‘Signatures’ promise quants a tool for all jobs

Little-known mathematical technique could find applications from pricing options to sniffing out alpha signals

Quantitative finance is an exercise in spotting patterns in a chaos of data – prices, cashflows, spreads, volatility. Myriad mathematical methods and models exist to detect these patterns. But a new idea known as signatures may soon transform how quants tackle the task.

The fledgling technique works by capturing the essence of a stream of information. Thus, users can compare time series data quicker and more efficiently, potentially helping speed up all manner of quant calculations from pricing