The McGinley Dynamic is a technical indicator that smooths a price series into a single line, in the manner of a moving average, but adjusts its own speed according to how far and how fast price is moving away from it. It was devised by the market technician John R. McGinley and published in the 1990s. Its aim is to follow prices more closely than an average of fixed length, with fewer false crossings.
McGinley presented the tool as a smoothing mechanism and not as a trading signal in its own right. The argument behind it is that any fixed-length average is fitted to one market speed. It trails badly when prices accelerate and sits too close when they drift, and the trader is left to choose a length that will be wrong much of the time.
Formula
The line is calculated recursively, each value from the one before:
- is the value of the indicator for the current period and is its value for the previous period.
- is the current price, normally the close.
- is the number of periods, chosen as for a moving average.
- is a constant, commonly 0.6, which makes the line track roughly as a moving average of periods would.
The first value is usually set equal to the first price or to a simple average. The key term is the ratio of price to the previous value, raised to the fourth power. When price is above the line the ratio exceeds 1, the divisor grows and the line moves up by a smaller fraction of the gap. When price is below the line the ratio is less than 1, the divisor shrinks and the line moves down by a larger fraction.
Worked example
Take and , with a previous indicator value of 100. If price rises to 104, the ratio is 1.04 and its fourth power is 1.170. The divisor is . The line moves by , to 100.57.
If price instead falls to 96, the ratio is 0.96 and its fourth power is 0.849. The divisor is . The line moves by , to 99.22. For the same four-point gap, the line travels further on the way down than on the way up. This asymmetry is deliberate: McGinley held that markets fall faster than they rise and that a tracking line should respond in kind.
Comparison with moving averages
A simple moving average weights every observation in its window equally. An exponential moving average closes a fixed fraction of the gap between price and the average each period. In the McGinley Dynamic that fraction is variable, since it depends on the ratio of price to the line.
It belongs with other adaptive smoothers, which try to reduce the lag that makes averages lagging indicators. Kaufman's adaptive moving average changes its speed according to how efficiently price is trending, and the Hull moving average uses weighted averages of different lengths to cancel lag. The McGinley approach is simpler than either and needs only one input besides price.
Use in practice
Chartists apply the line in the ways they apply any average in technical analysis:
- Trend filter. Price above a rising line is taken as an uptrend and price below a falling line as a downtrend.
- Dynamic support and resistance. In a steady trend, pullbacks often stop near the line, which is then treated as a moving support level or resistance level.
- Crossovers. Some traders pair two lines of different length, or the line and price, in the manner of a moving average crossover.
- Baseline. The line can replace the average at the centre of bands and oscillators.
The indicator is included in most charting packages. It is used across shares, futures and currencies and on any time frame.
Limitations
The McGinley Dynamic is still a smoothing of past prices and cannot anticipate a turn. In a sideways market it will be crossed repeatedly, as any average is, and the closer tracking can make these crossings more frequent, not fewer. The choice of remains a matter of judgement, and the constant and the fourth power were selected by the author on practical grounds.
Because the line depends on the ratio of price to its own previous value, it is sensitive to its starting point for the first stretch of data, and it is not defined for series that can be zero or negative, such as some spreads. Published evidence that it improves trading results over ordinary averages is thin. Claims for any smoothing method are best checked by backtesting with realistic costs, and such tests describe only the period examined.