On cars, old, new and future; science & technology; vintage airplanes, computer flight simulation of them; Sherlockiana; our English language; travel; and other stuff
I RECALL that my degree in mathematics came with a pledge to use this knowledge always for Good and never for Evil. Apart from occasionally torturing students with exam tie-breakers, I believe I’ve held true to this.
However, based on what I’ve learned from John Lancaster’s “Scalpers Inc.” in the London Review of Books, June 5, 2014, others of my training evidently didn’t take the pledge.
Lancaster’s review of The Flash Boys, by Michael Lewis, introduced me to High Frequency Trading, for short, HFT, a Wall Street concept that makes scads of money for some traders at the expense of others—including anyone owning any stock.
The Flash Boys dispels any notion that a stock exchange has anything to do with ticker tapes displaying prices, floors littered with Buy and Sell slips and ardent traders hustling around wearing ID tags.
Think of all this as a theater set for business TV reports. The real High Frequency Trading occurs in an ultra-high-security enclave of computers located in Mahwah, New Jersey, 25 miles north-northwest of lower Manhattan’s New York Stock Exchange.
By 2008, 65 percent of U.S. stock trading was performed by HFT algorithms, not people. Today, the percentage is even higher, giving the algorithm designers—the mathematicians—greater incentive to game these automated transactions.
Data travel at the speed of light, about 186,000 miles per second. Still, HFT firms spend big bucks to optimize fiber-optic transmission between, say, Chicago and Mahwah. It used to take 17 milliseconds (17/1000th of a second) for this round trip. HFT specialists have this down to 13 milliseconds, and they’re now talking in nanoseconds.
There are a billion nanoseconds in one second—more than there are seconds in 317 years.
As Michael Lewis observes in The Flash Boys, “The ‘haves’ paid for nanoseconds; the ‘have-nots’ had no idea that a nanosecond had value.”
As an example of HFT, one algorithmic game is called “front-running.” Suppose a trader executes a computerized Buy order for a particular stock. In the time between that trader hitting the Enter key and the NYSE at Mahwah receiving the order, the HFT firm’s computer “runs in front,” buying the stock first and then passing it on to the trader at a miniscule, but guaranteed profit.
The trader sees this as a Buy fulfillment, albeit at the higher end of the stock’s reported price range. He doesn’t even know he’s been had.
Two other techniques are called “rebate arbitrage” and “slow market arbitrage.” With either, for “arbitrage” read “gaming.”
As its name suggests, the rebate comes from an algorithm choosing an exchange that pays for the business; it’s a kickback for traffic.
Slow market arbitrage exploits tiny time lags in the reporting of stock prices between exchanges. Suppose the NYSE has a big seller, which thereby reduces the stock’s price. An HFT firm recognizes the sale, buys the stock on the NYSE and sells it on the other exchanges before the price gets updated.
As noted in The Flash Boys, “This happened all day, every day, and generated more billions of dollars than the other strategies combined.”
It’s enough to make me proud to be a mathematician. And a little embarrassed as well. ds
© Dennis Simanaitis, SimanaitisSays.com, 2014