Estimating the distribution of enterprise values with quantile neural networks

doi:10.1007/s00500-020-04726-w

機構典藏 > College of Engineering > Graduate Institute & Department of Civil Engineering > Journal Article > Item 987654321/119015

Please use this identifier to cite or link to this item: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/119015

Title:	Estimating the distribution of enterprise values with quantile neural networks
Authors:	Yeh, I-cheng
Keywords:	Quantile regression analysis;neural networks;enterprise value;distribution
Date:	2020-01-27
Issue Date:	2020-09-15 12:10:27 (UTC+8)
Publisher:	Springer
Abstract:	The probability density function of enterprise values may be more precise and useful in the cases of corporate investment, financing, or transactions. Although the quantile regression analysis can generate a set of models for a series of quantiles, it cannot generate the probability density function of the dependent variable. Therefore, this paper proposes a novel method of employing prediction results of the quantile neural networks to build probability density functions with which we can effectively assess enterprise values. Empirical evidence reveals that the estimated cumulative lognormal distribution curves of the Price-to-Book value ratio (PBR) and the data are well-matched. In addition, the corporate market value is equal to the PBR multiplied by the corporate stockholders equity. Thus, the corporate market value is also a lognormal distribution. PBR distributions of building and construction industries are more tilted to the left, implying that enterprise values of building and construction industries are lower than those of other industries with the same stockholders equity and return on equity.
Relation:	Soft Computing 24,　p.13085-13097
DOI:	10.1007/s00500-020-04726-w
Appears in Collections:	[Graduate Institute & Department of Civil Engineering] Journal Article

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	105	View/Open

Loading...