When backtesting you strategy, is there a minimum amount of time you traditionally test to validate your findings? I understand that a week or a month is obviously not long enough, but what do you use? 1 year? 5 years, or 10?
When you back test, it is important that you not only consider the length of time, but also the number of trades that you take. The number of trades that is included in your time period would be more important, IMHO, than the time period itself. For example, many of the automated traders that I have encountered consider 1000 trades to be a good number of backtested trades.