請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/74893
標題: | 預測臺灣股票市場熊市之相關變數探討 Factors predicting the bear stock market in Taiwan |
作者: | Po-Yi Li 李柏毅 |
指導教授: | 陳旭昇(Shiu-Sheng Chen) |
關鍵字: | 熊市,Probit 模型,Bry-Boschan 法則,投資人情緒,總體經濟變數,財務變數, Bear markets,Probit model,Bry-Boschan algorithm,Investor sentiment,Macroeconomic variables,Financial variables, |
出版年 : | 2019 |
學位: | 碩士 |
摘要: | 本文主要探討預測臺灣股票市場熊市 (bear market) 之相關變數,本研究除參考之前文獻實證上所採用的總體經濟變數及財務變數外,考量近期實證研究顯示股票報酬也可能受投資人情緒(investor sentiment)影響,本研究特別納入投資人情緒變數進行研究,希望透過諸多變數探討,
全面性瞭解影響臺股熊市的因素。本研究方法係使用 Bry-Boschan 法則定義熊市,並搭配 Probit 模型進行實證預測,透過樣本內與樣本外的預測進行實證分析,研究結果顯示總體經濟變數 : 期間利差、 通貨膨脹率及 M1b 成長率對於未來短期 (1 個月及 3 個月) 熊市具有顯著預測能力;財務變數 : 股市殖利率及股市報酬變動率具有較佳的預測能力;另投資人情緒變數 : 消費者投資股票信心變動率及波動率指數 (VIX)對於未來短期熊市具顯著預測能力,而賣買權比 (Put/Call Ratio) 對於未來長期 (12 個月及 24 個月) 熊市具有顯著預測能力。 This study investigates predictors for the occurrence of bear stock market in Taiwan. In addition to the macroeconomic and financial variables that have been discussed in empirical literature, this study referenced recent empirical studies that have verified the effect of investor sentiment on stock returns. Therefore, investor sentiment is included as a variable to comprehensively examine factors that may lead to a bear market. In terms of methodology, this study applies the Bry-Boschan algorithm for the definition of bear market and a probit model for empirical prediction. Through the empirical analysis of in-sample and out-of-sample testing results, this study concludes that macroeconomic variables including term spread, inflation rate, M1b growth rate exhibit significant predictive power for short-term (1 month and 3 months) bear markets; financial variables including dividend yield and the variance of stcok returns exhibit favorable predictive power; and investor sentiment variables including change rate in consumer confidence on stock investment and volatility index exhibit significant predictive power for short-term bear markets, while put-call ratio exhibits significant predictive power for long-term (12 months and 24 months) bear markets. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/74893 |
DOI: | 10.6342/NTU201904190 |
全文授權: | 有償授權 |
顯示於系所單位: | 經濟學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-108-1.pdf 目前未授權公開取用 | 1.2 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。