Courtesy of the author, we present to our readers the following excerpt from Dark Pools: High-Speed Traders, AI Bandits, and the Threat to the Global Financial System, by Scott Patterson, author of The Quants. Part 1 can be found here.
Haim Bodek thought practically nonstop for days about what the trade-venue representative had told him that night at the New York party.
The way that the abusive order types worked made him think back to a document he’d been given by a colleague that summer as he researched what was going wrong at Trading Machines. The document was a detailed blueprint of a high-frequency method that was said to be popular in Chicago’s trading circles.
It was called the “0+ Scalping Strategy.”
Bodek suspected that there might be a link between the order types and the strategy.
Riffling through his files, he quickly found it. While the document didn’t say which firm used the strategy, he’d been told by the colleague who’d given it to him that one of the most successful high-speed firms employed it, or something closely akin to it. Due to the sophistication of the strategy, he’d guessed from the start that it was probably written by a Plumber.
There was another giveaway that it had originated in Chicago, where Bodek had worked for several years at Hull Trading: “scalping.” To a trader, scalping didn’t mean the same thing it meant to most people—a suspicious-looking guy peddling tickets for a sporting event or rock concert outside a stadium. In trading, scalping was an age-old strategy of buying low and selling high—very quickly. It was a common practice on the floors of futures exchanges that populated the Midwest—the Kansas City Board of Trade or the Chicago Mercantile Exchange. The 0+ Scalping Strategy was apparently a futures-trading technique that had been transformed into a computer program.
Bodek started reading. Page two of the document laid out the purpose of the 0+ strategy. “Simple Goal: use market depth and our order’s priority in the Q to create scalping opportunities where the loss on any one trade is limited to ‘0’ (exclusive of commissions).”
He paused at that. Essentially, the author of the strategy was saying that its primary goal was to never lose money—the loss on any trade was “0.” In theory, this could be done through a scalping strategy. By being first in the “Q”—shorthand for the queue in which orders are stacked up, like theatergoers waiting in line for their tickets—the firm could always get the best trade at the best time.
But what happened when the firm didn’t want to buy or sell? Bodek kept reading.
“GOAL RESTATEMENT: use the market depth and our order’s priority in the Q to create scalping opportunities where the probability of a +1 tic gain on any given trade is substantially greater than the probability of a –1 tic loss on any given trade.”
Aha, Bodek thought, market depth. That was a reference to the orders behind this firm’s orders, the other theatergoers waiting in line. The 0+ trader is assuming that his firm is so fast and so skilled that it can almost always get priority in the trading queue—be the first to buy and the first to sell. The depth behind it, the other orders, is the rest of the market.
The author is saying I always want to win (or rather, I never want to lose). His probability of winning—a +1 tick—is “substantially greater” than a –1 tick loss.
The rest of the market—suckers like Trading Machines or every- day mutual funds—was insurance. Under the next heading, called SIMPLE PREMISES, the exact meaning of what insurance meant was spelled out.
“If we have sufficient depth behind our order at a given price level, then we are effectively self-insured against losing money. Why? If we get elected on our order, we could immediately exit our risk for a scratch by trading against one of the orders behind us.”
In other words, if the 0+ trader buys a stock (gets “elected”), and his algos suddenly detect that the price is likely to fall—they can see a large number of sell orders stacking up in the trading queue—he can flip and sell to the sucker standing behind him, resulting in a “scratch” (no gain and no loss). He can do this because his computer systems can “react fast enough to changing market conditions . . . to ‘always’ achieve, in the worst-case, a scratch or a cancel of our orders.”
It was the Holy Grail of trading. The 0+ trader was describing a strategy that effectively never lost. The rest of the market protected it whenever the firm’s algorithms detected the slightest chance that the market was moving against it.
It’s brilliant—and diabolical. A firm that has found a strategy that is virtually guaranteed to win on every trade has discovered a hole in the market. Trading is all about taking risk, but this author was describing a virtually riskless trade.
The situation confronting Bodek and other investors not using the 0+ strategy was challenging, to say the least. It was like driving a car down the freeway, and every time you tried to speed up, another, faster car was in front of you. No matter how many tricks you pulled, this car (a 0+ symbol stamped on its hood, of course) was always leading the pack. The only time you could get around it—when it would suddenly hit the brakes and vanish in the crowd behind you—was when a Mack truck was speeding right at you. Worse, the 0+ trader was the Mack truck!
The upshot: Regular investors, the suckers using those stupid limit orders, buy high and sell low—all the time.
The game had changed. Bodek became increasingly convinced that the stock market—the United States stock market—was rigged. Exchanges appeared to be providing mechanisms to favored clients that allowed them to circumvent Reg NMS rules in ways that abused regular investors. It was complicated, a fact that helped hide the abuses, just as giant banks used complex mortgage trades to bilk clients out of billions, in the process triggering a global financial panic in 2008. Bodek wasn’t sure if it was an outright conspiracy or simply an ecosystem that had evolved to protect a single type of organism that had become critical to the survival of the pools themselves.
Whatever it was, he thought, it was wrong.