Abstract
Abstract This paper evaluates the accuracy of forecasts for Polish interest rates of various maturities. We apply the traditional autoregressive Diebold-Li framework as well as its extension, in which the dynamics of latent factors are explained with machine learning techniques. Our findings are fourfold. Firstly, they show that all methods have failed to predict the declining trend of interest rates. Secondly, they suggest that the dynamic affine models have not been able to systematically outperform standard univariate time series models. Thirdly, they indicate that the relative performance of the analyzed models has depended on yield maturity and forecast horizon. Finally, they demonstrate that, in comparison to the traditional time series models, machine learning techniques have not systematically improved the accuracy of forecasts.
Highlights
Developing a model that would provide accurate forecasts for financial time series is the Holy Grail for financial markets participants
We apply the traditional autoregressive DieboldLi framework as well as its extension, in which the dynamics of latent factors are explained with machine learning techniques
In this group we consider a random walk (RW), autoregression (AR) and a random forest (RF) i.e. supervised machine learning model based on decision trees (Breiman, 2001)
Summary
Developing a model that would provide accurate forecasts for financial time series is the Holy Grail for financial markets participants. The first one contains univariate models which assume that the current levels of interest rates depend only on their lagged values. In this group we consider a random walk (RW), autoregression (AR) and a random forest (RF) i.e. supervised machine learning model based on decision trees (Breiman, 2001). We consider interest rate forecasts which are calculated using the yield curve interpolated with the Nelson and Siegel (1987) (NS) method assuming no arbitrage condition. We demonstrate that, in comparison to traditional time series models, applied machine learning techniques have not systematically improved the accuracy of forecasts. The final section is dedicated to the discussion and conclusions
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
