دانلود رایگان مقاله پیش بینی تحقق نوسانات مبتنی بر محدوده با روش میانگین مدل پویا

Abstract

In this study, we forecast the realized range-based volatility (RRV) using the heterogeneous autoregressive realized range-based volatility (HAR-RRV) model and its various extensions, which are called HAR-RRV-type models. We first consider the time-varying property of those models’ parameters using the dynamic model averaging (DMA) approach and evaluate the forecasting performance of three types: individual HAR-RRV-type models, combined models with constant weights, and combined models with time-varying weights. Our out-of-sample empirical results show that combined models with time-varying weights can not only generate more accurate forecasts, but also beat individual models and combined models with constant weights.

5. Conclusions

Since the realized range-based volatility (RRV) is proved to be a better measurement than the realized volatility (RV), this study uses RRV as the proxy for oil futures volatility. To investigate the timevarying property of HAR-RRV models’ parameters, this paper applies DMA approach as a combined model with time-varying weights into RRV framework. At first, this study constructs five individual HAR-RRV-type models including HAR-RRV-ONI considering overnight information. Moreover, based on the individual models, combined models with constant weights and time-varying weights are constructed. Finally, the performance of three types of the models is compared by various methods including MCS test. Our findings demonstrate that the HARRRV-type models can successfully capture the long-term memory behavior of volatility in oil futures market. Combined models systematically perform better than individual models. In particular, using DMA to combine the forecasts of HAR-RRV models can significantly improve the forecasting accuracy. DMA approach can beat both individual models and combined models with constant weights, including the combined model with equal-weighted average, which is usually used as the benchmark. The results highlight the significance of using DMA as a combined model with time-varying weights. By allowing both the models and their coefficients changing over time, DMA approach shows several benefits to be incorporated into RRV framework and opens a new path for forecasting technique in real time.