Best Execution, Trading and Market Centers

CFTC-SEC Joint Committee Report Notes

2011-02-19T17:34:00.009-05:00

YesYesterday the blue ribbon committee (membership list below) that the CFTC/SEC appointed to examine regulatory risks and make recommendations on how to avoid another flash crash released this report. Most of it reiterates support for measures that have already been proposed or implemented such as single stock pauses/circuit breakers, but there are a few interesting items.

Commends the SEC for their sponsored access approach (i.e., pre-trade risk mgmt) and pushes the CFTC to do the same. (Could be very positive for Sponsored Access providers like FTEN/Nasdaq, ULLink, Quanthouse, etc.)
The committee comes out in favor of incenting high frequency traders to make ‘real’ markets. (Could be very positive for Market Making 2.0 operations such as TradeBot & Virtu).
Recommends a ‘trade-at’ rule which would effectively penalize internalization and preferencing activities by 50 mils—possibly making it an unprofitable business model. (Very negative for traditional wholesale models like Knight, Citadel and Getco, as well as broker-owned dark pools. Potentially positive for non-broker-owned dark pools like LiquidNet, Pipeline and PDQ).

I think the most controversial aspect of the report will be the trade-at rule recommendation. Besides being completely outside the purview of the Committee’s charter, there are too many uncertainties, such as how many wholesale market makers (Knight, Citadel, etc.) might exit the business and what effect would that have on liquidity. So I doubt a rule will get passed or if one does, it will be watered down enough to be largely ineffective.

Committee Members
Brooksley E. Born, Retired Partner, Arnold & Porter LL.P, Former Chairman CFTC
John J. Brennan, Chairman Emeritus and Senior Advisor, Vanguard Group
Robert F. Engle, Michael Armellino Professor of Finance, Leonard N. Stern School of Business, New York University
Richard G. Ketchum, Chairman and Chief Executive Officer, FINRA
Maureen O’Hara, Robert W. Purcell Professor of Finance, Cornell University
Susan M. Phillips, Dean and Professor of Finance, The George Washington School of Business, Former CFTC Chairman
David S. Ruder, Professor of Law, Emeritus, Northwestern University School of Law, Former SEC Chairman
Joseph E. Stiglitz, Professor of Finance and Business Administration, Columbia University

How Low Is Low (Latency)?

2010-02-05T09:30:00.002-05:00

Historically, very few brokers or exchanges actually measured latency of their trading and matching systems, at least not in a way that invited comparison. In the cases where an entity talks about latency numbers they are usually expressed as turnaround times inside a particular system, for example in this article from Wall Street & Technology:

Nasdaq's Hyndman says that migrating to INET has allowed Nasdaq to offer the fastest possible transaction times. "We've reduced latency from 10 milliseconds down to 1," he says, referring to the time it takes once an order is placed on Nasdaq's system to acknowledge it electronically.

Unfortunately, the numbers can pretty much be made to say whatever you want them to say. The biggest problem (from a high-frequency trading standpoint) are that these types of measurements usually don't take into account peak loads on their systems, which tend to occur at the exact times that low-latency is most important to the trading strategy. I find it hard to believe that Nasdaq's expressed turnaround time is correct at 9:30AM, 12:00 noon, and 3:59PM. (Note that the above article was written in 2007 and I'm pretty sure Nasdaq--as well as their competition--have decreased their response times since then).

More recently, in order to attract high frequency order flow some markets have begun to figure out ways to bring more credibility to their metrics and measurements. This article mentions some of the initiatives from NASDAQ, NYSE and BATS. As you will quickly realize, however, one is pushing Correlix, one is pushing SeaNet and one is doing it themselves--good luck comparing those results.

So the fact is that there aren't really any 'standard' metrics for latency measurement of trading systems. This would seem to be a good area for the industry to create a working group to foster cooperation in. Until then, pretty much any reasonable approach can be used and defended by vendors. To be on the safe side and facilitate comparison many just settle for average turnaround.

If I were looking at a report, I would want to start with average times, but some statistics about maximums under some loading conditions would be very helpful as well. The latter would take into account the system's ability to handle increased market data traffic as well as order generation.

Early thoughts on the SEC Sponsored Access Proposed Rule

2010-01-21T10:54:00.003-05:00

I've been working on an analysis of the Sponsored Access rule, all 95 pages of it. The SEC wants firms that send trades to exchanges electronically (and which ones don't?) to implement electronic pre-trade checks that make sure their algos aren't going wild. The short story is that so far I don't see these requirements as overly onerous. Here are some preliminary thoughts:

1. The rule applies to all exchange and ATS connectivity, not just sponsored access. So all trading emanating from a B/D, including prop, program, algo, market maker, dma, etc., will be subject to the same risk checks. That means that virtually everyone will be on a level playing field (speed-wise) with the HFT guys being sponsored. Or phrasing it differently, no one will be at a competitive disadvantage by being subject to the risk management controls.

2. The CEO of the firm has to personally attest that proper risk controls are in place for all electronic trading. That's right, the CEO. So if you're running an HFT business you better be prepared to explain to him how your business works in very minute detail, because it's his butt on the line. Which by extension means your butt. (Can you imagine having to ask Jamie Dimon to sign off on this? Brian Moynihan?)

3. For the first time, the SEC has actually defined exactly what a broker should be monitoring for, i.e., capital/margin, relationship to last price, obvious errors, Reg SHO, Reg NMS, etc. I see this as going a long way toward relieving brokers' lawyers concerns about compliance, i.e., "If this list of checks is good enough for the SEC, it's good enough for me," effectively eliminating one of the existing blocks on sponsored access.

4. The fact that the SEC has put forward these rules will make it easier to extend sponsored access to other asset classes.

5. Winners:

Exchanges, because it will be much easier for clients to directly connect and there will be more brokers offering the service,
FTEN and their nascent competitors because many brokers will turn to them as an outsourced means of compliance with the rule. However, some CEOs may balk at being held responsible for the actions of a small technology firm.

6. Losers:

Firms that have previously spent lots of money building direct exchange access infrastructure, because clients will gravitate toward utility providers like FTEN
Electronic agency brokers because the value of the service they provide will become harder to differentiate.

Dark Pool Letters to the SEC

2010-01-16T21:37:00.001-05:00

Let's get 2010 off to a good start by looking at some of the letters received by the SEC about their proposed "Regulation of Non-Public Trading Interest" or as it is more commonly known, dark pool regulation. To summarize, it appears that most of the industry is supporting either dark pools or dark liquidity in one form or another, while most non-industry commentators are opposing it. Unfortunately, most of the non-industry letters don't specifically address the issues the Commission is looking to regulate, i.e., actionable IOIs, the display obligation threshold, and trade reporting. The public seems to overwhelmingly favor an "all-or-none" approach, that being something close to summary execution (and I don't mean order flow).

The buy-side is largely absent from the comments so far. Larry Tabb has taken up their cause with some interesting statistics he has gathered through his research. Liquidnet has admirably submitted one of the few letters that goes beyond its authors' own self interests.

