Sept 20, 2012 

High-frequency trading, which already has a sullied reputation, is even worse than the critics have charged, a new survey shows. 

The Federal Reserve of Chicago recently asked 30 firms associated with the industry — traders, exchanges, vendors and others — to evaluate where HFT stands in the wake of a series of high-profile blowups.

While the central bank's analysts knew there were issues, what they found exceeded their expectations.

Industry pros reported a rash of out-of-control computer algorithms that power the HFT platforms. They admitted that speed is more important than safety. And they even went so far as to say that they actually wish the industry was more regulated but want the rules to be applied equally, something that may not be happening today.

"Market participants at every level of the trade life cycle reported they are looking to regulators to establish best practices in risk management and to monitor compliance with those practices," Carol Clark, senior policy specialist at the Chicago Fed, wrote in a summary of the survey's findings.

The report came about as the Fed sought to uncover what happened in a series of well-publicized gaffes: A snafu that delayed the opening of the much-ballyhooedFacebook [FB  22.82    0.23  (+1.02%)   ] initial public offering; price spikes in dozens of stocks that happened after a glitch at Knight Capital; and problems at the BATS trading platform that squelched the exchange's own IPO.

What it found was an amalgam of firms desperately trying to stay ahead of each other even at the expense of their own businesses.

In theory, high-frequency trading, which is done through algorithms that process trades at lightning speeds in hopes of profiting by miniscule moves in prices, has a series of checkpoints that should prevent errors.

In practice, that's frequently not how it works.

"With the chance of an order passing though controls at so many levels, how can things go wrong?" Clark asked in her analysis. "One possibility Chicago Fed researchers found is that most of the trading firms interviewed that build their own trading systems apply fewer pre-trade checks to some trading strategies than others. Trading firms explained that they do this in order to reduce latency."

Latency is the time it takes to place an order to an exchange and have it confirmed. Making that process go as quickly as possible is at the heart of HFT.

The firms develop algorithms to speed the trades, but survey respondents reported the "algos," as they are more commonly referred, can run amok.

To address malfunctions, some firms said they fix bad algos simply by "tweaking old code" and putting new algos back into production "in a matter of minutes." One firm, though, said it had an out-of-control algo caused by a software bug inserted when it tried to fix an initial bad algo.

"Two-thirds of proprietary trading firms, and every exchange interviewed had experienced one or more errant algorithms. The frequency with which they occur varies by exchange," Clark said. "One exchange said it could detect an out-of-control algorithm if it had a significant price impact, but not if an algorithm slowly added up positions over time, although exchange staff had heard of such occurrences."

Clark proposes a variety of fixes for the system: Order and position limits, a "kill switch" that could halt bad trading immediately and at multiple levels; and profit-and-loss (P&L) limits that could restrict losses.

A broader question, though, could be why the nation's central bank is so concerned about irregularities in the stock market.

Nicholas Colas, chief market strategist at ConvergEx in New York, suggests one answer: That the Fed's quantitative easing programs hinge on an efficiently functioning market.

The Fed last week announced the third round of QE, an open-ended program to purchase $40 billion of mortgage-backed securities each month until the economy recovers enough to drive down the unemployment rate.

Central to the Fed's easing policies is the "wealth effect" created by rising stock prices.

"The upshot is that the Federal Reserve is right to be concerned about market structure, in that a strategy of boosting stock prices is predicated on the notion that people believe them," Colas said in a research note. "The Federal Reserve probably needs to rethink how stock prices 'transmit' information to consumers and businesses. The game is changing. Quickly."

http://www.cnbc.com/id/49102808