 2004).
`The remaining Russell 3000 securities are included as Control stocks. To elim-
`inate the potential confounding influence of index inclusion or index exclusion,
`we require that sample stocks be members of the Russell 3000 index after
`the June 2004 reconstitution and remain members of the Russell 3000 index
`
`15 A similar example can easily be created for the Nasdaq bid test.
`
`Coalition for Affordable Drugs IV LLC - Exhibit 1032
`
`

`
`46
`
`The Journal of Finance R(cid:4)
`
`Table I
`Sample Selection
`The sample includes all NYSE- and Nasdaq-listed stocks that are part of the Russell 3000 index
`on June 30, 2004 and are still part of the index as of June 30, 2005. Stocks that change listing
`venue, go private, are involved in a merger or an acquisition, or change ticker during the period are
`excluded. Nasdaq small cap stocks and stocks with a price exceeding $100 or an average quoted
`spread exceeding $1.00 for February 1 to April 30, 2005 are also excluded. To avoid undue influence
`of the open, we exclude data from 9:30 to 10:00 am. Pilot stocks are stocks that were designated as
`Pilot A securities by Reg SHO. All other NYSE- and Nasdaq-listed Russell 3000 index stocks are
`labeled Control stocks.
`
`Pilot Stocks
`
`Control Stocks
`
`NYSE
`Nasdaq
`Total
`
`448
`376
`824
`
`904
`757
`1,661
`
`Total
`
`1,352
`1,133
`2,485
`
`Percent
`
`54.4
`45.6
`
`after the June 2005 reconstitution.16 In other words, we exclude stocks that
`were added to the index (IPOs) during the period June 2004 through June
`2005 and stocks that were eliminated during the year due to corporate events
`such as mergers, bankruptcies, etc., plus stocks that were added or eliminated
`in the June 2005 index reconstitution. We also exclude 22 stocks that experi-
`enced ticker changes during the sample, 13 stocks that were listed on Nasdaq’s
`small cap market at any point during the sample, 17 stocks that were acquired,
`merged, or privatized during the year, and 4 stocks that changed listing venue.
`Further, we exclude 32 stocks with an average price above $100.00 or an av-
`erage quoted spread exceeding $1.00 during the February 1 to April 30 period
`to reduce the likelihood that our results will be influenced by outliers. Finally,
`we exclude the 34 stocks that were listed on AMEX due to the small sample
`size for this market. These filters bring the total sample down to 2,485 from the
`total of 3,402 stocks that appear on Russell