The award for most thoughtful comments, at least to date, goes to Steve Wunsch. He has turned the arguments of most dark liquidity critics inside-out, and in doing so makes sense of where we are, how we got here, and where we're going in terms of market structure. It's also pretty safe to say that Steve won't be working at the SEC anytime in the near future. To wit: "The Commission has, after all, been doing nothing but imposing transparency and fairness for three and a half decades, and it clearly isn’t working." Required reading!

For those who are fans of complete irrelevance, we're seeing some familiar themes like motherhood, apple pie, and short sales but we're also beginning to see the introduction of new 'worries' like after-hours trading and tax evasion. The power of God is invoked in some letters, something I'm sure the Commission is thrilled about. And to make things even more ominous, members of the plaintiffs' bar are now weighing in.

The deadline for submitting comments on this important proposal, release number 34-60997 is February 22, 2010. Just click here to make your views known.

Bill

"It is common in comment letters to praise the Commission and the work it is doing, complimenting its wisdom and courage for bringing up for discussion all the important and controversial issues of the day. It is also common to grant the Commission good intentions and to attribute any problems only to unintended consequences or to new technologies or to unanticipated economic developments. No one should take such letters seriously." -- Steve Wunsch, Wunsch Auction Associates, LLC, New York, New York

"By way of introduction, I am a plaintiffs' class action lawyer that has represented aggrieved investors in private securities class actions for the past ten years. It is from this perspective that I write. Without belaboring the point, the continued non-disclosure of large volumes of transactions in "dark pools" poses significant threat to efforts to prosecute private rights of action under the federal securities laws." -- Mary K. Blasy, Esq., Scott + Scott, LLP

"This is about hft, flash orders, and dark pools. IN truth I do not want any. I am also so sick of wall street firms telling the sec that I am the person who benefits from things I don't want...The very idea of dark pools is unamerican." -- David Blumenthal, MD

"Why, if you will prohibit or restrict dark-pool trading (DPT), do you allow after-hours trading (AHT)? ...AHT is every bit as discriminatory against a large segment of the market participants as is DPT... After-hours trading should be immediately abolished, or all markets and brokerages should be mandatorily extended to cover those hours." -- Malcolm L. Kantzler

"I am employed in a dusty old warehouse. I do not make much money. So I want the mutual fund managers to have the best pricing they can so I can make the most money on my humble investments. Dark pools provide that...These regulations come from the human need to control other people but do not control me. Or it is somehow evil for someone else to have a profit or more of a profit because that is greed but I can have a profit, even a large profit, and that is moral and ethical. We then conclude we need to stop this immoral greed. Then we use these type of regulations to stick it to the evil greedy people (the big boys - rich) we do not like. What happens is we hurt ourselves more than the people we are targeting. This regulation intended to stick it to the greedy ends up cutting our own throat. This is exactly the reason the Lord gave the tenth commandment Thou shalt not covet - Exodus 20:17. Regulation like this is coveting. Coveting always comes back and cuts the throat of the coveter. The Lord gave us the tenth commandment to protect us. The Lord will stick it to the greedy. I trust the Lord, not government regulation." -- Bob O., Plymouth, Minnesota

"... if the trades are not made public, it is a great opportunity for parties to make a secret profit that would not be taxed." -- William A. Thayer, San Diego, California

"The proposal to eliminate,almost(1/4 of 1%), "dark pools",should have been instituted long ago..Great work and may God bless" -- James A. Swiatek, Newaygo, Michigan

"Re Dark Pools: Don't regulate them ... ban them ... completely ..." -- Ken Blanton

"The number one recommendation from heads of buy-side desks on action to be taken to improve market structure is to move slowly, carefully, and with the utmost of care. What action would these traders like to see today as the politicians and regulators contemplate how to restore investor confidence? Twenty-five percent of head traders say three things: “don’t do anything precipitously,” get rid of an uneven playing field by banning flash orders, and leave short sales alone. Another 23% are calling for no restrictions on dark pools – no reporting requirements whatsoever and no action that would threaten the choice to go dark." -- Larry Tabb, CEO, TABB Group, LLC

"Call me naive, but since beginning my adventure into the stock market about three months ago I quickly came to the conclusion that there was some outside force manipulating stocks at a pace significantly faster than my ability to digest company information and react with a simple trade ticket to buy or sell. After looking into my suspicions via Internet searches I happened upon the issue of computer driven trading. I guess I was of the ignorant mindset that only a human would be allowed to initiate a trade. Basically I have come to the realization that the average individual has no chance what so ever against the algorithms and dark pools that are aligned to take his/her small savings a penny at a time at the speed of light. I can stare in amazement as the company stock charts bouncing up and down for no apparent real-world reasons, and now that I know that there is a mindless math program running at 50 gigahertz behind these movements I have given up and will chalk this experience up to the unrelenting Wall Street greed that will drive this nation into the ground. I have always considered myself a reasonably intelligent individual, but now I know what Chess Champion, Garry Kasparov must have felt like when he was defeated by IBM's Deep Blue Supercomputer...." -- Brian Klaus

"The stock you think you are buying in the stock market may not be that stock, but an inferior imitation. The inferior imitation is phantom stock, and it occurs in short selling. We propose procedures to close a perilous loophole for this fraud that the SEC rules permit." -- Tom Garcia, Ph.D., Professor (retired), University of Chicago, Booth Business School, Rancho Santa Fe, California

"Dark pools should be banned...along with all the other financial chicanery (derivatives, collateralized mortgages, erosion of credit standards) that has nearly caused the collapse of the world's economic systems...Why are some investors allowed to be in the markets and not abide the rules that are there for Everyman? ...Stop it. Stop all the infractions. Put the focus of investors upon creating wealth through the development of a tangible product, or a service that ennables those who do, or something that can be marketed to the world, rather than how to suck the economy dry in yet another financial shell game" -- Janice B. Campbell, Chief Compliance Officer, John A. Wolfe Associates, Inc., Portage, Michigan

"The Commission's proposal would apply restrictions on ATSs that have never been applied to other market participants performing the equivalent function, including agency trading desks, firms that execute as principal, firms that cross customer orders as agent, firms that internalize customer orders and floor brokers. Because of the disparity in application of the display requirement across different categories of market participants performing the same function, we expect market participants will structure around the rule restrictions to achieve the same result, but in a less efficient manner. ...attempting to address these alternative structures would mean the end of principal trading desks, agency trading desks and floor brokers, which would be a bad result for investors. We also disagree with the assertion that dark pools have created a "two-tier market." ...Since institutional and retail brokers representing long-term investors can participate in dark pools, dark pools do not create a two-tiered market. If the Commission has evidence to support the assertion that retail brokers are excluded from alternative trading venues, we think it should be presented and discussed." -- Seth Merrin, Chief Executive Officer; Anthony Barchetto, Head of Trading Strategy; Jay Biancamano, Global Head of Marketplace; VIad Khandros, Market Structure Analyst; Howard Meyerson, General Counsel; Liquidnet, Inc.

