Study On Investment Evaluation Of The Infrastructure Project Based On Modified Real Option

In order to overcome the weakness of real option investment evaluation methods in infrastructure project, such as the lack of pricing information and inaccurate result, this paper attempts to apply the neural network into investment evaluation of infrastructure project. Correcting the result from the real option model, neural network is able to estimate the infrastructure project value more accurately and can also improve the evaluation means of infrastructure investment project.The investment evaluation and decision-making method of infrastructure is mainly discussed. Firstly, the Real Option characteristics of infrastructure project were analyzed according to, huge flexibility and uncertainty of highway project. Based on the theoretical and practical experience, the shortcomings of the Real Options models such as lack of pricing information, inaccurate result are discussed. Then based on the discussion above, an Real Option model based on Artificial Neural Network is built up and the steps for using such a new model is demonstrated. Thirdly, concerning the value from the investment decision maker facing the uncertainty, we use the B-S pricing model to make quantitative analysis and sensitivity analysis of the investment opportunity. After that, we use the Artificial Neural Network to optimize the Real Option model and realize it by MATLAB. At last, we test the model by a case study of JS highway project. The result of the case study shows that the modified Real Option model can compensate the shortcomings of the Real Option Method, increase the accuracy of the investment evaluation of infrastructure project and give a truer picture of the infrastructure project evaluated. The new method is very meaningful for reference for investment evaluation of infrastructures.