Commodities Traders Increasingly Adopt Algorithms

Commodities Traders Increasingly Adopt Algorithms

Commodities traders increasingly are adopting algorithms, but these aren’t the equity markets’ automated trading strategies, as commodities algos often make room for human discretion. By Ivy Schmerken
Algorithmic trading is continuing its march through the asset classes. Proprietary trading shops with experience in other asset classes are bringing algorithmic trading to commodities markets, according to several experts.

Nigel Kneafsey, CEO of Options IT, a provider of high-performance trading technology infrastructure as a service, says automated trading strategies are expanding into commodities, such as energy and agriculture. “A lot of these firms are trading one part of the futures market and then moving to others,” he relates. After they “deploy a certain amount of capital into a strategy, they feel they have to move into another [asset class].” Most of the automated trading in commodities is taking place on futures and options exchanges in the Chicago area, though there is some activity cropping up in New York, notes Kneafsey, whose firm provides connectivity to the CME, CBOE and Intercontinental Exchange (ICE).

One of the reasons commodities are attracting new players and new strategies is that the markets are not as mature as equities. “It’s a less-trodden path than, say, equities, but the facilities are there to do it if one has the right strategy,” Kneafsey points out, suggesting that there is greater opportunity in commodities for electronic strategies to take advantage of market inefficiencies. Also, with more and more institutional money flowing into commodities, both through exchange-traded funds (ETFs) and listed futures and options, managers are looking for new ways to generate alpha in commodities.

“The asset class that has seen the most recent take-up in automation is commodities,” says Mike Reilly, head of commodities for Europe, the Middle East and Africa (EMEA) at Progress Software in London. Proprietary trading firms, he adds, don’t care about the underlying asset. “All they want is liquidity and price discovery and enough depth. They don’t want to be in positions [that they] can’t close within a reasonable time frame. Whether that’s U.S. Treasury bonds, oil or sugar — it doesn’t matter,” Reilly explains.

Though the exchanges don’t necessarily reveal how much of their volume originates from machines versus human traders, Reilly asserts that there is evidence that in the most-active front-end oil contracts (ICE Brent Crude Futures and WTI Nymex), as much as 35 percent to 50 percent of the volume is generated by algorithmic trading.

According to estimates, about 15 percent to 20 percent of all futures trading occurs in commodities futures contracts. Within commodities, there are various sub-types, including metals; agricultural products, such as sugar, soybean futures and grains; and energy, says Dan Hubscher, Progress Software’s principal, production marketing manager, for capital markets. “Most people talk about energy, and within energy, crude oil, natural gas, power and coal, and refined products such a heating oil,” he relates. “Automated trading is being applied to the most highly volatile [commodities], and active subsets of that tend to be energy futures contracts and specifically crude oil.”

Not Algorithmic Trading as We Know It

But while automation of commodities trading is growing, Hubscher says, “The actual trading strategies aren’t necessarily algorithmic the way we know them from equities and other futures.” Most of the trade automation in the commodities space is via electronic trading and risk management, or ETRM, systems, he explains. “This is more click-and-point trading, where the trade is executed by a human decision-making process.”

According to a recent survey, however, two-thirds of ETRM users are ready to rip out and replace their systems because they’re not keeping up with the markets, Hubscher points out. While ETRM provides screen-based trading and automated risk management hooks, Progress’ Apama division is promoting the concept of programmatic trading and risk management (PTRM) to commodities traders. With PTRM, algorithms support the trading decision and the risk management all in one integrated process that is supported by Apama’s complex event processing (CEP) technology, Hubscher says.

Some commodity trading advisers, however, rely on a combination of strategies that involve human discretion as well as programmatic rules. Nautical Capital in Boca Raton, Fla., is part of a new breed of trading advisers that focuses exclusively on commodities. “We are a commodity asset management firm that develops and manages multiple absolute return strategies,” says CEO David Henritze. For now, the firm trades with the managing principal’s capital, but Nautical plans to distribute its strategies through banks, Henritze reveals.

Nautical currently trades three strategies — two discretionary strategies driven by the commodity views of the portfolio managers and one systematic strategy. The first discretionary strategy capitalizes on relative-value opportunities through inter-commodity and calendar spreads; the second is based on directional views and is implemented via options. The systemic strategy, Dynamic Commodity Trading System, employs trend-following and fundamental algorithms to determine long, short or flat (cash) exposure to 15 commodities.

“On the futures side of the business, it’s very automated,” Henritze relates. “The way we put the trades on is done electronically though futures commission merchants.”

Ryan Carrier, Nautical’s chief investment officer, emphasizes that over time, the discretionary and systemic strategies perform differently in various markets. “With discretionary strategies, there is always human involvement in figuring out the weightings for the algorithms,” he says. “By having a range of strategies over time, investors will be diversified and make out the best.”

Carrier adds, “In spread trading, I don’t care if commodities go up or down, and both of those are discretionary strategies.” For example, sugar reached 40-year highs in late January. There was a supply shortage, and the near-term March contract went from 9 cents per pound to peaks of 31 cents per pound, according to Carrier. Meanwhile, May sugar contracts were selling at a 10 percent to 15 percent discount on the March contracts.

“We sold the March sugars at the highs, and we bought the May sugar. We would call that a market-neutral trade,” Carrier says. “We don’t care about the direction of sugar. We are focused on the relationship between the two contracts implemented through a calendar spread.” Carrier notes that the firm also trades inter-commodity spreads, such as crude oil to heating oil. “A lot of it is electronic,” he says, adding that everything Nautical trades is listed on an exchange and cleared through the exchange’s facility. “It’s all highly transparent and highly liquid.”

Algos Jump Trading Desks

Gary Stone, director of trading research and strategy at Bloomberg Tradebook, acknowledges that algorithmic trading is different in commodities than in equities. “In the equity market, you have a position and an order to fill, whereas in commodities, you have a strategy that you’re trying to implement,” he says. But, Stone points out, commodities algorithms are growing increasingly sophisticated.

Bloomberg Tradebook, a subsidiary of Bloomberg, began offering programmed strategies for futures about five years ago. Some customers, according to Stone, wanted to bring equity futures algorithms to commodities, and many of the strategies that Tradebook has rolled out for commodities are similar to those used on the proprietary trading desks of bulge-bracket banks. For example, Tradebook offers trailing stops, which allow a trader to trail the commodities market up while the computer automatically sets stop loss orders below the market price.

“We’ve done a lot of work with our clients in creating the order types they’re looking for,” says Stone. “If you are dealing with the energies and the foods, you can do crack spreads and crushes, or pairs algos to trade corn versus soybean, or you can do soybean versus oil.”

This year, Stone reports, Bloomberg has made a push into intra-commodities trading or arbitrage. “Our volumes have grown dramatically because we’ve been able to go across exchanges, both inter- and intra-exchanges,” he relates.

Ultimately, Stone suggests, “Commodities traders are doing things they couldn’t do before. And they can do it with significantly less slippage and more precision.”