Issue Brief: Volume 4,  Number 1

As American as Apple Inc.: Corporate Ownership and the Fight for Tax Reform

In October 2015, the largest U.S. pharmaceutical company by revenue, Pfizer, announced plans to merge with Dublin-based giant Allergan, noting that a primary driver of the deal is the ability for Pfizer to avoid U.S. corporate taxes.


  • Both supporters and critics of the current tax advantages enjoyed by U.S. multinational corporations bolster their arguments with appeals to patriotism: the MNCs and their political supporters argue that allowing inversions or other similar arrangements and instituting another tax holiday for “repatriating” overseas earnings are good for the American economy as a whole; opponents condemn these tax advantages as unpatriotic in depriving the U.S. of enormous sums of needed revenue.
  • But where, precisely, is the “home” to which profits held offshore return? For many purposes, home is where the shareholders are. Determining ownership of U.S. MNCs such as Apple and GE, however, is extremely hard to do. The available data are either incomplete or opaque.
  • It is clear that the current corporate tax regime needs to be improved. But appeals for policies that promote U.S. competitiveness by presuming U.S. ownership of U.S. incorporated parent companies rest, in the end, on very little.

As Pfizer explained in a press release, “the combined company would generate more than $2 billion in savings over the first three years and would enjoy a tax rate of 17 to 18 percent—far less than Pfizer’s current corporate tax rate of 25 percent” as a U.S. company.[1] The proposed Pfizer-Allergan deal is one of many corporate inversions in the past two decades undertaken by companies with the ability and size to shift profitable business lines to offshore, tax-friendly nations. These inversions raise important questions about the ownership of U.S. multinationals that will be addressed in this Issue Brief, based on a previous publication by Professor Chris William Sanchirico of Penn Law, “As American as Apple Inc.: International Tax and Ownership Nationality,” Tax Law Review, Vol. 68, No. 2, 2015, pp. 207-274 (available at 

A number of legislative proposals and administrative actions have been introduced with the intent to curb the practice of inversions, as recently as November 2015, when the Treasury Department issued actions (building on regulatory measures established in September of 2014) to further reduce the tax benefits of corporate inversions. Yet the race to end inversions begs a more fundamental question: When we speak of U.S. multinational corporations (MNCs), what do we mean by a “U.S.” company?[2] More specifically, to what extent are U.S. companies owned by U.S. shareholders and to whom are the tax benefits accruing? As we will see, no one—not the U.S. government or even the very companies benefitting from foreign tax shelters—can truly answer this question.

Firms that have undergone inversion or, like Apple, Cisco, GE, and Google, that have exploited other tax-advantaged structures such as the “double Irish, Dutch sandwich,” do so under the rallying cry of maintaining their international competitiveness.[3] These MNCs and their political supporters stress that being forced to pay comparatively high U.S. tax rates in an intensely competitive global environment places significant burdens on their businesses—and, by implication, on the U.S. economy as a whole. Likewise, these MNCs have lobbied for another “tax holiday” similar to the one they enjoyed in 2005, when U.S. multinationals were permitted to “repatriate” their earnings to the U.S. at a significantly reduced tax rate: 5.25 percent instead of 35 percent. Without a tax holiday, they argue, U.S. multinationals have no incentive to bring their accumulated overseas profits back to the U.S., where they could be used to hire more American workers and suppliers.

Many stakeholders, including politicians, taxpayers, and domestically-based U.S. businesses, wish to reform this status quo. They see the existing tax advantages as inefficient and unfair—for why should large, technology-intensive, multi-national companies be able to avail themselves of special tax benefits while other U.S. businesses, by virtue of their immobility or industry, pay full freight? But more than that, they condemn them as inherently unpatriotic in depriving the U.S. of enormous sums of needed tax revenue. 

However, if the argument of U.S. MNCs is that their international competitiveness, and the U.S. economy generally, will improve by allowing inversions or other similar arrangements and by instituting another tax holiday, and that these imperatives should guide the country’s tax agenda, it seems altogether reasonable to ask whom it is exactly these measures would benefit. As this Issue Brief will address in more detail, a clear picture of the ownership of these MNCs is extremely elusive.

