Forward testing means trading your written rules in real time on live prices, usually on a demo account or with very small size. It tests two things a backtest cannot: whether the rules work on data that did not exist when you wrote them, and whether you can follow them.
How to do it
- Use exactly the rules you backtested. Do not "improve" them on the fly.
- Take every valid setup, including the ones you do not like.
- Log each trade in R, with whether you followed the plan.
- Collect enough trades to be meaningful: dozens, not a handful.
What to measure
- Adherence. What share of trades did you take exactly as the rules say? If adherence is low, the problem is discipline, not the strategy.
- Expectancy in R per trade.
- Maximum drawdown and the longest losing streak.
- Differences from the backtest. If forward results are much worse, find out why: slippage, missed fills, or changes in the market.
Decide the next step in advance
Before you start, write what result would make you continue, change, or stop. For example: "After fifty trades, if expectancy is positive and adherence above 90%, I may trade with real money at 0.25% risk." Having the rule in advance prevents you from rationalising a poor result.
Scaling up carefully
If you move to real money, start with very small risk per trade and increase only slowly after you have shown, over a large sample, that you can follow the plan under real pressure.
Keep reviewing
Markets change. Review your numbers regularly, and be willing to retire a strategy that no longer works rather than defend it.