Working Paper no. 113
By Shamik Dhar, Darren Pain and Ryland Thomas
In this paper we estimate a structural empirical model of the UK monetary transmission mechanism, which can be used for policy analysis and forecasting. We model a small system of eight variables that theoretically have an important role in the transmission mechanism. The aim is to decompose the movements of each of these variables into a small number of independent underlying forcing processes or ‘shocks’, with a well-defined economic interpretation. To do this we estimate a statistical (VAR) model of the data, on which we impose a minimal number of identifying restrictions. Cointegration analysis is also used to distinguish between permanent shocks, which drive the stochastic trends of the system, and temporary shocks, which have purely cyclical effects. We find that, in addition to identifying shocks to productivity, domestic demand, external demand and the foreign exchange risk premium, we are able to distinguish between several types of monetary shock. In particular, we are able to make a distinction between ‘permanent’ monetary policy shocks, attributable to changes in the underlying nominal target of the authorities, and ‘temporary’ policy shocks, reflecting either policy ‘errors’ or transitory deviations from the authorities’ reaction function. We are also able to identify a financial intermediation shock, reflecting changes in the provision of credit by the banking system and the degree of financial liberalisation. We demonstrate some of the practical uses to which the model can be put. These include: (a) estimating the deviation of each of the variables from long-run equilibrium to generate measures of the output gap and the size of liquidity under/overhangs; (b) analysing the importance of different shocks for each of the variables over different periods in UK economic history; and (c) generating conditional inflation forecasts based on different paths for the stochastic trends and monetary policy.