2. EARNINGS PRESS RELEASES
3. TEXTUAL ANALYSIS OF OPTIMISTIC AND PESSIMISTIC LANGUAGE
3.1 Data Sample
3.2 Dictionary approach with DICTION 5.0
3.3 Descriptive Evidence
3.4 Correlation between language usage and future firm performance
3.5 Market response to optimistic and pessimistic language
4. A DIFFERENT APPROACH: NAÏVE BAYESIAN LEARNING ALGORITHM
5. SIMILAR STUDIES
6. EXCURSUS: A PROSPECTIVE BUSINESS RATIO
6.1 The Business Ratio
6.2 The Data sample
6.3 The Application of NETOPT ratio
6.4 Flaws of NETOPT ratio
APPENDIX: DECLARATION OF THE ACCOUNTING AND FINANCIAL VARIABLES
REFERENCES AND READING LIST
Earnings Press releases are the ticket to publicise firm’s performance. While numerous studies focus on the interpretation of numerical forms, this paper deals with the influence of optimistic and pessimistic language on firm’s future performance. In addition, it contrasts different approaches to measure the tone.
A closer look especially at propose and function of narrative form will be taken in Chapter 2: Rules and best practice guidelines to publish Earn- ings Press releases are presented and difficulties are pointed out.
Chapter 3 is based mainly on the study “Beyond the Numbers: An Analysis of Optimistic and Pessimistic Language in Earnings Press Re- leases” published by Angela K. Davis, Jeremy M. Piger and Lisa M. Sedor. They first examined the role language plays in the credible communication of information to investors. A textual analysis approach with DICTION 5.0 is presented and Results of other studies are added.
First, the preparation of a data sample for textual analyses of the con- nection between language and firm performance is represented. Subse- quently several accounting and financial market variables to measure firm performance are collected for each earnings press release.
Then, the technique of the dictionary-based content analysis program to categorized optimistic and pessimistic words is explained extensively. Afterwards, the pros and cons of this approach are discussed critically.
Results of the analysis like the increase in the average length of earn- ings press releases are evaluated. Also the descriptive data of the sample is listed. A baseline multivariate regression model verified a correlation between language usage and future firm performance. The used functions have been developed on sample data and accounting and financial market variables.
Next the market response to optimistic and pessimistic language is tested. It can be seen that an unexpected portion of optimistic and pessi- mistic language has a stronger influence on market returns than an ex- pected portion.
An alternative to textual analysis techniques is the naïve Bayesian learning algorithm presented in chapter 4. By comparison with the dictio- nary based approach advantages and drawbacks are shown.
The excursus shows a possible application of the found connection be- tween future firm performance and language in earnings press releases. Results of the previously introduced textual analysis are used to create a business ratio. In conclusion, Chapter 7 passes review of the paper and summarizes the results.
2. Earnings Press Releases
Press releases are a ticket to publicity and to announce firm’s perfor- mance in publications. Lewis et al. characterized them as “the major news event of the season” [Mahoney and Lewis (2004)] for companies and in- vestors, financial media, analysts and the market. After quarter-end earn- ings, press releases should be published as soon as practicable. Within the framework of investor relations (IR), they communicate the firm’s fi- nancial performance in numeric and narrative form to [cp. FEI / NIRI earn- ings press releases guidelines].
The U.S. Securities and Exchange Commission (SEC) established formal financial statements and regulates the way in which firms announce defined financial measures1. For firms listed on New York Stock Exchange (NYSE) and National Association of Securities Dealers Automated Quota- tions (NASDAQ) earnings press releases are mandatory.
Apart from analyses of operating results, historical data, positive and negative factors affecting key financial indicators, an accurate earn- ings press release should contain a realistic and truthful forecast for future quarters [cp. Trautmann and Hamilton (2003)]. “Promises of specific re- sults, exaggerated or unwarranted claims; or [...] opinions for which there is no reasonable basis” [NYSE Rules, Rule 472] should be avoided and a Safe Harbor language is to be used.
Despite these rules, language in earnings press releases ranges from a straight forward numerical view to a quite promotional presentation of the firm’s development. Also opportunistical behavior in writing press releases to delude investors is used. Therefore narrative form in disclo- sures is likely more difficult to analyze than the numerical form2.
Different ways to examine the language used in earnings press re- leases and their influence on stock markets and future firm performance will be described in the following sections. One approach is a textual anal- ysis with DICTION 5.0 implemented by Davis et al. They first examined “the role that language usage plays in the credible communication of in- formation to investors” [Davis et al. (2006), p. 25].
3. Textual Analysis of Optimistic and Pessimistic Lan- guage
3.1 Data Sample
Davis et al.’s data sample used for textual analysis consists of earn- ings press releases between January 1, 1998 and December 31, 2003 published by PR Newswire3. Diverse accounting, financial market and tex- tual-analysis variables, like return on assets (ROA) to identify firm perfor- mance in present and future, were collected for each file. [cp. Davis et al. (2006), p. 10]
The final data sample consists of 23,662 firm quarterly earnings press releases. The initial sample of 73,758 firm quarters was stripped of any non earnings related announcements: files with no essential data
1 Financial Executive Institute (FEI) and National Investor Relations Institute (NIRI) pub- lished practice guidelines for earnings press releases.
2 The numeric form in earnings press releases should be prepared under Generally Ac- cepted Accounting Principles (GAAP).
3 PR Newswire is hired by companies to deliver news and multimedia content.
- Quote paper
- cand. rer. pol. Marius Rombach (Author), 2011, An Analysis of Optimistic and Pessimistic Language in Earnings Press Releases, Munich, GRIN Verlag, https://www.grin.com/document/180864