This thread has brought up an interesting point about backtesting. The prevailing wisdom is very much test and optimise over long periods or a strategy cannot be seen as robust. But if market structure changes to the point where it makes sense to withdraw from the market because the strategy cannot cope, is this the best tactic?

Testing my strategies its become obvious there has been a big change in the indices, and to a lesser extent Forex, since the end of 2017. If I optimise a variable on a 1H TF from mid 2016 to now two outcomes are more likely than others: either the period from mid 2016 to end 2017 is somewhere close to break even and all the strategy profit is in 2018, or the profit occurs mainly in 2016/17 and 2018 shows break even or a loss. To my simplistic view this leads me to conclude that optimising for how the market is now is better than having a strategy that’s a compromise in all market conditions.  So all my strategies have been tested and optimised for 2018 only (all are on 1H TF or smaller so I don’t see it as a major issue given they all generate at least 60-100 trades this year).

This does mean a strategy is not something to set up once and forget. It will need monitoring and re-optimising as the market evolves. To me that’s a price worth paying for a strategy that makes a good profit now.

Note – I don’t believe this would be the case if I were trying to put together strategies on the 4H TF or bigger. These longer TF strategies would need to work over months and years, whereas my short term strategies can rely on weeks and months of trading to get a good set of data.

What does everyone think? Does it make sense to work with the prevailing market structure on the smaller time frames?