Toward Automatic Analysis of Financial Reports: Readability of Quarterly Reports and Companies' Financial Performance

Reading manually annual and quarterly reports helps financial analysts and decision makers to judge companies’ financial performance. Professionals infer from a report not only the current financial performance and strategic outlook of a company but also its prospective financial performance. They have insightful methods to find some intuitive indications of the future financial performance of a company from analyzing numbers from a report, reading its textual part and making professional guesses. The manual analysis of financial reports requires a lot of time, and time is a costly asset in a financial community. Automatically detected patterns by text mining methods characterize the writing style of the reports and can potentially provide analysts with the ground for professional guesses. One of the most attractive features of the successful report is its readability. Theoretically the readability of the reports is bond with company financial performance. This paper investigates the relationship between readability of financial reports and companies’ financial performance. We study the contribution of the readability toward the tone of a report. The patterns detected by one specific text mining method are characterized the tone of reports.