Abstract
There is little research on whether foreign real estate investors respond to domestic taxes, which is somewhat ironic given that real estate is such a tax-driven industry. This article investigates the impact of taxes on foreign investment in real estate by exploring the response to 2015 reforms to the Foreign Investment in Real Property Tax Act of 1980, a unique section of the US Internal Revenue Code that imposes real estate income and capital gains withholding taxes on foreign investors. The new law exempts qualified foreign pension funds, with the effect of reducing capital gains withholding taxes by up to 35% depending on investor position and home country tax treaties. The authors find that this reform led to a five to seven percentage point increase in annual investment growth from pension funds in countries that experienced a tax rate decrease, relative to general trends in foreign pension fund investment in US real estate investment trusts. In addition, the results suggest that pension funds in those countries responded almost immediately.
TOPICS: Real estate, legal/regulatory/public policy
Key Findings
• The results in this article suggest that domestic income taxes can be an important determinant of foreign investment in domestic real estate.
• Recent changes to the US taxation of foreign investment in US real estate were associated with increased annual growth in investment in US REITs from foreign pension funds in countries that received a tax rate decrease (relative to general foreign pension fund investment trends).
• Additionally, affected foreign pension funds appear to have adjusted their investment immediately after the tax reform went into effect.
In this article, we investigate how taxes influence foreign investment in US real estate by examining the response of foreign pension funds to an exemption from the Foreign Investment in Real Property Tax Act of 1980 (FIRPTA), effective December 2015. Because the tax impact of the 2015 FIRPTA reform varies by home country, we are able to explore the impact of tax reform in a way that controls for external economic factors. We focus on foreign portfolio investment rather than foreign direct investment, which also allows us to examine the effect of taxes without the concern of confounding investor property management and sale decisions. For years, investment advisors, trade groups, and lawmakers have asserted that the taxes levied under FIRPTA restrict foreign investment. However, there has been little to no evidence to support those assertions. Our results empirically support those assertions and suggest that domestic taxes play a role in shaping foreign real estate investment. In addition, to our knowledge, this is the first article to provide evidence on the magnitude and speed of the response from foreign investors to such a change in taxation. Our results also provide evidence on the effectiveness of different channels (i.e., country-specific tax treaties compared to broader tax legislation) of lowering the tax burden to encourage foreign investment in US real estate.
Foreign investors are generally not taxed on capital gains from the sale of US assets. However, foreign investors have been subject to a capital gains withholding tax on the sale of US real estate since FIRPTA was passed in 1980. For foreign investors in US real estate investment trusts (REITs), this includes up to a current rate of 15% tax on the sale of US REIT stock by foreign investors (increased from 10% in 2015) and up to a 35% tax on capital gains distributions. In contrast, US investors are taxed at their personal capital gains tax rate, which is currently capped at 20%, and US pension funds are generally exempt from taxation on capital gains.1 In addition to real estate tax liabilities, FIRPTA imposes burdensome documentation requirements on real estate transactions involving foreign investors.
FIRPTA thus established a system of asymmetrical tax treatment on capital gains distributions for domestic and foreign investors. In particular, US pension funds were exempt from taxation on both REIT capital gains distributions and the sale of REIT stock. Foreign pension funds were not. A foreign pension fund could avoid the withholding tax on the sale of stock or return of capital by owning no more than 5% of the shares of a publicly traded REIT or investing in a domestically controlled REIT.2 However, many foreign pension funds could not fully escape the withholding tax on REIT capital gains distributions. A foreign pension fund could reduce the 35% withholding tax on capital gains distributions to between 0% and 30% by owning no more than 5% of the shares of a publicly traded REIT. The reduction amount depended on the pension fund’s home country tax treaty with the United States, creating a higher tax burden for some foreign investors than for others.
In 2015, Congress enacted the Protecting Americans from Tax Hikes Act of 2015 (the PATH Act), which carved out a complete exemption from FIRPTA for qualified foreign pension funds (QFPFs).3 This means that QFPFs are no longer subject to FIRPTA on investment in any type of US real estate, whether through direct investment or portfolio investment. Specifically, a QFPF will not be subject to tax on the sale of REIT stock or return of capital distribution regardless of whether the REIT is domestically controlled or whether the fund stays under a de minimis ownership percentage. Additionally, QFPFs are no longer subject to the tax on REIT capital gains distributions, effectively changing the tax rate from a country-specific rate to 0% regardless of country. Also of importance, these foreign pension funds will no longer be subject to the possibility of having to file a US tax return because of their US real estate investment if they do not meet the standard FIRPTA exemptions.
