Forecasting economic times series by uncertain models

Research Project

Project/Area Number	16K03604
Research Category	Grant-in-Aid for Scientific Research (C)
Allocation Type	Multi-year Fund
Section	一般
Research Field	Economic statistics
Research Institution	Doshisha University (2017-2019) Ryukoku University (2016)
Principal Investigator	Maki Daiki 同志社大学, 商学部, 教授 (60423737)
Project Period (FY)	2016-10-21 – 2020-03-31
Project Status	Completed (Fiscal Year 2019)
Budget Amount *help	¥3,250,000 (Direct Cost: ¥2,500,000、Indirect Cost: ¥750,000) Fiscal Year 2018: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000) Fiscal Year 2017: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000) Fiscal Year 2016: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Keywords	誤って特定化されたモデル / 予測 / 不確実なモデル / 非線形時系列 / ボラティリティ / ARCH検定 / ノンパラメトリック検定 / ノンパラメトリック回帰 / 統計的特性 / 時変的特性 / サイズの歪み / モデル選択 / 不確実性
Outline of Final Research Achievements	This study has three main results. First is that nonlinear causality tests find spurious causality relationships in the presence of volatility spillover. Second is that tests for time-varying properties of the conditional mean by the bootstrap are robust regardless of the time-varying variance model, whereas tests for time-varying properties of the conditional variance do not perform well in the presence of misspesified time-varying mean. Third is that the ARCH tests based on the polynomial approximation regression approach have better statistical properties. From the results, we clear problems and improvements for forecasting time series by uncertain models.
Academic Significance and Societal Importance of the Research Achievements	本研究の成果により、不確実な状況で予測を行う際に、モデルやボラティリティ特定化の誤りの影響を明らかにするだけでなく、どのような手法を用いれば効率的に予測を行うことができるかを示せた。したがって、本研究の結果から予測精度を高めることができ、主要経済変数の予測やそれらを用いた政策等にも有益となりうる。そのため、本研究で明らかにされた事実は、経済時系列の予測に大きな役割を果たすと期待できる。