- Martingale is Evil?!
- FOREX - Trends, Forecasts and Implications (Episode 17: July 2012)
- Need advice automate my profitable strategy to MT5 EA and backtesting on multi-currency.
Hi,
For me it's important the Return/DD ratio (total profit / drawdown).
Also, Profit factor tells a lot (sum of winnings trade / sum of loosing trades). Success rate is not that important for me, especially when I am using breakout strategy, or any other strategy with RRR (return to reward ratio) 2:1 or higher.
To sum this up, I am looking at
number of trades (to know that I have statistically sufficient backtest, you can't evaluate stratyg with only 50 trades, I want 100-150 at minimum)
Profit factor > 1.3
Return/DD Ratio, at least 4.
Of course, you can have a good strategy with different numbers, if you are OK with bigger drawdown or other reason. These are only my thoughts.
You don't have to have one perfect strategy, you would most probably over-fitted it (optimalized strategy for the backtest data, which doesn't trade well in the future/unknown data) and it would not trade well in the future. Better solution is to have a lot of strategies combined to portfolio with focus on low drawdown and stable profits.
Hi,
For me it's important the Return/DD ratio (total profit / drawdown).
Also, Profit factor tells a lot (sum of winnings trade / sum of loosing trades). Success rate is not that important for me, especially when I am using breakout strategy, or any other strategy with RRR (return to reward ratio) 2:1 or higher.
To sum this up, I am looking at
number of trades (to know that I have statistically sufficient backtest, you can't evaluate stratyg with only 50 trades, I want 100-150 at minimum)
Profit factor > 1.3
Return/DD Ratio, at least 4.
Of course, you can have a good strategy with different numbers, if you are OK with bigger drawdown or other reason. These are only my thoughts.
You don't have to have one perfect strategy, you would most probably over-fitted it (optimalized strategy for the backtest data, which doesn't trade well in the future/unknown data) and it would not trade well in the future. Better solution is to have a lot of strategies combined to portfolio with focus on low drawdown and stable profits.
100-150 trades minimum, isn't that way too low? How many years of backtest is this for you?
No worries, i would never comment a commercial product here.
That depends on the timeframe. Sure, more number of trades is always better, and preferably from longer time period to reflect different market cycles.
You can also try Monte Carlo simulation, for analysis if your strategy behave same for changed market conditions. Because one thing is good strategy performance, another thing is whether the strategy will perform same in future or with different conditions, like higher spread, bigger volatility and so on.
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use