David G. Harris, The Impact of U.S. Tax Law Revision on Multinational Corporations' Capital Location and Income-Shifting Decisions, Journal of Accounting Research, Vol. 31, Studies on International Accounting (1993), pp. 111-140

We examine how tax legislation that restricted firms' deductions of CEO compensation above $1 million reduced the implicit contracting cost of compensation for firms that were expected to pay below that amount and that were not directly affected by the law change. We find that firms that expected to pay their CEOs less than $1 million actually increased their CEOs' cash compensation, contrary to Congress's expectations. Moreover, the magnitude of the unexpected increase in compensation is proportional to how far the CEO's expected compensation fell below Congress's new $1 million reasonable-compensation standard. Thus, our study provides evidence that some of the largest U.S. corporations responded in a manner contrary to policymakers' expectations. Our findings also support the theory of implicit contracting costs, by demonstrating that many firms reacted in an economically rational fashion when a change in the tax law decreased their implicit costs of CEO compensation.

In this study we examine whether banks owned by interstate multibank holding companies coordinate their security gains and losses to manage their tax, earnings, and capital management objectives. Specifically, we examine whether the realization of security gains and losses is related to the objectives of the individual bank, the consolidated group, or both. We find subsidiary banks manage their gain realizations not only to reduce their own state taxes, but also strategically to reduce their consolidated groups' tax expense. Specifically, members of consolidated banking groups shift gain recognition to lower-taxed group members and away from higher-taxed group members. In addition, we find evidence suggesting that banks realize security gains and losses to manage both their own and their groups' financial statement earnings.

Holthausen and Verrecchia's (1990) and Kim and Verrecchia's (1997) theoretical models predict that private information inferred at the time of an earnings announcement (private event-period information) is associated with greater trading volume. We provide empirical evidence consistent with these theories. Specifically, announcements that increase analysts' private information (as measured by Barron et al.'s [1998] empirical proxies) are associated with increased trading volume, consistent with some investors similarly acquiring private event-period information. In addition, announcements that decrease analysts' consensus are associated with more trading volume. Because consensus declines when private information increases, this finding provides reinforcing evidence that investors trade following earnings announcements because of private information that becomes useful only in conjunction with the information in the announcement and that this information is important enough to spur trading.