Exchange Traded Funds Market Structure and the Flash Crash 5

20

www.cfa pubs .org

?2012 CFA Institute

Financial Analysts Journal Volume 68 x Number 4?2012 CFA Institute

Exchange-Traded Funds, Market Structure, and the Flash Crash

Ananth Madhavan

The author analyzes the relationship between market structure and the flash crash. The proliferation of trading venues has resulted in a market that is more fragmented than ever. The author constructs measures to capture fragmentation and shows that they are important in explaining extreme price movements. New market structure reforms should help mitigate such market disruptions in the future but have not eliminated the possibility of another flash crash, albeit with a different catalyst.

he “flash crash” of 6 May 2010 represents one of the most dramatic events in the his-tory of the financial markets. Late that after-noon, major U.S. equ ity market indices

began to decline sharply. The Dow Jones Industrial Average (DJIA) dropped 998.5 points, the sharpest intraday point drop in history, followed by an astounding 600-point recovery within 20 minutes.The flash crash is distinguished from other market breaks—su ch as the one in October 1987—by its speed and rapid intraday reversal. Unlike other sharp intraday market breaks, such as the one on 28 May 1962, multiple securities traded at clearly unreasonable prices, including some (e.g., Accen-ture, 3M) that traded for pennies. Also notable was the disproportionate representation of exchange-traded products (ETPs) among the securities most affected, with prices diverging widely from their underlying net asset values.

Despite its short du ration, the flash crash affected many market participants. Exchanges ulti-mately canceled trades at prices below 60% of the 2:40 p.m. (EDT) price, bu t many retail investors with market stop loss orders still had orders exe-cuted at prices well below prevailing market levels earlier in the day. Professionals also suffered from the volatility: Liquidity providers who bought at distressed prices and hedged by short selling sim-ilar secu rities or fu tu res contracts incu rred steep losses as their long positions were canceled while the assets they had shorted rebounded in price.

It is difficult to overstate the potential negative consequences of another flash crash. Such an event cou ld dramatically erode investor confidence and participation in the capital markets for years to

come, leading to reduced liquidity and higher trans-action costs.1 A future flash crash toward the end of the day could severely disrupt the close and, hence,the pricing of index derivative produ cts, with follow-on effects for foreign markets and the subse-quent day’s open. Finally, the flash crash has already prompted several pu blic policy initiatives, and a repeat event could induce dramatic changes to mar-ket structure and the regulatory environment.

Given these concerns, a considerable effort has been made to understand and isolate the “cause” of the flash crash with a focus on the precise chronol-ogy of events. This article instead focu ses on the relationship between market structure and the flash crash withou t taking a view on its catalyst. My hypothesis is that equity market structure is a key determinant of the risk of extreme price changes.Today’s U.S. equity market structure is highly com-plex, with 12 for-profit exchanges (e.g., the NYSE)and some 30 odd dark pools competing for flow.Dark pools offer nondisplayed liqu idity and inclu de broker/dealer dark pools (e.g., Goldman Sachs’ Sigma X), exchange-owned pools (e.g., Direct Edge), and independent pools (e.g., ITG’s POSIT).

The resu lt of this proliferation of venu es is greater fragmentation of trading. Fragmentation u su ally refers to the actu al pattern of volu mes traded across different venues. In December 2011,based on trade-level data, the major market centers’share of total U.S. equ ity dollar volu me traded showed considerable dispersion, with NASDAQ at 23.8%, NYSE Arca at 16.5%, NYSE at 12.6%, BATS at 11.9%, Direct Edge EDGX at 8.1%, and the remainder accounted for by other exchanges and dark pools/broker internalization, included under FINRA’s Trade Reporting Facilities. Nor is this phe-nomenon limited to the United States or just equi-ties. In Europe, such entrants as Chi-X Global and

Ananth Madhavan is managing director at BlackRock,Inc., San Francisco.

T

Exchange-Traded Funds, Market Structure, and the Flash Crash

Turquoise have gained market share at the expense of traditional stock exchanges, and derivative trad-ing (e.g., options) is increasingly fragmented.

Although volume is a natural metric for frag-mentation, there is another dimension of interest—namely, a venue’s quotation activity at the best bid or offer. Quote fragmentation captures the compe-tition among traders for order flow and thus may be a better proxy for the dynamics of higher-frequ ency activity than a measu re based on the pattern of volu mes traded, which in tu rn cou ld reflect a variety of other factors, such as rebates.

Table 1 shows the venu e shares of the U.S. equ ity market for December 2011 based on all reported trades and quotes. The first column rep-resents the share of dollar volume. The second and third columns capture the market shares in quota-tion frequency and total dollar quoted depth (i.e., liqu idity available at the inside qu ote), respec-tively. In terms of the statistics reported above, markets are less fragmented from a quotation per-spective. In particular, the shares of NASDAQ and NYSE Arca as a fraction of all quotes at the best bid or offer are larger—34.7% and 21.9%, respectively, versu s 23.8% and 16.5% of total dollar volu me. Note that in terms of venues’ shares of total inside liquidity (i.e., total liquidity at the best bid or offer), the market is a little less concentrated compared with looking at just the frequency of quotes. At the single stock level, of course, these differences are not significant.2

■Discussion of findings. I conjectu red that prices are more sensitive to liquidity shocks in frag-mented markets becau se imperfect intermarket linkages effectively “thin ou t” each venu e’s limit order book. I began with a time-series perspective u sing intraday trade data from Janu ary 1994 to March 2012 for all U.S. equities and found that frag-mentation today is at the highest level ever. Frag-mentation was also mu ch greater than historical levels on the day of the flash crash. Cross-sectionally, I related fragmentation positively to company size and the use of intermarket sweep orders, typically u sed in aggressive liqu idity-demanding strategies by nonretail traders. I showed that ETPs are more concentrated than other equities.

With respect to the relationship between mar-ket stru ctu re and the flash crash, I fou nd strong evidence that secu rities that experienced greater prior fragmentation were disproportionately affected on 6 May 2010. This result is consistent with my hypothesis that market structure is important in understanding the propagation of a liquidity shock. Although quote fragmentation is related to volume fragmentation, the two measures are distinct, and they diverged on the day of the flash crash. Both volu me and qu ote fragmentation measu res are important risk factors in explaining the observed cross-sectional price movements in the crash.

My analysis provides insight into why ETPs were differentially affected (ETPs accou nted for 70% of equity transactions from 6 May that were ultimately canceled) even though ETP trading is less fragmented than that of other equ ities.3 For ETPs whose components are traded contempora-neou sly, widespread distortion of the prices of u nderlying basket secu rities can confou nd the arbitrage pricing mechanism for ETPs, thu s delinking price from value.

Table1.Market Share of Dollar Volume and Quotation Activity, December 2011

Venue

Share of

Dollar Volume

Share of

NBBO Quotes

Share of NBBO

Dollar Depth

Amex 0.15%0.34%0.03%

BATS BYX Exchange 2.80 1.960.69

BATS BZX Exchange11.9211.2710.94

BEX 2.27 1.250.53

CBOE 0.130.270.07

Chicago 0.690.39 2.99

EDGA 3.36 2.040.74

EDGX 8.058.04 3.62

FINRA 15.900.000.00

NASDAQ 23.7634.7143.18

NASDAQ OMX 1.43 1.46 6.00

National 0.45 1.910.96

New York 12.5814.4610.36

NYSE Arca 16.5221.8819.90

NBBO = National Best Bid and Offer.

