Dear All, now I am testing a new strategy. FYI I don't want to quit my job and make a living with trading, I simply felt in love with finance and want to see if of course I can make sone extra money for saving or just spending around. I started learning using finance for metal hedging and then I decided to study technical finance by myself. After trying several indicators, I went back to the base and set up this Trading System: - Fastest SMA 3x3 (A) - Slowest SMA 25x5 (B) - ENTER when the A>B AND the last close > previous high (and vice versa for SHORT) - EXIT when close < A I attach a picture (the chart is old, just for an example) Please all the critiques will be highly appreciated.
No edge here. Learn some basic programming skills, code the strategy and you will see the negative expectancy from a mile away. Make sure you have hundreds of iteractions, thousands better, before you can consider the report trustworthy. You got some of the brightest minds in the business trying to squeeze profits from the markets, you can't be that naive to think a simple moving average crossover system will make you profitable. If you get the right price action for it, it will look good, just like you did, but get the incorrect one, and it chops your head off and the losses wont make up for the winners. Sorry but it's not that easy, you gonna have to work much harder, much much harder.
The human eye/brain is drawn to things that look like what you're searching for, in this case it's winning trades. The result is something called "confirmation bias". It means you're getting excited about a chart that shows winning trades and not looking at charts with losers. For trend following systems like your example you're better off looking at long-only equity trades on the daily chart. A further warning about trend following systems: the success rate is < 50%, usually 30-40%. Statistically you have to expect long strings of losses with a hit rate like that.
As @Autodidact said, I would learn a little bit of programming and test my strategy using something like quantconnect. Simple doesn't mean that it won't work. In my experience, (smart + simple) usually gives me the best results. Though finding what to put inside 'smart' and inside 'simple' is not easy.
You can be profitable trading an MA cross strategy. However, you have to incorporate some discretion (filtering the entries and exits), and not some rigid rules like what you have shown (especially the exits).
Discretion sounds like BS when dealing with a mechanical strategy. It gives way to "I had to do (x), now I know, at least I'm learning" where (x) is always something new. If concrete rules don't work then don't bother with it.
Ya ya, let's retrace back to the newbie days, the MA crossover has negative expectancy, so let's add "filters" and more indicators, and bullshit, and divergence, and volume, and bla bla to make it profitable. It's like cooking pasta, trying it, taste bleh! lets add shit to it, still bleh! lets add cheese, bla bla, before you know it, you cooked a big pile of shit, congratulations noob, you suck at cooking and trading!
Nodoji and JackHershey are said to be the best of the best on this forum. See if you can find their old posts find a better method
Indeed, if you need to add filters/indicators etc, that's already curve-fitting. Never knew of anyone successfully trading anything focused on MAs. It's funny because I've done the exact same thing as the OP years ago, you can always find nice charts that bind to MAs beautifully but they're worthless in forecasting anything. Treat it as what it is - an average of past prices, then you might find some kind of use for them.