The Master Swing Trader:Moving Averages

Farley says that MAs are classic momentum tools that define both trend and natural pullback levels. He indicates that moving averages as an important Support/Resistance factor. And, as we know from a previous post, it takes more than a few S/R factors to add up to actually give a stock support (or resistance) After looking into MAs, I for sure concur that even mentioning a MA in a separate support context can be meaningless. Farley goes on to mention the 20 day MA is short-term, 50-day MA is intermediate term, and of course 200 day is long-term.

Farley mentions that you may want to adjust the MA periods for different time-charts. For example, he suggests a 5, 8, and 13 period MA for 1 through 60 minute charts. I'm going to demo a few MA concepts in the charts below, so I hope you get the idea of how MAs really work.

Farley also talks about MACD, or Moving Average Convergence/Divergence. This is simply how far apart two Moving Average indicators are, and whether the shorter timeframe MA is higher in price than the longer MA (which is a bullish-upward-moving chart). Here is a chart of GDX which shows a 20-day MA in blue, and a 50 day MA in black. It shows how the 50-day MA acts as a support indicator or “natural pullback” point, as Farley would say. Also indicated are the crossovers which are also 0 points on the MACD chart beneath. The blue line is the convergence/divergence factor on the MACD pane.. You should also notice that the MACD factors (20, 50) coincide with the moving averages I’ve selected for the price pane. This is important so the MACD will indicate the convergence/divergence of the Moving Averages you’ve chosen for the price chart.



For those who worry about the small errors in the crossover points, the MACD is calculated based on the close of the day, and the MAs are calculated based on the previous closes. So you would expect to see the MA crossover to happen before the MACD crossover in a downtrend, and expect the opposite in an uptrend. Anyway…

So I studied MACD, and moving averages. And they have a basic flaw. You cannot use one set of MA time periods. Here’s why. This next chart is a daily chart of CLR, and as you can see, the same 20,50 day moving averages show complete trash. Looking at it, I don’t see one support or resistance point where the MA was actually effective in pointing out one of those support/resistance points.



However, here is the same time-period chart, with the MA’s adjusted for 8 day (blue) and 13 day (black). I’ll leave it up to the reader to identify the numerous support and resistance points identified by the moving averages. Even the crossovers make sense now.



Why is this so? Our moving average theory obviously falls apart, unless we adjust the periods. The basic problem is volatility. The volatility of a stock, that is how much it rises or falls in a given timeframe, adjusts the meaningfulness of our moving averages. For a high volatility stock, a shorter MA period is required. For a lower volatility stock, a longer MA period is required.

After studying MA’s, I believe they are very difficult to manage. I have not yet come across a formula for setting the MA’s time periods because I haven’t come across a per-stock volatility index. And besides, to adjust the MA’s everytime I change to a different ticker is a time-consuming task. I kind of think MA’s are semi-worthless until you take the time to play around with the periods to see what fits that stock’s given volatility for the elapsed time you’re looking at.