Please use this identifier to cite or link to this item:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/72840
Title: | 使用基於彙總數據的市場份額估算進行股票投資 Stock Investing Using Market-share Estimation Based on Aggregate Data |
Authors: | Wen-Yi Luo 羅文義 |
Advisor: | 楊曙榮(Shu-Jung Yang) |
Keyword: | 隨機係數,邏輯式回歸,彙總數據,馬可夫鏈蒙地卡羅,貝式分析,股票市場, random coefficients,logit model,aggregate data,MCMC,Bayesian analysis,stock market, |
Publication Year : | 2019 |
Degree: | 碩士 |
Abstract: | 本研究運用BLP架構,探討投資人在購買股票時,基本面、技術面以及籌碼面的變化對於投資者在台灣股票市場的偏好影響程度。我們使用了一種貝氏估計方法,用於分析台灣證券交易所的不同股票的總體水平歷史交易數據。該貝氏分析方法基於Jiang,Manchanda and Rossi (2009)中所提出的演算法和具有隨機係數的彙總數據的多項logit模型 - 所謂的BLP模型,係以Berry,Levinsohn and Pakes (1995)命名。本研究主要關注元大投信所發行的元大台灣50ETF的成分股,以估算投資者的效用。研究之模型結果顯示投資週期為季度的投資者在做投資決策時,主要係依據收盤價,投信持股比例和股東權益報酬率。利用此投資決策做出歷史回測,並進一步加入停損機制,計算得到年化報酬率在2017Q1至2019Q1期間約為26.7%。 This study aimed to discuss the utility of investors when investors buy stocks in the stock market under the BLP framework. We applied a Bayesian approach to analyze different stocks based on the aggregate level historical trading data in the Taiwan Stock Exchange. This Bayesian analysis method was based on the algorithm in Jiang, Manchanda and Rossi (2009) and the multinomial logit model for aggregated data with random coefficients - the so-called BLP model, named for Berry, Levinsohn and Pakes (1995). This study mainly focused on the constituent stocks in Yuanta / P-shares Taiwan Top 50 ETF (0050.TW) and estimated utility for investors. The result indicates that the quarterly investors contemplated the closing price, SITE shareholding ratio, and ROE when making investment decision. The researcher further implemented a stop-loss strategy in the investment method and the annualized rate of return during 2017Q1 to 2019Q1 was about 26.7%. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/72840 |
DOI: | 10.6342/NTU201901777 |
Fulltext Rights: | 有償授權 |
Appears in Collections: | 商學研究所 |
Files in This Item:
File | Size | Format | |
---|---|---|---|
ntu-108-1.pdf Restricted Access | 1.63 MB | Adobe PDF |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.