I've read many traders recommend the 200 day MA. Some will recommend the EMA, others SMA. What in your experience is the most reliable?
The true answer is that it doesn't matter. MAs, as traditionally used, do not work. The conventional wisdom is that a 200 day SMA has some value as a self fulfilling prophecy. Do not believe the conventional wisdom. If you insist on going down the rabbit hole of moving averages, at least learn to view them as lowpass filters, and learn how to engineer them. John Ehlers is the man to study for filters in finance.
Yeah I'm familiar with low pass filters and signal processing in general. Also I agree there is almost definitely zero edge in just trading off the MA line itself. I just want to see what happens when price reaches there, in terms of tape, PA and volume, especially if there is confluence with support/resistance or round-numbers and we're already stretched in terms of daily range. If it looks like it's being respected and buyers are stepping up at the level then I might give it a deeper look.
Was about to mention this myself. I'm a ham radio operator and know these most well from the electrical engineering side of radio. A moving average is simply a filter on a signal. You should choose a filter that maximizes the signal to noise ratio without degrading the original signal too much. OP would be wise to pick up a book on digital signal processing if he wants to learn more. They are not good for low information signals like a stock ticker. It is generally applied to signals to decrease the amplitude of noise and increase the step response, however generally the higher you refine the moving average the less of the original signal you can recover. If you want to see this in action create a square wave and add random noise to it. Apply a 5, 10, 50, and 100 step moving average and watch how the original signal gets warped in the output. Noise is reduced by a factor of the square root of the sample size for the MA.
I've studied enough about DSP and time-series analysis. I've been working with things like wavelets, Fourier transforms, recursive MAs, and the usual toolbox for 10 years in my publications and general research. Even if I didn't have this knowledge, I don't believe it's very useful for trading (whether it's algorithmic trading or otherwise). I want to know how to practically apply the extremely basic stuff that traders talk about all the time (such as 200 MA) to make money, and want to leave aside the overcomplicated hocus pocus (e.g. online wavelets) that works in one domain (DSP) but shouldn't be applied to trading. Is there any edge whatsoever to be had with 200MA or not? If so, how to extract said edge? If enough people that have looked into it say no, I will trust that judgment and move on.
Excellent points. Most used is 200SMA, eg S&P 3000. Main reason to follow it is solely to be aware of what other traders are looking at. Remember as with all squiggly lines they're not useful in isolation, price action, volume and ranges/volatility generate actionable signals. The better you get, the more you realize it's all about price action and trade management, charts and squiggly lines are less important
decades ago, I started with hundreds of indicators and various MAs. so many till I can hardly see the candlesticks. Now I have zero except candlesticks.
I strongly recommend you use the simple moving average, as the exponential moving average - while more responsive to price changes - is a "repaint" indicator. In other words it can give you a buy/sell signal at time T1 and then 10 bars later (for example), at time T2, you will notice that this signal is no longer valid.
I hope I am not making up the following, but vaguely remember reading -the 200 day has something to do with a very common financial performance analysis period - that is a year. Ema approximates it better than simple - Paul Tudor Jones used them extensively. - given that so many old school traders use it, there is a self fulfilling prophecy angle to it and comes up with decent results in backtests (obviously not enough to have any edge)