High Frequency Trading as a Socially Worthless Fraud on the Marketplace

White Paper by R. Tamara de Silva -November 30, 2010

"Capitalism is the astounding belief that the most wickedest of men will do the most wickedest of things for the greatest good of everyone."

John Maynard Keynes

A few days ago, I came across an article in the New Yorker entitled, "What Good is Wall Street?; Much of What Wall Street Does is Socially Worthless" by John Cassidy, who writes that, "the most profitable industry in America has been one that doesn't design, build or sell a single tangible thing."[1] Socially worthless is an apt moniker for one opaque and esoteric corner of the financial markets called high frequency trading. High frequency trading firms make profits of almost $21 billion a year[2] because they get to see price and order information before every other market participant and can act on it. In other words, high frequency trading firms use inside information to cheat.

The next big financial fraud or crisis has nothing to do with the insider-trading probe that arose out of Galleon and is now making the news. So often it seems that frauds, bubbles and crises prosper because regulators are so much slower and seemingly inept compared to those they would watch over. The SEC is still considering financial reform for the credit crisis of last year, which took years to ferment and come to fruition. It is a cat and mouse game where the figurative cat is morbidly obese and cannot run after most of the mice, most of the time. Consider that even after the last credit crisis, we still have the three largest credit ratings agencies, S&P, Moody's and Fitch, Inc., that are still fraught with the same conflicts of interests that allowed them to make huge profits by giving AAA ratings to mortgage securities without any apparent basis other than the word of the companies that paid them and the bubble in real estate prices.

The past two to three decades have seen an unprecedented growth in the complexity of the financial markets as part of a larger trend, called financialization. Leaving bubbles and credit crisis aside, the growth in the complexity of financial instruments such as in the securitization of credit instruments with an alphabet soup of names from CDOs to Structured Investment Vehicles (SIVs) have heralded massive profits for investment banks. This level of complexity is not necessarily problematic unless as in the case of the recent credit crisis, Wall Street proves itself ill equipped to properly access the risk of these instruments.[3] While the complexity of financial instruments has exploded, risk models used to calculate their value and potential risk still utilize simplistic mathematical models based upon faulty underlying assumptions. Moreover, the lack of transparency, the inability to ascertain the actual value or price (these instruments are not marked the market, giving them only a notional value at best) and their opaqueness, also makes them even riskier than any existing risk model can calculate because their actual value can in theory be so removed from the calculable value of any "thing" -that is a "thing" possessing a market or discernible value, that they are the most contrived of trading vehicles with arguably fictional values.[4]

Financialization can be defined as the shift in the engines of economic growth away from manufacturing, production or even the service sector, to finance. Finance now accounts for the largest component of private sector GDP. GM and Ford make more from their financial services than from making cars, they have been losing money on making cars for some time. It is the financial firms that were considered "too big to fail."[5]

What also took place during financialization is a huge increase in private and public debt compared to GDP.

Has financialization and the increase in the complexity of financial instruments, while possibly making the markets more vulnerable to shocks like the credit crisis otherwise, led to the more efficient allocation of capital, or increased social welfare? Has it led to economic growth and increased aggregate good-in terms that are not measured by leading economic indicators or existing growth rates? Trading volume has exploded. But has more money been made by fewer people or is the larger "real" economy in better stead? Will we rue the demise of manufacturing and wish we had some new sophisticated manufacturing and research and development in sacrifice of some part of financialization? The answers depend on how the questions are framed and the jury remains out.

Regardless, financialization is not socially worthless. The financial industry encompasses the trading world and investment banking. Trading provides value to the world by its historical functions of price discovery, risk transfer and liquidity. Investment banking leads to the development of companies like Google, Amazon and biotech companies that may one day have a cure for cancer. All of these functions are socially beneficial. After the venture capitalists, the world needs Wall Street to supply its much larger pool of capital to new businesses with potentially valuable intellectual property that may benefit society as a whole. This is the classic example of self-interest affecting the greater good in capitalism. As Goldman Sach's CEO Lloyd Blankfein memorably remarked to an interviewer, last year, Goldman in providing capital to new businesses is "doing God's work."

But is this true? Capital formation, that is the financing of new businesses, is now only a small portion of what investment banks actually do. As John Cassidy points out, Wall Street's role in financing new businesses is now only a small part of what it does,

" The market for initial public offerings (I.P.O.s) of stock by U.S. companies never fully recovered from the tech bust. During the third quarter of 2010, just thirty-three U.S. companies went public, and they raised a paltry five billion dollars. Most people on Wall Street aren't finding the next Apple or promoting a green rival to Exxon. They are buying and selling securities that are tied to existing firms and capital projects, or to something less concrete, such as the price of a stock or the level of an exchange rate. During the past two decades, trading volumes have risen exponentially across many markets: stocks, bonds, currencies, commodities, and all manner of derivative securities. In the first nine months of this year, sales and trading accounted for thirty-six per cent of Morgan Stanley's revenues and a much higher proportion of profits. Traditional investment banking-the business of raising money for companies and advising them on deals-contributed less than fifteen percent of the firm's revenue. Goldman Sachs is even more reliant on trading. Between July and September of this year, trading accounted for sixty-three percent of its revenue, and corporate finance just thirteen percent."

