Developing a Method for Building Forward Multipliers of Companies Using Machine Learning

Abstract

In conditions of high volatility of commodity markets and the growing role of behavioral factors, traditional methods of comparative approach to business assessment are losing accuracy and predictive stability. This problem is particularly acute when using cost multipliers that do not reflect investors’ expectations and the future dynamics of companies’ financial performance. The article proposes a method for constructing forward multipliers based on the integration of machine learning and Monte Carlo simulation. The method allows you to separately predict the components of the multiplier – the market price of shares and net profit, taking into account fundamental, industry and behavioral cost factors. Special attention is paid to the problem of the lack of forecast values for exogenous variables, which is solved using scenario modeling. The method was tested on data from a public company in the oil and gas sector, which made it possible to identify structural differences in the factors determining price and profit. The results obtained demonstrate the high predictive ability of the models and confirm the expediency of switching from historical to forward multipliers in the practice of valuation. The developed approach expands the tools of the comparative approach and can be used in investment analysis and business valuation in conditions of uncertainty.