台灣與中國大陸營建業之企業估值模型的實證研究比較

Comparison of empirical study on enterprise valuation model for construction industry in Taiwan and Mainland China

楊明倫(Ming-Lun Yang)

指導教授 ： 葉怡成

淡江大學/工學院/土木工程學系碩士班/碩士(2020年)

https://doi.org/10.6846/TKU.2020.00626

摘要

論文提要內容：
成長價值模型認為企業價值可用公式 P=k∙B∙(1+ROE)^m估計，其中P=企業價值，B=淨值，ROE=股東權益報酬率，k=價值係數，m=成長係數，二者為待定係數。本文提出四個方法來估計上述係數，並以台灣股市(2008~2018年)與大陸股市(2009~2019年)為研究範圍。
研究結果顯示：
(1) 兩岸營建產業在成長係數m上相較其他產業均屬偏小。
(2) 兩岸市場之價值係數k均與規模成反比；成長係數ｍ與規模成正比。
(3) 台灣市場之價值係數k及成長係數m均與風險大小無明顯關係。
大陸市場則可發現價值係數k與風險成正比；而成長係數m與風險成反比。
(4) 台灣市場之價值係數k的變動隨著時間的變動相對平穩，大陸市場的變動較大，且價值係數k逐年漸減，成長係數m逐年上升，均有越來越接近台灣市場水準的現象。
此外，四種係數估計方法中：
(1) 股價之對數散布圖估計法誤差太大。
(2) 股東權益報酬率與股價淨值之對數散布圖估計法的k值估計偏大， m值估計偏小。
(3) 股東權益報酬率與股價淨值比之變數誤差模型法的k值估計偏小， m值估計偏大，但與前一方法接近，且較接近理論值，因此是最好的方法。
(4) 股東權益報酬率與股價淨值比之傳統迴歸分析法的k值估計大幅偏大，m值估計大幅偏小，但和各控制變數的關係與上述兩個方法相近。

關鍵字

企業價值 ； 成長價值模型 ； 變量誤差模型 ； 均值回歸

並列摘要

The Growth Value Model (GVM) considers that the enterprise value can be estimated by the formula P=k∙B∙(1+ROE)^m, where P=enterprise value, B= book value , ROE= Return On Equity, k=value coefficient, and m=growth coefficient. Both k and m are undetermined coefficients. This thesis proposes four methods to estimate the above coefficients, and takes the Taiwan stock market (2008~2018) and the Mainland China stock market (2009~2019) as the research scope.
The study revealed that (1) The growth coefficient m of the construction industry in Taiwan and Mainland China is relatively small compared with other industries. (2) The value coefficient k of the market on Taiwan and Mainland China is inversely proportional to the market value; while the growth coefficient m is directly proportional to the market value. (3) The value coefficient k and the growth coefficient m of the Taiwan market have no obvious relationship with the magnitude of risk; while in the Mainland China market, it can be found that the value coefficient k is proportional to risk; the growth coefficient m is inversely proportional to risk. (4) The change in the value coefficient k of the Taiwan market has been relatively stable over time; The changes in the Mainland China market have been relatively large, and the value coefficient k has gradually decreased year by year, and the growth coefficient m has increased year by year, all of which are closer and closer to the level of Taiwan market. In addition, among the four coefficient estimation methods, it can be concluded that (1) the estimation method of logarithmic scatter diagram of stock prices has rather large estimate error. (2) The estimation method of logarithmic scatter diagram of ROE and P/B ratio overestimates k and underestimates m. (3) The Errors-in-Variables model underestimates k and overestimates m, but they are close to the previous method and closer to the theoretical value, so it is the best estimate method. (4) The traditional regression analysis much overestimates k and underestimates m, but the relationship with each controlled variable is similar to the above two methods.

並列關鍵字

Enterprise valuation ； Growth Value Model ； Errors-in-Variables model ； mean reversion

參考文獻

參考文獻

1.Adsera, X., & Vinolas, P. (2003). FEVA: A financial and economic approach to valuation, Financial Analysts Journal, 59(2), 80-87.

2.CLAconnect (2011). Key Drivers of Value in Construction Companies, CLAconnect. https://www.claconnect.com/resources/articles/key-drivers-of-value-in-construction-companies

3.Chas. H. Kummell. (1879). Reduction of observation equations which contain more than one observed quantity. The Analyst, 6(4), 97-105.

4.Chen, A. S., & Lin, S. C. (2011). Asymmetrical return on equity mean reversion and catering, Journal of Banking & Finance, 35(2), 471-477.