1. NCCU Academic Hub
  2. Publications
  3. College of Social Science
  4. Theses
  5. Item

Publications-Theses

Article View/Open

Publication Export

Google ScholarTM

NCCU Library

Citation Infomation

Related Publications in TAIR

題名 利用高維度狀態轉換因子模型即時認定台灣的景氣轉折
A High-Dimensional Factor Model with Regime Switching for Real-Time Identification of Business Cycle Turning Points in Taiwan
作者 盧思云
Lu, Si-Yun
貢獻者 徐士勛
Hsu, Shih-Hsun
盧思云
Lu, Si-Yun
關鍵詞 景氣循環
狀態轉換因子模型
即時認定
EM演算法
Business cycle
Regime-switching factor model
Real-time identification
EM algorithm
日期 2025
上傳時間 4-Aug-2025 12:51:10 (UTC+8)
摘要 傳統景氣循環認定方法存在時間落後問題，如我國近3次景氣循環，國家發展委員會在轉折點發生約兩年後才會認定，降低景氣循環實務決策應用價值。本文採用國發會七項同時指標，資料期間涵蓋1999年1月至2024年12月，透過EM演算法估計模型參數，並結合機率門檻與景氣循環判定標準進行樣本外即時預測。研究設定景氣循環具有擴張與衰退兩種狀態，運用擴展窗口法模擬政策制定者在資訊不完全條件下的即時判斷情境。實證結果顯示，模型成功識別台灣樣本外所涵蓋的五次景氣轉折點，與國發會官方認定的平均誤差為0.8個月。結合景氣循環判定標準後，大幅降低假警報 (false positives) 問題。相較於官方發布時間，模型平均能提前21個月提供景氣轉折預警，其中特別在COVID-19疫情期間提前14個月預警第15次景氣循環衰退認定上。本研究為高維度狀態轉換因子模型相關領域供新的分析工具，並建立具實用價值的景氣轉折即時認定框架。
Traditional business cycle dating methods have a problem of time lag. For example, in the most recent three business cycles in Taiwan, the National Development Council (NDC) recognized the turning points approximately two years after their actual occurrence, which reduces the practical value of business cycle information in decision-making. This study uses the seven coincident indicators published by the NDC, covering the data period from January 1999 to December 2024. Model parameters are estimated through the Expectation Maximization algorithm, and out-of-sample real-time prediction is conducted by combining probability thresholds and business cycle dating rules. The study assumes that the business cycle consists of two states: expansion and recession, and applies the expanding window method to simulate the real-time judgment scenario of policymakers under incomplete information. The empirical results show that the model successfully identifies the five business cycle turning points in Taiwan covered by the out-of-sample period, with an average deviation of 0.8 months compared to the official dates recognized by the NDC. After incorporating business cycle dating criteria, the issue of false positives is significantly reduced. Compared to the official announcement dates, the model can on average provide early warning signals of turning points 21 months in advance, including a 14-month lead in predicting the 15th recession recognized during the COVID-19 pandemic. This study provides a new analytical tool for the field of high-dimensional regime-switching factor models, and establishes a practical real-time identification framework for business cycle turning points.
參考文獻 Bai, J. (2003). Inferential theory for factor models of large dimensions. Econometrica, 71(1), 135–171. Bai, J., & Ng, S. (2002). Determining the number of factors in approximate factor models. Econometrica, 70(1), 191–221. Banerji, A. (1999). The three Ps: Simple tools for monitoring economic cycles. Business Economics, 34(4), 72–76. Breitung, J., & Eickmeier, S. (2011). Testing for structural breaks in dynamic factor models. Journal of Econometrics, 163(1), 71–84. Bry, G., & Boschan, C. (1971). Programmed selection of cyclical turning points. In Cyclical analysis of time series: Selected procedures and computer programs (pp. 7–63). NBER. Burns, A. F., & Mitchell, W. C. (1946). Measuring business cycles. National bureau of economic research. Chauvet, M. (1998). An econometric