- Tytuł:
- Forecasting Yield Curves in an Adaptive Framework
- Autorzy:
-
Chen, Ying
Li, Bo - Powiązania:
- https://bibliotekanauki.pl/articles/483285.pdf
- Data publikacji:
- 2011
- Wydawca:
- Polska Akademia Nauk. Czytelnia Czasopism PAN
- Tematy:
-
interest rates
functional principal component analysis
local parametric model
Nelson-Siegel model - Opis:
- Forecasting yield curves with regime switches is important in academia and financial industry. As the number of interest rate maturities increases, it poses difficulties in estimating parameters due to the curse of dimensionality. To deal with such a feature, factor models have been developed. However, the existing approaches are restrictive and largely based on the stationarity assumption of the factors. This inaccuracy creates non-ignorable financial risks, especially when the market is volatile. In this paper, a new methodology is proposed to adaptively forecast yield curves. Specifically, functional principal component analysis (FPCA) is used to extract factors capable of representing the features of yield curves. The local AR(1) model with time-dependent parameters is used to forecast each factor. Simulation and empirical studies reveal the superiority of this method over its natural competitor, the dynamic Nelson-Siegel (DNS) model. For the yield curves of the U.S. and China, the adaptive method provides more accurate 6- and 12-month ahead forecasts.
- Źródło:
-
Central European Journal of Economic Modelling and Econometrics; 2011, 3, 4; 237-259
2080-0886
2080-119X - Pojawia się w:
- Central European Journal of Economic Modelling and Econometrics
- Dostawca treści:
- Biblioteka Nauki