美國聯準會會議紀要的文字探勘與台灣經濟變數預測
Text Mining of the FOMC Minutes and Forecasts of Taiwan Economic Variables
國立清華大學計量財務與金融學系(Department of Quantitative Finance, National Tsing-Hua University);國立台灣大學財務金融學系和國立台灣大學計量理論與應用研究中心 (Department of Finance, National Taiwan University, Center for Research in Econometric Theory and Applications (CRETA), National Taiwan University)
)
摘要
本文利用文字探勘的技巧,從美國聯邦公開市場委員會(Federal Open Market Committee, 簡稱 FOMC)所發表的正式官方文件:minutes of the FOMC 中萃取出重要的資訊,再利用情緒分析來判斷 FOMC 對三大經濟 「使命」 的正/負面看法,並建立指標, 以預測台灣的財經相關變數。相較於之前的文獻, 本文統一 FOMC 常用的專門術語,並且考慮了複合字的情況;這可以避免後續在計算文字字數時,產生重覆計算以及語意錯誤的情況。我們也利用統計模型 (MAP-PLSA)將每一份文件中的句子依據 FOMC 的經濟使命分成三個主題,再利用情緒分析技巧, 建立各類主題的指標數列。最後, 我們以迴歸模型來分析情緒指標與台灣相關財經變數之間的關聯性。
Abstract
In this paper we extract useful information from the minutes of the Federal Open Market Committee (FOMC) and examine how such information can help predict economic/financial variables. Based on the minutes during 1993--2016, we conduct sentiment analysis to determine the FOMC's attitude towards different topics, i.e., the mandates of the FED. Our approach is different from related studies in the following respects. First, we identify compound words which carry more specific meaning than do single words. Second, we adopt the topic model, MAP-PLSA, for estimating the conditional probabilities of these words/terms, which in turn can be used to classify sentences in the minutes under different topics. Third, the attitude towards each topic is determined by the ``tone'' of its sentences. We then proceed to evaluate whether the FOMC's attitude towards different topics can be used to improve economic forecasts.
關鍵詞
- 類型
- 論文
- 期刊
- 經濟論文叢刊
- ISSN
- 1018-3833
- 卷期
- 47卷3期
- 出版
- 2019年09月
- 頁碼
- 363–391
- 編號
- NTUEJ00006661
同期目錄
- 公開訊息對投票行為影響之探討吳震台(Cheng-Tai Wu)・頁 321–361
- 美國聯準會會議紀要的文字探勘與台灣經濟變數預測黃裕烈(Yu-Lieh Huang);管中閔(Chung-Ming Kuan)・頁 363–391
- 論家戶所得與資產對子女教育之影響 --- 以1993-1995 出生世代及其父母稅務資料為例沈暉智(Hui-Chih Shen);林明仁(Ming-Jen Lin)・頁 393–453
- 姓名筆劃數吉凶與運勢陶宏麟(Hung-Lin Tao)・頁 455–502