July/August 2012www.cfa https://www.wendangku.net/doc/f110362471.html,21

Financial Analysts Journal

From the public policy viewpoint, the fact that fragmentation is now at its highest level ever may help explain why the flash crash did not occur ear-lier in response to other liquidity shocks; the rapid growth of high-frequ ency trading and the u se of aggressive sweep orders in a highly fragmented market are recent phenomena. Current policy pro-posals will help mitigate future sharp drawdowns, but another flash crash, albeit with a different cata-lyst or in a different asset class, remains a possibility.

A Review of the Flash Crash

Initial speculation regarding the proximate cause of the flash crash varied widely, but a common theme was that an event so u niqu e in financial market history mu st itself have had an extraordinary cause. For example, early theories included incor-rect order entry by a trader (the so-called fat finger theory), a software bug at a major exchange, and a maliciou s, deliberate “denial of service” type of attack intended to damage the financial system. Yet, no evidence for these explanations has since come to light. Similarly, the disproportionate impact of the flash crash on ETPs led some early commentators to draw a connection between the sharp market moves on 6 May 2010 and the pricing and trading of these instruments.4 In a controver-sial report by Bradley and Litan (2010), the authors concluded that exchange-traded fund (ETF) pric-ing poses risks to the financial system, noting that “the proliferation of ETFs also poses unquantifiable but very real systemic risks of the kind that were manifested very briefly during the ‘Flash Crash’ of May 6, 2010” (p. 5). The authors proposed various ETF-related reforms and noted that in the absence of such rules, they “believe that other flash crashes or small capitalization company ‘melt ups,’ poten-tially much more severe than the one on May 6, are a virtual certainty” (p. 5). Ben-David, Franzoni, and Moussawi (2011) presented evidence from the flash crash that “ETFs served as a condu it for shock transmission from the futures market to the equity market” (p. 23).

The joint report of the U.S. Commodity Futures Trading Commission (CFTC) and the SEC pro-vided a detailed chronology of events on 6 May 2010 and suggested a possible catalyst for the flash crash (CFTC and SEC 2010). Specifically, at 2:32 p.m., a fundamental trader used a broker algorithm to sell a total of 75,000 e-mini contracts with a notional amount of approximately $4.1 billion on the Chicago Mercantile Exchange (CME). The trade was intended to hedge an existing equity position. The trader entered the order correctly and specified an upper limit on the amount sold as a percentage of volume but did not set a price limit for the trade. As a result, price movements were magnified by a feedback loop from the volume participation set-tings, precipitating the actu al flash crash.5 The CFTC/SEC report concluded that this single trade was the root cause of the flash crash.

The notion that the flash crash arose from an unlikely confluence of the factors discussed above is reassuring because it suggests that the chance of a recurrence is very low. It is also consistent with the absence of widespread and rapid price declines in any asset class or region in recent decades. However, seriou s qu estions remain. The CME e-mini futures order in question was large but not unusually so relative to the millions of contracts a day traded in e-mini futures and the actual volume traded during that period. Indeed, the fundamen-tal trader’s participation rate was about 9% of the approximately 140,000 e-mini contracts traded in from 2:41 to 2:44 p.m. The futures market did not exhibit the extreme price movements seen in equi-ties, which suggests that the flash crash might be related to the specific nature of the equity market stru ctu re. There have also been recent instances where individual stocks experienced “micro” flash crashes.6 For example, on 27 September 2010, Progress Energy—which had been trading at abou t $44.50 per share—inexplicably fell almost 90% in price before recovering in the next five minu tes. Other examples of very sharp price declines followed by rapid reversals without obvi-ous cause include such well-known names as Citi-group and the Washington Post Company. Unlike the “macro” flash crash of 6 May, these “micro”flash crashes did not cluster in time and affected only individu al stocks. Nonetheless, they are recent phenomena and su ggest the presence of more systemic factors.

Recent analyses have provided more insight into other, more fundamental potential triggers. In particu lar, considerable debate has occu rred regarding high-frequ ency trading activity and market quality. It is useful to distinguish between algorithmic trading, defined as ru le-based elec-tronic trading with specific goals for execution out-comes, and high-frequency trading, where orders are electronically routed to venues with a focus on minimal latency. The volu me attribu ted to high-frequ ency trading (inclu ding statistical arbitrage, liquidity provision, and “order anticipation” strat-egies) has grown rapidly in recent years; Zhang (2010) reported that high-frequ ency trading accounts for up to 70% of dollar trading volume in U.S. equities.

The increase in high-frequ ency trading has raised concerns, especially given order cancellation

rates of about 90% and the fact that these strategies

22www.cfa https://www.wendangku.net/doc/f110362471.html,?2012 CFA Institute

Exchange-Traded Funds, Market Structure, and the Flash Crash

are not well understood. For example, some high-frequency traders are alleged to use “quote stuff-ing” tactics—where they post and immediately cancel orders—in an effort to gain an advantage over rivals. Intentional qu ote stu ffing allegedly works by jamming the signal bandwidth of other fast traders, who must process quotation changes that only the trader posting the rapid quote changes can safely ignore.7 More generally, the term refers to sudden spikes in quotation activity that appear unrelated to fundamental news events or trading volumes. Egginton, Van Ness, and Van Ness (2011) provided an empirical definition of quote stuffing and found that during periods of intense quoting activity, affected stocks experience lower liquidity, higher transaction costs, and increased volatility.

But recent empirical evidence on the impact of high-frequency traders and faster trading is mixed. Hasbrouck and Saar (2010) examined low-latency strategies that respond to market events in millisec-onds using one month of data from 2007 and one month of data from 2008.8 They identified “strate-gic runs” that are a series of linked submissions, cancellations, and executions that are likely to have been parts of a dynamic strategy. Their results sug-gest that increased low-latency activity improves such market quality measures as short-term vola-tility, spreads, and displayed depth. Similarly, Hendershott, Jones, and Menkveld (2011) fou nd that algorithmic trading narrows spreads, lessens adverse selection, and decreases trade-related price discovery. Zhang (2010) conclu ded that high-frequency trading has a positive correlation with stock price volatility after controlling for exoge-nous determinants of volatility. The correlation is also stronger during periods when high-frequency trading volumes are high, which impairs price dis-covery. Hendershott and Moulton (2011) provided empirical evidence that au tomation has mixed effects. Faster trading increases bid–ask spreads but also results in more efficient prices.

Kirilenko, Kyle, Samadi, and Tu zu n (2010) examined the behavior of the e-mini S&P 500 Index futures market on 6 May 2010 using audit-trail transaction data. They classified more than 15,000 trading accounts that traded on the day of the flash crash into six subjective categories: high-frequ ency traders, intermediaries, fu ndamental bu yers, fu ndamental sellers, opportu nistic trad-ers, and noise traders. The authors concluded that “High Frequency Traders did not trigger the Flash Crash, but their responses to the unusually large selling pressu re on that day exacerbated market volatility” (p. 1).

Related concerns su rrou nd “venu e toxicity”and aggressive order tactics that cause rapid shifts in liquidity, which, in turn, could have led to the flash crash. Easley, López de Prado, and O’Hara (2011) measured venue toxicity on the basis of the estimated probability of informed trading in a stock. They argued that there is “compelling evi-dence” that the flash crash could have been antici-pated becau se increasing toxicity of order flow induces less liquidity provision by market makers. Similarly, Chakravarty, Wood, and Upson (2010) focused on the use of liquidity-demanding orders that sweep the entire book, known as intermarket sweep orders (ISOs). They found an increase in ISO activity in S&P 500 stocks du ring a short period around the time of the flash crash and concluded that these orders may have triggered the flash crash by aggressively taking bid-side liquidity.

