The Spot Price Forecasting and the Futures Markets Efficiency Analysis Based on Wavelet

摘要

This paper employs wavelet analysis technique to examine the wheat spot prices for forecasting and investigates the issue of market efficiency for wheat in America. This method is based on a wavelet multi-scaling approach that decomposes a given time series on a scale-by-scale basis. A wavelet-based prediction procedure is introduced and market data on wheat is used to provide forecasts for one period ahead. The results are compared with data from the wheat futures markets and the relative performance of this procedure is used to investigate whether futures markets are efficiently priced. The empirical results illustrate a persistent pattern indicating that the wavelet-based forecast procedure outperforms the futures markets on the average. Seemingly, these results support the idea that the wheat futures market might not be efficiently priced. It is worth mentioning that the predictive power of wavelet-based forecasting procedure is not sensitive to sample size. In contrast, the correlation coefficients between the future values and the actual values are sensitive to sample sizes.