Backtesting Crypto Trading Strategies

The main question of this talk is how can trading strategies be evaluated? In other words, how can I avoid investing in supposedly profitable trading strategies?

We use a Simple Moving Average strategy for cryptos as an example to show how multiple testing issues arise through data mining. We use methods proposed in Harvey and Liu (2015), Backtesting, The Journal of Portfolio Management, 42 (1) 13-28, to adjust our performance measures for multiple testing, for example, by calculating a “haircut Sharpe ratio”. The idea of this talk is to raise awareness that many supposedly profitable strategies, if “fair” backtesting is performed and high but reasonable testing standards are applied, are no longer profitable.