Least-squares forecast averaging
Least-squares forecast averaging
复制标题
DOI：
10.1016/j.jeconom.2008.08.022
复制
发表时间：
2008-10
影响因子：
6.3
通讯作者：
B. Hansen
中科院分区：
文献类型：
--
作者：
B. Hansen
提出了基于Mallows模型平均法的预测组合方法。该方法通过最小化马洛斯准则来选择预测权重。该准则是样本内均方误差（MSE）和样本外一步前均方预测误差（MSFE）的渐近无偏估计。此外，MMA的权重是渐近均方最优的时间序列依赖的情况下。我们展示了如何计算MMA权重的预测设置，并调查简单但说明性的模拟环境中的方法的性能。我们发现，MMA预测有较低的MSFE和有更低的最大遗憾比其他可行的预测方法，包括相等的权重，BIC选择，加权BIC，AIC选择，加权AIC，贝茨-格兰杰组合，预测最小二乘法，格兰杰-Ramanathan组合。
This paper proposes forecast combination based on the method of Mallows Model Averaging (MMA). The method selects forecast weights by minimizing a Mallows criterion. This criterion is an asymptotically unbiased estimate of both the in-sample mean-squared error (MSE) and the out-of-sample one-step-ahead mean-squared forecast error (MSFE). Furthermore, the MMA weights are asymptotically mean-square optimal in the absence of time-series dependence. We show how to compute MMA weights in forecasting settings, and investigate the performance of the method in simple but illustrative simulation environments. We find that the MMA forecasts have low MSFE and have much lower maximum regret than other feasible forecasting methods, including equal weighting, BIC selection, weighted BIC, AIC selection, weighted AIC, Bates–Granger combination, predictive least squares, and Granger–Ramanathan combination.