This chapter explains the testing of the consistency of the diffusion model with changes in the log-price of several commodities. It introduces the main issue of estimating diffusion processes by finite sample data and describes the estimation techniques to daily prices of crude oil, corn, copper and gold from May 2006 to March 2009. The chapter simultaneously detects whether daily time series of the price of such commodities are generated by a continuous diffusion model and identifies the functional forms in the dynamics of the corresponding log-price process. It reports the estimated values for the drift function, the diffusion function and the jump intensity function assuming that the dynamics of the log-prices is described by the diffusion model. The diffusion function is well represented by a linear decreasing function for all analysed commodities. It shows the highest variability across the analysed data sets.
FIGA TALAMANCA, G. et RONCORONI, A. (2015). Nonparametric Estimation of Energy and Commodity Price Processes. Dans: Handbook of Multi-Commodity Markets and Products: Structuring, Trading and Risk Management. 1st ed. Wiley, pp. 659-672.