Identifying What Shareholders Want and Who They Are

The appeal for corporate tax reform often manifests itself in a campaign for a repatriation tax holiday, or a reduction in the tax rate for “bringing home” the profits of U.S. companies’ foreign subsidiaries, ostensibly to reinvest the earnings in companies’ respective workforces, research efforts, physical capital, etc. In general, when a U.S. parent company wishes to extract profits from a foreign subsidiary, it must cause its subsidiary to pay a dividend that is generally taxable to the parent at the full U.S. corporate tax rate of 35 percent. With such accumulated profits in the trillions of dollars, it is clear why a reduction in the repatriation tax is attractive to many U.S. MNCs.[4]

But where, precisely, is this “home” to which profits held offshore will be returning? For many purposes, home is where the shareholders are. But where are they?

The first step in identifying the nationality of firm ownership begins undoubtedly with the economic research on “home country bias.” This well-known academic theory documents—and attempts to explain—the phenomenon that individuals tend to overinvest in home country stocks relative to foreign country stocks. For example, U.S. investors invest disproportionately in U.S. stocks.

Economists have offered many explanations for this phenomenon, citing causes like regulatory barriers, tax disincentives, transaction costs, foreign exchange risk, and local risks that might be better hedged by owning domestic stocks. But none of these seem capable of explaining the observed magnitude of overinvestment in home country stocks. 

The explanations that seem more viable concern corporate or national governance risk, information asymmetry (i.e., the possibility that investors have less information concerning companies in foreign countries), and irrational biases, such as the possibility that unfamiliarity with a foreign company leads to unjustified fear of investing—although it can be difficult to distinguish between these three explanations.

But there is a critical and overlooked point with respect to this literature that is prevalent in the messaging from U.S. MNCs and their lobbyists, namely that the theory itself does not distinguish between a purely domestic U.S. company and a multinational corporation. The multi-nationality of large U.S. MNCs presumably brings to potential foreign investors both rationally process-able information and irrationally process-able familiarity. Do information asymmetries and feelings of unfamiliarity apply to global companies like Apple and Google? After all, if a citizen of Germany, Brazil, or China wanted to learn about home country bias, they likely could Google it on their iPhone. Indeed, the most accepted explanations for “home country bias” seem themselves to argue for MNC exceptionalism because these firms are “cognitively domestic,” due to their global familiarity. Since U.S. MNCs frequently cite the home country bias literature as a defense of their domestic and patriotic ownership, it is important to know that this line of reasoning seems particularly inapplicable to their particular situation.

The Missing Numbers

The next recourse for breaking out firm ownership is to do so by scratch, using raw data. Unfortunately, this is an exercise in futility, as the data do not exist. The companies themselves almost certainly do not know who owns their shares since shareholders often do not register directly with an issuer. Using Wall Street parlance, these companies often know only the “street name,” which is to say the name of the brokerage used by the shareholder.

Various U.S. federal agencies do collect data on cross-border securities holdings, including foreign holdings of U.S. stocks. The Treasury International Capital (TIC) system collects data on “portfolio holdings” of foreign residents who hold less than 10% of any single company’s stock.[5] The Bureau of Economic Analysis, part of the U.S. Department of Commerce, collects data on “direct investment” (10% ownership or greater).

The TIC reported that the portion of U.S. equity (by value) held by foreigners in 2014 was 14.5%.[6] Despite the nearly maximal response rates to, and the thoroughness of TIC surveys, there are serious concerns with TIC data.[7] One such concern is so-called “custodial bias,” which highlights the reality that oftentimes only the country of residence of a foreign custodian bank is recorded in the surveys, not the country of residence of the account owner. Furthermore, the 14.5% figure is a ratio: foreign holdings of U.S. equity divided by total U.S. equity. TIC only collects data on the numerator. It would be a monumental and ultimately fruitless endeavor to attempt to obtain data for the denominator on the same basis as the numerator, and therefore this is not done. The results, then, can be assessed quite fairly as crude estimates. Finally, in the TIC-denominated industry, “Capital Markets (including Mutual Funds),” the TIC is unable to identify and separate the domestic owners from the foreign owners of U.S. equities underlying the mutual fund holdings reported in their surveys.[8] In other words, there is fund opacity. The combination of TIC data limitations and fund opacity make it nearly impossible to determine the ownership of large U.S. MNCs.

