Studies Of Earning Prediction Model Based On Earning Persistence

Accounting earning which can reflect a company’s operating results over a periodof time is the specific form of the company’s capital appreciation, and it plays animportant role in the performance evaluation and analysis about the company’s value.From the perspective of continuing operations, accounting earnings help investors topredict the company’s future performance and get the basic information to evaluate thecompany’s future cash flows. Meanwhile the outside investors can predict the furtherearning information based on the current earnings. According to the economic theory,the high persistence of accounting earnings will stabilize the optimal allocation of themarket resources and guide investors to the long-term investment, which will alsoreduce the speculative volatility of the element market. In view of the importance ofearning persistence to forecast for future-period earnings, this paper attempts to createan earning prediction model based on the earning persistence angle and help the listedcompany stakeholders analysis and forecast future earning situation of the listedcompanies.This paper is on the basis of studying the relevant literature of earning persistenceand earning prediction, and design earning volatility coefficient and earning changesfactor based on the earning persistence according to the conclusions of the study therandom walk model. Then add earning volatility coefficient and earning changes factorinto random walk model to establish this paper’s earning prediction model. In this paper,Shanghai A-share listed companies manufacturing are chose as the study samples andthe basic earnings per share is selected as the characterization of the accountingvariables. Firstly, this paper reviews the studies which are did on the history of earningspersistence by the domestic and foreign scholars; Secondly, this paper elaborates therelated concept of earning persistence, related accounting theory of earning informationand affecting factors of earnings prediction etc. which lay the foundation forestablishment of later earning prediction model. Thirdly, the process of establishingearning prediction model is started, the volatility of earning by itself and the changes inoverall level of the impact from previous earning on the future earnings are consideredrespectively, and earning volatility coefficient and earning changes factor are designedand added to the based prediction model, and the final prediction model of this paper iscompleted. By the end of this paper, Guodian Nanjing Automation Co. Ltd. is selectedas a case to test the feasibility and practicality of the earning prediction model.