A systematic commodity trading advisor has to have a very effective execution capability. The turnover of futures contracts for the Aspect Diversified Fund equates to 4,000 round turns (openings and closings) per million dollars of capital traded per year according to data from the Institutional Advisory Services Group (IASG). Table 3 shows comparison data for CTAs on the database of IASG with individual funds of a billion dollars or more. The turnover of positions at Aspect is at the high end of the range for managers of the largest funds. One reason for this is that Aspect Diversified Fund runs models with various time-frames, including more transaction-intensive short-term trading strategies. The longer the time-frame of analysis of the models the less transactions there are to execute. Some of Aspect's peer group have an average holding period for a position of more than 10 days whilst the fast systems at Aspect have an average holding period of 3-4 days and the average holding period across all frequencies is approximately 1 month. Another style point also comes into play, as the tightness of stop-loss levels is also a factor in turnover. So the tolerance for losses in the "money management" part of the CTA process is a factor in turnover numbers. Finally a manager that specialises in markets that have a propensity to whipsaw rather than trend also will have higher transaction costs. For whatever style of CTA there are a lot of transactions to handle.

For each transaction there is a bid-offer spread, a clearing and/or exchange fee and commission. Aspect has become a member of Eurex to help reduce the per ticket costs, and membership also gives a small timing advantage because orders can be sent directly to the exchange rather than via a broker's exchange connection. Another transaction cost is slippage in execution. This is the difference between the market price at time of trade generation and the price at which the trader executes. For larger CTAs the size of order can be the dominant factor in slippage: the most frequently traded lot-size in a market can be 1 or 2 contracts (lots) (traded by locals/scalpers/day traders/jobbing brokers) or maybe 5 lot transactions. The trade generation algorithm of a substantial managed futures program may spit out an order which is significantly larger.

The drag of costs for CTAs can be the difference between first or second quartile in rankings of performance. With so many transactions to implement each element of the transaction has to be addressed to minimise costs, and at Aspect that is the responsibility of James Udall, Trading Manager.

"We are organised into trading teams by asset classes, " says Udall, "that way we get a real depth of knowledge and expertise of the particulars of the market. Sometimes we have taken on someone who was a broker to Aspect, that way we really know them. So we have a commodity group, a bond group, and teams for equities, , currency and the over-night group which trades everything liquid within the Asian time-zone. We operate 24-hours a day."

Motivation is not a problem as traders are inherently competitive. For each trade they are benchmarked against the last trade done in the market (time stamped at the time the open order comes into the department). They compete within the team, as well as liking to "take on the market" periodically. The chance to measure themselves against others in the firm also comes in corporate events like the Aspect sailing day held last year. It was events like these that contributed to Aspect Capital being voted number three Small Company to work for in a "Sunday Times" nationwide poll last year. The competitiveness comes from the top of the firm as much as from trader psychology. The founders of Aspect want to beat their peer group of large managers, and they track their returns assiduously.

The trading group is 14 strong, and they execute 140-150 tickets that arise in a normal day and up to 250 in a busy period. For each asset class team in trading there is at least one member who acts as the link person with the research team. "We have regular meetings with the research staff so that we can feed them ideas for research, they can get an idea from us of how practical some of their ideas are, and we like to discuss where we can see liquidity we can trade, " says Udall. These links help the traders understand the reason why a trade has been generated by the systems. This allows the traders to go back to the research group and check whether an order is correct. The trading department has a tool to verify the type of order per market. So for example, in a market with upward momentum a sell order should not come in from a momentum model – that order can be queried by trading with research staff.

Some of the search for capacity is taking in markets which have become newly liquid, and markets which are commoditized (standard terms, freely traded and liquid) but don't happen to be exchange traded (like swaps). At the moment the ability to execute trades is not a constraint on Aspect's capacity – "I genuinely see a lot of scope to improve our execution capability further, " states a confident Udall. He sees other markets moving further towards electronic trading, and that can only reduce the costs for a firm like Aspect: "we execute 80% of our bond business electronically now," he says "and others can go the same way."