Forecasting and hedging in the foreign exchange markets. (Q838348)

The author investigates FX markets by using tools from machine learning and artificial intelligence and combining these tools with sophisticated financial modeling techniques. The book starts with a partial review of the classical concepts of exchange rate determination theory and financial theory. Then, the author addresses the problem of predicting daily exchange rate directions with support vector machines (SVM). The strategy is to build econometric models in order to extract statistical dependencies within an exchange rate time series that may be based on linear or nonlinear relationships. The underlying problem of forecasting exchange rates ups and downs is represented as a classification task. A supervised learning approach is used where the optimal target function is formally represented by a binary classification problem (BCP). Finally, exchange rate hedging in an optimization framework is considered.