Forecasting Methods and techniques
By:- Ramesh Murthy, Asst, Professor of Supply Chain, Dayananda Sagar University
Capt. Nagraj Subbarao, Dean & Professor of Leadership & Strategy, Dayananda Sagar University, Alok Chakravarthy, Asst. Professor of Business Analytics, Dayananda Sagar University
What is forecasting
Simply put, forecasting is the process of predicting the future requirement of resources. The resources could be raw materials, finished products, manpower or plant and machinery.
Importance of forecasting Forecasting is the most important activity based on which the organizations including supply
chain prepare their long, medium and short term plan.
Based on forecast:-
a) Supply Chain decides to set up or close manufacturing plants or increase the capacity of the existing plants.
b) Forecasting is used by marketing to plan their Sales Promotion and Price discounting plans
c) Based on forecast, the finance team plans their working capital requirement, fund raising plans through debt or equity etc.
d) Based on forecast, the procurement team enters into contract with the suppliers
e) Recruitment or retrenchment of workforce is planned based on forecasting
Characteristics of forecast
a) Forecast is never accurate. No forecasting technique results in accurate forecast. There will always be an error in the forecast and the role of different forecast techniques is to minimize the error.
b) Long term forecast is less inaccurate than the short term forecast.
c) Forecast at aggregate level is more accurate than forecast at the dis-aggregate level. E.g – Forecast for monthly sales of Lay’s Chips is more accurate than the forecast for the 50gm pack of Salsa flavor.
d) The farther the one is in the supply chain, the greater is the forecast in accuracy. This is explained by the Bull Whip effect which states that deviation in the fore cast gets magnified as it moves from the Retailer to distributor to the manufacturer.
Components of forecast
Forecast has 3 main and one additional component
a) Level – This is the de-seasonalized demand for a product.
b) Trend – This is the uniform increase or decrease observed in a demand over a period of time.
c) Seasonality – This is the periodic increase or decrease in demand which can be predicted.
d) Random component – As the name suggest it is random in nature and cannot be predicted.
The difference between the observed demand and forecast is called forecast error. For e.g. ,if the forecast for period T was 10 and the observed demand was 12, then the forecast error = 12 -10 =2. Similarly, if the forecast for the period T+1 was 20 and the observed demand was 18, then the forecast error is 18 – 20 = (-) 2.
a) Time Series method
b) Causal method using Regression Analysis
Read the full article Article on Forecasting Methods and Techniques