These analyses contribute deeply to our under-standing of the chronology of the flash crash and its possible catalysts. In contrast, this article focuses on analyzing the role of equity market microstruc-tu re in explaining the risk of an extreme price movement and remains agnostic about the specific trigger or spark. Specifically, I hypothesize that order book liquidity for securities that experience market fragmentation is more su sceptible to the effects of transitory order imbalances. Fragmenta-tion is normally measu red in ex post terms—by actual volumes traded across venues—but quota-tion activity may present a better idea of the true competition for order flow. Measures of fragmen-tation based on quotations (rather than volumes) captu re the competition among high-frequ ency traders and aggressive quote behavior that could cause the withdrawal of liquidity in times of market stress. The rapid growth of high-frequency trading, however, is a recent phenomenon—hence, my interest in both measu res. My hypothesis is that secu rities with greater fragmentation prior to 6 May 2010 were disproportionately affected during the flash crash.

Data Sources and Procedures

I turn now to my empirical investigation, beginning with a review of my data sources and procedures to select the sample universe.

Sample Selection. The sample universe con-sists of all 6,224 exchange-traded equ ity instru-ments in the United States for which a complete trading history is available from the NYSE Trade and Qu ote (TAQ) database and Bloomberg for 6 May 2010 and the 20 prior trading days (7 April to 5 May). I excluded stocks that experienced corpo-rate actions in the previous month, which reduced

the sample modestly to 6,173 names.

July/August 2012www.cfa https://www.wendangku.net/doc/f110362471.html,23

24

www.cfa pubs .org

?2012 CFA Institute

Financial Analysts Journal

The sample comprises 4,003 common stocks,968 ETPs, 602 closed-end funds, and 319 American Depositary Receipts, with the remainder being REITs and miscellaneous equity types. Regarding the primary exchanges, there are 2,560 NASDAQ National Market (Capital Market/Global Market/Global Select Market) names, 2,314 NYSE-listed stocks, and 917 NYSE Arca secu rities, with the remainder on Amex.9 It is also worth noting that the majority of ETPs (897 names) are listed on NYSE Arca.

Using the TAQ data, I computed a measure of how much a security was affected on the day of the flash crash. I defined the maximum drawdown as M , a continuous variable in the [0, 1] interval rep-resenting the largest price declines in the afternoon of 6 May 2010:

(1)That is, the drawdown is one less the ratio of the intraday low price to the intraday high price between 1:30 and 4:00 p.m. I collected data on a variety of stock-specific variables based on both daily and intraday data. These include equity type (e.g., ETP , REIT), market capitalization (in millions of U.S. dollars), primary exchange, Global Industry Classification Standard sub-industry, and average daily dollar volume for the 20 trading days prior to the crash. I also included volatility, which I defined as the standard deviation of five-minu te retu rns over the 20 trading days prior to the crash in the interval 1:30–4:00 p.m.

Each trade in the TAQ data is flagged with one or more condition codes, inclu ding intermarket sweep orders. ISOs are limit orders that are excep-tions to the order protection rule; they allow users to sweep all available liquidity at one market center,even if other centers are publishing better quotes.Traders using ISOs fulfill their Regulation National Market System (NMS) obligations to obtain the best price by simultaneously sending orders to all mar-ket centers with better prices. ISOs are most com-monly u sed by market makers and institu tional trading desks to sweep all available liquidity; they are rarely used by retail investors. I computed the percentage of dollar volu me of trades flagged as ISOs (identified by Condition Code F) for 6 May and separately for the previous month.

M e a s u r e s o f F r a g m e n t a t i o n. I used exchange codes at the trade- and qu ote-specific level to construct market structure metrics that cap-tu re the fragmentation of the market. Measu ring fragmentation in terms of traded volumes is natu-ral becau se it reflects the end resu lt of traders’

rou ting decisions across venu es. The simplest (inverse) measu re of fragmentation for a given stock is the k -venue concentration ratio, C k , defined as the share of volume of the k highest share market centers. So, C 1 is the volume share of the venue with the highest market share, C 2 is the volume share of the two largest venues combined, and so on, with C 1 < C 2 < C 3. Although simple, the concentration ratio may miss nuances of market structure from competition beyond the largest market centers, so I focused on the Herfindahl index, a broader mea-sure commonly used in the industrial organization literature. The volume Herfindahl index for a given stock on day t is defined as

(2)

where is the volume share of venue k on day t .The Herfindahl index ranges from 0 to 1, with higher figures indicating less fragmentation in that partic-ular stock.

Fragmentation can also be measured in terms of competition to attract flow by determining the frequ ency with which a venu e becomes the best intermarket bid or offer. Let represent the pro-portion of times that venue k had the best offer price of all posted National Best Bid and Offer (NBBO)quote changes on day (interval) t . is the ask-side Herfindahl index for a stock on day t , where

(3)

H a denotes the Herfindahl index averaged over the

20 trading days prior to the flash crash. Correspond-ingly , H b can be defined as the average bid Herfin-dahl index. Intuition suggests a very high correlation between bid- and offer-side qu ote fragmentation,bu t they can differ becau se of short-selling con-straints or other factors and over shorter intervals of time. For much of my analysis, I used the average quote Herfindahl index, H q = (H a + H b )/2.

Note that there are other sensible definitions of qu ote fragmentation. For example, if the cu rrent best bid for a stock is $34.48 (at venue A) and venue B improves that to $34.47, you would tally one for venue B. If one second later another venue, venue C, joins the best bid at $34.47, you would increment the count for venue C in computing Equation 3. As an alternative, you could exclude this latter quote change—because the best bid is unchanged—and restrict the count to only those observations offer-ing real price improvement. In the former case, the total count K in Equation 3 would be lower and you would see higher reported fragmentation. I gener-ally focused on straight quote competition at the

NBBO (i.e., based on Equation 3).

H s t v t k

k K | 2

1

,s k t n t k H t a H n t a

t k

k K

|

1

2

.

July/August 2012www.cfa pubs .org 25

Exchange-Traded Funds, Market Structure, and the Flash Crash

I computed the average daily Herfindahl indi-ces (volu me and qu ote) and concentration ratios from 1:30 to 4:00 p.m. for the 20 trading days prior to the flash crash and for the day of the flash crash.Qu ote fragmentation and volu me fragmentation are distinct, albeit closely related, economic mea-sures. Volume fragmentation reflects the outcome of order-routing decisions based on price and such factors as make/take rebates and dark pool liquid-ity. In contrast, qu ote fragmentation captu res the dynamic competition for order flow through quo-tation activity. Quote fragmentation is complemen-tary to the volu me-synchronized probability of informed trading (VPIN) measure used by Easley,López de Prado, and O’Hara (2011). Essentially,VPIN is a Level I metric (in that it uses time, traded volume, and price), whereas the measure proposed here is a Level II metric that uses the order book and its history. The correlation between the volume and quote Herfindahl indices is just 0.57, so it is evident that they capture different phenomena.

Empirical Analysis

In this section, I assess whether (controlling for other factors) fragmentation of trade and qu ote activity across exchanges played a role in the flash crash.Descriptive Statistics. The role of equity market structure is highlighted in Table 2, which compares measures of concentration based on daily sample means (not weighted by market capitaliza-tion or dollar value to avoid distortion by mega-cap stocks) for the baseline period of the 20 trading days prior to the flash crash (7 April–5 May 2010) and for 6 May 2010. The table also shows the data for ETPs and for non-ETP equity instruments.

The results confirm that ETPs were differen-tially affected during the flash crash, as measured by the steepness of the price declines they experi-enced. As shown in Table 2, the average drawdown for ETPs is 0.24, versus 0.08 for other equity assets.The second moment of drawdown is also mu ch larger for the ETP universe. Irrespective of the mea-sure of fragmentation used, ETP trading volume is more concentrated than that of other equities. The average top venue concentration ratio, C 1, is 0.56for ETPs, versu s 0.48 for non-ETP equ ities. This resu lt cou ld reflect the fact that the NYSE trades only NYSE-listed securities; most ETPs are listed on NYSE Arca. Across all asset types, fragmentation was significantly higher on 6 May 2010 than in the previous 20 trading days. Note that all asset types showed a marked increase in volume on the day of the flash crash, but the relative increase in the dollar volume in ETPs was much greater than that of other equities. This finding is consistent with the fact that ETPs typically account for a higher percentage of volume on volatile days.10

Table 2.

