Out-of-sample testing price discovery in commodity markets: the case of soybeans
提出一种多元框架下的样本外预测方法，检验期货价格对现货价格的预测能力，并以大豆市场为例，发现期货价格平均而言是最佳预测指标，但仅在特定时期成立。
Price discovery, a central function of futures markets, has been usually tested in-sample by studying the common stochastic trend between spot and futures prices. Instead, to evaluate futures as anticipatory prices, we develop a forecast approach to out-of-sample test price discovery in a multivariate framework. We apply it to the soybeans market. Results indicate futures prices as the best available “predictors” of future spot prices, although this finding holds only on average and for certain periods, other models show forecasting gains.