Waiting Times Distribution for Second-Scale Price Changes in FX Market

摘要

Using second-by-second data, we show that the lognormal distribution provides a good overall fit to the waiting time between price changes in FX market, if both short and long waiting times are included. For the subset corresponding to longer waiting-times, we show that the distribution is well described statistically by a power-law, with each currency pair having its own power-law exponent in the range 1.73.5. As part of our quest to develop a quantitative explanation of the entire waiting time distribution in terms of trading behavior, we offer an explanation for the subset of shorter waiting times using a simple microscopic agent-based model featuring competition for limited resources.