Market Structure Variables before and during the Flash Crash

Baseline Period

6 May 2010

Maximum Drawdown

C 1Volume Herfindahl

Quote Herfindahl

ISO Frequency C 1Volume Herfindahl

Quote Herfindahl

ISO Frequency Universe Mean 0.1060.490.360.310.270.460.330.330.37Median 0.0650.480.34

0.290.300.420.290.310.36

Std. Dev.0.170

0.12

0.12

0.09

0.13

0.14

0.13

0.10

0.15

Non-ETPs Mean 0.0800.480.350.310.280.450.320.330.36Median 0.0640.460.320.290.300.420.290.300.35Std. Dev.0.096

0.12

0.11

0.09

0.12

0.14

0.12

0.11

0.14

ETPs Mean 0.2430.560.450.350.210.500.380.370.40Median 0.0780.550.420.330.230.470.340.350.42Std. Dev.

0.338

0.14

0.15

0.10

0.16

0.16

0.15

0.12

0.20

Notes: The table provides summary statistics on market structure variables for the baseline period of the month prior to the flash crash (20 trading days) and for 6 May 2010 for the universe (6,173 stocks) and separately for 5,205 non-ETP equities and 968 ETPs. C 1 is the concentration ratio for the top venu e. Herfindahl concentration indices are reported for actu al volu me and for cou nts of qu ote improvement on the bid and ask sides. ISO frequency is the average dollar-weighted frequency of intermarket sweep orders. The data are based on the NYSE TAQ database.

Financial Analysts Journal

An increase in the use of aggressive tactics—based on the dollar volume with Condition Code F in the TAQ database—occurred on the day of the flash crash, and the increase was greater for ETPs than for other equities. For non-ETP equities, the mean frequency of Condition Code F (ISOs) was 0.36 on 6 May, versu s an average of 0.28 for the baseline period. For ETPs, however, the mean ISO frequency was 0.40 on 6 May, versus an average of 0.21 in the month before. The difference in means is statistically significant for ETPs but not for com-mon stocks. In both cases, the medians are close to the corresponding means, so results are not skewed by a few outliers.

I also examined whether other condition codes (e.g., stock option trades) showed marked differ-ences between the day of the flash crash and the previous month. None of the differences are eco-nomically or statistically significant. The resu lts show no clear relationship between drawdown and liquidity as proxied by market capitalization and trading volume, respectively. The concentra-tion indices, however, generally decline with drawdown—consistent with my hypothesis—but the relationship is not monotonic.

Separating the data by liqu idity is logical because fragmentation is likely to vary systemati-cally in this dimension. Table 3 provides means of key economic variables in the baseline period and the maximum drawdown on 6 May 2010 based on deciles of average daily dollar volume. Each decile contains about 622 stocks, so the standard error of the mean is relatively small. To avoid potential skew, the means are not weighted by size or vol-ume. Note that two of the concentration measures—the top venue share (C1) and the volume Herfindahl index—are strongly negatively related to volume, which is consistent with greater intermarket com-petition in more liquid stocks. The quote fragmen-tation measure does not vary much with volume, which su ggests different drivers. As volu me and company size increase, the ISO frequ ency (dollar weighted) monotonically increases, which is consis-tent with greater use of sweep orders in more liquid stocks and greater fragmentation. Finally, draw-down increases steadily to 13–14% in Deciles 5–7 before declining again to 10% in the top decile (i.e., the relationship with volume is not monotonic).

Time-Series Variation in Fragmentation. The time series of fragmentation can provide valu-able historical context. I used 18 years of intraday TAQ data for all U.S. equities from 3 January 1994 to 30 March 2012, a period that inclu des many important market structure changes. There are 4,587 trading days in the sample and a total of 42.6 million stock-days. For each stock on each day, I computed the volume Herfindahl index using all trades in the TAQ database for that stock on that day—a compu-tationally challenging task. I then compu ted the average stock’s Herfindahl index (u nweighted mean) to get an overall market concentration statis-tic for the day. Figure 1 plots the time series of the daily marketwide Herfindahl index along with a 50-day moving average. There is variation from day to day, bu t the mean individu al stock concentration was relatively constant from 1994 to the end of 2003,

Table3.Market Structure Variables by Trading Activity

Volume Decile Avg. Daily

Dollar Vol.

(millions)

Market

Capitalization

(millions)

ETP

Proportion

ISO

Frequency C1

Volume

Herfindahl

Quote

Herfindahl

6 May 2010

Maximum

Drawdown

All$42.7$3,665.115.6%26.9%0.4880.3630.3300.106 10.0101.412.7 6.30.6720.5730.3540.047 20.2272.522.219.50.5930.4690.3260.077 30.4180.820.725.50.5410.4100.3120.104 40.7374.416.728.10.5160.3810.3180.117

5 1.5755.317.728.50.5030.3650.3080.129

6 3.0971.014.529.20.4800.3420.3010.140

7 6.11,182.514.830.50.4530.3180.3060.129 814.92,878.412.132.20.4120.2840.3010.114 942.66,089.48.433.90.3710.2520.3020.102 10357.323,746.715.935.80.3430.2340.3120.100 Notes: The table provides summary statistics for the 20 trading days prior to the flash crash (7 April–5 May 2010) for the universe of 6,173 equity instruments (stocks and ETPs) and for the 10 deciles of average daily dollar volume. ISO frequency is the average dollar-weighted frequency of intermarket sweep orders, C1 is the concentration ratio or the share of the top venue, and the Herfindahl indices measure the concentration in traded volumes and quote competition (averaged on the bid and ask sides). The maximum drawdown

on the day of the flash crash is also shown.

26www.cfa https://www.wendangku.net/doc/f110362471.html,?2012 CFA Institute

Exchange-Traded Funds, Market Structure, and the Flash Crash

when the index was 0.752. A secu lar decline in concentration (an increase in fragmentation) is evi-dent beginning in the second quarter of 2003. Sev-eral market stru ctu re changes are likely to have increased fragmentation in the past decade. Deci-malization began in a phased manner starting in early 2001, when stocks began trading for the first time in minimum price increments of one cent. The ability of traders to u ndercu t qu otes by one cent versus an eighth or a sixteenth of a cent led to greater competition among venues. The introduction by the U.S. SEC of Regulation NMS in 2005 was also asso-ciated with a sharp increase in competition among primary exchanges from other venues, the entrance of many new venu es (dark pools and electronic commu nications networks, or ECNs), internaliza-tion of flow by brokers, and the growth of higher-frequency trading. Many higher-frequency traders in particular prefer to trade on ECNs rather than on traditional exchanges.

By the end of the sample period in March 2012, the average stock’s Herfindahl index was approxi-mately 0.306. Althou gh fragmentation has been increasing over much of the past decade, the new levels are unprecedented and may represent a tip-ping point in terms of the vu lnerability of stock prices to an order flow shock or other impu lse. Indeed, equity market fragmentation is now at its highest level ever and dramatically higher than it was 18 years ago.

