Market Commentaries and Stock Prices in Poland: A Text Mining Approach
Market Commentaries and Stock Prices in Poland: A Text Mining Approach

Summary/Abstract: From a theoretical point of view, the scope and quality of available information determines the market efficiency and, thus, investors’ decisions. However, an excessive amount of information leads to information overload. In the case of textual data, advanced analytical methods must be applied to identify some regularities and trends within the analysed text corpora. Text mining may be useful in supporting the decision-making process. The paper examines the interdependencies between market commentaries and stock prices. More specifically, it verifies the linguistic characteristics of opinions distributed by institutional investors (investment fund company) and their intertemporal links to the price movements on the Warsaw Stock Exchange. The results indicate that: 1) there is no significant linguistic difference between market commentaries written after weeks of relatively low and relatively high rates of returns on the Warsaw Stock Exchange; 2) the linguistic content of selected market commentaries does not have a predictive value for the Polish stock market; 3) commentaries with a one-week time difference linguistically differ less than the commentaries with two or more weeks’ time difference.

• Details
• Contents