"Dark pools-Just the given name should be enough to know they are wrong and should be done away with. One guarantee for sure, is that they are a way for the ones who operate them to hide the ways they manipulate the prices of stocks to their benefit and the real persons detriment. Price manipulation of any kind should be reason for jail and or fines and complete banishment from future participation in the markets of any type by any and all in a company at all levels with no exceptions. We are constantly told how the markets use fair pricing and fair participation standards to create level and fair playing fields. When dark pools are allowed to exist, this fairness is not anywhere present. Besides the unethical and immoral practice these dark pools cause and promote in the markets, they also allow for the hidden gains the owners of them receive to avoid the taxes the visible participants are forced to pay. How is this fair at all, in any form or amount? It is not. Period. The dark pools and any incarnation of them the owners of them come up with in the future must be outlawed with punishments mentioned above or harsher. With statutes that allow for ease of prosecution of offenders. These dark pools and their likes have caused the largest bubbles and their eventual bursting in the markets that has always provided the benefits to the owners and the losses to the individual. Always. Where is the fairness in that?" -- Matt K.

"Don’t surrender to the pressure from the large players and do not fall for the argument of lost employment ! There will be much more people employed if this malpractice is finally forbidden !" -- Aston Susilovic, Capital Alternative Investment Management GmbH, Switzerland

"Investors expect full transparency, honesty, and disclosure before and after selecting investments. There is no reason they shouldn't have that." -- Christopher Krause, Madison, Wisconsin, Morgan Stanley Smith Barney, Financial Advisor

"Because the Commission has had a virtually unobstructed ability to impose a flawed transparency on the stock market for over three decades, in many respects what appears to be good is bad and what appears to be bad is good." -- Steve Wunsch, Wunsch Auction Associates, LLC, New York, New York

Investors Say The Darndest Things, Part II

2009-12-31T08:23:00.006-05:00

I received so many emails from people thanking me for excerpting the comments on the SEC Flash Order ban proposal that I have decided to make it an ongoing project. In addition to providing food for thought, part of this exercise is to let you gauge the level of discourse that is shaping trading markets today and part of it is to, well, make you laugh.

The letters that clearly fall under the heading of, "You can't make this stuff up" make me wonder how much attention the SEC actually pays to them. That said, please understand that these are just the comments that struck me as interesting; there are many more that present balanced and well-reasoned viewpoints. In particular, I urge you to read the DirectEdge letter that makes the clearest case for flash functionality in the equity markets, and the CBOE letter that does the same for the options markets. For those of you who (like me) tend to doze off while reading SEC filings, the ISE letter is probably the most readable and entertaining.

Anyway, here are excerpts from the remainder of the letters to the SEC on the Flash Order ban proposal. I've also put links to the complete text of each letter on the SEC website following the quotes.

Stay tuned for comments from the Dark Pool proposal, they're sure to be interesting!

Bill

"Wall Street firms have become so large, rich and powerful that they have half of Congress (if not 3/4) in their pockets. And just about all of the U.S. Treasury...I, for one, would prefer to see them fail one after the other than be cheated on every trade I make...Thank you for looking and ENDING these awful and shameful practices. and don't stop there either. You need to defend the individual investor from these out-of-control monsters (I mean it)." -- Laurent Mayer, Individual Investor

"All trades should be required to trade and be reported upon no matter who, what or where the trade is arranged. This is the only way to provided market transparency. This include GS and "private" invester originization like KKK and Blackrock etc... I also like that "naked" options should also be prohibited." -- Bob G. Mall

"I can't believe you would allow false algo trading." -- Wallace Ungles

"The bottom line in this issue is not facts but emotion. Humans cannot tolerate the idea someone has more than they do. Or someone has an advantage over them. This is envy. This is jealousy. Making decisions based on these emotions always comes back to hurt and damage the one who is jealous or full of envy. Acting on these emotions is self-loathing resentment lust." -- Bob O., Plymouth, Minnesota

"Ban completely ... flash orders and high frequency trading." -- Kenton C. Blanton , Fremont, California

"To date, there are dozens of comments regarding the Proposal posted on the SEC website. They are almost all from individuals expressing anger at the industry over perceived unfairness in the markets. We recognize that the troubled economy has elevated fears and we appreciate the frustrations voiced by individual investors. However, we would like to set the record straight regarding flash trading. Unfortunately, flash trading has been mischaracterized as an unfair practice that benefits high frequency traders. This has confused investors and many industry professionals." -- William J. Brodsky, Chairman and CEO, Chicago Board Options Exchange, Inc., Chicago, Illinois

"BATS supports...the elimination of flash order types on rational policy grounds rather than on the exaggerated and often irrational concerns voiced widely in the media over the last several months. BATS briefly offered a version of the “flash” order (referred to as “BOLT”) earlier this year. When we implemented BOLT, we publicly stated that we did so for competitive reasons, but that we were concerned about the market structure implications of allowing equities and options exchanges as well as alternative trading systems to continue to offer such order types." -- Eric Swanson, BATS Exchange, Inc., Lenexa, Kansas

"...the Proposal would inflict unintended and substantial damage to the price transparency, liquidity, and execution quality currently enjoyed by retail customers." -- John C. Nagel, Managing Director and Deputy General Counsel, Head of Global Compliance, Citadel Investment Group, L.L.C.

"While a ban may appear to be a convenient response to populist sentiment, Direct Edge respectfully requests that the Commission carefully consider why and how such technology is used, how it can help investors achieve their objectives of accessing market-wide liquidity through a single market center, and whether any credible evidence of its impact on overall market integrity dictates an abolition of the practice as the only alternative. To date, the positive aspects of flash technology have been understated, the concerns have been generally overstated and under-supported, and less restrictive and comprehensive solutions to address such concerns have been ignored." -- Eric Hess, General Counsel, Direct Edge Holdings, LLC

"The business lobby group Businesses Aligned to Prevent Price Improvement, or BAPPI, today announced a drive to enact the Consumer Relief Initiative for Pride and Protection Legislative Engagement Act of 2010, or "CRIPPLE." CRIPPLE would prevent any business in the United States from asking its sales force to match a better price of a competitor. Explains BAPPI leader Rod Kanehl, "Why bother asking anyone to improve their prices? The government should require people to advertise their best prices. And if they don't have the best price, they must reject a customer's order. Better yet, the government should require a business to send the customer to a competitor who has a better price." -- Michael J. Simon, Secretary, International Securities Exchange, LLC, New York, New York

"Verbal representation...attracts significant institutional liquidity to the market, enhances price discovery, provides market information available to all participants through the floor brokers of member firms, and...provides one of the few real alternatives to automated markets for public investors today...it serves our markets well, and should not be circumscribed as part of the Commission's appropriate efforts to ban modern-day flash orders." -- Richard S. Rosenblatt, CEO, Rosenblatt Securities Inc.

