원문정보
A Study on the Forecasting Model of Price-Earnings Ratios Based on Neuro Network
초록
영어
This paper intends to study forecasting models of Price-Earnings Ratios(PER) by using accounting information in the Korean stock market. The use of forecasting models of relevant Accounting Information Systems(AIS) is helpful to change the investment method used the financial statements, core of accounting informations. This paper tried to represent that the investment method using the forecasting models of AIS would be helpful to invest. This study uses the Holthausen and Larcker(1991) methodology. Their study supports that the investment strategy forecasting PER directly and composing the portfolio will bring forth better results than those of Ou and Penman forecasting the accounting income first and composing the portfolio on the basis of that information. Therefore this paper analyzes the differences among the results of models applied various methodology to investment strategy of Holthausen and Larcker. This paper uses Artificial Neural Networks(ANN) model. The data examined span 14 years from 1983 to 1996. 58 fundamental variables of analysis are used from financial statements in Korean manufacturing companies. The result indicates that ANN model in forecasting accuracy is about 58%.
목차
II. 이론적 배경 및 선행연구
2.1 이론적 배경
2.2 선행연구
III. 연구설계
3.1 자료의 구성
3.2 변수의 측정
3.3 변수의 선정
3.4 모형개발기법
IV. 실증분석 및 결과
4.1 인공신경망 모형의 예측결과
4.2 인공신경망 모형의 예측정확도
V. 결론
참고문헌
Abstract