考慮應變數之因素模型：以預測台灣經濟成長率為例

Forecast Performance of Supervised Factor Models: An Application to Taiwan’s Economic Growth Rates

陳映君(Ying-Chin Chen)

指導教授 ： 管中閔

國立臺灣大學/社會科學院/經濟學研究所/碩士(2014年)

https://doi.org/10.6342/NTU.2014.03072

摘要

本論文討論三種考慮應變數之因素模型(supervised factor model): 偏最小平方回歸模型(Partial Least Square,PLS)、主共變量回歸模型(Principal Covariate Regression,PCovR)及組合預測之主成分分析模型(Combining Forecasts Principal Component Analysi,CFPC)。我們將上述三種考慮應變數之因素模型應用在台灣經濟成長率之預測，並以均方根預測誤差(RMSFE)及平均絕對預測誤差(MAFE)衡量其預測之優劣。我們發現，考慮應變數之因素模型通常較不考慮應變數的因素模型(Principal Component Analysis, PCA)在預測上有更小的預測誤差，其中又以CFPC表現最好。另外，我們也比較考慮應變數之因素模型與主計總處對經濟成長之預測。我們發現CFPC之預測與主計總處之預測能力不相上下，因此我們認為CFPC是一個能夠避免模型錯誤設定的簡化模型(reduce form)。

關鍵字

因素模型 ； 主成份分析 ； 偏最小平方回歸模型 ； 主共變量回歸模型 ； 組合 預測之主成分分析 ； 預測

並列摘要

This thesis discusses three supervised factor models: Partial Least Square (PLS), Principal Covariate Regression (PCovR) and Combining Forecasts Principal Component Analysis (CFPC). We apply the supervised and unsupervised factor models to forecast Taiwan’s economic growth rates with 77 macroeconomic variables. We evaluate the performance of different models by comparing their RMSFE and MAFE. We found that the supervised factor models usually outperform unsupervised factor model (Principal Component Analysis, PCA) and that CFPC performs the best among the three supervised factor models. Besides, the forecasts of CFPC and Directorate General of Budget, Accounting and Statistics (DGBAS) have similar performance based on the Diebold-Mariano (DM) test, so CFPC may be a good alternative when we want to avoid ad hoc models. PCovR and PLS also have smaller RMSFE and MAFE than PCA, but they are not statistically significantly better than PCA.

並列關鍵字

Factor Analysis ； Principal Component Analysis ； Partial Least Square ； Principal Covariate Regression ； Combining Forecast Principal Component Analysis ； Forecast

參考文獻

BAI, J., AND S. NG (2008): Large dimensional factor analysis. Now Publishers Inc.

principal components,” Journal of the American Statistical Association, 101(473), 119–

BARCLAY, D., C. HIGGINS, AND R. THOMPSON (1995): “The partial least squares

Technology studies, 2(2), 285–309.

BATES, J., AND C. GRANGER (1969): “The Combination of Forecasts,” Journal of the