Suppose it is the first week of April. You forecast that you will sell 15,300 un
ID: 449302 • Letter: S
Question
Suppose it is the first week of April. You forecast that you will sell 15,300 units of your product during the second week of April. When the second week of April comes along, the actual sales were 13,800.
a. Using an exponential smoothing model with coefficient a=0.3, what would your prediction be for demand during the third week of April?
b. Redo the calculation in (a) with a smoothing coefficient of a=0.8. What would your prediction be for demand during the third week of April?
c. Which coefficient (.3 or .8) leads to a forecast that is more responsive to changes in demand?
Explanation / Answer
Forecast as per exponential smoothing:
This can be written as:
St+1=St+t,
where t is the forecast error (actual - forecast) for period t.
In other words, the new forecast is the old one plus an adjustment for the error that occurred in the last forecast.
Forecast for next period = Forecast for previous period + (Actual of previous period - Forecast of previous period) x a
a)
So, Forecast for week 3 = Forecast for week 2 + (Actual of week 2 - Forecast for week 2) x a
= 15,300 + (13,800 - 15,300) x 0.3 = 14,850
b)
So, Forecast for week 3 = Forecast for week 2 + (Actual of week 2 - Forecast for week 2) x a
= 15,300 + (13,800 - 15,300) x 0.8 = 14,100
c)
0.8 is more responsive as it is creating less difference between the actual and forecast.