Who s Afraid of the Dark? Trading Off-Exchange

Size: px

Start display at page:

Download "Who s Afraid of the Dark? Trading Off-Exchange"

Jocelyn Short
5 years ago
Views:

(1 ( Who s Afraid of the Dark? Trading Off-Exchange Ana Avramovic 1--48 ana.avramovic@credit-suisse.

Increasing volume in dark pools and on the TRF generally raises some concerns, but it also provides benefits. We show that market quality in the TRF does not suffer compared to lit markets.

1 (1 ( Who s Afraid of the Dark? Trading Off-Exchange Ana Avramovic ana.avramovic@credit-suisse.com Market Commentary 17 April 01 % of all consolidated volume % % 1% Key Points Dark pools may sound scary, but that shouldn t be the case. Increasing volume in dark pools and on the TRF generally raises some concerns, but it also provides benefits. We show that market quality in the TRF does not suffer compared to lit markets. Instead, characteristics of the TRF are a direct consequence of our market structure rules. Thanks to competition and free markets, traders can vote with their volume. The best ideas and products tend to survive for a reason. Exhibit 1: US market share breakdown Other TRF 17.1% Dark Pools 1. Other Exchanges 4.1% DirectEdge 9.% Nasdaq BX.% Nasdaq 18. BATS 8.% BATS BYX.% Arca 11.% 11. Nasdaq Arca BATS BATS BYX Nasdaq BX DirectEdge Other Exchanges Other TRF Dark Pools "Other TRF" includes broker capital commitments and internalizations "Other Exchanges" includes Amex, CBSX, Chicago, National and PSX Data for February 01 Exhibit : Market share of the vs. TRF TRF % Mar 07 Mar 08 Mar 09 Mar Mar 11 Mar 1 Where Does Trading Happen? All over, thanks to fragmentation Fragmentation is not a new story for US equity markets. There are currently 1 exchanges and around 0 dark pools, numbers that have grown considerably over the years. But is that a good thing? Free market economics dictate that a wide array of choices for where to trade should ensure that the best offerings prosper, and in that competitive fight for market share, the end user will benefit. If only it were so simple. Concerns with the current structure Hard to find liquidity: It s too hard to find liquidity when it s disaggregated across so many venues. Sophisticated traders are able to avail themselves of SOR technology or relationships on cash trading desks at brokerage firms to find that liquidity, but where does that leave money managers without such resources at their disposal? Impedes price discovery: Only posted bids and offers convey the market s interest in a security and allow for price discovery (pretrade). As an increasingly large share of trading occurs away from lit markets, the absence of posted quotes distorts the supply/demand picture and impedes price discovery. Benefits Competitive forces: Trading venues can offer variations on pricing, order types and sizes, or other trade characteristics in order to cater to a particular type of flow. Minimizes impact: The anonymity of dark pools helps traders to minimize their impact. This is particularly useful for hard-to-trade stocks with wide spreads, or for large, information-based trades with high signaling risk. Technological advances: Lest we think that only the superconnected high frequency shops can avail themselves of the latest tools, and that they are the ones prowling in dark pools, it is interesting to point out that the most popular algorithm among longonly firms is actually one that never displays in lit markets. With dark pools recently capturing a record share of all US equities trading (just over 1% in January 01), and exchanges continuing to lose share, these issues are more relevant than ever. So, is fragmentation good?

