| In recent years,with the rise of "Internet plus","great health",the new energy,as well as the mergers and acquisitions wave of national economic restructuring and transformation and upgrading of traditional industries,the driving force which is called private equity gradually came into the public view.Valuation is the core of the management,and the valuation of the PE investment institutions is the basis of everything.But despite the fact that the current domestic theorists valuation for research are numerous,there are very few researches introduce the assessment practice in PE value method which is combined with the features of PE firms.This paper tries to fill this gap.It applies the principal component statistical analysis and data mining technology in the K-nearest neighbor with kernel classification algorithm to the relative valuation method in an effort to pick out the companies which have similar cash flow,growth potential and risk connotation with the valued company.And the purpose of this method is to make the results more objective and convincing.What's more,the comparative analysis,empirical analysis and inductive combined with cases are inevitable in this paper.In the empirical analysis of the cases,we first make use of the first principal component analysis to reduce the metrics dimensionality and remove the multicollinearity,and calculate the valuation metric weights,and then use K-nearest neighbor with kernel to select the most similar company from the comparable companies in the sample.Finally we get the valuation results.In the first two cases,the results of the improved relative valuation model are closer to the final transaction price than the results obtained from the classical theory;And in the last case,the improved relative valuation model has a smaller mean squared error.In summary,the improved relative valuation model makes a certain contribution to PE valuation in both theoretical and practical aspect. |
PDF Full Text Request