Wall Street is de-emphasing "doing God's work" because there is more money to be made elsewhere-in large part, high frequency trading. While the United States has a $1.4 trillion deficit, mergers and acquisition volume and capital formation is down, and the economy is by most accounts in perilous straits, investment banks and hedge funds have reported record quarters since the credit crisis from trading profits as result of high frequency trading. In the second quarter of 2009, Goldman reported that it had a record forty-six "$100 million" trading days.[6]

Several regulations set the stage for high frequency trading. By way of some background, in 1998, former SEC Chairman Arthur Levitt[7] advocated on behalf of the Regulation Alternative Trading Systems Act[8]. This regulation gave rise to electronic communications networks ("ECNs"), or electronic exchanges that would compete with existing exchanges like the NYSE by making markets in equities, and matching buyers with sellers. Remember this was in the apex of the Tech and day trading bubble. Seemingly everyone had an idea on what Internet stock was worth owning and day trading came into its own. The intent of the regulation was to allow anyone with a personal computer and an idea to participate in the market. What the regulation designed to create a level playing field could not have foreseen is the inability of investors with personal computers to compete with firms with supercomputers housed right next to the exchanges.

In 2000, the exchanges began quoting prices in decimals instead of fractions. This meant that minimum spread a market maker could pocket between a bid and offer became compressed from 6.25 cents, or a "teenie," down to a penny.

The next crucial piece of regulation came in 2005 in the form of the Regulation National Market System ("NMS")[9]. Prior to the NMS, stockbrokers and brokerages were obligated to offer clients the best possible execution of orders. As in the futures markets, brokers could not get a better fill for an order than their clients-this was also a prohibition against front-running or trading ahead of customer orders. In stocks, each brokerage could determine whether "best" price meant price gotten at the fastest fill or the most favorable to the customer. This provided some room for brokerages to match buy and sell orders internally and pocket the spread, or route the orders to exchanges. The second option allowed brokerages to profit from fees the exchanges gave for routing the orders because the brokerages were providing "order flow."

NMS changed the best execution rule to the best nationally available price rule. Under NMS, market orders would be posted electronically and executed almost instantly at the best available price.

One of the many strategies called rebate trading used by high frequency traders because of NMS is to continuously post both bids and offers on hundreds of stocks (collecting fees from the exchanges for providing order flow). Another strategy is to arbitrage the tiny price spreads that exist at any given moment between buy and sell orders.

High frequency trading firms ("HFTs") utilize a basket of algorithmically programmed market strategies executed by supercomputers that are housed directly next to the computers that drive marketplaces like the NYSE and NASDAQ. High frequency traders are not really conventional traders but quants and programmers that take advantage of their ability to have information before anyone else and "trade" on it. HFTs are more analogous to front-runners and insider traders than speculators or traders.

They use series of algorithms to take advantage of the computers' speed and proximity to the marketplaces to get information about orders and price before every other market participant. The physical exchanges like NYSE, NASDAQ and CBOE lease out space to HFTs that allows HFTs to place their supercomputers directly next to the supercomputers of the exchanges thereby giving the HFTs advantages of milliseconds and microseconds-to see price and order information (inside information) before anyone else that is not paying for co-location and does not have a supercomputer with algorithms at the physical exchange.[10] Their proximity to the servers at the physical exchanges give them an insurmountable advantage which they utilize to "trade," or effectively front-run everyone else's orders. Any argument that we have a level playing field in terms of price and order information in the market today is simply false.

The "trading" of HFTs is not trading per se in as much as it is having the ability to see information on orders and price before anyone else and arbitrage price differences between this information as it is disseminated to different market participants. The exchanges pay for order flow and permit co-location-essentially becoming complicit in the front-running of orders and permitting HFTs to pay for and use inside information unavailable to anyone else.

HFTs tend to deny that they are profiting from front-running and insider information by arguing that they are akin to floor traders and floor brokers who see orders and price information before it is disseminated to the public as real time price data. This argument is species. Both the SEC and CFTC have aggressively prosecuted front running on the part of floor brokers and specialists. But HFTs are operating under a separate set of rules. In part because HFTs operate in an opaque and essentially unregulated dark corner of the marketplace, these is no discernible mechanism for monitoring the activities of the HFTs. According to reliable sources, the SEC may not even have the technology to see what the HFTs do.

