2016 Fiscal Year Final Research Report
Portfolio optimization problem for high dimensional data
| Project/Area Number |
24730193
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Economic statistics
|
|---|
| Research Institution | Keio University (2014-2016)
Jikei University School of Medicine (2012-2013) |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2012-04-01 – 2017-03-31
|
| Keywords | 最適ポートフォリオ / 統計的推定 / 漸近理論 / 高次元データ / 縮小推定量 |
|---|
| Outline of Final Research Achievements |
In this research, we propose two types of optimal portfolio estimators when an investor aims to invest a huge number of assets. Suppose that high dimensional data with respect to the past asset return is available. Then we propose (1)Optimal portfolio estimator by using an unbiased estimator for the inverse of the covariance matrix, (2)Shrinkage type estimator for the plug-in portfolio weights estimator. First, we derive the theoretical properties and confirm numerical validity of these estimators. Then, we compare the performance (in terms of the sharp ration and the utility function) with existing portfolios based on the Japanese stock price data (200 issues) and realized the superiority of the proposed method.
|
| Free Research Field |
統計科学
|
