I'm 100% mechanical and it took me years to figure out something that worked. Since I can't program luck or intuition, I have to rely on numbers and stats. So these are some of my rules:
1) Trade only with a major trend. Use a short average (like 15) and a long average (like 50) to see whether to go long or short.
2) Once you know the trend from step 1, figure out WHEN to go long or short within that major trend. That's the easier part. Here comes the hard part.
3) You can't just reverse your entry parameters to get out of trades. That's why trading using ma crossovers doesn't work. Entries are usually decent if there's a good trend, but they get you out too late. You have to find some other parameters on which to base your exits (hint: look at volatility).
Forget backtesting and most indicators. Come up with a sound theory, code it, and then let it run in real time for days and weeks and months on a sim account. If successful, then backtest it if you want and optimize. But if you have too many things to optimize, forget it. Then trading becomes voodoo. A solid system should be "simple".
1) Trade only with a major trend. Use a short average (like 15) and a long average (like 50) to see whether to go long or short.
2) Once you know the trend from step 1, figure out WHEN to go long or short within that major trend. That's the easier part. Here comes the hard part.
3) You can't just reverse your entry parameters to get out of trades. That's why trading using ma crossovers doesn't work. Entries are usually decent if there's a good trend, but they get you out too late. You have to find some other parameters on which to base your exits (hint: look at volatility).
Forget backtesting and most indicators. Come up with a sound theory, code it, and then let it run in real time for days and weeks and months on a sim account. If successful, then backtest it if you want and optimize. But if you have too many things to optimize, forget it. Then trading becomes voodoo. A solid system should be "simple".