The key difference between floor trading and floor brokerage and electronic trading now is smart routing. Smart routing is used for all electronic orders. Smart routing is a practice that searches for an optimal combination of prices available for a stock or option order and seeks to immediately execute the order over all available electronic exchanges. The problem with smart routing is that it allows select market participants to see order flow. Floor brokers and specialists never got to see coming order flow. Specialists, floor brokers and floor traders could not see what their fellow traders and brokers were going to do next before they did it-but this is what is happening now. HFTs see order flow before everyone else.

Even though front-running is understood to be illegal, there is no mechanism to catch, or even expressly prohibit HFTs from profiting from their ability to see both order and price information before everyone else. HFTs have operated under the radar screen; their quants are not licensed or registered, as are floor traders and specialists. Their internal arbitrage strategies are not known or monitored. It is perhaps no coincidence that HFTs and investment banks like Goldman do not publicize high frequency trading. Fortunately, the issues involved are esoteric enough that they are not the fodder of mainstream media. All of this is an unfortunate real-time example of how regulators are seemingly always years behind the cheats.

Ironically, the exchanges continue to charge the public for real-time price data. This is also tantamount to a fraud on the market because what the public pays to see as real-time price data has already been seen and acted on by other people.

One example of unabashed front running acknowledged to be a "trading" strategy of the HFTs is the use of flash orders. The exchanges pay rebates to traders who post bids and offers to buys and sell shares of a stock.[11] They also charge fees to market participants who respond to these posted bids and offers. The systems of rebates creates an incentive for HFTs to post bids and offers, which they can almost instantly cancel or route their bids and offers to another marketplace, having collected the rebate, before they are accepted because they can see the order from the slower acceptor of the bid and offer coming. This allows the HFTs to collect rebates and artificially drive the price of a stock up or down.

What is far worse is that this practice, which is defended by the CBOE[12] and Goldman calls the integrity of price data into question. Simply stated, the displayed bid and offer may not exist and may be the results of privileged HFTs having price and order information unequal and unavailable to investors and market participants. HFTs can post bids and offers and trade ahead of responsive orders they see coming. HFTs can see a market order flashed to them before it is shown to anyone else and they can make a decision not based on a real quote from a real trade but based upon their own analysis of their risk (a more one sided and profitable analysis than anyone else gets to make)-they can re-route their bids and offers to another exchange and avoid interacting with orders they know exist before anyone else. The idea of real time price data is a myth.

HFTs through their ability to see orders before every other market participant have created a two tier-ed market place wherein they get real price data and all other investors and traders get delayed price data, previously viewed and assimilated, all the while being made to believe that they are still getting equal access to price information. This is nothing more than a fraud on the marketplace.

HFTs are so widely profitable because they are running a rigged game. Front running enables them to garner what are essentially risk-free profits. HFTs can see blocks of buy orders from slower market participants in a stock coming, they can then enter their own bids and drive up the price artificially, almost immediately later posting offers at much higher prices (having profited already from having driven up the price of the stock) than would have occurred if everyone had the same order information. A quant at a HFT who posts millions of bids and offers a day collects a hefty rebate from the exchanges all the while seeing all orders that would accept the his bids and offers beforehand.

Futures trading is a zero-sum game but equity trading is a negative sum game because of the costs of getting in and out of a trade, typically called the skid. When HFTs that front-run are able to garner risk-free profits, this translates into every other market participants' loss. Because the markets are a negative-sum game (due to trading costs) - any one entity's risk free profit comes at the guaranteed losses of someone else. In practice the public loses because it pays a tax on every market transaction. The tax is hidden and small (due to the driving up and down of prices as a result of front-running) but upon millions and millions of shares, it is substantial.

Co-locations fees and the money spent by HFTs on their supercomputers pay off in terms of speed. Speed, in advantages of milliseconds and microseconds is paramount because it enables the HFTs to see what is coming down the pipe, enabling them to act before the approaching order-simply stated to front-run.

Another defense of the HFTs is that because they account for almost seventy-three percent of trading volume, they are providing a great deal of liquidity. This is again a species argument. There is a fundamental difference between liquidity and volume. Liquidity is the ability of a market participant to get in and out of the market at a chosen price. HFTs create a tax on the market, which may well decrease liquidity. What HFTs do is trade millions of shares a days back and forth. Trading a 100 shares back and forth does not create liquidity-it merely creates volume.[13]

The arbitrage activities based upon a privileged position that is enjoyed by the HFTs are not to be confused with trading. Trading historically provided important social functions in the marketplace such as price discovery, liquidity and risk transfer. Price discovery does not occur and is antithetical to insider information. Creating volume and artificial price is not liquidity. Lastly, the HFTs because of their position to cheat are not assuming any market risk-the ability to see price and order information before everyone else and to act on it means that they "cannot lose."