characterization of business cycle dynamics with factor structure and regime switching. International Economic Review, 39(4), 969–996. Chauvet, M., & Piger, J. (2008). A comparison of the real-time performance of business cycle dating methods. Journal of Business & Economic Statistics, 26(1), 42–49. Cheng, X., Liao, Z., & Schorfheide, F. (2016). Shrinkage estimation of high-dimensional factor models with structural instabilities. The Review of Economic Studies, 83(4), 1511–1543. Diebold, F. X., & Rudebusch, G. D. (1994). Measuring business cycles: A modern perspective (Vol. 4643). National Bureau of Economic Research Cambridge, Mass., USA. Engle, R., & Watson, M. (1981). A one-factor multivariate time series model of metropolitan wage rates. Journal of the American Statistical Association, 76(376), 774–781. Filardo, A. J. (1994). Business-cycle phases and their transitional dynamics. Journal of Business & Economic Statistics, 12(3), 299–308. Hamilton, J. D. (1989). A new approach to the economic analysis of nonstationary time series and the business cycle. Econometrica, 57(2), 357–384. Hamilton, J. D. (1990). Analysis of Time Series subject to changes in Regime. Journal of Econometrics, 45(1), 39–70. Hamilton, J. D. (2011). Calling recessions in real time. International Journal of Forecasting, 27(4), 1006–1026. Kim, C. J. (1994). Dynamic linear models with markov-switching. Journal of Econometrics, 60(1), 1–22. Kim, C. J., & Nelson, C. R. (1998). Business cycle turning points, a new coincident index, and tests of duration dependence based on a dynamic factor model with regime switching. Review of Economics and Statistics, 80(2), 188–201. Kim, M.-J., & Yoo, J.-S. (1995). New index of coincident indicators: A multivariate Markov switching factor model approach. Journal of Monetary Economics, 36(3), 607–630. Lam, P.-S. (1990). The hamilton model with a general autoregressive component: Estimation and comparison with other models of economic time series: Estimation and comparison with other models of economic time series. Journal of Monetary Eco- nomics, 26(3), 409–432. Liu, X., & Chen, R. (2016). Regime-switching factor models for high-dimensional time series. Statistica Sinica, 26(4), 1427–1451. Pelger, M., & Xiong, R. (2022). State-varying factor models of large dimensions. Journal of Business & Economic Statistics, 40(3), 1315–1333. Stock, J. H., & Watson, M. W. (1989). New indexes of coincident and leading economic indicators. NBER macroeconomics annual, 4, 351–394. 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. Urga, G., & Wang, F. (2024). Estimation and inference for high dimensional factor model with regime switching. Journal of Econometrics, 241(2), 105752. 朱浩榜（2021），〈即時認定台灣的景氣轉折〉，《經濟論文叢刊》，49(3)，頁 335–370。 陳淑玲、黃裕烈（2015），〈臺灣景氣基準循環指數之檢討與改進〉，《Taiwan Economic Forecast & Policy》，46(1)。 陳惠薇（2009），〈我國第 11 次景氣循環高峰之認定與研析〉，《Economic Research》，9，頁 1。
描述 碩士
國立政治大學
經濟學系
112258020
資料來源 http://thesis.lib.nccu.edu.tw/record/#G0112258020
資料類型 thesis
dc.contributor.advisor 徐士勛zh_TW
dc.contributor.advisor Hsu, Shih-Hsunen_US
dc.contributor.author (Authors) 盧思云zh_TW
dc.contributor.author (Authors) Lu, Si-Yunen_US
dc.creator (作者) 盧思云zh_TW
dc.creator (作者) Lu, Si-Yunen_US
dc.date (日期) 2025en_US
dc.date.accessioned 4-Aug-2025 12:51:10 (UTC+8)-
dc.date.available 4-Aug-2025 12:51:10 (UTC+8)-
dc.date.issued (上傳時間) 4-Aug-2025 12:51:10 (UTC+8)-
dc.identifier (Other Identifiers) G0112258020en_US
dc.identifier.uri (URI) https://nccur.lib.nccu.edu.tw/handle/140.119/158277-
