Intelligent use of algorithms could be used to tackle some of the challenges market participants have encountered as a result of recent changes to the calculation methodology behind the WM/Reuters fix, a study by TD Securities has found.
“Technically, the WM fix cannot be perfectly replicated by any type of trading strategy. This is because a trading strategy will produce a weighted average effective price, whereas the WM fix is based, in part, on a statistical median price,” says Paul Aston, head of quantitative and algorithmic solutions at TD Securities.
“Only under very special circumstances will a weighted arithmetic average equal the median for a specific data sample, which in this context means there will always exist some amount of slippage or tracking error between an algo trying to replicate the WM fix and the official WM fixing price. We have done quite a bit of analysis to determine how far a trading strategy, such as the common TWAP, can practically and theoretically deviate from the WM Fix,” Aston explains.
The long-awaited changes to the fix methodology were implemented on February 15, following recommendations from regulators and consultation with the industry last year. Under the new framework, the calculation window has been widened from one minute to five minutes – in the case of the popular 4pm fix, the window now runs from 15:57:30 to 16:02:30.
“Statistically the new five-minute window introduces more exposure to price drift and volatility over the fixing window than under the old methodology. Hence, it will be more challenging for banks to cover their risk when promising clients the WM rate and it will require buy-side firms to become more sophisticated in the ways they seek liquidity around the fix. Algos could certainly play a big role in this space, but their effectiveness will depend on how well one can devise strategies to minimise slippage from the WM mid-price,” says Aston.