The intraday evolu tion of the fragmentation measu res on the day of the flash crash is also of interest. Figure 2 shows the evolution of the vol-ume and quote Herfindahl indices over the trading day on 6 May. For each one-minute interval, I cal-culated the volume and quote Herfindahl indices at the stock level and then estimated the sample mean across all stocks relative to the corresponding value for the same time interval on the day before (i.e., on 5 May). The figure shows five-minute mov-ing averages for both indices. The relative qu ote index decreased sharply at the time of the flash crash, and at its lowest point, it was almost 10% below the corresponding level the day before. This result suggests greater venue fragmentation in the late afternoon as off-exchange competition increased. In contrast, the corresponding volume fragmentation figu res instead show a marked increase in concentration at this time as intermarket linkages broke down and the NYSE entered its slow-trading mode. The divergent intertemporal behavior exhibited in Figure 2 is consistent with my previou s assertion that trade fragmentation and quote fragmentation capture different phenomena and can exhibit divergent behavior.

Examining the market shares of major venues over the day provides additional insight. Figure 3 shows the market shares of the major venu es—NYSE Arca, BATS, NASDAQ (combined), NYSE—as well as the trade reporting facilities (TRFs) and all other exchanges.11 For each five-minu te win-dow, I computed the market share of each venue as a percentage of dollar volume traded in the overall U.S. equity market. Market shares were relatively stable until the start of the flash crash, when inter-market linkages broke down and the NYSE’s mar-ket share dropped sharply. Later in the day, the market share of off-exchange venues declined and a return to normalcy occurred. The dynamic nature of competition among the venues is quite apparent.

Figure1.Daily Herfindahl Index for U.S. Equities and 50-Day Moving Average,

3 January 199

4 to 30 March 2012

July/August 2012www.cfa https://www.wendangku.net/doc/f110362471.html,27

Financial Analysts Journal

Figure2.Herfindahl Indices on 6 May 2010 Relative to 5 May 2010:

Five-Minute Moving Average

Figure3.Venue Market Shares on 6 May 2010 in Five-Minute Time Windows

28www.cfa https://www.wendangku.net/doc/f110362471.html,?2012 CFA Institute

July/August 2012

www.cfa pubs .org

29

Exchange-Traded Funds, Market Structure, and the Flash Crash

Determinants of Market Fragmentation.The complex interrelationships between the eco-nomic variables make it difficult to isolate the true determinants of fragmentation, necessitating a mul-tivariate analysis. I modeled the Herfindahl volume and quote concentrations, H i , in a given stock i as a fu nction of variou s asset-specific characteristics.Given that the dependent variable is directly related to the ratio of ou tcomes to trials (e.g., a venu e’s share in total volume or quote changes), I used a logistic regression and model fragmentation:

(4)

Here, is the logistic function and u i is a sto-chastic error term. The independent variables are chosen to captu re stock-specific factors and other controls. The most obvious of these are the primary listing exchange of the stock, proxies for trading activity (e.g., company size), and controls for asset type. I also included a measure of whether higher-frequency traders are active in the stock using inter-market sweep orders. Accordingly, I defined the independent variables as follows: NYSE , an indica-tor variable for whether the New York Stock Exchange is the primary exchange for the asset;log(MktCap ), the log of market capitalization (in millions of dollars); Volatility , the standard devia-tion of five-minute returns (scaled by 10–6) in the control period in the time window 1:30–4:00 p.m.;ISO , the average frequ ency of intermarket sweep orders over the 20 trading days prior to 6 May 2010;and ETP , an indicator variable for whether the asset type is an exchange-traded product.

Table 4 provides the logistic regression esti-mates for volu me and qu ote concentration, esti-mated separately. Because the dependent variable is a concentration measu re, negative coefficient signs imply more fragmentation. So, the negative sign on the NYSE indicator variable in both models implies more fragmentation for stocks whose pri-mary exchange is the NYSE than for those listed on other venues. Consistent with the earlier results, I found clear evidence that there is more concentra-tion in smaller-cap issues; across stocks, the volume Herfindahl index declines (i.e., fragmentation rises) as capitalization increases. Interestingly, this is not the case in the quote fragmentation model,which suggests that attributes other than size mat-ter in price competition. This result may be due to the fact that the underlying driver of competition (i.e., the profitability of qu ote-improving strate-gies) is complex and may have other determinants

beyond those captured in Equation 4. In both cases,there is no evident relationship between fragmen-tation and volatility. The ISO variable is highly significant and negative in both models, which indicates that ISO activity is positively associated with volume and quote fragmentation. This finding is consistent with the idea that high-frequency trad-ers use these types of orders to sweep limit order books to access all available liqu idity. The ETP dummy variable is not significant after controlling for other factors. I also estimated linear models for fragmentation and obtained the same results. Residual deviance for a logistic model is analogous to the residual sum of squ ares in a linear regression (it has a chi-squ are distribution), and I used it to assess the overall fit of the model. The results suggest that the model for volume fragmentation is a better fit than the equiv-alent quote model. It may be easier to interpret this “goodness of fit” in terms of corresponding adjusted R 2s in the linear specifications: 0.72 and 0.21 for the volu me and qu ote fragmentation models, respec-tively. The fact that some of the key variables are statistically significant means the logistic model (Equation 4) provides valuable information linking stock-specific factors to fragmentation.

H F u NYSE MktCap Volatili i i i i i c c (),log()

x x E E E E E E and

0123t y ISO ETP u i i i i E E 45.

F x i c E Table 4.

Cross-Sectional Determinants of Fragmentation

(standard errors in parentheses)

Volume Concentration

Quote Concentration

Intercept 0.646**–0.701**(0.091)(0.092)NYSE –0.189**–0.429**(0.068)(0.070)log(MktCap )–0.131**0.035*(0.018)(0.018)Volatility –0.142–0.077(0.547)(0.586)ISO –1.516**–0.514**(0.253)(0.260)ETP

0.0270.016(0.079)(0.080)Residual deviance 109.185164.255Null deviance 394.180211.626Degrees of freedom

6,155

6,155

Note: The table presents logistic models of Herfindahl concen-tration indices for volume and quote activity.*p < 0.05.**p

< 0.01.

30www.cfa pubs .org ?2012 CFA Institute

Financial Analysts Journal

Analysis of Drawdown. I turn now from the determinants of market fragmentation to an analy-sis of the role of market structure in explaining the pattern of maximum drawdown on 6 May 2010.12 As noted previously, I wanted to examine both quote and volume fragmentation measures (H q and H v ) in prior periods because my hypothesis is that stocks with greater fragmentation were more exposed to impulses that could trigger abrupt price declines.

The discu ssion in the previou s section high-lights the need for su itable controls at the stock-specific level, including liquidity, volatility, rout-ing behavior, and asset type. I used dollar volume as the proxy for liquidity. Because this variable is highly skewed to the right and approximately log-normally distributed, I used a log transformation in my models to dampen the impact of large-volu me ou tliers. For volatility, past research has shown a strong intraday seasonality that varies across stocks. Accordingly, I used an intraday mea-sure (five-minute return volatility) estimated over the afternoons in the control period. Given that the drawdown is essentially a retu rn, I inclu ded the inverse of the opening price on 6 May 2010 to

capture any bid–ask spread or other microstructure effects related to price level. I also included past ISO activity as a control for the propensity for higher-frequ ency traders to trade that particu lar stock u sing aggressive order techniqu es. I expected greater ISO activity to be associated with more fragmentation. Finally, regarding asset type, the results from Tables 2 and 3 indicate that it is impor-tant to control for whether the stock in question is an ETP or another type of equity.

Table 5 contains estimates of multiple regres-sion models where, for stock i , the dependent vari-able is the maximu m drawdown, M i , and the independent regressors inclu de market stru ctu re variables and other control variables:

(5)

The control variables are compu ted over the 20days prior to the flash crash and include

?log(ADV ), the log of average daily volume in

millions of dollars;

M H H ADV Volatility InvPrice i v i v q i q i i E E E E E E 011234,,log()

i i i i ISO ETP u E E 56.

