Simplest Strategy Ever. Buy If first hour candle's green Else sell. Close @ End Of Day. Doesn't always work but still. Thought I discovered a more complex strategy Then I tested this and I actually underperform ... Ahahahaha That's SPY less interesting But it makes money long and short
That last strategy could work if you put on a longer term filter, to keep it from going long during one of the (rare) drawdowns in the index. If close > (ma, 200) then ....
I found that certain styles of moving average buy and sell signals do, in fact work in bear markets, but the parameters need to be set independantly from the others. I never tried using volume profile to code an algo, but it would give a way to possibly buy in one of Scatty's support bands, since the volume puts in a steep hill on supports / resistance zones.
Now if you could just get a read on the greed and fear on losing account holders, it would be an absolute milk machine!
Daily bars, and if you are working on a shorter term chart just use a conversion factor. My old trend following algos always went flat / losing when trying to go long in a bigger down trend. This is one of the oldest and most fundamental rules of directional trading. Swim with the current, but you already know that.
Certain people think a rabbits foot will work. Moving averages might work for you if you spend the rest of your life curve fitting the lookback periods. Moving averages get you in and out late. Why am I saying this? Because I'm stupid attempting to educate which works against me.
I’ve backtested multiple strategies lately and adding SMA as filter never helped improve the bottom line. I posted (then removed) the results of the basic strategy with a 200 daily SMA as filter and it did no good. Even worsened (?) it actually. Backtesting is an eye opener. Lots of shit can quickly be debunked. Now I’ve tried QuantConnect and I have not yet been able to run a backtest successfully but the hassle is worth the time. Those who say it’s not worth it are flat earther. Backtesting is a far better point of departure than most alternatives. I think it even beats the “10,000 hrs screen time” but it won’t replace it that’s for sure. However 10,000hrs can be drastically more productive if backtesting is done on the side. I am thankful for the low barrier to entry (GPT + TV or whatever) that’s been achieved for the past few years.
Moving averages may have worked decades ago. The markets are smarter now. Moving averages have no logic. As a rough rule of thumb a 200 SMA could be used as a trend indicator, only buy stocks above a 200MA, but its still only a very crude indicator. You'll see plenty of opportunities missed when stocks trade below a 200MA.
i think it improved risk adjusted returns such as mean daily return / max drawdown when i last investigated. like buying a stock after a good earnings vs speculating on the earnings even tho a large % of gains occur on earnings. not really something ive tried to trade on tho.