Extreme Returns without News: A Microstructural Explanation

What triggers extreme exchange-rate returns? Though news is the source of volatility in standard theoretical models, in reality volatility is often unrelated to news. This paper shows that extreme exchange-rate returns -- and, more generally, high kurtosis of returns -- are statistically inevitable even in the absence of news. We identify four microstructural sources of return kurtosis in price-contingent order flow: (1) high kurtosis in the distribution of price-contingent order sizes; (2) clustering of price-contingent order executions at certain times of day; (3) clustering of order executions at certain price levels; and (4) the tendency of positive-feedback trading to propagate trends. Using simulations calibrated to price-contingent orders placed at a major foreign exchange dealing bank we show that when each factor operates in isolation, the one that contributes most to kurtosis in returns is kurtosis in the order-size distribution. When the factors operate simultaneously, however, their interactions prove far more important. Extreme returns in the absence of news should be viewed as natural rather than anomalous.