A study on estimation of volatility using big data analysis
Project/Area Number |
15K03406
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Research Field |
Economic statistics
|
Research Institution | Kwansei Gakuin University |
Principal Investigator |
|
Project Period (FY) |
2015-10-21 – 2018-03-31
|
Project Status |
Completed (Fiscal Year 2017)
|
Budget Amount *help |
¥3,640,000 (Direct Cost: ¥2,800,000、Indirect Cost: ¥840,000)
Fiscal Year 2017: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2016: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Fiscal Year 2015: ¥2,340,000 (Direct Cost: ¥1,800,000、Indirect Cost: ¥540,000)
|
Keywords | ビッグデータ / オンラインニュース / 動的トピックモデル / トピックスコア / 予測 / 実現ボラティリティ |
Outline of Final Research Achievements |
1. Data manipulation: We used Reuters Japan's online news article as a proxy variable for information supply. Then, we executed morphological analysis on the acquired online news articles and converted these into data for each word class. 2. Model formulation: We implemented time series models with lag values of topic score calculated using a dynamic topic model which is one of stochastic generating models of documents and volatility as explanatory variables. 3. Outcome: By using the model including the topic score, we could demonstrate empirically that the predictive power of volatility improves.
|
Report
(4 results)
Research Products
(16 results)