Did Burger King submit to acquisition by a Canadian donut chain for tax reasons? Or, at least, once Burger King and Tim Hortons decided to merge, did they choose to have a Canadian parent for tax reasons? A recent empirical study by Andrew Bird, Alexander Edwards and Terry Shevlin suggests that one tax factor—the existence of “locked out” offshore earnings—increases the likelihood that a non-U.S. acquirer will acquire a U.S. target. Bird, Edwards and Shevlin analyze thousands of merger transactions, without regard to whether the transactions might be labeled “inversions.” Their paper contributes to the considerable literature that tests the idea that accounting and tax disparities affect firm prices and transaction decisions.
Bird, Edwards and Shevlin examine several thousand public company firms with a parent corporation incorporated in the United States, for example under Delaware law. Each of the firms in the sample was acquired between 1995 and 2010. The paper considers the possibility that these target U.S. firms might have been acquired by U.S. or non-U.S. acquirers. Bird, Edwards and Shevlin find that when a U.S.-parented target corporation has more offshore “locked-out earnings,” the target firm is more likely to merge with a non-U.S. acquirer rather than a U.S. acquirer.
The concept of “locked-out earnings” has meaning in both tax law and tax accounting vocabularies. From a U.S. income tax law perspective, the presence of “locked-out earnings” generally means that a firm has achieved a low tax rate on non-U.S. earnings housed in non-U.S. subsidiaries of a U.S. parent. The earnings are “locked-out” because the U.S. parent would have to pay significant residual U.S. corporate income tax (as well as, perhaps, non-U.S. withholding tax) upon a dividend of the earnings to the U.S. parent. Such a residual U.S. tax would be reduced by credits for foreign income taxes paid, but a firm that has achieved a low tax rate on non-U.S. earnings will generally not have significant foreign tax credits available.
Financial accounting for taxes provides a required disclosure related to locked-out earnings. A firm can designate and report earnings that are indefinitely reinvested in a foreign jurisdiction as permanently reinvested, or PRE. A firm need not record any income tax expense with respect to PRE, on the theory that the designated earnings are earmarked for offshore investment and will never come home. Other work has shown that both tax and nontax incentives drive PRE designation. In other words, there is some overlap between PRE in accounting and the idea of locked-out earnings in tax, but the overlap is not complete.
Having a non-U.S. parent—rather than a U.S. parent—alleviates both the tax law and the accounting problems associated with locked-out earnings. From a tax law perspective, it may be possible for a firm to repatriate pre-transaction earnings without residual U.S. tax using a “hopscotch loan” or other “out-from-under” technique. (The Obama administration recently promised regs that would treat hopscotch loans as deemed dividends, but only for a firm that had engaged in a transaction treated as an inappropriate inversion.) Also, future non-U.S. earnings should not produce the tax-lockout problem.
A favorable accounting result follows from favorable tax results, to the extent available. In other words, if there is no residual tax upon the repatriation of earnings, a foreign parent need not report a tax expense upon a distribution of foreign earnings.
Bird, Edwards and Shevlin use three proxies for locked-out earnings. One is the PRE disclosed in a target firm’s financial statement footnotes. A second is a binary PRE indicator variable that equals 1 if there is any disclosure that PRE exists. A third is a repatriation cost measure derived from the difference between U.S. corporate tax that would be imposed on foreign income (at the U.S. statutory rate) and current foreign income tax expense already paid (to account for foreign tax credits).
They find that the more PRE a U.S. firm has, the greater the likelihood that the firm will be acquired by a non-U.S. firm. A one percentage point increase in PRE translates to a 0.58 percentage point increase in the likelihood of a foreign acquirer. A one percentage point increase in the repatriation cost measure proxy for locked-out earnings has a smaller effect—it translates to a 0.02 percentage point increase in the likelihood of a foreign acquirer. Both are highly significant, at the 1% level, meaning that there is a 99% probability that the relationship described in the sample of firms also applies for a larger population. When the authors control for the existence and proportion of foreign earnings at a firm, a one percentage point increase in PRE translates to a 0.36 percentage point increase in the likelihood of a foreign acquirer, with significance at the 5% level. These changes compare to 23% overall likelihood of a non-U.S. acquirer for a U.S. target with non-U.S. operations.
The authors also divide non-U.S. acquirers into two groups according to whether they operate under territorial systems, meaning that the jurisdiction of incorporation only taxes domestic income; or worldwide systems, meaning that the jurisdiction of incorporation taxes (at least in theory) both domestic and foreign income. They find that the observed correlation between locked-out earnings and the likelihood of non-U.S. acquisition is driven by non-U.S. firms incorporated in territorial jurisdictions. A one percentage point increase in PRE translates to a 0.32 percentage point increase in the likelihood of acquisition by a territorial non-U.S. firm, but lacks a statistically significant correlation with the likelihood of acquisition by a worldwide non-U.S. firm.
This work has relevance for international tax policy because it suggests a general reason why non-U.S. acquirers may have an advantage over U.S. acquirers for some U.S. target firms. Bird, Edwards and Shevlin do not claim that their results show that accounting and tax considerations cause firms to seek an acquirer; the paper’s sample does not include any control group of firms that did not merge at all. But their results may have relevance for proposals to adopt territoriality for the U.S. system.
More generally, their results emphasize the importance of working out how to tax non-U.S. parented firms with U.S. subsidiaries. This project has long been a sideshow in the U.S. The paradigm case in regulations, white papers and case law features the U.S. parent of a multinational group. Earnings stripping legislation has been stalled for years. This project needs to move forward in order to reach appropriate results for non-U.S. parented multinationals, including those that result from acquisitions described in the Bird, Edwards and Shevlin paper.