← All methods

Theta · Pro

Theta

SES plus a drift line — the M3-famous combination (Pro).

What it assumes

SES combined with a drift (Theta-line) — M3-famous.

When to use

Macro / mildly trended series.

When to avoid

Strong seasonal retail without a seasonal wrapper.

Knobs

ŷ = SES + 0.5·h·drift. Pro.

How it works

The classic Theta method combines a smoothed level (SES) with a drift (“theta line”). It did remarkably well on the M3 competition for many mildly trended macro series. Strong retail seasonality without a seasonal wrapper is a known weakness. In WhichForecast it is a Pro family; Free users still see the DNA row, soft-locked.

Sample forecast question

A quarterly indicator drifts up with little calendar shape. What is the Theta forecast for the next four quarters?

Step-by-step on these numbers

Sample series (index)
PeriodActualFitted
Y1 Q11414
Y1 Q21514
Y1 Q31614.95
Y1 Q41615.95
Y2 Q11816.00
Y2 Q21917.90
Y2 Q32018.94
Y2 Q42119.95
Y3 Q12220.95
Y3 Q22321.95
Forecast
PeriodForecast95% interval
Y3 Q323.4522.34 – 24.56
Y3 Q423.9522.38 – 25.52
Y4 Q124.4522.52 – 26.37
Y4 Q224.9522.72 – 27.17

Parameters the engine found

  • alpha = 0.950
  • last = 22.947
  • drift = 1
  1. The question

    A quarterly indicator drifts up with little calendar shape. What is the Theta forecast for the next four quarters? Sample quarters: Y1 Q1=14, Y1 Q2=15, Y1 Q3=16, Y1 Q4=16, Y2 Q1=18, Y2 Q2=19, Y2 Q3=20, Y2 Q4=21, Y3 Q1=22, Y3 Q2=23 (index).

  2. SES + drift

    Classic Theta (M3) combines a SES level with a drift line. SES α = 0.950 gives last level 22.95. Drift = (23 − 14)/9 = 1.

  3. Forecast rule

    ŷ_{n+h} = SES_last + 0.5 · h · drift. The ½ is the usual Theta-line weight in this engine. Y3 Q3 = 22.95 + 0.5×1×1 = 23.45; Y3 Q4 = 22.95 + 0.5×2×1 = 23.95; Y4 Q1 = 22.95 + 0.5×3×1 = 24.45; Y4 Q2 = 22.95 + 0.5×4×1 = 24.95.

  4. Interval

    Residual σ = 0.567. For h = 1 the 95% band is [22.34, 24.56] around 23.45. The engine labels this a residual-Gaussian heuristic (σ√h), not a simulation interval.

Graph of this sample

Y1 Q1 · 14 Y3 Q2 Y3 Q3
Theta on the canned sample: actuals, fitted, and a 4-step forecast. Actual Fitted Forecast
The chart plots the canned table on this page (actuals, in-sample fitted, forecast, 95% interval). Not your Excel series. Free analysis never uploads raw data.

Educational only. Not investment, weather, or operational advice. In the add-in, rank is rolling-origin MASE — “best supported among candidates on this series.”

Use it in the Excel Add-in