股票訂價折現模型-以台灣股票市場為例

Taiwan Stock Valuation With Time-Varying Discount Factor

曹家銘(Chia-Ming Tsao)

指導教授 ： 周雨田

國立交通大學/管理學院/經營管理研究所/碩士(2012年)

https://doi.org/10.6842/NCTU.2012.00135

摘要

有別於Chow [2] 在假定折現因子為常數的情況下，考慮了現金股利和現金股利成長率所建立的傳統股票訂價模型，在適應預期的基本假設下，本文嘗試將此非線性模型，推廣至一個更一般化的股票評價模型。本篇論文包含了四組模型，藉由將折現因子納入模型中，以及透過不同類型股票的時間序列資料，我們欲探討，現金股利、現金股利成長率、名目無風險利率以及市場風險溢酬是否對於台灣地區指數型基金和傳產類股之成份股的股票價格產生影響。本文除了使用指數型基金的成份股為樣本資料外，更首次將類股資料導入非線性的股利折現模型中。研究結果顯示：(1) 在我們選擇的三種指數型和八種傳產類股中，只有台灣高股利指數、食品類、機電類以及金融類股資料在我們使用的一般化股利折現模型中，符合適應預期的基本假設；(2) 由部分樣本資料不符合適應預期的研究結果，我們可以推測其個別投資人較不受過去財務訊息所影響；(3) 在我們所討論的成份股中，我們普遍發現其個別投資人對於股利成長率與股價的的關係存在著悲觀觀點，而這樣的觀點恰與Chow [29] 以香港恆生指數為樣本之研究結果不謀而合；(4) 針對傳統的八大類股而言，由於相似的股利政策和市場風險溢酬，使得僅有機電和建築類股的股票價格能被模型所解釋；(5) 我們進一步發現，在傳統的股利折現模型中，水泥窯業、食品以及機電類股之非限制方程式的係數具有相似的現象，儘管這些類股分屬不同產業，但這樣的結果隱含著這些類股的股價受到相似的因素所影響; (6) 透過統計檢定方法，我們發現名目無風險利率的期望值以及市場風險溢酬的期望值對於台灣股票價格具有顯著的影響，同時也說明了考量折現因子後的一般化股票評價模型更能捕捉台灣地區的股票價格。

關鍵字

適應預期 ； 非線性折現模型 ； 股票訂價模型 ； 折現因子 ； 市場風險溢酬

並列摘要

In contrast with the model of Chow [2], which implied that the logarithm stock price is a linear function of expected log dividends and the expected rate of growth of dividends under the assumption of the adaptive expectation, we have attempted to provide a general approach to estimation of models with stock price in this paper. This research includes four models designed to investigate how dividends, growth rate of dividends, nominal risk-free rates and risk premiums affect individual stock prices by using the different kinds of data for stocks. Following the theoretical framework of Chow [2], our researches use the individual stock of the stock market index as well as the individual stock of the eight major sectors as data in four models. The preliminary findings are: (1) Only the individual stock of TWSE Taiwan Dividends+ Index, Cement & Ceramics, Foods, Electric & Machinery, Construction and Finance sectors are consistent with the assumption of adaptive expectation. (2) The data which are not fit the adaptive expectations may suggest that the investors of these data do not take the historical information into consideration. (3) Furthermore, we discover that the coefficients α for Etdt are practically zero in the data, which are consistent with the adaptive expectations. Similar to the results of Chow [29], which used the Hang Seng Index, the empirical phenomena suggest that the overall pessimistic view of investors in these data. (4) For individual stock of the eight major sectors, merely the individual stock of the Electric & Machinery and Construction are consistent with the adaptive expectation hypothesis and can be explained by the expected level of log dividends. (5) We further discover that the unrestricted β coefficients are similar in the Cement & Ceramics, Foods, and Electric & Machinery sectors in model 1. This result indicates that behaviors in these sectors are identical. (6) According to the statistical test, we have strong evidence that the expected nominal free-risk rates and expected risk premiums have significant effect to contribute the current pricing. Besides, we find statistical evidence supporting the general model of stock price formation.

並列關鍵字

Adaptive Expectations ； Nonlinear Present-Value Model ； Stock Valuation Model ； Discount Factors ； Risk Premiums

參考文獻

[3] Todaro, M. P.” A Model of Labor Migration and Urban Unemployment in Less Developed Countries”, American Economic Review, Vol. 59, No. 1, pp. 138-148, 1969.

[4] Brennan, M. J. and Eduardo S. Schwartz. “Evaluating Natural Resource Investments”, Journal of Business, Vol. 58, No. 2, pp. 135-157, 1985.

[6] T.A. Lloyd, A.J. Rayner and CD. Orme. “Present-value models of land prices in England and Wales”, The European Review of Agricultural Economics , Vol.18,

pp. 141-166, 1991.

[7] Fairfield, P. M. ”P/E, P/B and the Present Value of Future Dividends”, Financial Analysts Journal, Vol. 50, No. 4, pp. 23-31, 1994.