2 % of day's volume Got a Better Idea? If you could design a new market structure from scratch, what would it look like? Would you allow dark pools? Would you impose minimum obligations on high frequency traders? Slow down markets and restrict algo usage? Before answering, consider the objective. For longonly institutional investors, the role of the market is to allow them to implement trade ideas without others benefiting from their proprietary research. However, if the efficient market hypothesis holds, the market will detect and incorporate their research almost immediately, stripping the long-only manager of his alpha potential. How does this happen? Any order even those in the dark exposes a trade to signaling risk since it must be printed to a single consolidated tape immediately after execution. The only way to avoid signaling intention before a trade is to stay in the dark where all traders are anonymous and do not advertise quotes. The buy-side seems to have caught on to this, because Guerrilla an algorithm designed to avoid making its presence known is the most popular algo among long-only accounts. And how is Guerrilla able to do this? By refraining from posting in lit markets and taking advantage of dark liquidity. 1 1% 6% % Exhibit 4: Volume curve by exchange (S&P 00 stocks) 9:0 :00 :0 11:00 11:0 1:00 1:0 1:00 1:0 :00 :0 :00 :0 Trading on the Primary Long gone are the days when was the place to trade. Nasdaq stepped onto the scene and created a bifurcated market for listings, but even holding the designation of listing venue no longer counts for much. shows that the amount of trading occurring on the primary (listing) exchange is only around, with less liquid stocks trading a higher percentage on the primary. % on PRIMARY Exhibit : % of Volume traded on primary exchange vs stock s ADV Average Daily Volume (shares) Russell 000 S&P 00 Furthermore, most of the trading that does happen on the is clustered around the open and close (exhibit 4), since the primary venue still sets the official MOO and MOC prices (note that the volume curves for other venues may be influenced by their pricing structure; see p.). For stocks with low liquidity, a significant portion of their ADV occurs in just the single MOC print (>), because it may be hard to find enough liquidity scattered throughout the day to be able to trade in any meaningful size otherwise. That single MOC print is a big part of why the primary venue is more relevant for less liquid stocks (exhibit ). This skew is only present for the %traded on the primary, not the % traded on the TRF (Trade Reporting Facility; see box on page ). As exhibit 6 on page shows, most stocks trade around % on the TRF, regardless of the stock s liquidity. BATS ARCA EdgeA EdgeX Nasdaq TRF Exhibit : Average Market-On-Close volume by stock s liquidity MOC volume on Primary / Total Volume 1% 6% % listings Nasdaq listings <0K 0K mil mil mil >mil month ADV (shares)

3 Exhibit 6: % of volume traded on TRF vs stock s ADV g y % on % PRIMARY on TRF Dark Pools % of US Volume 7 6 Russell 000 S&P 00,000 0,000 1,000,000,000,000 Average Daily Volume (shares) Exhibit 7: Dark pool market share vs VIX Index 16% 1 1% 6% % Jan 08 Apr 08 Jul 08 Oct 08 Jan 09 Apr 09 Jul 09 Oct 09 Jan Apr Jul Oct Jan 11 Apr 11 Jul 11 Oct 11 Jan 1 Dark Pools market share VIX Index VIX Index The Alternatives and the Trade Reporting Facility (TRF) Dark pools shouldn t be as scary as they sound. The only real difference between dark pools and exchanges is that dark pools are not required to display bids and offers. After a trade is consummated, you will see it printed to the tape immediately, whether it occurred on an exchange or in the dark. Another distinction is that dark pools do not identify themselves directly on the consolidated tape and all print using the TRF moniker, as opposed to exchanges which each have their own mnemonic. The trade is still printed publicly for all to see (see box at left for more). Volumes confirm popularity The fact that dark pools maintain anonymity and do not display quotes is the reason they are popular with long-only managers. When the buy-side wants to trade a high conviction idea, they do not want to lose precious alpha by signaling their presence. Exhibit 1 on page 1 showed the dramatic growth in TRF activity currently more than 1 in every US trades which confirms the strong demand to trade off-exchange. If traders didn t realize the benefits, they would not do so. We also note that TRF trading tends to be inversely related to volatility. When vol is high, traders are less willing to take on the opportunity cost of waiting for a cross. Consistently lower volatility since last August has contributed to the shift towards dark pools and the TRF (exhibit 7). Regulatory actions The SEC introduced two proposals in 009 that would i) prohibit actionable IOIs, and ii) lower the threshold for mandatory public dissemination of quotes. The Commission has been focused on Dodd Frank implementation and has not taken action on these proposals in the. years since introduction, but regulatory risk remains high. The other main part of TRF volume, internalizations, also came under scrutiny in 011 when the proposed a retail liquidity program that would offer price improvements to selected retail customers. The proposal was an attempt to combat wholesalers who, they believe, are taking liquidity away from the exchanges. The SEC has extended the time period in which they plan to take action until this summer. Source: Rosenblatt Securities What s in a name? Defining the terms A dark pool, as the name implies, does not post bids and offers before a trade. But, immediately after execution, the trade is printed to the consolidated tape (same as for exchanges) Dark pools do not report directly to the tape, as exchanges do. Instead, they report to a Trade Reporting Facility (TRF), which then prints to the tape. The TRF also collect trades from broker-dealer internalizations and other Alternative Trading Systems (ATSs) that are not dark. All dark pools are ATSs, but ATSs do not have to be dark pools. Note that some exchanges also offer dark/hidden order types, which are undisplayed orders held on the exchange. These have the same implications on price discovery and market impact as orders in a dark pool would.

