Revisiting the choice between incentive stock options and nonqualified stock options: A study of tax and non-tax costs

Corporations can grant stock options to their employees in the form of either incentive stock options (ISOs) or nonqualified stock options (NQOs). ISOs are more tax favorable for employees because they are able to recognize income from the ISO at capital gains rates. NQOs are more tax favorable to the employer because the company can take a tax deduction for stock option expense. This tax difference is the primary difference between the two types of options, yet previous research has not found strong significance for taxes in examining why companies choose one plan over the other.; This dissertation extends the prior research by incorporating non-tax costs not included in other models, as well as introducing other tax measures into the model. Specifically, corporate governance costs are added to the model. Financial reporting costs and agency costs are used as control variables. The other tax measures included in the analysis are the variance of the marginal tax rate, Alternative Minimum Tax status, and the interaction between the marginal tax rate and a proxy for the individual capital gains rate. A logistic regression model is used to analyze the data.; Results of this dissertation are similar to results of previous studies in that no strong evidence is found to support the hypothesized relationship between taxes and option plan choice. The 1998 results show marginal significance in the expected direction on the marginal tax rate, but 1999 results present strongly significant results in the wrong direction. No significance is found for Alternative Minimum Tax, tax variance, or corporate governance costs. A potential reason for these results is excessive noise in the dependent variable and weak measures for some of the independent variables.; The main contribution of this dissertation is the use of a more comprehensive model to examine stock option choice. Previous studies in this area did not include controls for corporate governance costs and also did not include financial reporting costs and agency costs together. Even though the significance on the marginal tax rate is only marginal, this dissertation also adds to the growing literature by using Graham's (1996) simulated marginal tax rate measure.