Exponential Trends in Statistical Forecasting

This model assumes that the underlying trend in demand for the item has been rising (or falling) by an approximately constant percentage per month during the period of history selected by the user. Graphically this would produce a curved trend that can be rising or falling.

• IFP will obtain a trend curve of overall best fit to the adjusted historical data
• This trend curve is then projected into the future
• Note that for this model all selected historical periods are given equal weight
• Forecast values are then obtained by multiplying by the average seasonal factor for each month

Rising Exponential Trend

Rising exponential trends are suitable for products in the rapid growth phase of their life cycle.

Falling Exponential Trend

Falling exponential trends are suitable for products in the decline phase of their life cycle.