There are, however, more data sources to evaluate. Googling “Who owns Google?” provides some useful information from SEC Form 13F about the “ownership” shares of large institutional investment managers (IIM) in the company. But there are significant 13F reporting gaps that are not immediately clear when looking a table of IIM shares on Yahoo! Finance or NASDAQ. For a company like Apple, that represents a roughly 40% reporting gap (as of March 2015). Also, as it turns out, the enforcement of 13F reporting is relatively weak compared to TIC reporting. None of this finally matters, as Form 13F was designed to reveal the portfolio choices of large portfolio managers, not characteristics like the nationality of the ultimate owners. The IIM reports merely the total holdings of a given stock across all the funds and accounts over which it has investment discretion. It does not reveal the nationality of those for whom it managing investments.

The SEC administers two other sources of potentially useful data for determining the foreign ownership share of large U.S. MNCs. The first of these two reporting regimes is for registered management investment companies, or mutual funds, on Forms N-Q and N-CSR. The second is for registered investment advisers (RIA) on Forms ADV and PF. While it is not unreasonable to assume that U.S. mutual funds have predominantly U.S. ownership,[9] there appears to be a significant amount of ownership of U.S. equity that is not accounted for by U.S. mutual funds – about 72%, according to the Investment Company Institute, the national association of investment companies.[10] As for registered investment advisers, Form ADV requires only that they list their total “regulatory assets under management (AUM),” a single figure representing both domestic and foreign clients that includes stocks, bonds, and other types of securities.

Additionally, RIAs tend to manage a substantial amount of their regulatory AUM in “separate accounts” available to wealthy individuals and large institutions. A recent Treasury report found that about 43% of assets managed by RIAs reside in these separate accounts, which is problematic for our present purpose. Not only do the top five asset management companies alone manage $5.5 trillion in separate accounts, but also data for separate accounts are not reported publicly.[11] Then there is the issue of private funds, inclusive of hedge funds, private equity funds, and venture capital funds. According to the Treasury, these appear to be about half the size of separate accounts in terms of regulatory AUM. The Dodd-Frank Act required RIAs to release a wealth of insightful information about private funds, including the foreign status of fund ownership. Unfortunately, little is revealed about a fund’s actual portfolio.

The final potential source of foreign ownership data comes from tax reporting and arises in connection with the tax paid by foreign investors on U.S. company dividends. The tax in question is remitted by the U.S payor of the dividend rather than the foreign payee, hence the more common name of this tax: the withholding tax. For every foreign payee each year, a U.S. withholding agent (e.g., a commercial bank or a brokerage firm) must file a Form 1042-S listing the dollar amount of dividends paid to such a payee during the year. At this point, it is perhaps unsurprising to discover that problems akin to custodial bias—observed in TIC data—plague this IRS data. Then again, it was never the intention of 1042-S to record the foreign ownership share of U.S. MNCs. Many of these reporting regimes are cleverly designed and extremely detailed, but none ever sought to answer the foreign ownership question, and none offers answers serendipitously.

A Healthy Brainstorming Exercise

Given all of the data limitations detailed above, consider this question: can it be said with reasonable confidence that at least 50% of shareholders of large U.S. MNCs are U.S. citizens? No, it seems that it cannot. We simply do not know. Accordingly, the U.S. MNCs arguing for lower tax rates on the basis of spillover benefits to the U.S. economy make a specious case when they assume U.S. ownership, especially when one considers the outcome that the United States could receive less revenue overall, as lower rates would apply to all domestically domiciled businesses. The same is true of a repatriation tax holiday. This policy option assumes that profits will come into and improve the U.S. economy. But what if that does not happen? Most of the repatriated profits may well go to shareholders and not workers, which is what happened during the 2005 repatriation tax holiday, so knowing who shareholders are and where they live would aid significantly in determining which national economies will reap the greatest shares of the earnings from another holiday.[12]

Moving past a nationalistic line of reasoning based on competitiveness and acknowledging that there is a lack of data to support the ownership claims of many large U.S. MNCs, what remains is a basic Keynesian question: if the goal is to stimulate the U.S. economy, is a repatriation tax holiday or lower corporate tax rate the optimal solution? Why not labor incentives or infrastructure policies or greater R&D tax breaks? The burden of proof seems to rest with those championing another holiday or lower rates. Proponents should be able to prove U.S. ownership, but that is not possible under the current reporting regime. The TIC, SEC, and IRS would have to synchronize their reporting requirements with the explicit goal of seeking to answer this question of ownership, but that is unlikely to happen. Therefore, proxies for U.S. ownership should be robust if the appeal is “help us compete.” None qualify at the moment, including the often-cited 14.5% foreign-ownership figure from the TIC.

