2010年7月9日 統計数理研究所 オープンハウス 確率モデル推定パラメータ値を用いた市場木材価格の期間構造変化の探求 Searching for Structural Change in Market-Based Log Price with Regard to the Estimated Parameters 吉 本 敦 数理・推論研究系 教授 We investigate how estimated parameters change with different sub-set of the whole time series data in order to seek any structural change in the market. We apply the geometric Brownian motion as well as geometric mean-reverting process. The data are those gathered in one local log auction market in Fukuoka, Japan. Conclusions ・ Structural change can be searched by the set of estimated parameters －1991-1997 ・ Parameters of the model are random variables ・ Possibility of stochastic volatility model?