The Securities Acts came into being with provisions against insider trading and front-running because it was recognized that cheating and unfair advantage harms a free market and is at war with any well-functioning mechanism for price discovery. If the HFTs are allowed to continue cheating because of the established use of smart routing, in time everyone else will be unable to compete, including all investors and traders other than large volume HFTs.

The SEC and the United States Attorneys Office should look at the arbitrage practices of the HFTs for what they are-cheating. The fact that seventy-three percent of trading volume results from the arbitrage activities of HFTs should not lead them to be considered another industry "too big to fail." Yet, there is a very real possibility that even the SEC and un-sophisticated legislators may believe a bankrupt argument that regulating a de facto illegal enterprise that accounts for so much of trading volume may harm the markets.

The idea that markets are self-correcting and tend toward equilibrium may, all bubbles and crisis aside, still be true but an unequal playing field and entrenched insider information almost never self-corrects. What is at stake is nothing other than the integrity of the financial markets. ©

R. Tamara de Silva

November 30, 2010

Chicago, Illinois

[1] "What Good Is Wall Street? Much of what investment bankers do is socially worthless" By John Cassidy, The New Yorker, November 27, 2010

[2] "Stock Traders Find Speed Pays, in Milliseconds" by Charles Duhigg, The New York Times, July 24, 2009

[3] Current risk models are based on over-simplistic mathematical models, which fail catastrophically when they are most needed to work. There are many discussions of what constitutes risk in the financial markets but not surprisingly, there is no one definition. Typically, discussions of risk revolve around the concepts of Value at Risk (VAR), beta, delta, the capital asset pricing model (CAPM) and the Black-Scholes options pricing model (BSM). All these ways of quantifying risk are based on inarguably faulty assumptions. The recent credit crisis and mortgage meltdown is another example of how Wall Street's methods of modeling market behavior and accounting for risk, are fundamentally flawed. The financial models of investment bank analysts assign likelihood to the possibility of certain events occurring. Financial models assume a normal distribution (a bell curve) of asset returns or risk. The belief that most market events occur in a normal distribution is the single key assumption made by many financial models, including the capital asset pricing model (CAPM) and the Black-Scholes option pricing model (BSM) and VAR. Using a normal distribution, events that diverge from the mean or center of the bell curve, by five or more standard deviations, known as a five-sigma event, are very rare and ten-sigma events are nearly impossible. However, the 1987 market crash represents a change of 22 standard deviations. The odds of such a 22 standard deviation event occurring are so low as to deemed impossible.

[4] The value of transparency in the marketplace is possibly best explained by its absence-opaqueness. The lack of transparency is called opaqueness. The environment that led to the credit crisis was opaque. In the mortgage debacle, few of the players knew what the baskets of mortgages they were packaging, buying and selling were actually worth. The participants in instruments that led to this current crisis operated in a very opaque if not downright murky environment. The mortgage related securities being traded from brokers to banks and between banks were not pegged to the value of anything tangible. One could make the case that they were not even derivatives because their value was effectively not derived from an underlying anything. But if they were, their value was not discoverable, or perhaps not verifiable. The values of mortgage securities were not marked to market, they were not pegged to an underlying asset and if they were, no reasonable allowance was made for unfavorable movements in the value of the underlying assets. Price is one of, if not in many instances the single most relevant information provided in a transparent market.

[5] High frequency trading may be the next "too big to fail" because it accounts for approximately 73% of all trading volume.

[6] The success rates of high frequency "traders" appears at first blush to be statistically impossible but as discussion will reveal-they essentially cannot lose because of advantages they have over everyone else-specifically, the ability to see price data and order flow before everyone else, be paid for order flow and the absence of any apparent enforcement mechanism or arguably even prohibition against front-running or insider trading

[7] Now an advisor to Getco, LLC (one of the largest high frequency trading firms) and Goldman Sachs.

[9] June 6, 1934, c. 404, Title I, § 11A, as added June 4, 1975, Pub.L. 94-29, § 7, 89 Stat. 111, and amended Nov. 8, 1984, Pub.L. 98-620, Title IV, § 402(14), 98 Stat. 3358; Dec. 4, 1987, Pub.L. 100-181, Title III, §§ 313, 314, 101 Stat. 1256; Dec. 21, 2000, Pub.L. 106-554, § 1(a)(5), 114 Stat. 276

[10] http://www.nasdaqtrader.com/Trader.aspx?id=colo

[11] It is unclear, due to the opaque nature of this industry whether the exchanges pay for bids and offers alone or only those bids and offers that are filled.

[12] http://www.futuresindustry-digital.com/futuresindustry/201009?pg=56#pg56

[13] The market's volatility has increased (measured by aggregate computations of beta) over time coincident with the increase in trading volume of 164% since 2005. It is possible that a significant part of the increase in aggregate volatility may not be driven by fundamentals, but by high frequency trading, which already drives price information akin to the tail wagging the dog.