Publications-Theses
Article View/Open
Publication Export
-
Google ScholarTM
NCCU Library
Citation Infomation
Related Publications in TAIR
題名 台灣股價加權指數報酬率與重要總體變數選擇探討
Exploration of the Relationship Between Taiwan’s Weighted Stock Index Returns and Key Macroeconomic Variables作者 洪子逸
Hung, Zih-Yi貢獻者 蔡致遠<br>李浩仲
Tsai, Chi-Yuan<br>Li, Hao-Chung
洪子逸
Hung, Zih-Yi關鍵詞 台灣股價加權指數超額報酬
因子
樣本外預測
機器學習
TAIEX Excess Returns
Factors
Out-of-Sample Forecasting
Machine Learning日期 2025 上傳時間 4-Aug-2025 12:51:21 (UTC+8) 摘要 本文蒐集2004年1月至2023年12月共80項對於台灣股價加權指數具潛在影響力總體變數，探討不同降維方法所萃取之潛在因子對模型預測力的影響。本文以擴散指數模型為框架，第一階段將變數以不區分類別與區分類別的兩種方式，進行四種降維演算法，分別是主成分分析、核主成分分析、偏最小平方迴歸與偏分量迴歸，第二階段採取遞迴式最小平方法建立預測模型，並以隨機漫步模型為基準，衡量樣本外預測能力。結果顯示，核主成分分析在大多數設定下皆具最佳預測表現，進行類別區分可進一步提升監督式降維方法的預測能力，大宗資產與就業類因子對台灣股價加權指數超額報酬率之邊際貢獻最為顯著。
This study compiles 80 macroeconomic and financial variables that may affect the Taiwan Stock Exchange Capitalization Weighted Index (TAIEX) for the period from January 2004 to December 2023 and investigates how latent factors extracted through alternative dimension reduction techniques influence out of sample predictive performance. Within a diffusion index framework, the analysis proceeds in two stages. In the first stage, the variables are processed under two schemes (without prior classification and with pre-classification), and four dimension reduction algorithms are applied: principal component analysis (PCA), kernel PCA (KPCA), partial least squares regression (PLS) and partial quantile regression (PQR). In the second stage, rolling recursive ordinary least squares models are estimated, using a random walk specification as the benchmark for forecast evaluation. The results show that KPCA provides the highest predictive accuracy in most settings, while pre-classification further improves the performance of the supervised methods. Factor importance analysis indicates that commodity related and labour market factors make the largest marginal contributions to forecasting TAIEX excess returns.參考文獻 徐士勛、管中閔與羅雅惠（2005）。以擴散指標為基礎之總體經濟預測。台灣經濟預測與政策。36（1），1-28。 劉祥熹與涂登才（2012）。美國股市及其總體經濟變數間關連性與波動性之研究—VEC GJR DCC-GARCH-M 之模型應用。《經濟研究》，48（1），139–189。 張天惠（2012）。我國金融情勢指數與總體經濟預測。中央銀行季刊，34（2），11-42。 李偉銘、吳淑貞與黃啟泰（2015）。總體經濟變數對臺灣股市之大盤及類股熊市預測表現之探討。經濟研究，51（2），171-224。 葉錦徽與潘宗麟（2024）。探索預測台灣通膨隱而未現的重要因子－監督式降維模型的實證。管理評論，43（3），19-40。 Bai, J., & Ng, S. (2008). Forecasting economic time series using targeted predictors. Journal of Econometrics, 146(2), 304–317. Camacho, M., & Sancho, I. (2003). Spanish diffusion indexes. Spanish Economic Review, 5(3), 173–203. Chen, N., Roll, R., & Ross, S. (1986). Economic forces and the stock market. Journal of Business, 59(3), 383–403. Cooper, J. P. (1972). The predictive performance of quarterly econometric models of the United States. In B. G. Hickman (Ed.), Econometric models of cyclical behavior (pp. 813–947). NBER. Dhakal, D., Kandil, M., & Sharma, S. (1993). Causality between the money supply and share prices: A VAR investigation. Quarterly Journal of Business and Economics, 32(3), 52–74. Giglio, S., Kelly, B., & Pruitt, S. (2016). Systemic risk and the macroeconomy: An empirical evaluation. Journal of Financial Economics, 119(3), 457–471. Giacomini, R., & White, H. (2006). Tests of conditional predictive ability. Econometrica, 74(6), 1545–1578. Kelly, B., & Pruitt, S. (2015). The three-pass regression filter: A new approach to forecasting using many predictors. Journal of Econometrics, 186(2), 294–316. Lo, A. W., & MacKinlay, A. C. (1988). Stock market prices do not follow random walks: Evidence from a simple specification test. Review of Financial Studies, 1(1), 41–66. Lo, A. W., & MacKinlay, A. C. (1989). The