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) 112258020zh_TW
dc.description.abstract (摘要) 傳統景氣循環認定方法存在時間落後問題，如我國近3次景氣循環，國家發展委員會在轉折點發生約兩年後才會認定，降低景氣循環實務決策應用價值。本文採用國發會七項同時指標，資料期間涵蓋1999年1月至2024年12月，透過EM演算法估計模型參數，並結合機率門檻與景氣循環判定標準進行樣本外即時預測。研究設定景氣循環具有擴張與衰退兩種狀態，運用擴展窗口法模擬政策制定者在資訊不完全條件下的即時判斷情境。實證結果顯示，模型成功識別台灣樣本外所涵蓋的五次景氣轉折點，與國發會官方認定的平均誤差為0.8個月。結合景氣循環判定標準後，大幅降低假警報 (false positives) 問題。相較於官方發布時間，模型平均能提前21個月提供景氣轉折預警，其中特別在COVID-19疫情期間提前14個月預警第15次景氣循環衰退認定上。本研究為高維度狀態轉換因子模型相關領域供新的分析工具，並建立具實用價值的景氣轉折即時認定框架。zh_TW
dc.description.abstract (摘要) Traditional business cycle dating methods have a problem of time lag. For example, in the most recent three business cycles in Taiwan, the National Development Council (NDC) recognized the turning points approximately two years after their actual occurrence, which reduces the practical value of business cycle information in decision-making. This study uses the seven coincident indicators published by the NDC, covering the data period from January 1999 to December 2024. Model parameters are estimated through the Expectation Maximization algorithm, and out-of-sample real-time prediction is conducted by combining probability thresholds and business cycle dating rules. The study assumes that the business cycle consists of two states: expansion and recession, and applies the expanding window method to simulate the real-time judgment scenario of policymakers under incomplete information. The empirical results show that the model successfully identifies the five business cycle turning points in Taiwan covered by the out-of-sample period, with an average deviation of 0.8 months compared to the official dates recognized by the NDC. After incorporating business cycle dating criteria, the issue of false positives is significantly reduced. Compared to the official announcement dates, the model can on average provide early warning signals of turning points 21 months in advance, including a 14-month lead in predicting the 15th recession recognized during the COVID-19 pandemic. This study provides a new analytical tool for the field of high-dimensional regime-switching factor models, and establishes a practical real-time identification framework for business cycle turning points.en_US
dc.description.tableofcontents 1 前言 6 2 文獻回顧 7 2.1 景氣循環轉折即時認定 7 2.2 狀態轉換因子模型 9 3 模型設定 11 3.1 狀態轉換因子模型設定 11 3.2 識別條件 (Identification Conditions) 11 3.3 參數估計 12 3.3.1 Λ 和 𝜎2 的一階條件 13 3.4 EM 演算法 14 3.4.1 EM 演算法參數估計 15 3.4.2 EM 演算法小結 17 3.5 景氣循環判定方式 18 4 實證分析 18 4.1 資料處理與變數選擇 18 4.2 模型估計結果與景氣轉折點認定 20 4.2.1 因子數量估計20 4.2.2 因子序列與模型初始值設定 22 4.2.3 樣本外預測結果 22 4.2.4 模型預測轉折點 24 5 結論與討論 27 A 附錄 32zh_TW
dc.format.extent 760755 bytes-
dc.format.mimetype application/pdf-
dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0112258020en_US
dc.subject (關鍵詞) 景氣循環zh_TW
dc.subject (關鍵詞) 狀態轉換因子模型zh_TW
dc.subject (關鍵詞) 即時認定zh_TW
dc.subject (關鍵詞) EM演算法zh_TW
dc.subject (關鍵詞) Business cycleen_US
dc.subject (關鍵詞) Regime-switching factor modelen_US
dc.subject (關鍵詞) Real-time identificationen_US
dc.subject (關鍵詞) EM algorithmen_US
dc.title (題名) 利用高維度狀態轉換因子模型即時認定台灣的景氣轉折zh_TW
dc.title (題名) A High-Dimensional Factor Model with Regime Switching for Real-Time Identification of Business Cycle Turning Points in Taiwanen_US
dc.type (資料類型) thesisen_US
