Data for Taylor Rule Exchange Rate Model with QE
Description
The data relates to a study on the Taylor rule exchange rate model which incorporates the effects of alternative monetary policies such as QE and is related to the paper 'Taylor Rule Exchange Rate Models with Alternative Monetary Policies' by Rudan Wang and Bruce Morley. The aim is to determine if this model outperforms the random walk and other Taylor rule models and UIP in out of sample forecasting. Although it beats the random walk in all countries, it doesn't consistently outperform the forecasts from a UIP model. The data consists of quarterly time series from 1990 to 2019 for four countries: the UK, US, Japan and Eurozone. It comprises the standard Taylor rule variables, such as the interest rate, inflation rate and output gap, as well as the bilateral exchange rates and returns on government bonds. The real time output data was collected from the OECD Real-Time Data and Revisions Dataset and Real Time Dataset for the OECD – Dallas Fed. The inflation data is from the historical World Economic Outlook forecast database. The remaining data are drawn from the IMF’s International Financial Statistics and the FRED Economic research data base. Money market rates (or “call rate”, Central Bank of Japan) are used as a measure of the short-term interest rates that the central bank sets every time period. The interest rate on 10-year maturity bonds have been used as the long-term interest rates. The nominal exchange rate is defined as the U.S. dollar price of foreign currency and is taken as the end-of-month exchange rate
Files
Steps to reproduce
The model estimated is the standard Taylor rule exchange rate model, augmented with the long-term government bond yield. The aim is to determine if the model outperforms the standard Taylor rule model and UIP using out of sample forecasts, where the forecasts are measures using a number of tests including the Clark West test. RATS and Matlab used. The real-time forecasted inflation data are constructed from the CPI inflation data from the historical World Economic Outlook forecast database. The forecast is updated every six months, therefore, we use the first six months of data for the forecast for Q1 and Q2, the second six months of data for Q3 and Q4. The shadow rates for the Eurozone, Japan, UK and the USA, as developed by Wu and Xia (2016) have been obtained from the Reserve Bank of New Zealand and Wu and Xia. The data refer to the rates at the end-of-month for each quarter. The output gap is constructed as the percentage change of actual output from a quadratic trend, using an expanding window. Therefore for the first vintage 1989:Q4, the trend is calculated using data from 1985:Q1 to 1989:Q3. For each subsequent vintage, we update the trend by one quarter.