THE BLOG

Has Computer Trading Made the Stock Market a "Crapshoot"?

02/08/2011 01:16 pm ET | Updated May 25, 2011

Jim McTague contends in his forthcoming book, Crapshoot Investing: How Tech-Savvy Traders and Clueless Regulators Turned the Stock Market into a Casino, that the present use and regulation of computerized trading, High Frequency Trading (HFT) have severely harmed investors and the economy. McTague, currently the Washington Editor of Barrons, has decades of experience reporting on the regulation and operation of the financial sector. In this book he takes on people in an industry with immense profits and top government officials who tried to regulate equity (stocks and other financial assets) trading . McTague presents a picture that was well known by professional traders while largely uncovered and unknown by the general public, especially retail investors, until May 6, 2010, the day of the Flash Crash.

McTague's views confront long standing academic theories first advanced at the University of Chicago after intensive study of the closing prices of all stocks back to 1929. When it was found that throwing a dart at the Wall Street Journal stock page would have produced better results than listening to the professional stock pickers, Professor Eugene Fama produced the theory of efficient markets in 1965. It held that all public information is already discounted into stock prices so that, at any given moment, the future change in the price of a stock is best described as a random walk.

At least one professor at the University of Chicago did not believe the efficient market theory. Milton Friedman, my teacher and friend, said that traders would not buy a membership on the New York Stock Exchange in order to trade on their own account if stock prices were a random walk.

The efficient market theory became a central part of the curriculum in many business schools throughout the country for decades including the present. Now McTague turns this theory around. Human traders are getting beaten and manipulated by machines that are not on random walks, at least in the short run.

Many people are familiar with the TV broadcasters interviewing people on the trading floor of the New York Stock Exchange, which McTague cites as a backdrop for CNBC and Fox Business News. But that is no longer where most of the trading is done. McTague says that "In May 2010, the NYSE, which at its peak boasted of an 80 % share of all the volume in NYSE listed stocks, reported a 21 % market share. The volume could include the NYSE's own electronic trading computers housed at an undisclosed location. The backdrop that McTague cites with its trading booths and order makers called "specialists," handles a small and shirking volume of market volume.

The main momentum for the decline is the increased use of computers with software (algorithms) programs that now trade from 50% to 70% of equities. Machines which trade in milliseconds (1 millisecond = 0.001 seconds) have replaced human traders. McTague describes conditions on May 6, 2010, the day of the Flash Crash. There were 11 registered exchanges and more than 70 alternative trading systems (ATS's). McTague says that "thirty of the ATS's were so called 'dark pools' that did not trade in the lit market and thus did not display there quotes."

If the HFT's firms are near an exchange, such as using the NYSE's large computer center in its undisclosed location, its computers can buy and sell equities much faster than human traders. McTague says that in a single day of trading, HFT trading is equivalent to 30 years in a natural market and that 2% of all market participants generate 70% of the equities volume.

The HFT algorithms search the market for price discrepancies. Within milliseconds they exploit this discrepancy buying low and selling high. The HFT proponents, including many traders, welcome this arbitrage operation because it keeps the bid and ask prices for stocks very close together. It is very useful for traders to know an exact price rather than a range between bid and ask prices.

The algorithms can be programmed to detect buzz words from news makers, such as Fed officials, allowing HFT computers to react ahead of the human traders. This is a form of "front running". The HFT firms can indulge in "momentum ignition" or "spoofing" where they buy millions of shares of a stock to attract interest; then sell in a few milliseconds benefiting from human traders who have taken the bait and pushed prices higher. HFT firms can also use "quote stuffing" or "pinging:" entering many orders to see if there is interest and then canceling them.

McTague attacks government regulators and regulations that he says created many of the problems with HFT's. The Securities and Exchange Commission under Chairman (1993 to 2001) Arthur Levitt required the ATS's "to provide investors a fair opportunity to participate in the system." McTague says the exceptions the SEC allowed in the "fair excess rules" created the ATS's "dark holes" and benefited hedge funds.

SEC Chairman (2003 -- 2005) William H. Donaldson had the SEC adopt a National Market System that required "a new trade-through rule for the electronic age that said you could trade through the lowest displayed quote at an exchange if the exchange responded within 1 second." The rule caused the NYSE to become an electronic exchange itself to respond in 1 second and its share of volume in its listed stock fell to 24%. The advent of many HFT firm's began in 2005.

McTague said it set the stage for the Flash Crash. On May 6, 2010 at 2:30 P.M EST the Dow Jones average of stock prices fell 998 points wiping out $1 trillion in equity. In just 10 minutes the market suddenly reversed itself and the whole episode took 20 minutes. There were huge problems that occurred. Two hundred stocks traded at zero, a nice pick for algorithmic commuters that could pick them up in milliseconds . These and other trades were later rescinded. The HFT firms had a 35 second advantage during this crisis, a huge advantage.


Some people tried to blame the crash on one firm which placed a large sell order, $4.1 billion. The firm used manual trading and an "execution algorithm" [or 'sell algorithm'] which took into account price, time, and volume. " It took over 5 hours to execute the first 75,000 contracts." The "relatively small orders... were both legitimate and consistent with market practices." McTague blames much of the sudden contagion that drove stock prices down on the algorithmic computers that responded . One HFT strategy called "sniping" attempts to identify large orders to see if a group of small orders are coming from the same place.

McTague describes similar spike events on individual stocks had already occurred. For example, Washington Post stock jumped 99% in one second "from $462.84 to $929.18" on June 16, 2010, "yet another public embarrassment for the SEC and the exchanges."

The staff of the Commodity Futures Trading Commission" contributed language about HFT problems in the 2010 Dodd-Frank Wall Street Reform and Consumer Protection Act ,according to McTague. It prohibited practices deemed "disruptive of fair and equitable trading" called "spoofing". I believe there is a serious problem about how this kind of broadly defined regulation will be applied, especially in light of the poor effects of the SEC regulations McTague describes.

McTague's book is an indispensible read for everyone interested in the problems and benefits of replacing human traders with algorithmic machines in one of most important markets in the world. The equity markets play a central role in the allocation of scarce resources to many of the most productive enterprises that provide jobs for workers, and goods and services for consumers. Retail investors should also consider Jim McTague's advice about short-term stock trading: "It has become a shark tank and we are the anchovies."