Forecasting Methods and Techniques for Businesses
Forecasting
Forecasting is predicting or estimating future events.
Forecasts are made on many different variables. Operations and supply chain managers need forecasts to plan output levels, purchases of services and materials, workforce and output schedules, inventories, and long-term capacities.
Demand Patterns
The repeated observations of demand for a service or product in their order of occurrence form a pattern called a time series.
There are five basic patterns of most demand time series:
- Horizontal
- Trend
- Seasonal
- Cyclical
- Random
Key Decisions on Making Forecasts
- Deciding what to forecast: Level of aggregation, Units of measurement
- Choosing the type of forecasting technique: Judgment methods, Quantitative methods, Causal methods (linear regression)
Time-Series Methods
- Naive forecast
- Simple moving averages
- Weighted moving averages
- Exponential smoothing
- Multiplicative seasonal method
- Trend projection with regression
Forecast Error
For any forecasting technique, it is important to measure the accuracy of its forecasts.
Forecast error for a given period is simply the difference found by subtracting the forecast from actual demand:
Et = Dt – Ft
Forecast performance is determined by forecast error.
Measures of Forecast Error
- Cumulative sum of forecast errors (Bias error): CFE = ΣEt
- Average forecast error: Ē = CFE/n
- Mean squared error: MSE = ΣEt2 / n
- Standard deviation of errors
- Mean absolute deviation: MAD
- Mean absolute percent error
Linear Regression
A dependent variable is related to one or more independent variables by a linear equation.
The independent variables are assumed to *cause* the results observed in the past.
The simplest linear regression model is: Y = a + bX
where:
- Y = dependent variable
- X = independent variable
- a = Y-intercept of the line
- b = slope of the line
Time Series Methods
Naïve Forecast
The forecast for the next period equals the demand for the current period: Ft+1 = Dt
Simple Moving Averages
Weighted Moving Averages
Exponential Smoothing
The forecast for period t + 1 can be calculated at the end of period t (after the actual demand for period t is known) as:
Ft+1 = Sum of last n demands / n = (Dt + Dt-1 + Dt-2 + … + Dt-n+1) / n
where:
- Dt = actual demand in period t
- n = total number of periods in the average
- Ft+1 = forecast for period t+1
Weighted Moving Averages
In the weighted moving average method, each historical demand in the average can have its own weight, provided that the sum of the weights equals 1. The average is obtained by multiplying the weight of each period by the actual demand for that period, and then adding the products together.
Example: A three-period weighted moving average model with the most recent period weight of 0.50, the second most recent weight of 0.30, and the third most recent period weight of 0.20 is:
Ft+1 = 0.50Dt + 0.30Dt-1 + 0.20Dt-2
Exponential Smoothing
Requires only three items of data:
- Current period’s forecast (Ft)
- Current period’s actual demand (Dt)
- A smoothing constant, alpha (α), where 0 ≤ α ≤ 1
Next period’s forecast = α(Current period’s actual demand) + (1 – α)(Current period’s forecast)
Ft+1 = αDt + (1 – α)Ft
Trend Patterns: Using Regression
A trend in a time series is a systematic increase or decrease in the average of the series over time.
Trend projection with regression is a forecasting model that accounts for the trend with simple regression analysis.
Principles of the Forecasting Process
- Forecasts are rarely perfect.
- Short-term forecasts are usually more accurate than long-term forecasts.
- Forecasts of aggregated demand are usually more accurate than forecasts of demand at detailed levels.
- Forecasts developed using multiple information sources are usually more accurate than forecasts developed from a single source.
- Better forecasts result in better customer service and lower costs, as well as better relationships with suppliers and customers.