Table 5.

Multivariate Analysis of Maximum Drawdown (standard errors in parentheses)

7 April–5 May 2010 5 May 2010

I

II III IV Intercept

0.147**0.166**0.089**0.130**(0.015)(0.016)(0.010)(0.012)Volume Herfindahl –0.182**–0.150**–0.043*–0.015(0.033)(0.031)(0.020)(0.020)Quote Herfindahl —–0.108**—–0.143**—(0.024)—(0.020)log(ADV )–0.0000.0010.003**0.002*(0.001)(0.001)(0.001)(0.001)Volatility –0.010–0.010–0.008–0.010(0.043)(0.043)(0.043)(0.043)InvPrice 0.0100.0080.001–0.002(0.006)(0.006)(0.006)(0.006)ISO –0.019–0.0130.0120.010(0.022)(0.022)(0.021)(0.021)ETP 0.181**0.182**0.171**0.178**(0.006)(0.006)(0.006)(0.006)Adjusted R 20.1310.1340.1410.136F -statistic

155.6

136.5

135.4

135.4

Notes: The table presents cross-sectional regression models of maximu m drawdown on 6 May 2010regressed on control and market structure variables (6,151 observations):

Here, is the Herfindahl concentration index for volume (quote) activity in the control period,7 April–5 May 2010, in Models I and II and on 5 May 2010 in Models III and IV.*p < 0.05.**p < 0.01.

M H H ADV Volatility InvPrice i v i v q i q

i i E E E E E E 011234,,log()i i i i ISO ETP u E E 56.H

H i v i q

Exchange-Traded Funds, Market Structure, and the Flash Crash

?Volatility, the average volatility measu red by the 20-day average of the daily standard devi-ation of five-minute return intervals scaled by 10–6 in the period 1:30–4:00 p.m.;?InvPrice, the inverse of the opening price on

6 May 2010;

?ISO, intermarket sweep order activity mea-sured by the dollar-weighted proportion of vol-ume accounted for by Condition Code F orders;?ETP, a dummy variable that takes a value of

1 if the asset is an ETP and 0 otherwise; and ?u i, a stochastic error term.

Table 5 presents four models, two each for the control period (7 April–5 May 2010) and the day before the flash crash (5 May 2010). I estimated Model I using only volume fragmentation (i.e., I set E1,q = 0), whereas for Model II, I allowed both volu me and qu ote fragmentation effects, where fragmentation is measu red over the previou s month. Models III and IV are identical to Models I and II, respectively, except that they use the most recent measure of fragmentation (i.e., the Herfin-dahl indices based on the day before the flash crash).

Recall that larger values of the measures H q and H v mean more concentration, so a negative coefficient implies that more fragmented stocks are associated with larger values of the drawdown coefficient. For Models I and III, with volume frag-mentation alone, the coefficient is negative in the control period, which is consistent with the hypothesis that more fragmented stocks experi-enced greater drops on 6 May after controlling for other factors. The coefficient is statistically signifi-cant at the 5% level for the control period (Model I), bu t it is not significantly different from zero u sing the fragmentation estimates from the day prior (Model III).

Of particu lar interest is that the inclu sion of qu ote fragmentation (Models II and IV) adds explanatory power. Both volu me and qu ote frag-mentation measu res are statistically significant at the 1% level in Model II, which u ses the control period data. For Model IV, which u ses the most recent day for the Herfindahl compu tations, the coefficient for volume fragmentation is statistically insignificant and the coefficient for quote fragmen-tation is negative and significant at the 1% level. This result confirms that volume fragmentation and qu ote fragmentation are different economic phe-nomena. Quote fragmentation is an important risk factor in explaining the propagation of the original liquidity impulse, consistent with the thinning out of order books in those stocks with the most aggres-sive quotation activity by higher-frequency traders. The importance of quote fragmentation in explain-ing the cross-sectional impact of the original liquid-ity shock highlights the importance of imperfect intermarket linkages, which are the root cau se of fragmentation. In contrast, O’Hara and Ye (2011) did not find evidence that market fragmentation harms market quality, possibly because imperfect intermarket linkages matter most in times of stress.

Volatility does not appear to be a predictor of drawdown, which indicates that the events of 6 May 2010 were not related to the normal patterns of risk. The inverse price variable is positive in Models I–III, which suggests larger drawdowns in lower-priced stocks, but it is not statistically signif-icant. Average daily volume effects are weak cross-sectionally; other control variables, including price, may capture the effect of liquidity. Prior ISO activ-ity has a negative coefficient in Models I and II but is not significant. As documented in Table 5, this variable is positively associated with fragmenta-tion, so the presence of the fragmentation variables already captures the impact of ISO activity. Note that omitting this variable has no real impact on the estimated coefficients or their significance levels.

The ETP indicator variable is positive and sig-nificant after controlling for other independent variables. Note that this resu lt does not reflect a failu re of ETF pricing. Rather, u ncertainty in the qu oted prices of component stocks makes it increasingly challenging for market makers as the normal arbitrage pricing mechanism breaks down. Of cou rse, market makers rou tinely make tight markets in ETPs where quotes on the underlying component secu rities are not available or timely (e.g., international ETPs), but in such cases, they are not exposed to risk arising from those securities’being traded simultaneously. The iShares Russell 1000 Growth Index Fund (IWF) provides an illus-trative case stu dy. Figure 4 plots the cu mu lative continuously compounded returns of the ETF and its intraday net asset valu e (NAV) in 60-second increments from 12:00 to 4:00 p.m.13 Prior to the flash crash, the ETF price closely tracked the intra-day net asset value of its constituent stocks, reflect-ing the smooth operation of the intraday creation/ redemption arbitrage mechanism. The tight rela-tionship of price and intraday net asset value held until about 2:45 p.m., at which point the constitu-ents of the underlying basket themselves could not be correctly priced. The uncertainty caused a tem-porary delinking of price and value. The ETF then experienced a sharp price decline bu t recovered rapidly, with its price again closely tracking its intraday net asset value by 3:10 p.m.

The multivariate results are robust to a number of alternative specifications and controls. Specifi-

cally, I estimated logistic regressions to account for

July/August 2012www.cfa https://www.wendangku.net/doc/f110362471.html,31

Financial Analysts Journal

the limited range of the dependent variable and reached the same conclu sions.14 I also estimated models that include primary exchange but found no evidence that primary exchange listing is a fac-tor in explaining the cross-sectional patterns of price declines; the same is tru e for models that inclu de asset type (other than ETP). Overall, the goodness of fit as measured by the adjusted R2 is more than 13%, which is relatively high given the dispersion in the dependent variable. Conclusion

Late in the afternoon on 6 May 2010, the sharpest intraday point drop in the history of the Dow Jones Indu strial Average occu rred. The so-called flash crash is distinguished from other market breaks by its speed, its rapid intraday reversal, and the fact that many stocks and ETPs traded at clearly unrea-sonable prices. This article highlights the role of equity market structure and the changing nature of liquidity provision in exacerbating the impact of an external liquidity shock, without taking a view as to its catalyst.

Specifically, I showed that the impact of the flash crash was greatest in stocks experiencing frag-mentation prior to 6 May. Both volume fragmenta-tion (which represents the actual pattern of trading activity across venu es) and qu ote fragmentation (which captures the dynamic competition for flow) are important in explaining the propagation of the crash. Using tick data for all stocks traded in the United States in 1994–2012, I showed that fragmen-tation is now at its highest level ever. This fact may partly explain why a similar flash crash did not occu r previou sly in response to some other cata-lyst.15 In particular, market structure may matter less when the markets are fu nctioning normally than in times of stress.

