ABSTRACT
We compare the market values of executive stock option (ESO) trades with their Black & Scholes (1973) model values calculated following the major accounting standards, SFAS No. 123r and IFRS2. Our results show major underpricing compared to the traditional B&S method values. This should be considered while applying SFAS No. 123r and IFRS2 for estimating fair values. Especially time to expiration has a major influence on the undervaluation suggesting that the possibility of a change in corporate structure lowers the cost of ESOs to shareholders.
Acknowledgements
The authors thank Kari Lukka, two anonymous referees, David Aboody, Panu Jousimies, Panu Kalmi, Markku Kaustia, Matti Keloharju, Juhani Linnainmaa, Lisa Meulbroek, Timo Mäkelä, Tapio Pekkala, Jonathan Rogers, Jack Stecher, Håkan Thorsell, Tuomo Vuolteenaho, as well as seminar participants at the 12th European Financial Management Association, 27th European Accounting Association, GSFFA Research Seminar, London School of Economics, University of St Gallen, Helsinki School of Economics and University of Vaasa and American Accounting Association 2004 Annual Meeting for helpful comments and suggestions. Financial support from Kauppakorkeakoulun Tukisäätiö is gratefully acknowledged.
Notes
1. OTC data would be available on the Microsoft ESOs, where J. P. Morgan purchases employee options. The price level shows a discount of 80% relative to the value of similar traded options (Bettis et al., Citation2005).
2. Similar trends could be found in the USA during the 1980s (Aboody, Citation1996), in the UK in the late 1980s (Buck and Bruce, Citation1991), in Canada in the early 1990s (Klassen and Mawani, Citation2000) and in Germany in the late 1990s (Winter, Citation1999).
3. If the following trade with shares takes place on the following day, we selected the share price based on the previous trade prior to the ESO trade.
4. As a robustness check, we also used a fixed time period of 250 days prior to each ESO trade for volatility estimation. Volatility figures were slightly different, in three companies and nine ESOs, the fixed 250-day volatility resulted in higher volatilities and in another three companies and five ESOs they were lower. All the regression results were qualitatively the same irrespective of the volatility estimation choice.
5. No mean-reversion tendency is considered since all companies have been listed for a longer time period.
6. We also used a fixed 250-day volatility estimation period to control the influence of volatility estimations. All signs and significance levels remain practically the same, and only minor changes exist in the multipliers. The major difference is in the size of the constant. All constants are smaller when using fixed 250 days for volatility estimation, and the difference varies between 16.52 and 13.96 percentage points.
7. Such differences are, for example, the probability of changes in corporate structures, taxation of ESOs and policies on whether executives can hold the options when leaving the company. In Finland, executives leaving the company typically can hold their ESOs after vesting.