I'm trying to read up on these, but most don't seem to be all that great off hand. Anyone have any thoughts?
According to Brian Millard,
"moving averages are…extremely powerful tools, but unfortunately the majority of investors have no idea of how to harness this power."
I would not dare to make such a generalized statement, but I have in fact found that how to best put moving averages to use for me personally is unlike most of the available advice that’s out there.
Some traders, such as Nick McDonald of Trade with Precision, recommend employing the same set of moving averages, irrespective of the time frame in which one is trading. For example, Investopedia suggests that the 5-, 8- and 13-bar simple moving averages offer perfect inputs for day traders seeking quick profits on the long and short sides.
Norm Fosback, the former head of the Institute for Econometric Research, states that
"there are no magic numbers in trend following... It should be a basic requirement of any moving average trend following system that practically all moving average lengths predict successfully to a greater or less degree."
Indeed, I have often read that the moving average one chooses is not as important as getting familiar with the way in which price interacts with it (or something to that effect).
All of this notwithstanding however, I have arrived at a somewhat rare view of moving averages most closely resembling the thoughts of those who promote the use of a “baseline.”
To paraphrase Global Prime, a baseline is
a mechanical way of entering the market and keeping you disciplined in terms of engaging under the right conditions.
But my personal definition starts with cycle theory, which holds that cyclical forces, both long and short, drive price movements, and can be used to anticipate turning points. (According to Brian Millard, Jim Hurst’s work on cycle theory was based in part on a belief that some 23% of price motion is based on cyclic movements which are additive in nature and can be seen clearly if envelopes are constructed around the price movement.)
It also incorporates Edgar Peters’ fractal market hypothesis, which views financial markets as fractal in the sense that they follow a cyclical and replicable pattern. (Fractals might be defined as "fragmented geometric shapes that can be broken down into parts which replicate the shape of the whole.")
So to generate a baseline, I conduct an analysis to uncover these cyclical waves which are formed in the wake of price action, then define their general frequency and magnitude, and finally plot a centered moving average that comes as close as possible to approximating the zero amplitude of the corresponding waves/cycles.
Consequently, the claim that there is no "best" moving average is not a notion by which I operate—opting instead to use baselines which I calculate in the manner just described.