SEOs in a ‘Hot Market’: evidence of timing

Abstract

This study analyses the financing decision of raising equity through a rights issue in a developing market, the Athens Stock Exchange (ASE), during a particular emerging period. Specifically, this study examines the information content of accounting items derived from published financial statements the year prior to a ‘hot’ period in explaining post-issue stock price performance. We are using data from listed companies in the ASE during the ‘hot period’ of year 1999 when stock prices burst and an unusual large number of seasoned equity offerings (SEOs) took place. Our empirical results do not verify a statistically significant relationship between discretionary accruals in the year preceding the issue and post-issue stock returns. Moreover, historical accounting items do not provide value relevant information and cannot be used to explain post-issue stock returns. Market trend prior to the issuing is proved to be the only significant variable in explaining post SEO returns. The overall findings are in line with the market timing theory which claims that managers just time their equity issues in an upward moving market in order to increase the offering proceeds.

Acknowledgments

The authors gratefully acknowledge useful comments and suggestions from Professor Dimitrios Ghicas. An earlier version of the paper has been presented at the 2004 EAA Conference (European Accounting Association) and at the 2004 HFAA Conference (Hellenic Financial Accounting Association).

Notes

¹ As ‘Hot issue’ markets are defined markets at periods of high new issue volume and high level of initial returns (Ibbotson and Jaffe, 1975; Ritter, 1984).

² When information asymmetries exist, managers have private information about firm quality and may pursue their own objectives at shareholders expense.

³ For example, firms’ reported earnings could be affected by accelerating revenue recognition and deferring expense recognition. This effectively shifts earnings to the current period from subsequent periods. Alternatively, firms’ earnings can be affected by changing methods of inventory accounting, revising estimates of bad debt expense, etc.

⁴ An overview of important dates of institutional and structural changes is summarized as follows: In 1991 Law 1969 (Gov Gaz. A. 167) established the Capital Market Commission as a supervisory authority of the ASE. In year 1992 the presidential Degree 53/1992 (Gov Gaz. 22/14-2-1992) established the legal framework for the dissemination of confidentiality and/or inside information; In h1995 Law 2324 (Gov Gaz. A. 146/17-7-1995) allowed over-the-counter transactions and short selling. Later on in year 1997 Law 2533/1997 provided the legal framework for the privatization of the ASE and also for the establishment of the Athens Derivatives Exchange.

⁵ Ministerial Decision 2063/B.69/1999 set up the basic listing requirements for the Parallel Market; Law No. 2733/1999 provided the regulatory framework for the New Market (NEHA), where small and medium sized companies (fast growing or innovative) can be listed. Regulation No. 18/15-1-1999 on the methods and procedures regarding the trading transaction through the new Electronic Trading System (OASIS); Regulation No. 19/15-1-1999 on the obligations of the listed companies in cases of mergers, takeovers, changes in the main activity or spin-off the sector; Regulation No. 20/5-3-1999 on the procedures for the listing of shares in the ASE; Regulation No. 21/16-6-1999 on the requirements for the listing of construction companies on the ASE.

⁶ On 1 January 2001.

⁷ Equity issues through private placement may also happen.

⁸ Recent articles argue that the Jones (1991) model suffers from the omitted variable problem and thus generates imprecise parameter estimates (Bernard and Skinner, 1996). Dechow et al . (1995) find that DA from the Jones model is imprecise due to the large variation of the estimated coefficients, even though the Jones model exhibits more power than other existing accrual models in detecting earnings management.

⁹ We classify this later group of companies as nonSEOs.

¹⁰ Approximately 252 trading days.

¹¹ Market return has been calculated by using the analogous methodology as of computing post-issue returns in each regression model, respectively.

¹² We have excluded from our analysis banks, assurance and leasing companies.

¹³ For testing the relation between the coefficients we have performed a Wald test.

¹⁴ The looking forward model cannot be used to explain post-issue returns since it requires data which are not known at the time of the offering (i.e GR_Sales).

¹⁵ The lagged model also captures the possibility of earnings management even 2 years prior to the offering.

¹⁶ Approximately 252 trading days.