Thursday, May 5, 2011

Perez, The Speed Traders

High-frequency trading has mesmerized people in search of easy money, has challenged regulators, and has been an easy target for the press. In The Speed Traders: An Insider’s Look at the New High-Frequency Phenomenon That Is Transforming the Investing World (McGraw-Hill, 2011) Edgar Perez tries to set the record straight. Alternating between an annotated timeline of the development of high-frequency trading and interviews with top high-frequency traders, Perez illuminates the world of speed.

We normally think of high-frequency trading as super-fast order processing, measured in microseconds. And, yes, it is. But a trader’s data feed (whether market quotes or news feed) also has to be lightning fast, as does his database management system. Adam Afshar, one of the interviewees, explained this in the starkest of terms: “High-speed database management, in my opinion, is the linchpin of any computational or robotic trading system. … To put this into perspective, a test done in Microsoft Excel that takes about five months takes five days in SQL … and less than one second in a database management system that is specifically designed for stock market and numerical analysis. … Of course, having low-latency data is important; having an analytics system, be it a genetic algorithm, neural network, or basic network, these are all things you cannot do without, but none of these things makes any sense or has any value, monetizable value at least, if you do not have a very sophisticated high-speed database.” (p. 133) Reading this, I felt—at least for one wrenching moment—that I had been consigned to the dustbin of history. Somehow neither Zeno (Achilles and the tortoise) nor Aesop (the tortoise and the hare) provided consolation.

But speed does not automatically translate into profit. Manoj Narang, the founder of Tradeworx, speaks to this point. “Profitability,” he explains, “can have two distinct meanings in the realm of high frequency, and these meanings are almost mutually exclusive. When applied to high-frequency trading, profitability usually refers to the consistency of profits, because consistency is what makes high-frequency trading so appealing. … However, to most people profitability means something entirely different, which is the overall amount of profit generated. In this regard, the highest-turnover strategies are at a decided disadvantage.” A typical trade will return only about 0.1 cent per share. “And that only includes trading costs, not the costs of running such an expensive and technology-intensive business to begin with!” (p. 197)

Moreover, he continues, assume that high-frequency traders account for the entire market volume, which has averaged 8 billion shares per day, and that they make 0.1 cent per share of profit. The yearly profit would come to $2 billion. “This might sound like a lot of money, but let’s keep in mind that this is the total amount for an entire industry. There are many other corners of the financial industry, such as derivatives trading or hedge funds, that generate hundreds of times this amount of profit in one year.” (p. 197)

Admittedly, not all shops run incredibly high turnover strategies. At the slower end of the scale, high-frequency trading more closely approximates automated day trading where positions can be held for minutes or even hours. Consistency suffers, but overall profitability can increase (or, of course, decrease).

Perez explores the charges that have been brought against high-frequency trading and for the most part, with the help of his interviewees, tries to dispel them. Did high-frequency trading cause the Flash Crash? No. As David Cummings, chairman of Tradebot Systems, wrote: “Who puts in a $4.1 billion order without a limit price? The trader at Waddell & Reed showed historic incompetence.” (p. 155) Some traders also point the finger at the practice of busting trades, which can punish liquidity providers.

The Speed Traders offers a sympathetic portrait of an industry that is often demonized. By tracing its roots (going back to Instinet which began operating under an unwieldy name in 1970), Perez demonstrates it to be a logical development of technological change in the financial markets. By separating out some controversial practices, such as flash orders, from high-frequency trading Perez makes it more defensible. All in all, an enlightening book.

No comments:

Post a Comment