A Study On The Application Of Neural Networks To Income Method For Parameter Determination

With the economic globalization accelerating,economic activities have become increasingly frequent.The assessment of enterprise value will be used in the field of stock transfers,asset restructures,mortgage,stock issues,therefore it plays an increasingly important role in business management and investment decisions.No doubt that the correct scientific evaluation of the enterprise value is very urgent.Income approach is one of the common business valuation methods,which brings the time value of capital into it,can reflect the potential profitability objectively with more reasonable costing;compared with the market approach,it is more persuasive.The income approach is the essence of the company's future benefits,using the discount rate converted to the valuation point of time.Therefore,business valuation method depends on the implementation of revenue earnings years,the discount rate and the amount of revenue.The method of determining the discount rate includes the capital asset pricing model(CAPM),the weighted average cost of capital model(WACC),the risk of cumulative method and so on;but for different companies and different evaluation purposes,these methods also have some disadvantages;it cannot completely apply to all enterprises.Research on methods of determining the amount of revenue doesn't have unified approach;a method to predict they usually use is the subsection prediction method which is mainly based on historical data in financial statements to forecast the amount of revenue,but it ignores other factors.In order to make the income approach more objectively reflect the enterprise value,how can we scientifically determine the amount of revenue and the discount rate is especially important.This paper analyzes the method and its applicability in the profit amount and discount rates;combining with the characteristics of neural network the scope of application,this paper puts forward BP neural network proposed to simulate the discount rate,the amount of earnings forecast thought.BP neural network has strong learning ability,which is widely used in simulating and forecasting complex problems.When we use the BP neural network simulation discount rate,we can fully analyze the impact factors of the discount rate,overcome the deficiencies of the prior determination method;because of considering their own situation and the objective economic environment affects on the amount of a company's future earnings,the amount of revenue forecasts become more scientific,rational and convincing when we use BP neural network to predict the amount ofincome.We based on XX enerrh company equity transfer to explain it in detail about how to use the BP neural network model in simulating the discount rate and forecasting the earnings,using the discount rate simulation and revenue forecast value based on BP neural network to assess the company's value.The results showed that the evaluation of enterprise's value by using this method gets closer to the actual transaction price.This method can cover the shortage of existing methods,thus have a higher accuracy.Of course there are still some shortcomings,but I hope this research will provide a new way to determine the income method parameters.