| dc.contributor | 國貿系 | |
| dc.creator (作者) | 郭炳伸 | |
| dc.creator (作者) | Yang, Zheng;Wu, Haocheng;Kuo, Biing-Shen;Ma, Yongkai | |
| dc.date (日期) | 2026-05 | |
| dc.date.accessioned | 20-Apr-2026 10:20:51 (UTC+8) | - |
| dc.date.available | 20-Apr-2026 10:20:51 (UTC+8) | - |
| dc.date.issued (上傳時間) | 20-Apr-2026 10:20:51 (UTC+8) | - |
| dc.identifier.uri (URI) | https://ah.lib.nccu.edu.tw/item?item_id=182127 | - |
| dc.description.abstract (摘要) | Using a high-dimensional dataset comprising 993 macroeconomic predictors, we develop a dimensionality reduction combination forecast framework to examine the out-of-sample predictability of the Chinese equity premium. We compare forecasts across two aspects: (1) 14 predictor groups versus the full set, and (2) 15 individual dimensionality reduction models versus three combination methods. Our findings indicate that the full set offers richer information and that combining forecasts across dimensionality reduction models yields statistically and economically out-of-sample gains. Encompassing tests and MSPE decomposition explain the benefits of the dimensionality reduction combination forecast. These findings survive a series of robustness checks. | |
| dc.format.extent | 106 bytes | - |
| dc.format.mimetype | text/html | - |
| dc.relation (關聯) | Journal of Economic Dynamics and Control, Vol.186, 105308 | |
| dc.subject (關鍵詞) | Equity premium; Macroeconomic variables; Dimensionality reduction; Machine learning; Combination forecast | |
| dc.title (題名) | Forecasting Chinese equity premium: A dimensionality reduction combination approach | |
| dc.type (資料類型) | article | |
| dc.identifier.doi (DOI) | 10.1016/j.jedc.2026.105308 | |
| dc.doi.uri (DOI) | https://doi.org/10.1016/j.jedc.2026.105308 | |
