Triple Exponential Moving Average (TEMA)
Type
Trend follower
Short introduction
TEMA, like DEMA and DEMA2, was introduced by Patrick G. Mulloy as a further modification of the exponential moving average.
The DEMA, TEMA and DEMA2 variants differ in the trend model used. While DEMA is based on a linear regression model (i.e. trend lines), TEMA is based on a quadratic regression model, which means that the trend term can still include curved components that depend quadratically on time.
Formula/calculation
Like DEMA, TEMA is a linear combination of single, double and triple exponential averages. The double and triple averages are used to estimate the trend-related correction.
EMA1 =EMAn(S)
EMA2 =EMAn(EMAn(S))
EMA3 =EMAn(EMAn(EMAn(S)))
DEMA = EMA1 + (EMA1 - EMA2) = 2 * EMA1 - EMA2
TEMA = EMA1 + DEMA(S-EMA1) = 3 * EMA1 - 3 * EMA2 + EMA3
where:
S = time series to be smoothed
n = Period number for DEMA and TEMA
Statement/Interpretation
For information and interpretation of the TEMA, please refer to the literature recommended in the chapter on Double Exponential Moving Average (DEMA) .
Default setting
- none; depending on trend periods
Basic trading systems
- MACD-TEMA