A New Formula for Intercommodity Spreads

Borrowing a page from chemistry, CME Group has taken the complex art of intercommodity spreads trading and made it more efficient and less risky for customers.

For about a decade, CME Group has been trading non-implied intercommodity spreads – where both components of the spread trade the same month and the same year at a specific ratio – across many different contracts. Only recently did CME Group begin moving into the more complex realm of implied intercommodity spreads, beginning in 2006 with crack spreads – the differential between the crude oil price and petroleum products extracted from it.

According to Donna Nehila, director, software engineering and match engine development at CME Group, moving to implied intercommodity spreads proved to be extremely challenging, as the technology had to be customized for every new spread type.

“There was very little reuse of logic,” Nehila says. “Every implied combination had to be coded separately. It was very labor intensive.”

The original design, which was complex, inflexible and tough to maintain, posed a challenge for new developers on the system. With plans afoot to expand the CME Globex electronic trading platform’s ability to accommodate additional crack and Treasury spreads, Nehila and her colleagues realized something needed to change.

“Trying to accommodate these spreads using the current code base would have been too complex,” she says. “We recognized we needed to move forward to redesign implied functionality.”

Nehila and her team decided to “look in-house and outside of our walls” and found a solution in the seemingly distant realm of chemistry. Using chemical-like abbreviations – think compound formulas like “H2O” – Nehila and her team were able to create a “template for spreads,” allowing CME Group technologists and traders essentially to plug in futures without having to do all the customized heavy-lifting. This in turn ensures that the exchange is competitively positioned for
new opportunities in implied markets.

CME Group began its proof of concept testing in the summer of 2007, which entailed a complete rewrite of the implied combinations and writing algorithms to handle what had been purpose-built unique functionality. The massive task took roughly 18 months of development and testing. By the fourth quarter of 2008, the new system was launched. It was worth the wait, developers and users say.

“The problem with trading the yield curve has historically been the execution or legging risk,” says Xavier Laurens, a partner and trader at Chicago’s Port 22 LLC, who trades CME Group’s implied intercommodity spreads. The difficulty in intercommodity spread trading had been in “buying one side and missing the other, therefore adding to your execution costs.” But, with the new products, Laurens says, “the intercommodity spreads have taken the legging risk out because you get both legs of the spread at once. It allows traders to focus on what really matters – being right on the direction of the yield curve.”

Back to Spring 2009 Issue Home