Date of Original Version
The purpose of this study is to examine returns on Japanese equities over nearly a four-decade period and to compare results among the four decade and the entire period of the study. “Long memory” modeling of time series developed to predict slowly moving time series is a method to predict long time components of time series data. Previous, other studies indicated some progress in producing results of predictability by these “long memory” analyses. The authors examined statistically for some of the reasons why long memory forecasting may not be very suitable for predicting equity returns over lengthy periods of time. Data secured from a source that collect information on Japanese equity returns, enabled a study of possible explanations of why lengthy predictions are difficult. The analysis is of an application to financial time series and does not dispute the use of long memory modeling in other applications. The conclusions made are therefore not universal but only to the use in financial engineering and time series analysis. Future work should consider the cost effectiveness of long-memory modeling in other forms of financial time series analysis.
Jeffrey E. Jarrett and Yifei Li. 2018. "ANALYSIS OF EQUITY RETURNS IN THE JAPANESE FINANCIAL MARKET; TIME SERIES METHODS." International Journal of Business Management and Economic Review 1, no. 6: 183-202.
Available at: http://ijbmer.org/link.php?id=73
Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.