Constructing a financial conditions index for the United Kingdom: A comparative analysis

摘要

We investigate the optimal constituent variable weighting method for a UK financial conditions index (FCI) using a small number of financial indicators. The criterion for choosing the optimal weighting model concentrates on the index's ability to predict economic activity. We develop a "two-step" process as a new weighted-sum method and show that it is superior to other existing weighted-sum models in creating an FCI. For comparative purposes, we create another FCI using a time-varying parameter factor-augmented vector autoregressive (TVP-FAVAR) with stochastic volatility model as a principal-component method. The results suggest that the TVP-FAVAR model is the best variable-weighting model to create an FCI in relation to its purpose of forecasting developments in the economy.