Developing a Strategy So It Trades Like It Back-Tests
Developing a tradeable strategy involves two steps. The first is to define the risk/reward balance the trader can live with, and the second is to develop a strategy that meets those guidelines and trades in real time like it back-tests. Most of the rest of this book will address how to develop strategies that meet your trading objectives. This chapter addresses something much more important: the pitfalls a developer will run into in developing a strategy that will trade like it backtests. There are many “how to develop” ways; this book only shows mine, but there is only one way to make sure your developed strategy trades in real time like it backtests. That one way involves minimizing curve-fitting in your development and making trading assumptions in your development that work as realistically as possible. Far and away, the biggest pitfall is curve-fitting. This chapter uses statistics to help clarify the concept, but don’t skip over it because of the math. These are important concepts, and if you don’t understand them you will continually develop systems that fail in real-time trading and you will never understand why.
When you’re developing a trading strategy, you test trading ideas against historical data to see if the idea is profitable. If you had an infinite amount of historical data, you could be sure that an idea that proves profitable on the historical data would be just as profitable in the ...