abstract: Stock market news and investing tips are popular topics in Twitter. In this dissertation, first I utilize a 5-year financial news corpus comprising over 50,000 articles collected from the NASDAQ website matching the 30 stock symbols in Dow Jones Index (DJI) to train a directional stock price prediction system based on news content. Next, I proceed to show that information in articles indicated by breaking Tweet volumes leads to a statistically significant boost in the hourly directional prediction accuracies for the DJI stock prices mentioned in these articles. Secondly, I show that using document-level sentiment extraction does not yield a statistically significant boost in the directional predictive accuracies in the presence of other 1-gram keyword features. Thirdly I test the performance of the system on several time-frames and identify the 4 hour time-frame for both the price charts and for Tweet breakout detection as the best time-frame combination. Finally, I develop a set of price momentum based trade exit rules to cut losing trades early and to allow the winning trades run longer. I show that the Tweet volume breakout based trading system with the price momentum based exit rules not only improves the winning accuracy and the return on investment, but it also lowers the maximum drawdown and achieves the highest overall return over maximum drawdown. / Dissertation/Thesis / Doctoral Dissertation Computer Science 2016

京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第22574号 / 情博第711号 / 新制||情||122(附属図書館) / 京都大学大学院情報学研究科社会情報学専攻 / (主査)教授 吉川 正俊, 教授 鹿島 久嗣, 教授 田島 敬史, 特定准教授 JATOWT Adam Wladyslaw / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DGAM

This thesis examines the effect of coronavirus-related sentiment on Swedish stock market returns during the coronavirus pandemic. We study returns on the large cap and small cap price indices OMXSLCPI and OMXSSCPI during the period January 2, 2020 – April 30, 2020. Coronavirus sentiment proxies are constructed from news articles clustered into topics using latent Dirichlet allocation and scored through sentiment analysis. The impact of the sentiment proxies on the stock indices is then measured using a dynamic multiple regression model. The results show that the proxies representing fundamental changes in our model — Swedish Politics and Economic Policy — have a strongly significant impact on the returns of both indices, which is consistent with financial theory. We also find that sentiment proxies Sport and Coronavirus Spread are statistically significant and impact Swedish stock prices. This implies that coronavirus-related news influenced market sentiment in Sweden during the research period and could be exploited to uncover arbitrage. Finally, the amount of sentiment-inducing news published daily is shown to have an impact on stock price volatility. / Denna studie undersöker den effekt coronavirus-relaterat sentiment haft på avkastningen på svenska aktieindex under coronaviruspandemin. Vi studerar avkastningen på large cap- och small cap-prisindexen OMXSLCPI och OMXSSCPI under perioden 2 januari 2020 – 30 april 2020. Proxier för coronavirus-sentiment konstrueras från nyhetsartiklar som klustrats i ämnen genom latent Dirichlet-allokering och poängsatts genom sentimentanalys. Sentimentproxiernas påverkan på aktieindexen mäts sedan med en dynamisk multipel regressionsmodell. Resultaten visar att proxierna som representerar fundamentala förändringar i vår modell — svensk politik och ekonomisk policy — har en starkt signifikant inverkan på avkastningen på båda indexen, vilket är konsekvent med finansiell teori. Vi finner även att sentimentproxierna sport och spridning av coronaviruset är statistiskt signifikanta i sin påverkan på svenska aktiepriser. Detta innebär att coronavirus-relaterade nyheter påverkade marknadssentiment i Sverige under undersökningsperioden och skulle kunna användas för att upptäcka arbitrage. Slutligen visas mängden sentimentframkallande nyheter publicerade per dag ha en inverkan på aktieprisvolatilitet.