My research provides a framework to evaluate recent market structure debates in equities, deriva-tives, and other asset classes as new venu es and technologies erode the notion of a single, primary market for a security. This is not to say that I sup-port policies designed to increase market concen-tration at the possible expense of competition. Rather, my view is that recent policy proposals should be evaluated in the context of whether they address the root cau ses of fragmentation in the form of intermarket linkages that are inadequate or prone to failure in times of stress.16

?Uniform mechanisms across exchanges to curb extreme price volatility. Su ch mechanisms inclu de individu al secu rity circu it breakers and price bands (which use a “limit up/limit down” for price movements similar to those used in futures markets worldwide) to pause trading du ring disru ptions, allowing for contra-side liqu idity to emerge. Althou gh stock-specific circuit breakers would, by defini-tion, prevent extreme price movements, they

Figure4.Cumulative Intraday Returns and Net Asset Value for iShares

Russell 1000 Growth Index Fund, 6 May 2010

32www.cfa https://www.wendangku.net/doc/f110362471.html,?2012 CFA Institute

Exchange-Traded Funds, Market Structure, and the Flash Crash

pose some challenges. For example, restrictions on price movements at the single-stock level intended to protect investors could complicate the pricing of basket instru ments, su ch as ETPs, as shown in Figure 2. Further, the tiers of the limit u p/limit down are consistent between stocks in a basket and the corre-sponding ETF. So, trading in an ETF could be temporarily halted even thou gh the basket constitu ents continu ed to trade.17 The con-verse is also possible. Circuit breakers also do not resolve the underlying issue of fragmenta-tion, which remains a risk factor. Bethel, Lein-weber, Rübel, and Wu (2011) su ggested that market fragmentation signals gave warnings ahead of the flash crash on 6 May 2010. They noted that a more graduated approach relying on an early warning system for unusual mar-ket conditions based on such indicators could be more effective than a hard circuit breaker.?Clearer guidelines for intermarket order routing.

My results highlight the role of quote fragmen-tation and intermarket sweep orders. Policies designed to redu ce the likelihood of orders being routed to venues with little liquidity are thus critically important. My results that show

a relationship between ISO activity and frag-

mentation su pport the notion of a circu it breaker type of approach (see, e.g., Chakra-varty, Wood, and Upson 2010) to limit the use of aggressive sweep orders in times of market stress. A widely discussed idea is a “trade-at”

rule that would require off-exchange trades in dark pools or other internalization venues to be execu ted at prices better than the cu rrent national best bid or offer. The trade-at rule is controversial because it could reduce the prof-itability of brokers who internalize their flow.

Critics of the trade-at rule cite unintended con-sequences in the form of higher execution costs from information leakage, greater latency, and fewer crossing opportunities. The trade-at rule cou ld be difficu lt to implement and enforce because public quotes may not always be reli-able indicators of prices actually obtainable in the market. Without a better understanding of how the trade-at rule would operate, it is diffi-cu lt to predict whether the ou tcome wou ld affect fragmentation; it might simply redu ce intermarket competition.

?Greater transparency regarding trade error cancel-lation rules. Clear ru les regarding when exchange trades will be canceled may prevent the withdrawal of liquidity provision in times

of stress. If cancellation rules are arbitrary and nontransparent, liquidity providers, including market makers and some hedge fu nds, may fear that one side of their hedged trades may be canceled, exposing them to risk when they bu y a secu rity that has fallen in price while going short a related asset.

?Clearly defining the obligations of lead market mak-ers. There was previou sly no gu idance con-cerning minimu m qu oting standards for market makers to maintain two-sided markets.

Consequ ently, market makers commonly relied on “stub quotes” (i.e., offers to buy or sell at a su bstantial premiu m above or discou nt below the best bid or offer). Stub quotes were not intended to be execu ted bu t, rather, pro-vided a way for a dealer to participate only on one side of the market. The SEC eliminated stub quotes and implemented new rules that forced market makers to maintain continuous two-side quotations that are within a defined percentage around the best bid or offer. Given my results, this is a sensible approach to pre-venting the extreme trades (e.g., at pennies) that occurred during the flash crash.

?Audit trail. On 26 July 2011, the SEC, motivated by concerns that regulators lacked a complete view of the sequence of events during the flash crash, u nanimou sly passed the “large trader reporting rule.” The rule is intended to allow the SEC to reconstruct market events and aid investigations and enforcement actions. It is an important step toward a consolidated au dit trail that will give regulators the tools neces-sary to monitor trading patterns across multi-ple exchanges and improve enforcement, althou gh there are many technical qu estions that still need to be addressed, as noted in Bethel, Leinweber, Rübel, and Wu (2011).

In su mmary, my resu lts show that the flash crash can be linked directly to cu rrent market structure—in particular, the pattern of volume and qu ote fragmentation—which su pports the argu-ment that a lack of liquidity is the critical issue that requ ires the greatest policy attention. Consider-able progress on this issue has been made since the dramatic events of 6 May 2010. The safeguards and reforms that have been implemented in the U.S. equity markets should help slow down a potential future market disruption similar to the flash crash. Bu t they have not eliminated the possibility of another flash crash, albeit one with a different

catalyst or in a different asset class.

July/August 2012www.cfa https://www.wendangku.net/doc/f110362471.html,33

Financial Analysts Journal

The views expressed here are those of the author alone and not necessarily those of BlackRock, its officers, or its directors. This article is intended to stimulate further research and is not a recommendation to trade particular securities or of any investment strategy. Information on iShares ETFs is provided strictly for illustrative purposes and should not be deemed an offer to sell or a solicitation of an offer to buy shares of any funds that are described in this article. I thank Hering Cheng, Jeff Dean, Jessica Edrosolan, Michael G ates, Joe G awronski, Bhavna Kapoor, David Leinweber, Marcia Roitberg, Richard Rosenblatt, and Mike Sobel for their helpful suggestions.

This article qualifies for 1 CE credit.

Notes

1.See, for example, Barr (2010), who stated, “Whatever their

cause, the frequent market outages only feed the sense that the entire market is either a casino rigged by the money never sleeps crowd or a hou se of cards on the verge of collapse. Neither view, it seems safe to say, is apt to restore investors’ dwindling confidence.”

2. A common (inverse) measure of fragmentation is the Her-

findahl index, which is simply the sum of the squares of market shares. For a particular stock, if the shares in dollar volume of venues A, B, and C are 50%, 40%, and 10%, the Herfindahl index is 0.42 (= 0.25 + 0.16 + 0.01). If the relative frequencies with which venues A, B, and C become the best bid or offer (as a fraction of all quote changes) are 80%, 10%, and 10%, the index wou ld be 0.66 (= 0.64 + 0.01 + 0.01), indicating less fragmentation. Ignoring the sub-breakdown within dark pools, the overall Herfindahl indices for vol-u me, qu otes, and depth in December 2011 for the U.S.

market as a whole were 14.8%, 21.0%, and 25.5%, respec-tively. Clearly, the average of stock-level Herfindahl indices would produce a higher number because some stocks trade primarily in one venue.

3.Exchange-traded funds (ETFs) and exchange-traded notes

(ETNs) are subsets of exchange-traded products. In an ETF, the underlying basket securities are physically represented, whereas an ETN is senior, unsecured, and uncollateralized debt that is exposed to credit risk. ETPs account for up to 40% of U.S. trading volume.

4.See, for example, Wu rgler (2011). Ramaswamy (2010)

examined the operational frameworks of exchange-traded fu nds and related them to potential systemic risks. The role of leveraged ETFs has also been discussed (see, e.g., Cheng and Madhavan 2009) in the context of end-of-day volatility effects.

