Contents of Volume 18

Volume Contents

Contents of Volume 18     

Estimating unknown join points: Determination of the yen-dollar exchange rate

We propose a Bayesian procedure to estimate a switching regression in which the number of switching points (i.e. join points) is not known. We apply the Bayesian procedure to a regression model for the yen-dollar exchange rate using monthly data from January 1973 to June 1992. We identify three join points in January 1978, September 1988, and March 1990. We compare the post-sample forecast performances of our switching regression model to those of other regression models. The post-sample forecasts show that the Bayesian switching model performs better than the other models.     

The Effectiveness of Social Media for Business Activities in Japanese Firms

A large number of firms use social media for various business activities. This paper clarifies how social media has become effective to sales in Japanese firms. To this purpose, the following are addressed: (i) whether social media is useful for company activities, (ii) what kinds of business activities are affected using social media, and (iii) how social media improves the final outcomes of sales for firms. This study uses data of business activities and social media conducted in 2014. Probit analysis and structural equation modeling (SEM) are employed. Probit analysis clarifies the relationships between the effects of social media and its use in sales. The SEM model clarifies the relationships among the effects of social media use in business. In the SEM, a model is examined by taking sales caused by use of social media as an explained variable, while the explanatory variables are the first and intermediate effects of social media. The results indicate that social media is effective for business activities such as promoting sales. This study may help to promote social media in Japanese firms.     

Encoding of high-frequency order information and prediction of short-term stock price by deep learning

Predicting the price trends of stocks based on deep learning and high-frequency data has been studied intensively in recent years. Especially, the limit order book which describes supply-demand balance of a market is used as the feature of a neural network; however these methods do not utilize the properties of market orders. On the other hand, the order-encoding method of our prior work can take advantage of these properties. In this paper, we apply some types of convolutional neural network architectures to order-based features to predict the direction of mid-price trends. The results show that smoothing filters which we propose to employ rather than embedding features of orders improve accuracy. Furthermore, inspection of the embedding layer and investment simulation are conducted to demonstrate the practicality and effectiveness of our model.