size and power of the variance ratio test in finite samples: A Monte Carlo investigation. Journal of Econometrics, 40(2), 203–238. Sims, C. A. (1980a). Macroeconomics and reality. Econometrica, 48(1), 1–48. Stock, J. H., & Watson, M. W. (1998). Diffusion indexes (NBER Working Paper No. 6702). National Bureau of Economic Research. Stock, J. H., & Watson, M. W. (2002). Forecasting using principal components from a large number of predictors. Journal of the American Statistical Association, 97(460), 1167–1179. Chiang, T. C. (2023). Real stock market returns and inflation: Evidence from uncertainty hypotheses. Finance Research Letters, 53, 103606. Keswani, S., Puri, V., & Jha, R. (2024). Relationship among macroeconomic factors and stock prices: Cointegration approach from the Indian stock market. Cogent Economics & Finance, 12(1), 2355017. Humpe, A., McMillan, D. G., & Schöttl, A. (2025). Macroeconomic determinants of the stock market: A comparative study of Anglosphere and BRICS. Finance Research Letters, 75, 106869. 描述 碩士
國立政治大學
經濟學系
112258021資料來源 http://thesis.lib.nccu.edu.tw/record/#G0112258021 資料類型 thesis dc.contributor.advisor 蔡致遠<br>李浩仲 zh_TW dc.contributor.advisor Tsai, Chi-Yuan<br>Li, Hao-Chung en_US dc.contributor.author (Authors) 洪子逸 zh_TW dc.contributor.author (Authors) Hung, Zih-Yi en_US dc.creator (作者) 洪子逸 zh_TW dc.creator (作者) Hung, Zih-Yi en_US dc.date (日期) 2025 en_US dc.date.accessioned 4-Aug-2025 12:51:21 (UTC+8) - dc.date.available 4-Aug-2025 12:51:21 (UTC+8) - dc.date.issued (上傳時間) 4-Aug-2025 12:51:21 (UTC+8) - dc.identifier (Other Identifiers) G0112258021 en_US dc.identifier.uri (URI) https://ah.lib.nccu.edu.tw/item?item_id=177839 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 經濟學系 zh_TW dc.description (描述) 112258021 zh_TW dc.description.abstract (摘要) 本文蒐集2004年1月至2023年12月共80項對於台灣股價加權指數具潛在影響力總體變數，探討不同降維方法所萃取之潛在因子對模型預測力的影響。本文以擴散指數模型為框架，第一階段將變數以不區分類別與區分類別的兩種方式，進行四種降維演算法，分別是主成分分析、核主成分分析、偏最小平方迴歸與偏分量迴歸，第二階段採取遞迴式最小平方法建立預測模型，並以隨機漫步模型為基準，衡量樣本外預測能力。結果顯示，核主成分分析在大多數設定下皆具最佳預測表現，進行類別區分可進一步提升監督式降維方法的預測能力，大宗資產與就業類因子對台灣股價加權指數超額報酬率之邊際貢獻最為顯著。 zh_TW dc.description.abstract (摘要) This study compiles 80 macroeconomic and financial variables that may affect the Taiwan Stock Exchange Capitalization Weighted Index (TAIEX) for the period from January 2004 to December 2023 and investigates how latent factors extracted through alternative dimension reduction techniques influence out of sample predictive performance. Within a diffusion index framework, the analysis proceeds in two stages. In the first stage, the variables are processed under two schemes (without prior classification and with pre-classification), and four dimension reduction algorithms are applied: principal component analysis (PCA), kernel PCA (KPCA), partial least squares regression (PLS) and partial quantile regression (PQR). In the second stage, rolling recursive ordinary least squares models are estimated, using a random walk specification as the benchmark for forecast evaluation. The results show that KPCA provides the highest predictive accuracy in most settings, while pre-classification further improves the performance of the supervised methods. Factor importance analysis indicates that commodity related and labour market factors make the largest marginal contributions to forecasting TAIEX excess returns. en_US dc.description.tableofcontents 第一章 緒論 1 1-1前言 1 1-2研究架構 3 第二章 文獻回顧 4 2-1總體經濟變數預測方法 4 2-2總體經濟指標與股市 5 第三章 研究方法 7 3-1資料來源及變數分類 7 3-2資料處理 14 3-3研究方法 14 3-3-1擴散指數架構 14 3-3-2第一階段：降維演算法 15 3-3-2-1主成分分析（Principal Components Analysis, PCA） 16 3-3-2-2核主成分分析（Kernel Principal