Figure 1

It does not suffice merely knowing the percentage of U.S. equity issued by large U.S. MNCs, on the one hand, and the percentage of U.S. equity owned by foreigners, on the other. Rather, one needs information that narrows focus along both dimensions simultaneously. Given what we know about home country bias and how it argues against itself in regard to MNCs, the 57% foreign-ownership proxy implied by the international revenues of U.S. MNCs known to have large holdings of foreign earnings (see Figure 1) may be just as valid a figure and just as flawed.[13] The point simply is that numbers may not offer a way forward in this debate.


The current corporate tax regime needs improvement. The structures that large, technology-intensive companies can and do exploit through the double Irish, Dutch sandwich continue to allow for low-taxed MNCs to build up their IRAs (using the analogy from above) and benefit from tax breaks that companies of smaller size and/or different industry cannot access. But appeals for policies that promote U.S. competitiveness and that rely on U.S. ownership of U.S. incorporated parent companies rest, in the end, on very little. If Congress decides to pick winners, they should at least know who owns the horses in the race.


  [1]“Pfizer and Allergan to merge in $160 billion inversion,” Washington Post, November 23, 2015.

  [2]Many bills related to corporate tax reform have already been introduced in the 114th Congress. HR 415, S 198 and S 174 all treat mergers as U.S. firms if the U.S. shareholders maintain control of the resulting entity; HR 1809 and S 975 would keep inverted firms from receiving federal contracts, and S 922 and HR 1790 would restrict inversions. There have been many other similarly targeted legislative proposals in prior Congresses.

  [3]This structure describes the repositioning of highly profitable business lines of U.S. companies in foreign countries via the incorporation of new, offshore subsidiaries with the singular purpose of dramatically lowering corporate tax bills. In recent years, several MNCs (notably Apple) have reincorporated some of their most profitable subsidiaries in Ireland, which has maintained lax tax statutes well-suited for the exploitation of low-tax or no-tax islands like Bermuda or the Caymans. It is critical to note, however, that the laws and rules governing these loopholes are voluminous, scattered across several countries, and constantly changing.

  [4]It may be helpful to think of this structure as similar to an Individual Retirement Account (IRA). Income placed within an IRA is untaxed upon earning, grows tax free, and is taxed, along with the growth, upon withdrawal. The present discounted value of a total tax bill, under both the double Irish, Dutch sandwich and an IRA, may be significantly reduced if earnings accumulate and are reinvested over a long enough time period.

  [5]The TIC is a joint initiative of the Treasury, the Federal Reserve, and the Federal Reserve Bank of New York, and it collects the majority of its data from U.S.-resident broker-dealers and custodian banks. The TIC shares some of its data with the International Monetary Fund’s Coordinated Portfolio Investment Surveys (CPIS), but unfortunately several economically advanced countries do not reciprocate, including China and many Middle East oil-exporting nations, limiting the accuracy of TIC data.

  [6]TIC, Foreign Portfolio Holdings of U.S. Securities, April 2015, available at

  [7]See note 96 in Sanchirico’s paper ( for a hypothetical scenario illustrating TIC data limitations.

  [8]This industry amounts to a full 25% of all foreign holdings in U.S. equities.

  [9]See notes 165-166 in Sanchirico’s paper ( for more on mutual fund reporting. 

  [10]Investment Company Institute, 2013 Investment Company Fact Book, available at

  [11]U.S. Treasury department, Office of Financial Research, 2013, available at

  [12]The Tax Policy Center estimates in their models that shareholders bear between 60-70% of the corporate tax burden, so the implicit assumption is that shareholders would receive 60-70% of the benefit from a repatriation tax holiday.

  [13]Offshore Profit Shifting and the U.S. Tax Code—Part 2 (Apple Inc.): Hearing Before the Permanent Subcommittee on Investigations of the S. Comm. on Homeland Security & Governmental Affairs, 113th Cong. 35-37, 152-91 (2013), available at