dc.relation.reference (參考文獻) Bai, J. (2003). Inferential theory for factor models of large dimensions. Econometrica, 71(1), 135–171. Bai, J., & Ng, S. (2002). Determining the number of factors in approximate factor models. Econometrica, 70(1), 191–221. Banerji, A. (1999). The three Ps: Simple tools for monitoring economic cycles. Business Economics, 34(4), 72–76. Breitung, J., & Eickmeier, S. (2011). Testing for structural breaks in dynamic factor models. Journal of Econometrics, 163(1), 71–84. Bry, G., & Boschan, C. (1971). Programmed selection of cyclical turning points. In Cyclical analysis of time series: Selected procedures and computer programs (pp. 7–63). NBER. Burns, A. F., & Mitchell, W. C. (1946). Measuring business cycles. National bureau of economic research. Chauvet, M. (1998). An econometric characterization of business cycle dynamics with factor structure and regime switching. International Economic Review, 39(4), 969–996. Chauvet, M., & Piger, J. (2008). A comparison of the real-time performance of business cycle dating methods. Journal of Business & Economic Statistics, 26(1), 42–49. Cheng, X., Liao, Z., & Schorfheide, F. (2016). Shrinkage estimation of high-dimensional factor models with structural instabilities. The Review of Economic Studies, 83(4), 1511–1543. Diebold, F. X., & Rudebusch, G. D. (1994). Measuring business cycles: A modern perspective (Vol. 4643). National Bureau of Economic Research Cambridge, Mass., USA. Engle, R., & Watson, M. (1981). A one-factor multivariate time series model of metropolitan wage rates. Journal of the American Statistical Association, 76(376), 774–781. Filardo, A. J. (1994). Business-cycle phases and their transitional dynamics. Journal of Business & Economic Statistics, 12(3), 299–308. Hamilton, J. D. (1989). A new approach to the economic analysis of nonstationary time series and the business cycle. Econometrica, 57(2), 357–384. Hamilton, J. D. (1990). Analysis of Time Series subject to changes in Regime. Journal of Econometrics, 45(1), 39–70. Hamilton, J. D. (2011). Calling recessions in real time. International Journal of Forecasting, 27(4), 1006–1026. Kim, C. J. (1994). Dynamic linear models with markov-switching. Journal of Econometrics, 60(1), 1–22. Kim, C. J., & Nelson, C. R. (1998). Business cycle turning points, a new coincident index, and tests of duration dependence based on a dynamic factor model with regime switching. Review of Economics and Statistics, 80(2), 188–201. Kim, M.-J., & Yoo, J.-S. (1995). New index of coincident indicators: A multivariate Markov switching factor model approach. Journal of Monetary Economics, 36(3), 607–630. Lam, P.-S. (1990). The hamilton model with a general autoregressive component: Estimation and comparison with other models of economic time series: Estimation and comparison with other models of economic time series. Journal of Monetary Eco- nomics, 26(3), 409–432. Liu, X., & Chen, R. (2016). Regime-switching factor models for high-dimensional time series. Statistica Sinica, 26(4), 1427–1451. Pelger, M., & Xiong, R. (2022). State-varying factor models of large dimensions. Journal of Business & Economic Statistics, 40(3), 1315–1333. Stock, J. H., & Watson, M. W. (1989). New indexes of coincident and leading economic indicators. NBER macroeconomics annual, 4, 351–394. 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. Urga, G., & Wang, F. (2024). Estimation and inference for high dimensional factor model with regime switching. Journal of Econometrics, 241(2), 105752. 朱浩榜（2021），〈即時認定台灣的景氣轉折〉，《經濟論文叢刊》，49(3)，頁 335–370。 陳淑玲、黃裕烈（2015），〈臺灣景氣基準循環指數之檢討與改進〉，《Taiwan Economic Forecast & Policy》，46(1)。 陳惠薇（2009），〈我國第 11 次景氣循環高峰之認定與研析〉，《Economic Research》，9，頁 1。zh_TW