This study uses Shikiho data and ChatGPT to assess the impact of stale, factual information on stock prices, revealing potential market inefficiencies as stock performance is benchmarked against risk factors and suggesting a need for further exploration of public information in price discovery.

This study uses a Convolutional Neural Network to analyze Japanese stock market charts, finding that two-dimensional historical data can predict future returns, revealing unique market characteristics beyond common momentum effects.

This research investigates the forecasting performance of financial professionals and university students across artificial lab-based tasks and real-world field tasks, revealing nuanced differences in predictive abilities and highlighting the importance of task context in evaluating financial forecasting expertise.

This study links investor sentiment in news text to the "sell in May and go away" calendar anomaly. By analyzing a large dataset of newspaper articles from 1986 to 2010, the researchers identify a pattern of optimism in the first half of the year and pessimism in the latter half, suggesting a synchronization with market seasonality.

This study investigates the impact of analysts' signals disseminated from the Bloomberg Terminal on short-term market reactions in Japan from 2000 to 2010. Findings suggest a significant market response to rating changes rather than the ratings themselves. Additionally, while investors can capture some abnormal profits by trading on these signals, transaction costs limit the feasibility of large-scale investments based on this strategy.

This study analyzes the impact of 2016 guidelines by the Japan Securities Dealers Association, banning sell-side analysts from accessing earnings previews before official release. Results indicate heightened stock return volatility post-guideline, signaling improved information dissemination, with analysts' earnings forecast accuracy remaining unchanged, likely due to reliance on alternative information sources.