"First, let me say that this practice is insidious...This practice ensures that even if I can make wise, well grounded choices, the brokerages are always going to one-up me on the buy price, and sell just before I do...Every service agreement I have ever signed (I'm assuming) whether with a full service brokerage such as Merrill-Lynch or a deep discount brokerage such as TradeKing GIVES THE BROKERAGE MY CONSENT AND BLESSING TO ALLOW THEM TO FRONT-RUN MY ORDERS WITH FLASH TRADING." -- Ryan Fisher, New York, New York

"While I have nothing against the maker/taker exchanges or the high frequency traders who are proposing that flash orders be banned these options exchanges and traders should expect OptionsHouse to do everything in its power to execute OptionsHouse customer orders away from them. OptionsHouse is doing what brokers are supposed to do with payment for order flow and zero exchange fees in that it is passing on its cost savings directly through to its customers by offering low commissions. Flash orders enable OptionsHouse to continue to keep its commissions low thereby providing an ongoing cost savings to its customers." -- George Ruhana, CEO, OptionsHouse, LLC

Let the Games Begin

2009-11-05T09:11:00.004-05:00

As I write this note it has been over two weeks since the SEC met on October 21 to approve some measures relating to dark pools. Interestingly, the Commission has not yet published these proposals in the Federal Register nor on their website so other than some meeting debate and press release information we still don't know exactly what they contain. This seems like an inordinate amount of time to leave people to speculate about the impact of these important proposals and I wonder if there isn't a bit of intrigue going on behind the scenes in Washington. Of course, that hasn't stopped industry participants from talking about potential ways around the dark pool proposals should they become rules.

The most onerous proposal seems to be the one lowering the display requirement from 5% to 0.25% of average daily volume. This would mean that most dark pools that transmit IOIs would have to show a quote which would effectively eliminate their "darkness." But maybe not.

The whole issue of the SEC trying to eliminate actionable IOIs is somewhat confusing. These messages tend to create greater liquidity for orders by knitting together undisplayed interest in different market centers. Why would the regulators want to stop that? As an investor, shouldn't I have the right to show my order to whoever I choose? Do I really want regulators telling me under what conditions I can access markets in the name of supposed fairness?

Investors Say The Darndest Things!

2009-10-29T15:35:00.003-04:00

The SEC recently discussed a ban on flash trading and published a proposed rule addressing their concerns. As is usual, they allowed people to submit letters and comments about the practice and the proposed rules. (You still have a couple of weeks left if you want to submit your own letter.)

Many of the submissions are from people who clearly do not understand flash trading and some are from those with obvious self-interests, but nearly all reflect a deep distrust of either Wall Street, the SEC or Goldman Sachs. Also, there is a recurrent theme about naked shorting, which of course has no relationship to flash trading. Even Bernie Madoff has been thrown in for good measure.

Here are some of my favorite quotes excerpted from the letters; the full text of each is available at the SEC website.

_________________________________________________

"After revelations over the past year of how corrupt and unfair the capital markets have become, the SEC should have shut down all the stock markets."

"How long are we going to be manipulated by mathematical formulas until we realize that the capital markets have collapsed because there are no longer any "suckers" on the other side?"

"As long as the SEC is a partner with this criminal element the SEC and the Obama administration will go down in history as no better than Bernie Madoff. PS the only reason Goldman Sachs pays these high bonuses is because their chosen few are willing perpetrators and keep their mouths shut. Its like the Cosa Nostra. After your a 'made' assassin your character must be silent about the real facts."

"any profits made on transactions that take 24 hours or less to complete should be taxed at nearly 100%. under 1 hour would be taxed at a maximum rate of 100%."

"As a general rule, I do not want anything a wall street firm says benefits me."

"I see nothing wrong with limiting trading to a maximum of once a day just as I can change my portfolio inside a 401K. I see everything wrong with an etf trading as much as every 15 seconds."

"Stocks that look good when I can buy them re:20 minute delay in what I see verses the microsec. that people with flash order access have.By the time I buy,they have gone down and I instantly loose money"

"I have read that Goldman Sachs acknowledges they will make $8 billion from their trading software this year, and it is incredible to me that they know in advance how much their HFT software will skim from the markets."

"This can't happen. Institutional investors are taking advantage of small investors. Have you people learned nothing?"

"Having said that, I believe this move to ban Flash trading is like fixing kitchen cabinet first, when there is a whole in roof of the house. There are much more dangerous issues in practice that call for SEC's immediate attention, like naked short sale."

"Please return the exchanges to their normal practice of serving as exchanges for individual investors and not ATMs for the big firms."

"S.E.C., It is nice you are doing something but Flash Orders only really effect The Big Traders us little people are being killed by naked short selling."

"I believe the ban on flash order would enhance the equility [sic] of US stock markets."

"You people at the SEC have a ton to answer for... No less than allowing the endangerment our very existence as a Republic."

"According to a recent article in Securities Industry News, Goldman, Morgan, and a handful of other firms are actually using the NYSE real estate for a fee to place their machines closest to the exchange. These are not advantages available a small broker dealer and/or a small investor."

"Just ask the hedge funds what they think and do the opposite. Stop asking the public and doing the opposite."

"I for one have lost much of my faith in our trading system - Since America only represents about 22% of the world's collective GDP, 78% of my new investments have been going overseas."

"Someone needs to be on top of this. As in Monday."

"There needs to be a randomized lag inserted into the ordering process in order to truly level the playing field this would also dampen some of the speculation driven volatility in the marketplace, as well as discouraging some churn....A randomized delay of even up to a minute is not going to negatively impact the overall flow of the markets."

"Disclaimers: I am not now, nor have I ever been an employee of a brokerage firm."

"I don't know what is SEC motto/principal anymore."

What is their edge?

2009-10-13T09:00:00.004-04:00

In considering the validity and sustainability of the High Frequency Trading (HFT) model we need to understand what drives the profitability of the strategies with some certainty. When asked to explain the success of HFT many (uninformed) observers point out attributes of successful firms such as sophisticated technology, highly educated personnel, and a flexible research & development environment. But these are not guarantors of success in HFT, and in fact they can often lead to ruinous trading decisions (see, "When Genius Failed: The Rise and Fall of Long Term Capital Management," by Roger Lowenstein). Instead, I believe it is only necessary to examine one simple question to understand HFT profitability, "What is their edge?"

One can examine each firm or perhaps each trading strategy in the HF space and determine at least one unique factor that leads to their success--their edge. For marketmakers, these factors often revolve around access to order flow. The basic strategy behind marketmaking is to buy at the bid and sell at the offer while managing risk between trades. In this way they can yield some portion of the bid-ask spread. How do the various market participants accomplish this?

Let's start with proprietary marketmaking firms like Tradebot or Getco. The opportunity to buy or sell at the NBBO is somewhat dependent on being "first in line" at a particular market center. With no order flow of their own these firms are forced to install the fastest infrastructure capable of vaulting them to the top of the book at each venue that they want to trade in, when they want to trade. Similarly, they often need to withdraw their orders quickly when they sense the market is turning against them. These communication lines and the hardware/software to process them are expensive to build and maintain but once they are in place, they give their owners the ability to access those markets faster than other traders. That is their edge. Although these participants sometimes characterize practices like Flash trading, fragmentation and internalization of order flow as being detrimental to the market, the reality is that those practices tend to level the playing field between the firms with that edge and those who do not. For example, if a flash order mechanism pauses an order for 20ms to give multiple participants an opportunity to respond, the value of being able to respond in 1 or 2ms is greatly diminished. And in the case of internalization, a brokerage firm gets first crack at the orders before the HF firm(see below). It's therefore unsurprising that they would oppose these practices.

