Awesome Oscillator as a drift gauge: 14.5 bars of lag between the 5 and 34 SMA

[Fairman]

Awesome Oscillator as a drift gauge: 14.5 bars of lag between the 5 and 34 SMA

The Awesome Oscillator is the 5 period simple average of the median price, (high + low) / 2, minus the 34 period simple average of the same series. Most of the time it's read for colour changes and zero crosses. There's a simple way to read its value as a speed.

In a steady ramp where the median price rises by a fixed amount each bar, a simple moving average of n bars lags the price by (n minus 1) / 2 bars. For the 5 period that's 2 bars. For the 34 period it's 16.5 bars. The gap between the two averages is the difference in lag, 14.5 bars, times the rise per bar.

So AO is about 14.5 times the drift per bar. If AO reads 0.00145 on EURUSD, the drift is about 0.0001 per bar, which is 1 pip a bar. On M5 that's 12 pips an hour. A reading of 0.0006 means roughly 6 pips divided by 14.5, or 0.41 pips a bar, about 5 pips an hour.

Divide the reading by 14.5 and then by 0.0001 and you have pips per bar. For yen pairs divide by 0.01.

This only holds for a smooth trend. In choppy conditions the averages cross and the AO swings around zero without any steady drift behind it. Still, it turns a unitless wiggle into something you can compare with a spread: if the estimated drift per bar is 0.2 pips and the spread is 0.8, the oscillator is positive and nothing tradeable is happening.

[Fairman]

To see it for yourself, put the AO on an M5 chart of EURUSD and find a clean 10 bar run where price rose steadily, then read the value at the end. Take the net rise in median price over those bars and divide by the number of bars. Multiply by 14.5 and compare it with the value on the indicator. The two will be close in a smooth run and apart when the run was jerky. That exercise takes about five minutes and it's the quickest way to learn what the histogram height means in pips.

[Fairman]

The conversion also shows why AO lags. Both averages are late, the longer one much more so, and AO crosses zero only when the 5 period average falls below the 34 period one. After a steady rise, price can turn down while the 34 average is still climbing on its own lag, and the 5 average has to drop under it before the zero cross appears. So the cross arrives several bars into the reversal, and the larger the drift was before the turn, the later it comes. For a scalper that delay is part of the cost of using the zero line as a trigger.

[Fairman]

A related use is to set a minimum reading. If you want a drift of at least 0.5 pips a bar before taking a trade in its direction on M5, then you need an AO value of at least 0.5 x 14.5 = 7.25 pips, or 0.000725 on EURUSD. Mark that level as two horizontal lines on the indicator window, at plus and minus 0.0007. Readings between them are 'not enough drift'. It's a rule I can test and fit on past data, and it doesn't depend on the colour of the bars.

[Fairman]

The same reasoning works on other symbols with the unit changed. On GBPJPY the AO reading is in yen, and 0.145 is 1 pip a bar, since a pip is 0.01. Drift per bar is the AO divided by 14.5 in every case, and what you compare it with is the spread in the same unit. So 0.145 on GBPJPY M5 against a spread of 2 pips means one bar of drift covers half a spread, so it takes two bars of that drift just to pay for the spread once.