Trends in the value-relevance of quarterly earnings announcements over 1972-2015

No Thumbnail Available
Journal Title
Journal ISSN
Volume Title
School of Business | Master's thesis
Ask about the availability of the thesis by sending email to the Aalto University Learning Centre
Degree programme
In this thesis I study the trends in the value-relevance of quarterly earnings announcements covering years 1972-2015. The topic has been of interest to many researchers and has maintained its attractiveness since the value-relevance has changed over the years and the reasons for it remain somewhat unambiguous. The motivation to a topic this actively researched strives from the fundamental changes in the business models of the companies included in the sample. The portion of technology and service business companies has increased rapidly during the latest years, and in contrast to traditional manufacturing companies, their financial reporting and accounting rules differ quite a lot. This causes differences in the value-relevance of earnings announcements. In the previous research, the timelines used do not cover the latest 5-10 years, indicating that the effect of these newly listed technology and service centered companies does not reflect to the results. As a results, prior research is lacking a consistent understanding on the effects of the latest years of the data available from Compustat and The Center of Research in Security Prices (CRSP) data base. The dataset used in this study extends from January 1972 to December 2015 and comprises firms listed in NYSE, NASDAQ and AMEX. The quarterly data for firms is retrieved from Compustat and the daily return data is from CRSP. Hypotheses are formulated based on prior research and they are formulated to each income statement line item used to measure the value-relevance. These income line items are net income (NIQ), operating income before depreciations (OIBDPQ) and operating income before extraordinary items (IBQ). Return model is used to capture the instant market reaction to quarterly earnings announcements around the announcement date, and in contrast, the price model aims to capture the effect of the accumulated information from the whole quarter. Fama-MacBeth method (Fama & MacBeth, 1973) is used to run in total six different cross-sectional time series regressions, out of which three are based on return model and three are based on price model. Prior research has shown that some discretionary items, like extraordinary items in income statement, reduce the value-relevance of earnings (Donelson et al., 2011). The results gained in this thesis are somewhat in line with those prior results. Net income has the lowest value-relevance or the strongest decrease in value-relevance in both of the models. Additionally, the results indicate that the value-relevance enhances when moving up in the income statement from net income towards pure earnings.
Thesis advisor
Jarva, Henry
earnings, earnings announcement, quarterly earnings announcement, value-relevance, earnings response coefficient, price model, return model
Other note