Waste of time. You're far better off with a single moving average along with some extra logic to deal with frequent crossings when price is wobbling around the MA. For most purposes that's also a waste of time though. (Since I wasted some time in my days: ) I've run scanning algorithms to exhaustively test (a shitload) of MA crossover combinations. No dice (as in working out of sample). Interestingly enough, I've also once read some mathematical argument (over my head) that an MA(M) with MA(N) crossover can be replicated by a MA(M) with MA(1) = P crossover, i.e. one of the two parameters is entirely redundant. And, a more simplistic argument is that the second MA merely introduces more lag.
Two different moving averages work well, HOWEVER, not as a stand alone strategy as 100% of traders try. They can be good as a way to define trend. And one should study a few months, 200-400 hours to understand what is normal slopes, abnormal slopes, counter to each other, narrowing and expanding of averages in normal conditions and volatile conditions. You should have least 500 sample size of each to form stats just on trend, what is price action doing on above "occurrences". Sounds like much work and you can not do by programming cause you will not learn much. Remember, entry signals means the least in most systems. But you need to an idea of risk management by how you will entry, breakouts often require large stops and waiting for price to retrace has to be well defined so that MAE and MFE can be applied to define risk and possible profit. One can just use price and develop trend, but I have often found way too many rules that contradict each other because of no volatility or too much volatility, when HFT's are present they present falseness of over all market, so unless you very good at price action, easier to learn the ins and outs of what 2 moving averages do in relationship to price. Most likely not what you wanted to hear, but if you are going to compete in the toughest skilled business on the planet, best be ready. Good luck.
Since the invention of computers, indicators started to flourish. I have deleted all indicators years ago as it is worthless useless. learn to read chart with just candlesticks.
I can't stand candlesticks, started doing own charts way before home computers were sold most likely reason. Candles have too much width and colors, too noisy for me. I believe Candles are an indicator? I think some indicators are very useful, easier, but hardly anyone looks deep down on what is normal and abnormal of relationship between price and the indicator itself, most people don't know the formula's of the indicator and the why's of it. Hey, if all people stopped using them, be much tougher to collect ?
Not using the crossover, just using the 30,60 ema mov avg to show trend direction of the longer timeframe. Could use a 4,9ema crossover to trade in the direction of the trend shown by the 30,60ema.
There's a guy named Steve Burns on Twitter who periodically shows results of backtests suggesting that MA crossovers can produce good results. He's talked a lot about 10/50 EMA crossovers and occasionally about others during the past year. Like a lot of traders posting on Twitter, he sells a book and/or a trading course and I don't want to promote them because I haven't seen them. His main Twitter site, however, has a lot of interesting content and I check it out from time to time. Main Twitter site: https://twitter.com/SJosephBurns Twitter page with a link to one recent set of backtests:
right. some people don't like candlesticks. way before computers were invented, we had to use graph paper, and plot the price by marking x on the chart.