Predicting Takeover Target and Analysis of Portfolio Performance

碩士 === 國立臺灣大學 === 財務金融學研究所 === 105 === This thesis studies the influence factors of takeover prediction and the potential to generate excess return based on the prediction model in Taiwan market. A takeover prediction model is developed using logit regression and a sample of 180 targets and 13,705 n...

Full description

Bibliographic Details
Main Authors: I Hsu, 徐懿
Other Authors: 邱顯比
Format: Others
Language:zh-TW
Published: 2016
Online Access:http://ndltd.ncl.edu.tw/handle/60078606461619651734
id ndltd-TW-105NTU05304009
record_format oai_dc
spelling ndltd-TW-105NTU053040092017-03-26T04:24:34Z http://ndltd.ncl.edu.tw/handle/60078606461619651734 Predicting Takeover Target and Analysis of Portfolio Performance 預測併購標的與投資組合績效分析 I Hsu 徐懿 碩士 國立臺灣大學 財務金融學研究所 105 This thesis studies the influence factors of takeover prediction and the potential to generate excess return based on the prediction model in Taiwan market. A takeover prediction model is developed using logit regression and a sample of 180 targets and 13,705 non-target observation from Taiwan listed companies from 2001 to 2014. Research period across 2003-2007 bull market and 2008 financial crisis, a completed economic cycle. The empirical analysis finds significant negative influence of ROE, earning before tax as well as positive relations of EBITDA, sales growth, book value, PE ratio, 3 month return on takeover probability. An out-of-sample test for 2014 found a predictive accuracy of 83.5% for the model, identifying 66.67% of target and 83.4% of non-targets correctly. The portfolio get return of 35.6%, higher than market portfolio with return of 8.1% in 2014. For clarifying the extension of the model, the thesis form a rolling test from 2006-2013 to test portfolio performance. The portfolio realized compound annual return 14.71% compared to market portfolio return of 3.91%. 邱顯比 2016 學位論文 ; thesis 27 zh-TW
collection NDLTD
language zh-TW
format Others
sources NDLTD
description 碩士 === 國立臺灣大學 === 財務金融學研究所 === 105 === This thesis studies the influence factors of takeover prediction and the potential to generate excess return based on the prediction model in Taiwan market. A takeover prediction model is developed using logit regression and a sample of 180 targets and 13,705 non-target observation from Taiwan listed companies from 2001 to 2014. Research period across 2003-2007 bull market and 2008 financial crisis, a completed economic cycle. The empirical analysis finds significant negative influence of ROE, earning before tax as well as positive relations of EBITDA, sales growth, book value, PE ratio, 3 month return on takeover probability. An out-of-sample test for 2014 found a predictive accuracy of 83.5% for the model, identifying 66.67% of target and 83.4% of non-targets correctly. The portfolio get return of 35.6%, higher than market portfolio with return of 8.1% in 2014. For clarifying the extension of the model, the thesis form a rolling test from 2006-2013 to test portfolio performance. The portfolio realized compound annual return 14.71% compared to market portfolio return of 3.91%.
author2 邱顯比
author_facet 邱顯比
I Hsu
徐懿
author I Hsu
徐懿
spellingShingle I Hsu
徐懿
Predicting Takeover Target and Analysis of Portfolio Performance
author_sort I Hsu
title Predicting Takeover Target and Analysis of Portfolio Performance
title_short Predicting Takeover Target and Analysis of Portfolio Performance
title_full Predicting Takeover Target and Analysis of Portfolio Performance
title_fullStr Predicting Takeover Target and Analysis of Portfolio Performance
title_full_unstemmed Predicting Takeover Target and Analysis of Portfolio Performance
title_sort predicting takeover target and analysis of portfolio performance
publishDate 2016
url http://ndltd.ncl.edu.tw/handle/60078606461619651734
work_keys_str_mv AT ihsu predictingtakeovertargetandanalysisofportfolioperformance
AT xúyì predictingtakeovertargetandanalysisofportfolioperformance
AT ihsu yùcèbìnggòubiāodeyǔtóuzīzǔhéjīxiàofēnxī
AT xúyì yùcèbìnggòubiāodeyǔtóuzīzǔhéjīxiàofēnxī
_version_ 1718435380251852800