| In recent years,artificial intelligence technology and related industries have been developing rapidly,the value of artificial intelligence applications has become increasingly prominent,and new products and services are gradually moving from the concept to the commercialisation.Based on the existing research on Ohlson model,this paper establishes an improved valuation model,taking into account the characteristics of the AI industry and its enterprise value,in order to provide AI enterprises themselves and capital market investors with a more accurate reference basis for value determination.Firstly,based on an introduction to the AI industry and enterprises,the influencing factors,characteristics and key points of enterprise value from a value perspective are analysed to provide ideas for the selection of subsequent valuation methods and model improvements.Secondly,the applicability of several traditional valuation methods in the valuation work of AI enterprises is analysed,the Ohlson model is introduced according to the shortcomings of the existing methods,and the Ohlson model is compared and analysed with the traditional valuation methods.Again,starting from the "other information" item in Ohlson model,we combine the characteristics of enterprise value and value influencing factors to find "other information" that is more appropriate and can reasonably reflect the value of AI enterprises.By changing the calculation method of residual income and selecting alternative variables that can quantify "other information",the difficult-to-predict duration of future earnings and the discounted value of residual income are transformed into calculations using simple and available historical financial statement data,which are then introduced as panel data for hypothesis testing and empirical analysis.The ’other information’ item in the Ohlson model formula is then specified and refined.As a result,the calculation process of the original model was improved,resulting in three improved Ohlson models that were tested by factor analysis and linear dynamic information hypothesis,and were more suitable for the valuation of AI companies.Afterwards,Hikvision,a representative company in the field of artificial intelligence,was selected as the case company to study the applicability of the improved Ohlson models in the valuation of the case company through specific cases,to clarify the basic matters of asset valuation,to assess the value of the case company using the three improved Ohlson models respectively,and to compare the valuation results calculated using the improved models with the total market value as at the valuation benchmark date.A comparison was made to analyse the differences and reasons for the valuation results,reflect the applicability and advantages of the improved Ohlson model,and provide a new and feasible idea for the valuation of AI enterprises.Ultimately,the paper draws the following research conclusions:1.In the case where it is difficult to reasonably predict the future earnings period and residual earnings of an enterprise,the use of the original Ohlson model for the valuation of AI enterprises may result in a significant underestimation or significant overestimation of the future residual earnings of the enterprise,making it difficult to use the original Ohlson model directly in the actual valuation work.By choosing specific parameters to replace the "other information" items in the original model according to the value characteristics of AI enterprises,the errors in the calculation process can be reduced,making the calculation results more refined and making the improved Ohlson model more applicable in the valuation of AI enterprises.2.The improved Ohlson model simplifies the calculation of residual income,which can be calculated directly using the historical period financial data of enterprises,making the calculation process of the improved model more objective and reducing the subjectivity in the prediction process of residual income and its discounted value calculation process.Using the financial data of 233 A-share listed companies in the field of artificial intelligence from 2017 to 2021 as panel data,the linear dynamic information hypothesis under each improved approach was tested,and on this basis,separate valuation models were constructed for empirical analysis,and based on the results of the hypothesis testing and empirical analysis,the coefficients of the variables in the improved models were obtained,resulting in an improved Ohlson valuation model applicable to artificial intelligence companies.The model was finally developed.3.In the valuation study of the case company,the valuation result of the original Ohlson model before the improvement was CNY124,330.2635 million.The improvement to the Ohlson model,when replacing the ’other information’ item with R&D expenditure,resulted in an improved residual income of CNY586,247.3578 million using the Du Pont analysis metric and CNY632,981.5537 million using the net profit metric,which took into account These improvements take into account the role played by R&D expenditure in the value formation process of AI companies;when the weighted value of R&D expenditure and intangible assets is used to replace the "other information" item,the result of improving residual income using the Du Pont analysis indicator is CNY437,248.4773 million,which takes into account the role played by R&D expenditure and intangible assets in the value formation process of AI companies.Overall,the improved Ohlson model has improved the potential underestimation or overestimation of enterprise value and residual income in the original model,reduced the subjectivity of the valuation process and improved the accuracy of the valuation results. |
