Bullish or bearish: A Bayesian dichotomous model to forecast turning points in the foreign exchange market

While most practitioners in the foreign exchange market always try to detect whether the market is “bullish” or “bearish,” there is no application of statistic method to predict the dichotomous sentiment. This paper presents a Bayesian trading system to forecast market trends and reversals. By estimating the market price distributions, we can apply the optimal stopping rules to answer the question and control the trading system according to the a priori probability and risk preference of the investor.