Abstract
Divisia money is a monetary aggregate that gives each component asset an assigned weight. We use an evolutionary neural network to calculate new Divisia weights for each component utilising the Bank of England monetary data for the U.K. We propose a new monetary aggregate using our newly derived weights to carry out quantitative inflation prediction. The results show that this new monetary aggregate has better inflation forecasting performance than the traditionally constructed Bank of England Divisa money. This result is important for monetary policymakers, as improved construction of monetary aggregates will yield tighter relationships between key macroeconomic variables and ultimately, greater macroeconomic control. Research is ongoing to establish the extent of the increased information content and parameter stability of this new monetary aggregate.
Original language | English |
---|---|
Place of Publication | Birmingham |
Publisher | Aston University |
ISBN (Print) | 9781854497390 |
Publication status | Published - Dec 2008 |
Bibliographical note
RP0818Keywords
- Divisia money
- monetary aggregate
- component asset
- assigned weight
- evolutionary neural network
- quantitative inflation prediction
- inflation forecasting performance