What about customer-driven market makers such as Citadel, Knight and ATD? These firms have in-house, "captive" order flow that derives from associated retail brokerage units or through relationships with firms that have their own order flow. (These relationships may or may not involve so-called "payment for order flow.") In any case, these market makers get the "first look" at these orders before they are routed elsewhere. If they want to trade with them (at the National Best Bid/Offer of course) they have that option, if not they merely route the orders to an exchange for best execution. (This practice is known colloquially as "cherry picking.") Operating in this manner is a tremendous edge for at least three reasons:

The market maker doesn't have to compete with other market makers for the order, thus technology requirements are eased
The market maker is able to judge the quality of the order flow on a continuing basis, which allows him to avoid getting picked off, or "bagged."
Since the market maker trades only with the orders he desires, his risk management regimen can be greatly simplified. In other words, there is no reason for this type of market maker to take on a position he is uncomfortable with.

So great is the value of the order flow edge that in the case of ATD, a firm that began without client order flow but after realizing the benefits of trading with it, sold itself to Citigroup in order to trade with the Smith Barney retail flow. Unfortunately, that flow now resides with Morgan Stanley and ATD does not.

The biggest lesson to be learned from this is that success in one asset class does not guarantee or even imply success in another. Unlike manufacturing automobiles where you can design an engine for one car that will work fine in another, HF trading demands highly specialized systems to take advantage of different opportunities. Even if you have the fastest equity market making system, that doesn't mean that you will be profitable in trading options because in the options markets, legacy market makers have the edge regardless of the quality of their technology. This explains why firms like Tradebot haven't entered the options trading space and won't unless things change. With the advent of MiFid, Europe is definitely rising as a trading venue but HFT success has been limited.

SEC Publishes Flash Proposal

2009-09-23T13:36:00.002-04:00

Last week, the SEC published their long-rumored proposals concerning Flash orders. As many people had suspected it is clear that the Commission is leaning toward doing away with this order type. I completed a quick read through the document and jotted down some notes which I will summarize here. I'm considering writing a formal comment letter in response but it seems like this horse has left the barn and it doesn't seem like a good use of time to try to lead it back in.

Analysis of SEC Flash Order Release

In the proposals, the SEC mentions what they perceive to be problems with flash.

1. Flash orders are dark, therefore they hinder price discovery

Flash order themselves are actually market orders which are not revealed to the public. Only the liquidity responding to flash orders is dark. The orders themselves would not be displayed in any circumstance as they would be marketable and therefore executed immediately. It is doubtful that the responding liquidity will ever be transparent because those participants would not play under the proposed rules. More likely outcome is that the liquidity being provided to flash orders will evaporate under the proposal.

According to the SEC's own statistics, there are far worse problems with dark liquidity than flash orders which only account for 0.8% of total volume. In contrast, the amount of hidden liquidity in dark pools alone is more than 8.0% and there is surely much more when taking into account reserve and hidden order on traditional exchanges. Why are they focusing on this relatively small corner of market structure? It smacks of rearranging deck chairs on the Titanic.

2. Flash orders can only be accessed by high speed computers

I'm shocked. Shocked to learn that people are using computers to trade. In a previous comment letter written by Getco (a leader in the high-frequency trading space) they mention some revealing trading statistics. For example, during a typical 1/2 second (500 millisecond) period recently JP Morgan stock traded more than 250 times. The quotes in the stock changed dozens of times in that same period. It is virtually impossible to select any single order in the National Market System and trade with it without a computer. Even then it's doubtful that any single order could be targeted. So why should Flash orders be any different?

Shame on the SEC for trying to polarize this into a debate about high-speed trading vs long-term investing. There are enough smart people there to know that this is all about competition between markets, not whether investors are being disadvantaged. Perhaps we should eliminate the use of the telephone as well because those with a phone have an obvious advantage over those without.

3. Flash order quotes are only available to subscribers, not on the SIP

This is something that can be easily remedied by allowing markets to place locking quotes on the SIP.
But the SEC has prohibited locking quotes in the past. Maybe they should rethink that position. What's so bad about locked markets? Isn't a locked market a 'perfect' price? No spread for the investor to pay.
Even if the Flash order were displayed as part of the (locked) consolidated NBBO for 20 milliseconds would anyone even notice it? Would anyone care?

4. Effect on manual markets (read: We're worried about the NYSE)

The Rule 602 exemption that allows Flash orders was originally put in place due to concerns that manual markets would not be able to negotiate prices without displaying to the SIP. The SEC believes removing the exemption would alter or eliminate the practice of floor broker negotiation.
The sad truth is that very little floor broker negotiation takes place anymore at the NYSE or anywhere else. When CNBC used to do a live shot from the floor you'd see Maria Bartiromo getting jostled around by harried brokers trying to get orders executed. Have you seen the CNBC reports from the floor lately? Most of the time it appears that the Exchange is closed! And that's during the day.

The SEC-Proposed solution

The SEC proposes to do three things

Eliminate the exemption from quoting requirements for flash orders, i.e., orders that are executed or withdrawn immediately after communication to a market center.
Require flash quotes to conform to existing rules, in particular the rule that a quote may not intentionally lock an existing quote.
Conform above requirements for (which are mainly for exchanges only) to all markets, i.e., ATSs and ECNs.

I'll write more about this when I get a chance but in the meantime please post your comments on this very important piece of rulemaking.

DirectEdge Exchange Application Published

2009-09-14T15:04:00.002-04:00

Traders Magazine is reporting that the SEC is publishing DirectEdge's exchange application in the Federal Register. This is an important step for DE because it signals the final stage of the give-and-take negotiations that go on behind the scenes before a new exchange is launched.

Interestingly, DirectEdge has apparently not removed the rules which enable flash trading from it's application and therefore we expect to see them published for comment as well. This seems strange; if the SEC was going to prohibit flash trading soon (as has been reported and outright declared by Senator Schumer) why would they let DirectEdge's application include it? One thought is that they might be using this filing as a lightning rod designed to elicit more public comment on the practice. Another is that the Commission is prepared to stand up to the politicians on this issue.

More to come RSN (Real Soon Now)!

SEC to release flash proposals

2009-09-14T07:19:00.003-04:00

I have learned that the SEC will put out a directive this week that focuses on so-called "Flash" trading. Among the proposals will be the banning of the practice, which is clearly the approach favored by some of our elected officials. And while it's unfortunate that the Commission may do this without adequate public comment perhaps it's for the best in order to quell the follow-on effect this witch hunt has had--the excoriation of high-frequency trading and market making in general.

