Opinion is divided about the usefulness of MAs. Fact is, sometimes they work and sometimes they don't. They can be useful to traders who understand when they do and when they don't. (Continue on next entry as editing blocks my view...)
50MA... up until late Feb market was simply below the downsloping line and was meaning only "negative tone" to the market. 1. Late Feb, tested @ 91.52 and acted as support. 2. Through Mar it sort-of acted as support and resistance in a sloppy fashion. 3. ~Apr 10, support... and then 5 times in Aug. 4. In Sep, SP gapped below. As no longer "support", expect to be "resistance" if tested from underside. 5. ~Oct 13, acted as resistance ~ 134.48. Notice also... 200MA was resistance at 114 in Feb, and support at 118 just this last week. Bottom Line... (1) MAs "often act as S/R". When they hold as such they are good places to act. (2) Often the market moves back and forth across the MA "like it's not even there"... recognize that and don't try to trade it... but keep an eye on it... as it WILL again act as S/R. FWIW...
My first futures trading "system" was a crudely thought of MA crossover trigger. I have learned since that indicators on their own are completely worthless...being derivatives of price (or volume) they put you in the trade too late and much of the meat has been left on the bone...you get in right before consolidation or area of contention and get wrecked in the chop that ensues. What backtests like a dream simply fails when live. However, when used as conditional support I really believe indicators can help. I am mainly focusing on volatility indicators now and am finding them extremely useful...but always secondary to PA. Using MA for nothing more than a directional bias isn't a bad idea. Quick glance then verify with other directional conditions like trend(s), s/r, etc... Just brainstorming but with moving averages I would watch the gap more than the crossover but idk...worthy of some research probably. IMO (for intra day trading at least), the goal is to be positioned BEFORE THE MOVE...if you are wrong, get out or change sides, if you are right then you can use indicators to help keep you in trade and alert when to start looking for an exit via PA.
There is a reason moving averages work sometimes and other times not. MAs are in reality lowpass digital filters. The two basic types are finite impulse response (FIR) and infinite impulse response (IIR). FIR filters are your typical MA that everyone knows, and IIR filter are what people call exponential moving averages (EMA). Price can be modeled as a complex wave containing many waves of different frequencies. FIR and IIR have different frequency responses as waves appear and disappear in price movement. But, what they have in common is that waves of higher frequency are attenuated in "energy" and lower frequency waves pass through the filter. Also, both types of filter introduce what is called group delay, or more commonly called lag. So, lag and frequency response are the reasons MAs have a checkered track record. John Ehlers is person whose work you want to read about. Technical Papers (mesasoftware.com)
I do similar stuff like this guy (have a math degree). You can use a Recursive Median Filter (EMA of price medians). An adaptive moving average uses the efficiency ratio to reduce lag (Kaufman's). The standard for detrending is a Linear Regression moving average. You can also use the sum of price minus an average and the difference of averages, prices, vols, etc. You can use indexation techniques (aggregated normalized data). People should remember, that the definition of volatility is the term variance of returns. So, if you index returns and compare them (in real-time), you will have insight into the vol market.
Linear regression gives the appearance of "working" when it aligns with the current wave...otherwise you will be left scratching your head. The market does not move randomly...it has no free will. You have to align yourself with its determinism.
. Ahem... some individuals have found value in utilizing the slopes of medium term moving averages like EMAs and WMAs to try to trade with the trend . My preference is to only accept long entry signals when the slope is up and opposite for shorts. Combining them with a good momentum indicator can point out low points in price where one can look for long entries as well as peaks in price where one could look to place sell orders. Other factors should be taken into consideration too like candle structure, position of the close, risk:reward ratio, etc. The chart below is the same one shown in post #1. It shows a 40 period EMA plus a momentum indicator called the Ehlers Stochastic, which is similar to the fast line of the regular Stochastic indicator but better because it can still point out extremes when prices are moving in a long linear trend. Conventional Stochastics and RSI and some other momentum indicators don't show extreme values very well in strong trends, though these extremes are often the preferred places to initiate buy and sell trades.
I use a cyclical indicator "Whiter is Brighter". I have yet to find an MA I am happy (and profitable) with. For those who do offline analysis, I think the ultimate MA is an IIR in forward-backward mode, ala Gustafsson, a true zero group delay filter.