1 6 78 9 1% Average %Volume on TRF Exhibit 8: Average g % of volume yon TRF by share price 4% % % 1% "regular" stocks % stocks priced >$0 Average %Volume on TRF Size quoted on best bid + best offer

Exhibit : Average volume and share price in TRF by market cap 0 4 40 0 0 1 $7 9% 9% $9 % $0 7% $ Exhibit 11: Average quote size at NBBO (S&P 00 stocks priced >$) 004 00 006 007 008 009 0 011 1 4 6 7

because stocks that report more through the TRF tend to exhibit wider spreads, higher volatility, and higher impact costs.

Market structure matters It is true that cheaper stocks tend to trade more on the TRF (exhibit 8). This should not be surprising because 1 cent is currently the minimum tick size in the US.

So, minimizing spread costs by executing at the mid as opposed to the quoted bid/ask becomes more significant for low priced stocks.

4 % Average %Volume on TRF Exhibit 8: Average g % of volume yon TRF by share price 4% % % 1% "regular" stocks % stocks priced >$0 Average %Volume on TRF Size quoted on best bid + best offer 11% 41% 7% <$ % 11% % $ 0 $0 0 7% $0+ Stock Price Exhibit 9: Average tick size of S&P stocks in 011 Average tick size of S&P 00 stocks in % % 8 9 % 7% 11% More % 4 % % 6 8 % of stocks Exhibit : Average volume and share price in TRF by market cap $7 9% 9% $9 % $0 7% $ Exhibit 11: Average quote size at NBBO (S&P 00 stocks priced >$) $60 $4 $0 $1 $0 large mid small micro Market Cap Size More Average Share Price Market Quality on the TRF Some academics have reported that fragmentation and internalization erodes market quality because stocks that report more through the TRF tend to exhibit wider spreads, higher volatility, and higher impact costs. However, we argue that all three of those points are directly related, and they re also a consequence of confusing cause with effect. Market structure matters It is true that cheaper stocks tend to trade more on the TRF (exhibit 8). This should not be surprising because 1 cent is currently the minimum tick size in the US. Since many stocks already trade 1 penny wide (exhibit 9), the spread is much larger relative to price for low priced stocks. So, minimizing spread costs by executing at the mid as opposed to the quoted bid/ask becomes more significant for low priced stocks. The only way to achieve sub-penny pricing for a 1-cent-wide spread is by using hidden/dark order types. Thus, stocks that have higher TRF volume do tend to have wider spreads (in basis points), just as the studies claim, but this is logical because a midpoint execution is more meaningful when the spread is wider, and spreads (in bps) tend to be wider for low priced stocks. Another way to examine off-exchange activity is by looking at the breakdown by market cap. Largecaps tend to have higher average share prices and less activity in the TRF. Conversely, microcaps have much lower prices and trade more on the TRF (exhibit ). One final point to note as to why stocks with more activity off-exchange are more volatile is that low priced/small cap stocks tend to have a higher beta. By definition, they will fluctuate more than the market s average. Now we ve established that stocks that trade more in the TRF tend to be: Low priced Smallcaps High beta Given that, it is not surprising that they would exhibit wider spreads, higher volatility, and higher impact costs. This is largely a function of our current market structure (namely, tick sizes), not a function of market quality. Displayed size remains strong Interestingly, even as volume is migrating away from the lit markets, we haven t seen displayed size suffer. As exhibit 11 shows, posted size at the NBBO has actually been increasing over time (008 being the understandable exception, as prudent risk management would mandate scaling back size). Even more important for our argument here, the data holds up for both regular stocks (S&P stocks priced >$), and for smallcaps (Russell 000). European study finds that price discovery is unharmed In an earlier study of dark pools in Europe (Measuring Dark Pools' Impact), we found no evidence that they damage price discovery. Our detailed report also highlights the cost of posting on lit books and the performance benefits of dark pools. 4

