Pivot Mean Oscillator
- Indicators
-
Marco Calabrese
As Ph.D. in Computer Science I have been dealing with numbers, timeseries and Artificial Intelligence since a long time.
My early interests were in pattern analysis of signals from gas sensors. Subsequently, I switched to Iot data streams, especially for Industry 4.0 applications. - Version: 1.0
Pivot Mean Oscillator (PMO) is a unbounded zero-centered oscillator that provides a quick and normalized measure of divergence between two spreads. Detailed description of the mathematical formulation along with some comments and experiments can be found in this article.
Trading aspects
This version of PMO is tuned on spreads relative to Close and Open price signals against their cumulative moving average (CMA). Positive PMO(m,n) values means that last (normalized) m Close prices are higher than last (normalized) n Open ones, thus detecting long positions while the opposite accounts for short positions. Upward and downward PMO reversals can be considered similar to oversold and overbought conditions found by RSI.
A very basic and simplified set of IF THEN rules for trading with PMO is the following:
- IF upward reversal happens below zero centerline THEN early warning BUY
- IF downward reversal happens above zero centerline THEN early warning SELL
- IF upward crossing zero centerline THEN recommended BUY
- IF downward crossing zero centerline THEN recommended SELL
Furthermore:
- as a rule of thumb, when a spike occur in PMO values it is highly probable that market will react in the opposite direction.
- in order to reduce lag, it is useful to minimize the number of last Close prices to consider (e.g. PMO(1,8)). As a drawback, this reduces smoothness of the PMO signal.
Key features
- applicable in all timeframes
- applicable to all symbols
- useful both for scalping and intraday strategies
- useful for finding spikes, local maxima and minima
- useful for comparing any couple of FX signals (due to normalization, see article for details)
- easy-to-use with minimum setting
- free
Input Parameters
- startingTime: datetime of the first bar to consider for calculation
- MA_close: number of last close prices to average
- MA_open: number of last open prices to average
- data_logging: flag to enable data recording from startingTime to now for data analysis purposes
- filename: name of the file used to archive data
Good