Extrapolative methods in forecasting for Operations

Extrapolative methods seek to identify patterns in past data. Most of these patterns depend on four components of demand horizontal, trend, seasonal, and cyclical. The appropriateness of an extrapolative method will depend on which components of demand are operating in a given situation.

Components of Demand:

The horizontal component of demand exists when the demand fluctuates about an average demand. The average demand remains constant and does not consistently increase and decrease. The sales of a product in the mature stage of the product life cycle may show a horizontal demand pattern.

The trend component of demand refers to a sustained increase or decrease in demand from one period to the next. For example, if the average monthly demand for a product has increased 10 to 15% in each of the few years, then an upward trend in demand exists. The sales of products in the growth stage of the product life cycle tend to show an upward trend, whereas those in decline tend to show a downward trend.

The seasonal components of demand pertains to the influence of seasonal factors that impact demand positively or negatively. For example, the sales of snow blowers will be higher in winter months and lower in summer every year indicating a seasonal component in the demand for snow blowers.

The cyclical component of demand is similar to the seasonal component except that seasonality occurs at regular intervals and is of constant length, whereas the cyclic component varies in both time and duration of occurrence. For example, the impact of a recession on the demand for a product will be reflected by the cyclic component. Recessions occur at irregular intervals and the length of time a recession lasts varies. This component is present in most economic data, such as GNP, personal income, and industry sales of such consumer durables as automobiles and major appliances.

Moving Average Method:

The simplest extrapolative method is the moving average method. In this method, two simple steps are needed to make a forecast for the next period from past data.

Step 1: Select the number of periods for which moving averages will be computed. This number, N, is called an order of moving average.
Step 2: Take the average demand for the most recent N period. This average demand then becomes the forecast for the next period.

To illustrate this method, consider the demand data for a product for which, in the months of February, March, April, May, and June, the demand was 90, 80, 120, 100, and 80 units, respectively. Interest is in making a forecast for the month of July.

Step 1: Choose N =4. Other values can also be chosen. Larger N values will have a greater smoothing effect on random fluctuations in demand. Smaller N values will emphasize the more recent demand history. Notice that N = 1 will result in the present period’s demand being the forecast for the next period.

Step 2: Find the average demand for the most recent 4 periods, N=4

Moving average= Demand for March, April, May, June / 4

= 80 + 120+ 100+ 80 / 4

= 95 units

The forecast for July is therefore 95 units.

Once N is selected, the new moving average for each future period is computed by taking the average demand for the most recent N periods. The disadvantage of this method is that it requires the storage of demand data for N periods for each item. In a production where forecasts for large number of items are to be made, these storage requirements could be significant. Further, this method will not provide good forecasts if the demand data reflect trend or seasonal components. For example, if there is an upward trend in the data, then a forecast made using the moving average method will underestimate the actual demand.

The moving average method gives equal weight to the demand in each of the most recent N periods. We can modify the method, however, by assigning a different weight to each previous period. Exponential smoothing methods, which are discussed next, are convenient for accomplishing the differential weighting of demand in previous periods. Further, these methods can incorporate trend and seasonality components of demand in forecasting.