We investigate the return predictability in news using a large sample of electronic news article diseminated from the Bloomberg. Using Naive-Bayes classifier, we create models that predict return of the Nikkei 225 futures. Our test reveals two models that beat the market by substantial margin. Based on the two selected models, we conduct out of sample money management simulation for the period between 2001 to 2011 and achieve 11.3% return per annum with the Sharpe ratio of 0.693. The result is better than the top 5% of the 1000 random models.
One is numerical information, including past stock prices, currency exchange rates, and interest rates. The other is textual information, mainly news stories covering statements of government dignitaries, consumer trends, and miscellaneous events. Although numerical information has been proven useful for predicting stock prices, its predictive power is limited in a sense that much information, such as the statements of government dignitaries mentioned above, resides only in textual information. Given a stock for which one would like to predict the future price, textual data provides different?but much wider coverage?of information from numerical information, which may be beneficial in prediction. This study exploits public Web news articles and attempts to estimate the residue that cannot be explained only by numerical information through a simple additive regression model. In addition, to distinguish between different types of news articles, such as those specific to particular companies or types of industry and more general top news, the framework of multiple kernel learning is adopted. The validity and effectiveness of the proposed approach is evaluated on the real-world data consisting of share prices of Nikkei 220 companies and 47 thousand Web news articles.
The effect of option markets on their underlying market has been studied intensively since the first option contract was listed. Despite considerable effort including theoretical and empirical approaches, we still don't have a conclusive view toward that problem. We look the effect of option market, especially the effect of dynamic hedging, on underlying market by using artificial market. We proposed two market model in which option market and underlying market interacts. In this study, we could confim that dynamic heding increase or decrease the volatiliy of underlying market under some conditions.
The purpose of this research is to discuss some investment risk management issues for unexpected fluctuations such as "unintended consequences" in markets. "Unintended consequences" are illustrated as phenomena based on mutual-interaction between the macro and micro level, known as macro-micro linkage. From this perspective, I am interested in the changes of risk model methodology after the financial crisis and then tighter restriction by regulator in financial industry. I am focusing on how to generate better risk estimation based on the statistical model under the current circumstances.
This study presents an ANWSER model (asset network systemic risk model) to compute a bankruptcy reproductive ratio in analyzing the risk of bank failures in a financial crisis. We present the bankruptcy reproductive ratio as a function of the average number of interbank loans, the share of big banks, the risk exposure and diversity of investments which banks make.
A method to build trading rules at algorithmic trading by Genetic Algorithm (GA) is presented. The trading rules are built by adapting parameters of technical indices. New fitness functions in GA, which aim at a robuster behavior to change of market trends, are proposed. Furthermore, We have found that the effectiveness of technical indices varies depending on market trend. We propose a hybrid dealing method which switches the technical indices by market trend. The result of experiment shows that profits through dealing are improved by our proposed method.