As corporate tax reform emerges as a national legislative priority, a key proposal by the Trump administration and others is to scrap the United States’ current worldwide system of taxation. In that system, income earned abroad by subsidiaries of U.S. multinationals is subject to federal taxes when brought home.
Instead, the proposal favors a territorial tax system that would enable U.S. parent companies to receive profits from their overseas subsidiaries without incurring a federal levy. At present, U.S. companies have left an estimated $2.6 trillion outside the country in order to avoid the high U.S. taxation rate.
Minus a considerable disincentive to bringing foreign profits home, it is hoped that there would be a positive impact on total shareholder return. However, that effect is likely to be modest, new research suggests.
The research also provides evidence, based on a large sample of companies, of a phenomenon that until now has been noted only anecdotally: multinationals’ propensity to fund shareholder payouts through heavy borrowing in order to avoid repatriating foreign-earned profits for that purpose.
Of course, not all companies leave all of their foreign profits offshore. A study published in the current issue of The Accounting Review, an American Accounting Association journal, finds that repatriation tax costs have indeed reduced the two principal types of payouts to shareholders: dividends and share buybacks.
Yet in recent years that relationship appears to have weakened — and even before that, repatriation taxes substantially constrained dividends but diminished the likelihood of share repurchases only slightly and the volume of repurchases not at all.
The paper’s author, Michigan State University professor Michelle Nessa, investigated the impact of repatriation tax costs over two periods chosen to skirt a pair of extraordinary events — namely, the one-year repatriation tax holiday authorized by the American Jobs Creation Act of 2004 and the economic dislocations caused by the Great Recession of 2008.
The main portion of the study draws on data from many hundreds of U.S. multinationals during the 18 years prior to the American Jobs Creation Act — a trove of 12,444 firm-years.
Nessa found that, in general, the greater the cost of repatriation, the less likely firms were to pay dividends to shareholders and the less lucrative the dividends they did award were likely to be.
For example, a multinational firm with a single-year repatriation tax obligation at the payout group’s mean level (0.61% of assets) was 4.35% less likely to issue a dividend than a firm owing no repatriation tax. Further, the size of the dividend as a portion of company assets averaged 14.32% less than that paid by a non-owing firm, leading the author to comment that the “economic magnitude of the effect is substantial.”
In contrast, no significant relation was found between repatriation tax costs and the amount of companies’ stock repurchases, which overtook dividends as the principal form of shareholder payout over the 18-year period. Although such a relation did emerge for multinationals that were financially constrained (judged by the proportion of negative words in company annual reports), those firms constituted a minority of the sample.
To explain why repatriation tax costs limited share repurchases less in the overall sample than in financially constrained companies, the professor writes that the latter will be less able to borrow and/or will face higher borrowing costs. Investigating further, she finds that high stock repurchases strongly correlated with high debt, “a result consistent with U.S. multinationals incurring borrowing costs to avoid U.S. repatriation taxes.”
As to why the taxes might affect dividends and repurchases differently, Nessa writes, “U.S. multinationals could occasionally incur costs (e.g., borrowing, utilizing tax attributes, engaging in complex transactions) that allow them to access the wealth represented by their foreign cash without triggering U.S. repatriation taxes. If these cash inflows to the U.S. parent are non-recurring, they are likely to be distributed through share repurchases, because [unlike dividends] share repurchases do not implicitly commit the firm to similar future payouts.”
Yet, the differing patterns between the two kinds of payouts disappear in a later period investigated in the research, 2009 through 2014. Over those six years, neither dividends nor share buybacks were significantly related to repatriation tax costs.
Why that occurred is beyond the scope of the study. “It may have to do with the extremely low cost of borrowing in the post-recession years,” Nessa surmises. “There were not any changes in U.S. tax laws related to repatriation that would account for the different results from those of the earlier period studied.”
She adds, “If shifting to a territorial system results in more payouts to shareholders, it will most likely occur at financially constrained firms, which constitute a minority of U.S. multinationals. But will such a reform occasion a bonanza for the majority of multinationals’ shareholders? That does not seem likely. When it comes to payouts, many companies seem to have worked their way around this particular tax barrier already.”