Component Analysis, KPCA） 16 3-3-2-3偏最小平方迴歸法（Partial Least Squares, PLS） 17 3-3-2-4偏分量迴歸（Partial Quantile Regression, PQR） 18 3-3-3第二階段：遞迴式最小平方法（Recursive OLS） 18 3-4模型衡量方法 19 3-5變數重要性衡量方法 19 第四章 實證結果 21 4-1樣本外分析 21 4-1-1最適降維演算法 21 4-1-2關鍵類別 23 4-1-3關鍵變數 26 第五章 結論 29 參考文獻 31 附錄 33 zh_TW dc.format.extent 1338522 bytes - dc.format.mimetype application/pdf - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0112258021 en_US dc.subject (關鍵詞) 台灣股價加權指數超額報酬 zh_TW dc.subject (關鍵詞) 因子 zh_TW dc.subject (關鍵詞) 樣本外預測 zh_TW dc.subject (關鍵詞) 機器學習 zh_TW dc.subject (關鍵詞) TAIEX Excess Returns en_US dc.subject (關鍵詞) Factors en_US dc.subject (關鍵詞) Out-of-Sample Forecasting en_US dc.subject (關鍵詞) Machine Learning en_US dc.title (題名) 台灣股價加權指數報酬率與重要總體變數選擇探討 zh_TW dc.title (題名) Exploration of the Relationship Between Taiwan’s Weighted Stock Index Returns and Key Macroeconomic Variables en_US dc.type (資料類型) thesis en_US dc.relation.reference (參考文獻) 徐士勛、管中閔與羅雅惠（2005）。以擴散指標為基礎之總體經濟預測。台灣經濟預測與政策。36（1），1-28。 劉祥熹與涂登才（2012）。美國股市及其總體經濟變數間關連性與波動性之研究—VEC GJR DCC-GARCH-M 之模型應用。《經濟研究》，48（1），139–189。 張天惠（2012）。我國金融情勢指數與總體經濟預測。中央銀行季刊，34（2），11-42。 李偉銘、吳淑貞與黃啟泰（2015）。總體經濟變數對臺灣股市之大盤及類股熊市預測表現之探討。經濟研究，51（2），171-224。 葉錦徽與潘宗麟（2024）。探索預測台灣通膨隱而未現的重要因子－監督式降維模型的實證。管理評論，43（3），19-40。 Bai, J., & Ng, S. (2008). Forecasting economic time series using targeted predictors. Journal of Econometrics, 146(2), 304–317. Camacho, M., & Sancho, I. (2003). Spanish diffusion indexes. Spanish Economic Review, 5(3), 173–203. Chen, N., Roll, R., & Ross, S. (1986). Economic forces and the stock market. Journal of Business, 59(3), 383–403. Cooper, J. P. (1972). The predictive performance of quarterly econometric models of the United States. In B. G. Hickman (Ed.), Econometric models of cyclical behavior (pp. 813–947). NBER. Dhakal, D., Kandil, M., & Sharma, S. (1993). Causality between the money supply and share prices: A VAR investigation. Quarterly Journal of Business and Economics, 32(3), 52–74. Giglio, S., Kelly, B., & Pruitt, S. (2016). Systemic risk and the macroeconomy: An empirical evaluation. Journal of Financial Economics, 119(3), 457–471. Giacomini, R., & White, H. (2006). Tests of conditional predictive ability. Econometrica, 74(6), 1545–1578. Kelly, B., & Pruitt, S. (2015). The three-pass regression filter: A new approach to forecasting using many predictors. Journal of Econometrics, 186(2), 294–316. Lo, A. W., & MacKinlay, A. C. (1988). Stock market prices do not follow random walks: Evidence from a simple specification test. Review of Financial Studies, 1(1), 41–66. Lo, A. W., & MacKinlay, A. C. (1989). The size and power of the variance ratio test in finite samples: A Monte Carlo investigation. Journal of Econometrics, 40(2), 203–238. Sims, C. A. (1980a). Macroeconomics and reality. Econometrica, 48(1), 1–48. Stock, J. H., & Watson, M. W. (1998). Diffusion indexes (NBER Working Paper No. 6702). National Bureau of Economic Research. Stock, J. H., & Watson, M. W. (2002). Forecasting using principal components from a large number of predictors. Journal of the American Statistical Association, 97(460), 1167–1179. Chiang, T. C. (2023). Real stock market returns and inflation: Evidence from uncertainty hypotheses. Finance Research Letters, 53, 103606. Keswani, S., Puri, V., & Jha, R. (2024). Relationship among macroeconomic factors and stock prices: Cointegration approach from the Indian stock market. Cogent Economics & Finance, 12(1), 2355017. Humpe, A., McMillan, D. G., & Schöttl, A. (2025). Macroeconomic determinants of the stock market: A comparative study of Anglosphere and BRICS. Finance Research Letters, 75, 106869. zh_TW