While Flash trading is not the same thing as high-frequency trading or electronic market making, the ignorant tend to lump them together. Some (including those who would also have us return to the "good old days" of privileged floor participants having unfair advantages over the public) would undoubtedly like to see all electronic trading banned. But the chance of that happening is about as likely as banning horseless carriages from our roads.

An important fact not remember about this debate is the following: everyone involved has an axe to grind.

. The traditional exchanges are losing their life blood--order flow--to market centers that offer Flash order trading.
. Newspaper reporters, especially the popular media who generally don't understand market microstructure, are prone to being manipulated by those who do.
. The politicians receive substantial campaign donations from exchanges. They also generally like getting their name printed in the newspaper when a story is associated with "protecting the public."
. The SEC must ultimately answer to the politicians, often in front of a televised Congressional investigatigatory panel. The questions posed of these well-meaning public servants are often inflammatory and unfair. And quite often they are not even given a chance to respond.
. Some brokers are too small or too undercapitalized to be able to spend the money it takes to provide electronic trading to or for their clients. They are at a competitive disadvantage.
. Some brokers have spent vast sums building and maintaining algorithmic and high-frequency trading infrastructure that forms the basis of an extremely lucrative business.

So when following the debate which will ensue in the next few weeks keep the above list handy. It may help when forming a reasoned opinion.

A Time To Trade

2008-10-25T15:44:00.001-04:00

Our country's current economic crisis has many causes, none of them easily solved. But dramatic changes in the structure of our equity markets of the past ten years have contributed far more to the turmoil than has been ackowledged by regulators, brokers and academics. Through a patchwork of new rules, exchange demutualization, and technology deployment we have permanently changed the nature of the most basic transaction in our markets: how buyer meets seller to trade stock. And the results of these changes are impacting investment decisions in many ways, some known and others not, some good and others bad.

One of the most fundamental--and fundamentally flawed--premises underlying our stock markets is that share prices continuously reflect the correct prices for their companies. Defining 'correct' in this sense is tricky but most would agree that if one person is willing to pay $50 to buy a share and another is willing to sell a share for $50, the correct price for that stock is $50. But what if these two people didn't come together to trade at the same time and place? If the seller tried to sell just a few minutes before the buyer showed up he might receive a much lower price. To make matters worse, when the buyer eventually placed her order after the seller was finished she could end up paying a much higher price. And although this example assumes they miss each other by just a few minutes, the reality is that in today's turbo-charged equity markets this unfortunate outcome often occurs when they miss each other by just microseconds. Unfortunately, when it comes to trading sometimes slower is better.

Another way to think about this is to imagine what would happen if a lot of investors placed a lot of orders to cumulatively buy 50,000 shares of a stock beginning at 3:00 PM and lasting until 3:59 PM. Common sense tells us that if there are more buyers than sellers the price would begin to rise fairly quickly, increasing the last price by, let's say, $2. By the time 4:00 rolls around all the buyers orders are finished and there is one more person who just wants to sell a paltry 1,000 shares. Unfortunately since there are no buyers left willing to pay the last price the seller's order might move the stock price down by the entire $2 or even more. So even though thousands of shares traded at the higher price just one minute prior, the financial websites and newspapers will all show a questionably low closing price for the stock based on this 1,000 shares.

History provides us with a couple of solutions to this problem. Students of market structure are aware that until 1871, the NYSE held only a single auction each day at a specific time for each stock. Everyone who was interested in buying or selling shares of the Erie Railroad would know that at 3:00 PM an auction would be held for those shares. All investors would place their orders and at 3:00 a clearing price would be calculated based on the the maximum number of shares that could be traded at that price. For example, if orders representing 5,000 shares were willing to buy at up to $20 and 2,000 shares were willing to sell at $20, the price of the stock would be set at $20 and 2,000 shares would change hands. There was no possiblity of a buyer and seller not interacting because of time differential; any order that was present by 3:00 PM was represented in the auction.

While efficient, this process didn't allow for continuous trading. Some market participants thought it would be a good idea to provide liquidity for stocks at any time during the day, not just at an appointed time. The NYSE and it's member-owners, the brokers, saw this as a huge opportunity to satisfy their clients desires. (Presumably they were also pleased at the thought of being able to charge commissions throughout the day rather than just once.) But the members realized that if buyers and sellers didn't meet at the same time the customers would be unhappy--exactly the situation we find ourselves in today. What was needed was an auctioneer, a coordinator, an arranger who's main function would be to bring together as many participants to trade as possible. When there wasn't enough buying or selling interest that arranger would be required to step in an trade with his own capital, to make a market in the stock. And so the Exchange created the Specialist, a job that has existed until today.

Rather than pore over the specific requirements and responsibilities of a specialist, investors should understand their job holistically. The specialist system is designed in some ways to push investors to think about what they are doing rather than reacting in a knee-jerk fashion. By forcing a moment of reasoned thought many price swings can be avoided and millions of dollars saved. A recent example of this was the preciptious drop of United Airlines stock when an erroneous news story made it onto the Internet. Because there was no specialist for the stock (upon exiting bankruptcy United decided to move their listing from the NYSE to Nasdaq) who could slow things down the stock traded to a fraction of its correct value within minutes.

Even more important is the specialists job of centralizing and controlling trading interest. By convincing a participant to wait briefly for the other side of a trade to appear he is capable of saving that investor a lot of money. This function has saved investors countless amounts of money over the years. If a mistake was made by the New York Stock Exchange it was in not adequately communicating the benefits of this system to investors.

A reasonable person would think that regulators and market participants would want to foster a process where investors with buying and selling interest in a particular stock could meet at the same time and place as much as possible. But the tortured process of market deregulation has actually lead to pretty much the opposite result. To wit: in the U.S. there are more than 40 different venues to trade the same equity shares; you can buy or sell GE in over 40 different places. The SEC, prodded by self-interested electronic trading venues enabled the creation of this Rube Goldberg-esque marketplace over the past 7 years in the interest of promoting competition. While it has certainly acheived that goal, this program has also succeeded in fragmenting liquidity beyond all recognition.

Today, extremely advanced (and expensive) technology is required just to figure out the best place to send a stock order. Some brokers have spent money on these systems, known as Smart Order Routers, but many have not. The rapidly expanding marketplace has also greatly increased the profitability of so-called "black-box" traders who develop computerized algorithms to take advantage of fleeting arbitrage opportunities between venues. It doesn't require a great leap of faith to understand that these profits are coming at the expense of less-sophisticated investors and their brokers.

Another important consequence of this market deregulation is that the NYSE specialist--who once controlled most trading activity in his stocks--has pretty much been rendered impotent. He now has very little ability or motivation to slow down trading in his assigned stocks. Many people believe that the NYSE retains him only to keep open their floor and thus provide a ready-made backdrop for financial news reporters. And so the NYSE is in the process of doing away with the most important functions of the Specialist. Investors will be poorer because of it.