5.Faced with increased volume, the NYSE entered “slow-

trading mode” while stocks continu ed to trade in elec-tronic venu es, su ch as BATS, resu lting in price distor-tions. Liqu idity providers began to withdraw their liqu idity, given concerns that some trades wou ld be canceled u nder the “erroneou s trade ru le,” resu lting in some market sell orders—inclu ding stop loss orders—being executed at pennies.

6.See Bowley (2010b) and Barr (2010).

7.See Bowley (2010a).

8. A millisecond is one-thou sandth (10–3) of a second.

Brogaard (2010) noted that some high-frequ ency traders execute trades with round-trip execution times measured in microseconds (i.e., in one-millionth [10–6] of a second).

9.The NASDAQ Capital Market (1,285 names) consists of the

smaller companies traded on the NASDAQ National

Market. The NASDAQ Global Market (900 names) com-prises the middle-tier companies on the NASDAQ National Market. The NASDAQ Global Select Market (375 names) represents the highest-cap companies on the NAS-DAQ National Market.

10.Borkovec, Domowitz, Serbin, and Yegerman (2010) argued

that ETF market makers withdrew liquidity after suffering severe losses.

11.Alternative execution facilities, such as ECNs and broker/

dealers, are required to report U.S. equity trades away from exchanges through TRFs. Other exchanges include Amex, the Boston Stock Exchange, the Chicago Stock Exchange, and the National Stock Exchange.

12.The results also hold for other measures of the magnitude

of the flash crash, inclu ding intraday volatility after 2:40 p.m. on 6 May 2010 relative to benchmark afternoon (1:30–4:00 p.m.) volatility for the previous 20 days.

13.Intraday prices are from the TAQ database; net asset values

are compu ted u sing market capitalization weights at the beginning of the day.

14.Note that there are differences in the empirical distributions

of the market structure and return variables that informed my choice of model. That is, the fragmentation variables reflect the outcome of different trials (e.g., a venue’s share in volu me), whereas the drawdown measu re is a retu rn, albeit one constrained to a certain interval.

15.Of cou rse, secu rity prices can exhibit sharp price move-

ments over short time intervals even in a centralized market stru ctu re. For example, on 28 May 1962, the DJIA fell sharply and IBM’s stock price fell 5.3% in 19 minutes. The decline was broad (1,212 issues declined and only 74 rose), but unlike the flash crash, stocks did not trade at absurd prices. The cause of that event is unclear—perhaps a short-lived panic.

16.Some of the more fundamental market structure recom-

mendations are from the Joint CFTC-SEC Advisory Com-mittee on Emerging Regulatory Issues—a body created by legislation to investigate the flash crash (see CFTC and SEC 2010).

17.The cu rrent limit u p/limit down proposal has an u pper

band and a lower band beyond which trading cannot take place. For Tier 1 stocks—which include those in the S&P 500 and the Russell 1000 Index as well as 344 ETFs—the upper and lower bands are 5% of the average price of the security over the preceding five minutes. For Tier 2 stocks (all other securities), the upper and lower bands are 10%. If one of these bands is reached and all orders above or below the band limit are neither canceled nor execu ted within 15

seconds, then a five-minute pause will occur.

34www.cfa https://www.wendangku.net/doc/f110362471.html,?2012 CFA Institute

Exchange-Traded Funds, Market Structure, and the Flash Crash References

Barr, Colin. 2010. “Progress Energy Joins Flash Crash Crowd.”Fortune (27 September): https://www.wendangku.net/doc/f110362471.html,/2010/ 09/27/progress-energy-joins-flash-crash-crowd.

Ben-David, Itzhak, Francesco Franzoni, and Rabih Moussawi. 2011. “ETFs, Arbitrage, and Contagion.” Working Paper 2011-20, Dice Center, Ohio State University.

Bethel, E. Wes, David Leinweber, Oliver Rübel, and Kesheng Wu. 2011. “Federal Market Information Technology in the Post Flash Crash Era: Roles for Supercomputing.” Working paper, Lawrence Berkeley National Laboratory (September). Borkovec, Milan, Ian Domowitz, Vitaly Serbin, and Henry Yegerman. 2010. “Liquidity and Price Discovery in Exchange-Traded Funds: One of Several Possible Lessons from the Flash Crash.” Journal of Index Investing, vol. 1, no. 2 (Fall):24–42. Bowley, Graham. 2010a. “Stock Swing Still Baffles, with an Ominous Tone.” New York Times (22 August).

———. 2010b. “The Flash Crash, in Miniature.” New York Times (8 November).

Bradley, Harold, and Robert E. Litan. 2010. “Choking the Recov-ery: Why New Growth Companies Aren’t Going Pu blic and Unrecognized Risks of Future Market Disruptions.” Research report, Ewing Marion Kauffman Foundation (November). Brogaard, Jonathan. 2010. “High Frequ ency Trading and Its Impact on Market Quality.” Working paper, Kellogg School of Management, Northwestern University.

CFTC and SEC. 2010. “Findings Regarding the Market Events of 6 May 2010.” Report by the U.S. Commodity Futures Trading Commission and the U.S. Securities and Exchange Commission (30 September).

Chakravarty, Sughato, Robert A. Wood, and John Upson. 2010.“The Flash Crash: Trading Aggressiveness, Liquidity Supply, and the Impact of Intermarket Sweep Orders.” Working paper. Cheng, Minder, and Ananth Madhavan. 2009. “The Dynamics of Leveraged and Inverse Exchange-Traded Funds.” Journal of Investment Management, vol. 7, no. 4 (Fourth Quarter):43–62.Easley, David, Marcos López de Prado, and Maureen O’Hara. 2011. “The Microstructure of the ‘Flash Crash’: Flow Toxicity, Liquidity Crashes, and the Probability of Informed Trading.”Journal of Portfolio Management, vol. 37, no. 2 (Winter):118–128. Egginton, Jared F., Bonnie F. Van Ness, and Robert A. Van Ness. 2011. “Quote Stuffing.” Working paper.

Hasbrou ck, Joel, and Gideon Saar. 2010. “Low-Latency Trad-ing.” Working paper, New York University.

Hendershott, Terrence J., and Pamela C. Moulton. 2011. “Auto-mation, Speed, and Stock Market Quality: The NYSE’s Hybrid.”Journal of Financial Markets, vol. 14, no. 4 (November):568–604. Hendershott, Terrence J., Charles M. Jones, and Albert J. Menk-veld. 2011. “Does Algorithmic Trading Improve Liqu idity?”Journal of Finance, vol. 66, no. 1 (February):1–33.

Kirilenko, Andrei, Albert S. Kyle, Mehrdad Samadi, and Tugkan Tuzun. 2010. “The Flash Crash: The Impact of High Frequency Trading on an Electronic Market.” Working paper, University of Maryland (May).

O’Hara, Maureen, and Mao Ye. 2011. “Is Market Fragmenta-tion Harming Market Quality?” Journal of Financial Economics, vol. 100, no. 3 (June):459–474.

Ramaswamy, Srichander. 2010. “Market Stru ctu res and Sys-temic Risks of Exchange-Traded Funds.” Bank for International Settlements Working Paper 343.

Wu rgler, Jeffrey. 2011. “On the Economic Consequ ences of Index-Linked Investing.” In Challenges to Business in the Twenty-First Century. Edited by Gerald Rosenfeld, Jay W. Lorsch, and Rakesh Khurana. Cambridge, MA: American Academy of Arts & Sciences.

Zhang, Frank. 2010. “High-Frequency Trading, Stock Volatil-ity, and Price Discovery.” Working paper, Yale University

(December).

July/August 2012www.cfa https://www.wendangku.net/doc/f110362471.html,35

Reproduced with permission of the copyright owner.Further reproduction prohibited without permission.