National base forecast

The Commission has chosen to use an econometric model for its first national forecast. The econometric model takes historical data on drivers for electricity demand, and then projects forward using forecasts for those same drivers provided by other agencies. The drivers used are population, GDP, number of households, temperature, electricity price, and a correction for shortage years.

The model was developed in Matlab. The Matlab scripts, historical data, forecast drivers, and tables of forecast values are provided.

• National forecast files

The Matlab forecast model calculates error bounds on the forecast using a technique called Monte-Carlo simulation. This yields confidence limits on the forecast. There is a 90% confidence that future demand for electricity in New Zealand will fall between the bounds shown on the following graph.

• Forecast with confidence limits
• Forecast components
• Forecast values table

A sensitivity analysis [PDF, 69KB] of the forecast indicates that the future immigration rate and GDP growth are the main uncertainties affecting the forecast.

The national forecast was reviewed by NZIER [PDF, 147 KB], which resulted in additional analysis [PDF, 29 KB].