Instead of quixotic attempts to ban short selling, perhaps the regulators should consider a better alternative. On a temporary basis, the SEC should order that a group of stocks (the financial sector?) be traded at their primary venue once and only once each day, at 4:00PM. Trading could still take place intra-day at regional markets if an investor specifically requested, but everyone would know that a vast amount of liquidity would be available at the primary markets at the close. If they wanted that closing price it would be easily obtainable. Brokers wouldn't have to do much thinking about where to route their retail orders and investors could still request that their orders be routed for immediate execution at their own risk.

While this idea sounds extraordinary, some smarter financial professionals are already beginning to understand and practice it. One example is James Cramer of CNBC's Mad Money, who has advised his viewers to save their orders for the close in order to skirt the tremendous intraday volatility. If more investors told their brokers to mark their orders "On Close" it would be come a self-fulfilling prophecy.

This is much more than an academic argument. The future of our markets depends on people having confidence in its structure. The most widely known barometer of market level is the Dow Jones Industrial Average and that index is most frequently reported on television, radio and newspapers at its closing price. But the Dow is worthless if the prices of its individual component stocks are not fairly determined. Whipsaw trading near the close may account for much perceived volatility and as the Dow falls, so too will investor confidence in the markets.

Blogysteria!

2008-10-25T15:17:00.004-04:00

Ok, let's stop and take a deep breath. Pause the hysteria machine. Every time the markets become unruly we get a spate of blog posts from semi-informed "experts" that tend to ignore or miss various facts about their chosen subject matter. In 1987 it was all about program trading. This time around it seems to be about electronic and/or algorithmic trading although we've still got some fossils around that blame program trading for the ills of the world.

A good example of what I'm talking about is this post by Joseph L. Shaefer from "Seeking Alpha" which is a site I normally have a lot of respect for. Mr. Shaefer's article contains many inaccuracies that taken together will undoubtedly get a lot of people excited. Unfortunately it is typical of a lot of articles that have appeared in the past few weeks.

Let's look at the facts.

Fact 1. So-called "Dark Pools" including Goldman's Sigma-X are required to print each and every trade to the consolidated tape in real time. Any claim to the contrary is just wrong. The only thing that is dark about them is that they do not publish quotes before the trade.

Fact 2. Dark Pools that are operated by brokers are no different from the "old days" when Mr. Shaefer was running the trading desk at Schwab (I didn't think Schwab even had a trading desk in 1987--before they purchased Mayer & Schweizer). In those days, you had to call a dealer to find out what his market was in INTC. Today, you just send an order to his dark pool. Much more efficient and requires fewer humans to be paid to answer the telephone.

Fact 3. Program trading means many things but it is not a license to manipulate the market. In fact, the NYSE publishes weekly statistics on program trading activity that available for anyone to peruse, including regulators of all stripes and colors. Paradoxically, program trading desks are often the most aware of the rules and regulations because computers are phenomenally good at keeping extremely detailed logs of all their activity, and are mandated to do so by FINRA, the NYSE and the SEC. It would be kind of hard to dispute that you did some kind of 'evil' trade when the evidence is right there in your database.

Fact 4. One of the by-products of algorithmic trading may be the slicing of large orders into small orders. But this is a necessary step to protect those orders from being gamed by market participants. Think about it. If you needed to buy 100,000 shares of a stock that trades 50,000 shares/day would you let everyone know about it by placing an order to buy the whole thing at once? If your mutual fund or pension fund manager did that you would sue him for gross incompetence.

Fact 5. The traders that Shaefer disdains are often the only ones left who are supplying liqudity to the markets. Does anyone believe it's the hedge funds and mutual funds who are buying when the market is down 500-700 points? More likely it is the electronic market makers who are putting their capital at risk.

Let's try to tone down the blogysteria, please.

Electronic Markets Aren't Always 'Better'

2008-09-09T09:17:00.008-04:00

Ok, I know what you're thinking. "Here's another old-timer who's yearning for the good old days of the horse and buggy. Or the Buttonwood Tree." But you'd be wrong. I'm firmly in the camp that automation of our stock markets is generally a good thing. However, there are times when electronic markets stumble and when they do, watch out below!

Yesterday, two separate events occurred that remind me of an old maxim (are there any other kind?) that I first heard a long time ago from my friend, Peter Madoff: "When you live by the computer you die by the computer." First, the London Stock Exchange was down for most of the day on what could have been one of the busiest days in its history. (Here's a link to the Wall Street Journal article). And on this side of the pond, United Airlines (didn't) file for bankruptcy protection. But the avalanche of trades executed through Nasdaq when people thought it did quickly drove the price of UAUA down by 75%--wiping out about $1B of market value in minutes! Oops.

Here's a chart.

So what have we learned from these two events? Could anything have been done to change the outcome? Are there market structure issues that need to be addressed?

The LSE problem is really nothing new. The fact that it happened on a busy day should surprise no one. These things almost always happen on busy days. Busy days tax communication lines, disk drives, and the people that are responsible for monitoring them. Until we know exactly what caused this problem it would be unfair to blame the programmers or network engineers or support personnel. But people make mistakes and we have to live with that fact.

The United Airlines snafu points out some fundamental market structure differences between Nasdaq and the NYSE, where United was listed before it emerged from bankruptcy. (Why the company decided to switch exchanges is unclear.) When a stock trades on Nasdaq or most other electronic markets there are really no brakes on what price the stock can trade at. Rule #1 at these markets is generally, "The Price Is Right" and Nasdaq very rarely steps in. If someone with nefarious intent puts out erroneous information about the company, see Rule #1. In this case, it seems like just a dumb human error but Rule #1 still applies. That's why Nasdaq has refused to correct or "break" (rescind) any of the crazy trading in UAUA.

On the other hand, the NYSE has long had a system where reasoned human judgement is brought into play at the point of trade execution. That human is called a Specialist, soon to be changed to a Designated Market Maker. If United was trading at the NYSE, you can be certain that once the sell orders started pouring in the specialist would have immediately halted the stock pending an investigation. How long did it take Nasdaq to notice that something was wrong and halt trading? My guess is that the answer is, "not as fast as someone who has their own money at risk."

There are other instances where you want a human to handle your trade. I will address an important one in my next post.

IOI Soup

2008-04-10T06:59:00.005-04:00

Indications Of Interest (IOIs) have been around nearly as long as trading in some form or another. Historically an IOI was a way for traders to say, "I'm willing to buy or sell this security but you should communicate with me so together we can figure out how many shares and at what price." This differs from what happens on an exchange or ECN in one important respect: the ability to trade immediately and without reservation. An exchange quote implies that you can trade at that price for that number of shares without having to communicate with the trader who entered it; and IOI does not. So why do people use IOIs?

The exchange model is not always the most efficient way to trade, even if it is the fastest and most assured. That's because traders don't always want to reveal the full size of their orders for fear that other market participants will use that information against them. This is especially true when the order size is large relative to the National Best Bid or Offer (NBBO). So the IOI mechanism evolved to give traders a way to alert the market that they were interested in a stock pending further discussion.

