Liquidity Risk, Speculative Trade, and the Optimal Latency of Financial Markets

Abstract

How fast should a security trade? To answer this question, we model the trading of a security via periodic batch auctions and study how market quality depends on the clearing frequency of the auctions. In our model, the optimal clearing frequency is determined by four factors: (1) the volatility of the security’s value; (2) the dispersion of investors’ reservation prices; (3) the intensity of trading in the security; and (4) the correlation of the security’s value with other securities. We find that for most parameter values, market quality is maximized at intermediate auction intervals, when markets are neither too fast nor too slow. Based on simulations, we demonstrate that an analogous intermediate time delay is also optimal for order placements in continuous double auction markets. Finally, using proxies of the model parameters, we estimate that the ideal clearing interval for a typical S&P 500 stock is on the order of a few seconds currently, compared with several minutes in the 1990s.