摘要:Forecasting agricultural prices can be challenging. The price is affected by so many factors that their combined effect may be a random walk process. The purpose of this paper is to verify the random walk hypothesis on the futures contracts for wheat on the CBOT and MATIF exchanges. The research is supposed to identify the type (random, persistent or anti-persistent) of time series represented by futures contracts. The fractal dimension estimated by the surface area division method (SDM) was the tool used to determine the statistical properties of the futures contract time series. This paper presents a simplified calculation scheme for this dimension and a section of significance tables. Usually, research findings do not enable rejecting the hypothesis of a random walk of futures contracts for wheat, which suggests they follow an unpredictable pattern. On the other hand, the CBOT (unlike MATIF) witnessed a slightly different behavior of December and March contracts (which moved closer to a process with a strengthened trend) and of May, June and September contracts (which moved closer to a process of adjustment to the average). This can be used for speculative purposes. However, all the results fell within the limits of random walk. The length of the time series is an interesting question: it turns out that as the time horizon becomes longer, the results are more stable, i.e. the number of inconsistencies with the random walk process decreases. The random walk hypothesis is difficult to reject. It is confirmed that the determinants of agricultural prices drive their random motion.
其他摘要:Modelowanie i prognozowanie cen w rolnictwie jest procesem niezwykle trudnym. Liczba czynników wpływająca na cenę jest na tyle obszerna, że łącznym efektem ich oddziaływania może być proces błądzenia losowego. Celem artykułu jest weryfikacja hipotezy błąd
关键词:cereal market; futures prices; random walk; persistent series; anti-persistent series
其他关键词:rynek zbóż; ceny futures; błądzenie losowe; szeregi persystentne; szeregi antypersystentny