With the advent of high speed data communications several systems were developed to transmit IOIs between traders and investors. The grandfather of these systems is known as AutEx. Nearly forty years after being launched AutEx is still considered the premier IOI system for institutional block trading. But for various reasons other competitors to AutEx have been developed, each gaining some significant market share. Traders Magazine recently ran a good cover story with different perspectives about this business.

Dark Quotes

Lately, the concept of IOIs has been extended to more types of trading. In particular, dark pools have seen the need to link with each other in order to tap the greatest possible amount of liquidity. But they encountered a problem: what if Dark Pool 'A' routes an order to Dark Pool 'B' but 'B' doesn't have a contra-side order? 'A' would then have wasted time on an unproductive liquidity probe. Multiply that by dozens of dark pools and you have a problem. IOIs have ridden to the rescue.

In a transparent market scenario, 'B' could have just looked at 'A's quote before routing the order but the whole idea behind dark pools is to not have transparency. Instead, Dark Pool 'A' can continuously transmit IOIs for orders on it's book to Dark Pool 'B' and if 'B' receives an order that it can't fill, it can route it to 'A' with fairly high certainty that 'A' will be able to execute the order.

An observer might wonder, "What's the difference between an IOI in this context and a quote?" Mainly it's that since the IOI is not immediately executable (or maybe a better way to say it is that it's not guaranteed) it doesn't have to be shown to the whole world, thus preserving the hidden aspect of the dark pool. Maybe we should rebrand this form of IOI to reflect the true nature of what it is--a seemingly anomalous "dark quote."

Regulators have struggled to come up with an all-encompassing definition and their thinking is still evolving. As more liquidity goes dark the distinction will become more and more important.

The Future of Traders

2008-04-10T06:08:00.004-04:00

On May 9, 2008, I'll be moderating a panel at the SIFMA Market Structure Conference that will examine the roles of various participants in the institutional trading process, especially with respect to algorithmic and electronic trading. Here's a link to the program. The industry has a lot of conferences about various topics but this one is a perennial favorite and always well attended, so I can recommend it without reservation.

When I think about how to structure these panels I always try to come up with topics that will cause some degree of controversy, otherwise you just end up putting the audience to sleep, especially when you have the after-lunch spot. Here's a short overview that summarizes my panel, "Algorithmic Trading in a World With Fewer Traders."

In many ways, algorithmic trading is both a cause of, and an answer to, the reality of a greatly diminished equity trading workforce. This panel will focus on past changes in market structure and technology that have led to the rise of electronic trading and the changes to come that will ensure its continued growth. In particular, the panel will discuss the future roles of different participants in the trading process - market makers, traders, sales traders, and programmers - and how each participant will add value in the future.

Any time you imply that people will lose their jobs because of technology you can be pretty sure there will be a fair amount of debate on the subject. That's what makes it interesting!

Besides me, the people on the panel will be:

Brian Fagen, Head of U.S. Program Trading and Electronic Trading Sales for Lehman Brothers; David Leinweber, Haas Fellow in Finance at U.C. Berkeley; Dan Mathisson, Head of Advanced Execution Services (AES) for Credit Suisse; David Weisberger, head of Global Electronic Market Access for Lava Trading (part of Citi); and Jarrod Yuster, Managing Director at Merrill Lynch.

I've known all these folks for a long time, either as colleagues or competitors and I know they will make for a lively discussion. They are mostly representative of the new order of things as their livelihoods depend on the success of electronic trading but I think they have enough perspective within their own organizations to understand the ramifications of what they do on the rest of the business.

Please try to attend.

Exchange Fee Analysis, Part II

2008-03-08T19:01:00.007-05:00

Let's assume that a firm knows two things: how many shares of market orders it executes and how many shares of limit orders it executes. My first crack at a model to calculate costs looked like this:

TotalCost = (MarketShareVolume * LiquidityTakingCharge) - (LimitShareVolume * LiqudityProvisionRebate)

Notice something missing? This formula assumes that all market shares actually get executed against a particular market center's book. In reality, that rarely happens. Unless a market center has 100% of the liquidity in a given stock there's no way that it can satisfy all liquidity requests, i.e., market orders. To be accurate the equation should include a third term that represents the amount of volume that gets routed to other markets and how much that would cost. Something like this:

TotalCost = (MarketShareVolume * LiquidityTakingCharge) - (LimitShareVolume * LiqudityProvisionRebate) + (SharesNotExecuted * RoutingCharge)

How do we calculate SharesNotExcuted pre-ante? Most markets don't publish figures on how much business they route away so figuring this out is tricky. One way would be to look at the amount of shares executed as a percentage of total shares handled by a particular market center. For example, if a market receives 1,000,000 shares but executes 500,000 shares we can assume that, all things being equal, we might get 50% of our volume executed there. This is somewhat unsatisfying statistically but we may not have any better information.

Coming in Part III: Putting it all together.

Exchange Fee Analysis, Part I

2008-03-08T17:58:00.005-05:00

This project started at the request of a client who wanted advice on where to route their order flow. I explained that there are many factors that should go into the decision with the somewhat nebulous concept of "best execution" being paramount. After some discussion, I though we could at least start with something simple: where the least expensive market center would be. And I say "least expensive" in terms of explicit costs--mainly transaction fees. We'll leave interesting things like market impact for later. But I cautioned that best execution means a lot more than what a broker pays in transaction costs and that we would revisit the concept in the future.

Anyway, I thought the transaction fee calculation would be fairly simple. Most market centers charge brokers three different rates per share:

Liqudity taking charge. The amount a broker pays when an order (usually a market or marketable limit) removes liqudity from a market's order book.
Liquidity provision rebate. Market centers generally rebate money back to brokers who provide liquidity on their book, usually in the form of limit orders.
Routing charge. If an order cannot be immediately matched on a market's book it can be routed to another market that has a better priced contra-side order.

With those three things in mind I started to collect the different charges from each of the market centers. (Usually they post them on their web site but some require you to call to find out). In the course of doing this I remembered that different markets have different pricing tiers such that the more order flow a firm sends the more attractive the pricing. This required normalizing the data into a tabular form with prices for each tier. And just to confuse the issue, some markets have other "special deals" like offering discounts to specialists or market makers, and I had to figure out how to include that in my database.

So after a not considerable amount of work I developed a database that contained all the data for 13 different exchanges and ECNs. I then thought I was ready to start calculating the cheapest place to trade. I soon saw that I was wrong.

In my next blog entry: how to calculate exchange fee costs.

Introduction

2008-02-24T07:53:00.004-05:00

If you're here you are probably interested in some aspect of trading. I'm going to use this forum to write about various aspects of the trading business with an emphasis on:

Best Execution
Methods and techniques for broker/dealers to insure that they are providing great service for their investor clients.

Electronic and Algorithmic Trading
The state of the art in trading methodology. Ideas about how to use technology to streamline workflow, manage difficult orders, and increase alpha.

Market Centers
Thoughts on the business of exchanges, ECNs, ATSs, MTFs, and other places to trade.

If you haven't already visited my web site here is a link.