Estimation and Forecasting of Russian Money Market Yield Curves

Abstract

The paper analyses an approach to forecasting the trajectory of the RUONIA money market rate (key rate proxy), based on a linear combination of the values of the ROISfix swap yield curves, which reflect market expectations about the future trajectory of the rate, and the forecasts of a vector autoregression model incorporating macroeconomic variables. The Nelson–Siegel and Svensson models are used to construct yield curves. According to the results obtained, for horizons of a year or more, the application of the proposed combination improves the accuracy of forecasts compared to market forecasts, while for shorter horizons, market expectations are more accurate. The study also analyses the impact of a monetary policy shock on yield curve parameters using the local projections method and shows that a monetary policy shock changes the shape of market forecasts, affecting the yield curve at all time horizons and raising long-term rate expectations by one percentage point. To test the applicability of the model in practice, a simulation of the monetary shock of 28 February 2022 was conducted.