BACK-TESTING OF ALGORITHMIC TRADING STRATEGIES ON FOREX MARKET, OVER THE RISKY PERIOD OF FINANCIAL CRISIS
BACK-TESTING OF ALGORITHMIC TRADING STRATEGIES ON FOREX MARKET, OVER THE RISKY PERIOD OF FINANCIAL CRISIS

Summary/Abstract: An algorithmic trading system is a set of precise rules that automatically define, without any human intervention, trading decisions (buy or sell) in the markets. It is a completely autonomous investment system on the financial markets. The challenge assumed in this paper is to back-test algorithmic trading strategies over the risky period of financial crisis. We optimized these trading systems on the exchange rates of EUR / USD, based on Technical Analysis rules (the crossing of moving averages, and the Bollinger bands). The parameters of these technical indicators are allowed to vary freely on a given domain and genetic algorithms are used to maximize the Sharpe ratio with respect to those parameters. In order to asses the overall performance of the trading system we consider several cases, by alternating the two technical indicators mentioned above and by using two methods of computing the cumulated profit (equity function):without or with commission and slippage fees.

• Details
• Contents