what killed you was not statistics but the way you didn't interpret them. clearly you know nothing about building systems at all - or - interpreting their results either. it's always human to blame everything but yourself for failure, this is why no one can trade. m
Stop it. You know nothing about markets. You can not win 65-75% with 4-12% DD. Don't fool others here. It's impossible to do that. You would become a billionaire. So stop it. We all know that it's not possible.
mark brown is a LARPing idiot but you can with 65-75percent with a small drawdown. the reason is that the win rate is factor of your strategies effectiveness (with your upside and downside potential) and the drawdown is a factor of the amount of leverage you use.
If there is a strategy like this: it closely follows the Dow Jones Index and goes long if the market falls sharply (assuming the magnitude can be defined), it has been profitable for the past 30 years and has surpassed many fund managers. Questions to think about: A: This is an effective strategy B: This is not a strategy Why?
It depends on what you want. If you look only riskadjusted returns then yes, it can make sense to a limited point. Do not expect one hit wonders when you are searching for good algos. You can be happy with a Sharpe 1.5 to 2.0 longterm. The easiest to code is probably the seasonal strategy. Here you have the sell in may effect, the turn of month effect and the turnaround tuesday effect. That is it here. There is no need to look for any finetuning as there are also no intraday seasonal effects which can be exploited. With that seasonal you can get Sharpe 1.5 to 2, which is said is doable also for the longterm. But in any case you are just chasing the equity indices and want more than those buyandhold returns, there is only one option for you. You invest here with a leverage of 2 in the SP500 or Nasdaq100 index and then you beat those index returns. There is no other strategy that can beat those buyandhold returns with no additional leverage longterm.
If you only want to reduce the drawdown a lot by not reducing the average returns too much, the easiest what you can do on a longterm base is just to apply the 10 month or 200 day moving average on the SP500. If it closes below then you exit your long position, if it goes again above you go entry long again. With that you can cut your drawdown almost by half, somewhat in that region, without reducing your returns too much in the longterm. And now you have a better sharpe algo strat made just with one easy input. It is also backtest on many decades or even centuries to be an effective and working momentum strategy. Even there is scientifiy study showing that it works on global commodities applied over the last 800 years based on the Global Financial Data database (which has all the required past data). It can be that easy.