5 Conclusion As regulations have changed and technology has improved, the equity market has evolved. As with any competitive market, the competition has attracted new entrants, each hoping to differentiate themselves and win market share with new offerings. While the changes may not appeal to everyone, traders can express their preferences with trading volumes, and the best ideas should win out. For those who have lost confidence in today s markets and simply call for increased regulation, we must remember that the current complexity does not easily allow for targeted legislation, as so many parts of our market s microstructure are so closely related. Any new laws will almost certainly have implications beyond what is written. As one example of the tradeoff: to critics hoping to reign in technological developments, we point out that a slower market would expose market makers to more risk, leading to wider spreads and less depth. Rather than resisting the market s natural evolution, we should have faith that in a free market, competition drives innovation. Our current market structure is the way it is because customers will embrace the best ideas, and as Darwin teaches us, those are the changes that should survive. May the Best Model Win A 1- combo of new competitors and declining volumes have conspired against exchanges and upped the ante in their fight for market share. As we would hope would happen in a competitive market, venues have approached this problem by introducing new order types and pricing schemes, some of which have fared better than others. Key moments in this market evolution: Hidden orders (1999): Reg ATS permits hidden order types: liquidity that rests undisplayed within a lit market. Functionally, these are the same as orders in dark pools. According to data from Rosenblatt Securities, exchange dark liquidity accounts for of all consolidated volume today. Maker-taker pricing (1997): in a dramatic departure from the traditional access-fee-only model, Island ECN started charging users for taking liquidity and rebating those who provided liquidity. Inverted pricing (006): Taking the innovation further, BATS launched an aggressive campaign to buy volume. They inverted the maker-taker model and actually paid traders to take liquidity (what could be better?) This became a popular promotion and was subsequently copied by other venues looking to ramp up market share quickly. Flash orders (007): This controversial practice was invented by the CBSX and employed by a handful of ATS s. The process worked as follows: if the venue could not fill an order and had to route it away, they first flash it to a select group of participants, hoping that someone would agree to step up and take the order. The venues argued that this process saved on routing fees and resulted in price improvement, but others argued it gave an unfair look to privileged traders and created information leakage. The SEC introduced proposals to ban the practice in 009, but any exchanges that once used flash orders have voluntarily stopped doing so.

6 Disclaimer: Please follow the attached hyperlink to an important disclosure: Structured securities, derivatives and options are complex instruments that are not suitable for every investor, may involve a high degree of risk, and may be appropriate investments only for sophisticated investors who are capable of understanding and assuming the risks involved. Supporting documentation for any claims, comparisons, recommendations, statistics or other technical data will be supplied upon request. Any trade information is preliminary and not intended as an official transaction confirmation. Use the following links to read the Options Clearing Corporation s disclosure document: Because of the importance of tax considerations to many option transactions, the investor considering options should consult with his/her tax advisor as to how taxes affect the outcome of contemplated options transactions. This material has been prepared by individual traders or sales personnel of Credit Suisse Securities Limited and not by the Credit Suisse research department. It is provided for informational purposes, is intended for your use only and does not constitute an invitation or offer to subscribe for or purchase any of the products or services mentioned. The information provided is not intended to provide a sufficient basis on which to make an investment decision. It is intended only to provide observations and views of individual traders or sales personnel, which may be different from, or inconsistent with, the observations and views of Credit Suisse research department analysts, other Credit Suisse traders or sales personnel, or the proprietary positions of Credit Suisse. Observations and views expressed herein may be changed by the trader or sales personnel at any time without notice. Past performance should not be taken as an indication or guarantee of future performance, and no representation or warranty, expressed or implied is made regarding future performance. The information set forth above has been obtained from or based upon sources believed by the trader or sales personnel to be reliable, but each of the trader or sales personnel and Credit Suisse does not represent or warrant its accuracy or completeness and is not responsible for losses or damages arising out of errors, omissions or changes in market factors. This material does not purport to contain all of the information that an interested party may desire and, in fact, provides only a limited view of a particular market. Credit Suisse may, from time to time, participate or invest in transactions with issuers of securities that participate in the markets referred to herein, perform services for or solicit business from such issuers, and/or have a position or effect transactions in the securities or derivatives thereof. The most recent Credit Suisse research on any company mentioned is at 01, CREDIT SUISSE 6

July 2012 Chartbook The Halftime Report

July 2012 Chartbook The Halftime Report Average Daily $VA LUE Traded ($Billions ) $Billions (212 ( US China Japan CHI-X London Hong Kong Germany France Canada Korea Australia Brazil Taiwan Spain India Italy $billions Switzerland Sweden Amsterdam