Please enable JavaScript.
Coggle requires JavaScript to display documents.
Forecasting (Characteristics of forecasts (all strategic and planning…
Forecasting
Characteristics of forecasts
all strategic and planning decisions in a supply chain
All of these decisions are interrelated
always wrong
should include a measure of error
Long-term forecasts are less accurate than short-term forecasts
Aggregate forecasts are more accurate than disaggregate forecasts
Forecasts should not be used to the exclusion of known information
Improving forecast accuracy
early warning mechanisms
the law of large numbers
as volume increases, relative variability decreases
Increase the level of aggregation
Gather the sales history at detailed level
Forecast at different hierarchy levels
Beware of product proliferation
Reduce information delays and replenishment lead times
Reduce demand volatility
Reduce the “hockey stick effect”
Continuous performance monitoring
Critically assess value added from temporary price reductions and promotions
Forecasting basics
Basic approach to forecasting
Forecasting methods
Qualitative: primarily subjective forecasting
Quantitative: objective forecasting
Simulation
Evaluation of forecast methods
Residu = forecast – actual
Forecast error measures
Demand variability vs. forecast (in)accuracy
Time series forecasting methods
Moving averages and exponential smoothing
Used when demand has no observable trend or seasonality
Current forecast for all future periods is always based on the current estimate of the level
Dealing with a trend
Holt’s model: double exponential smoothing
Dealing with seasonality
Deseasonalizing demand
Estimating level and trend
Estimating seasonal factors
Box-Jenkins (ARIMA) models
Conclusions