However, in a change to the quantitative trading strategies launched to date, a manager has been added to monitor the strategy and check the algorithm is still appropriate. Diapason Commodities Management, a Lausanne, Switzerland-headquartered commodity asset-management firm, will review the strategy on a six-monthly basis, and propose any changes to a strategy committee, which includes members of the asset manager and JP Morgan.
The changes to the strategy could include changing parameters of the algorithm, adding or removing commodities from the universe of available assets, or tweaking the roll mechanism.
“We might propose some change to the algorithm because of extreme market conditions in terms of risk,” said Frederic Hervouet, head of sales at Diapason. “Also we want to be as diversified as possible, so as other commodity markets become more liquid, we can add them as components to the basket.”
Structured products based on quantitative trading models have proved popular with institutional, high-net-worth and retail investors over the past year. As the payouts are dependent on rules-based algorithms, the products are considered more transparent than the hedge fund strategies many seek to replicate. As trading decisions are made by a quantitative model rather than by a manager, the products also typically charge lower management fees than hedge funds.
The Diapason Commodity Rotator charges an annual fee of 130 basis points. But Lionel Semonin, managing director and global head of commodity investor products at JP Morgan in London, says comparing the Diapason Commodity Rotator to other quantitative trading strategies is not comparing like with like.
“Investors want the transparency in how the basket components are chosen and the overall trading strategy, but they also want advice, since the commodity market is opaque and the pace of change is quick,” said Semonin. “This offers the best of both worlds – the discipline, the framework and the quantitative aspect of it, combined with some discretionary asset management. The fees are cheaper than for a fully managed product.”
Based on 12 years of backtesting, the internal rate of return of the product is 20.7% net of all fees, with a daily volatility (annualised) of 14.6% and a Sharpe ratio of 1.1. The maximum exposure of any commodity in the basket is 10% – that means if only one out of the 25 commodities meets the performance and mean reversion tests set out by the algorithm, the remaining 90% of the investment would be held in cash rather than being invested in the single performing commodity.
“The algorithm works well in the sense that there is under-investment when there is a bear market. The model aggressively re-invests up to 100% when the market is coming back to a bull situation,” said Hervouet. “The effect is that you have a high correlation with the commodity indexes when the market is rising, but a low correlation on the downside of the market.”