Stocks News

Tech helps commodity funds outsmart market

NEW YORK (Reuters) - Artificial intelligence is helping trend-following commodity hedge funds triumph in treacherous markets when human brains alone aren’t enough.

Artificial intelligence is helping trend-following commodity hedge funds triumph in treacherous markets when human brains alone aren't enough. REUTERS/Vivek Prakash

With industry data showing the average hedge fund down 20 percent or more this year due to strategies messed up by plunging commodity and stock prices, some in the game called Commodity Trading Advisors are up 50 percent or more.

But the trend-following CTAs say it isn’t all their work: credit should also go to their computerized trading systems.

“It’s like a computer playing chess against an excellent individual chess player,” said Bernard Drury, president and chief executive at New Jersey’s Drury Capital, a CTA relying entirely on systemic trading or trading without discretion.

“The computer can be counted on not to make human errors, not to make a miscalculation, not to be tired and not to have a bad day,” said Drury, whose systemic flagship fund is up 57 percent on the year managing about $155 million (98 million pounds) in commodity and financial investments.

CTAs count themselves as part of the $1.7 trillion hedge fund industry, trading in a wide array of markets that include energy, metals, agriculture, financial futures, bonds, stock indexes and currencies. Like hedge funds, they have management fees of 2 percent and performance fees of 20 percent and strive for alpha, or performance beyond market expectations.

But while hedge funds operate with little transparency, investors in CTAs can see where their money is invested and work out profits and losses at the end of each trading day. Since CTA trades are mostly on exchanges, they are also known as the “managed futures” industry and are better regulated than hedge funds that often trade over-the-counter.

A Reuters poll of 12 managers that invest in a basket of hedge funds found October could be a horrific month for the industry, with seven of out of 14 strategies poised to lose money, as the global credit crisis worsened.

The survey found that funds best suited to weather the storm were those deploying global macro strategies -- mixed bets on commodities and financials based on global economic conditions -- which are what some top-performing systemic CTAs have been following.

BarclayHedge, a database of hedge funds and CTAs, put average year returns for systemic CTAs at 10.74 percent versus the 8.97 percent for those trading on discretion.

“There is no model that is best in all situations,” said Sol Waksman, BarclayHedge’s founder.”But in the current environment, looking back, it looks like systemic traders have done better than discretionary traders.”

“How many people would have had the foresight to think that when crude oil was trading at over $140 a barrel earlier this year it would go to below $70?” Waksman asked.

Non-discretionary CTAs religiously follow buy and sell signals coughed out by their individual computer models. This can also lead them astray, especially during market turns or when markets are range trading with no clear trend.

BarclayHedge notes that May to August 2006 was particularly rough for CTAs when oil and other commodities were in a range.

“The most apparent weakness is that you could be right on your money but exit a trade too quickly, limiting the potential for gains,” said Ernest Chan, a Chicago trader and author who calls automated trading strategies “artificial intelligence”.

Drury concurs. “What you lose is the ability to have some idiosyncratic insight that makes you say ‘let’s put all our money, or increase the bet size, on a particular trade.’

But shortcomings in a system can be fixed by “refining, implementing new ideas and adding new markets,” said Ernest Kaiser, head of client relations at Switzerland’s Progressive Capital Partners, whose flagship systemic fund is up around 30 percent year-to-date.

Using computerized trading models to make bets on markets is not new. Hedge funds and other major market players reportedly made profits last year using electronic algorithmic trading models to not only follow trends but also speculate on them as both stocks and commodities rallied.

But as the credit crisis knocked U.S. stock values down by a third and prices of some commodities like oil by half, hedge funds with decidedly long, or bullish, positions suffered when they could not turn short, or bearish, as deftly as the trend-following CTAs.

“The role of a hedge fund is to offer long, short risk premium in a smart way. What the events of this year have proven is that a lot of hedge funds were purely long on risk premium,” said Marc Malek, managing partner at Conquest Capital, a New York CTA, whose systemic fund has returned 35 percent year-to-date.

“Also with hedge funds using discretion to bet on relative values, like shorting one stock while going long on another based on historical relationships, a deleveraging situation like this throws that model out of the window,” Malek said.

Editing by Marguerita Choy