For years, real estate experts, investment advisors, and lawmakers have all speculated on how FIRPTA severely restricts inbound investment in real estate as part of their call to loosen those restrictions. For example, in June 2015, the Real Estate Roundtable (2015) submitted an official statement to the Senate Finance Committee on behalf of over a dozen national trade organizations, noting that “[t]he FIRPTA regime is an anti-competitive outlier that deters and deflects capital to other markets.” After the passage of the PATH Act, these same groups similarly speculated on how the PATH Act would result in increased foreign investment in US real estate, sometimes citing that alleged increased investment in support of the complete repeal of FIRPTA (DeBoer 2017; Rosen and Sakamoto 2017). Even Congress (2017) has relied on this assumption, stating in a 2017 letter to Secretary of the Treasury Steven Mnuchin that the FIRPTA reform in the PATH Act is “estimated to have injected billions of dollars in foreign investment into the US real estate market.” However, there is little to no empirical evidence on whether the PATH Act actually did lead to increased foreign investment in US real estate. In this article, we attempt to provide some initial evidence on the impact of the PATH Act, specifically on foreign pension fund investment in US REITs. Our results suggest that foreign pension funds responded immediately to the PATH Act, increasing their relative investment in publicly traded US REITs by the end of 2015. Specifically, that increase in investment occurred only in countries that experienced a tax rate decrease as a result of the Act.
We exploit the 2015 exemption from FIRPTA to test the responsiveness of foreign pension fund investors to changes in REIT capital gains distribution tax rates. Foreign pension funds are reportedly extremely sensitive to FIRPTA taxes and have developed sophisticated blocking structures to minimize US taxes when investing in US REITs (Bear, Lillis, and Naylor 2010). Therefore, we predict that we will see an increase in direct investment into US REITs from foreign pension funds after the PATH Act. However, any measured increase in investment could also be attributable to attractive US real estate markets and/or changing global economic factors. To parse out the influence of the PATH Act capital gains distribution tax decrease, we examine the impact of tax rate changes on country-specific investors by determining whether there is a differential effect between those countries that experienced a tax decrease after the PATH Act (i.e., whose capital gains tax rate decreased from above 0% to zero) and those countries that did not experience a tax decrease after the PATH Act (i.e., whose tax rate was already zero).
The link between taxes and foreign real estate investment has been largely unexplored.4 There is recognition that taxes may matter, but little evidence exists. Clayton et al. (2015) pointed out that “because tax treaties frequently do not treat all capital equally, cross-border investors often face a higher tax drag.” Falkenbach (2009) found that over two-thirds of property investors surveyed considered taxes important in selecting investment locations but did not analyze types of taxes or the magnitude of the effect of taxes. Mauck and Price (2017) examined determinants of foreign versus domestic real estate investment, including two general tax variables, but those variables did not capture the actual tax burden to an investor. Ziobrowski, McAlum, and Ziobrowski (1996) considered whether differences in country income tax rates could explain interest in foreign real estate investment and concluded that there were no significant after-tax benefits of investing in US real estate for foreign investors.
There is also little evidence that any type of pension fund (domestic or foreign) takes taxes into account when investing in real estate. Webb (1984) is the exception. He conducted a survey of pension fund managers and found that some managers considered taxes in their investment decisions, but over half of the responding managers did not use any after-tax measures in evaluating investment options.5 Given the date of the study and its qualitative rather than quantitative nature, whether taxes matter to foreign pension fund investors in real estate is still an open question.
Howard, Pancak, and Shackelford (2016) were the first to show that taxes do matter to foreign real estate investors by examining a 2004 change to FIRPTA. However, that analysis excluded pension fund investors. We expand on their findings and specifically examine whether taxes matter to foreign pension fund investors by looking at a more recent FIRPTA reform. We provide confirmation that taxes are in fact a determinant of portfolio investment by foreign pension funds in US real estate. We also provide evidence on the magnitude and speed of investor responses and the effectiveness of different tax policy channels in encouraging investment.
In our empirical tests, we estimate the amount of investment in each publicly traded US REIT from foreign pension funds in eight foreign countries from 2006 through 2015.6 For example, we find that Norwegian pension funds held just over 3% of the outstanding shares of Prologis at the end of 2014. Any capital gain distributions for these Norwegian investors in 2014 would have been subject to a 15% withholding tax (as long as an individual investor owned less than 5% of the REIT shares outstanding).7 However, in 2015 (after the tax law change became effective December 18), any capital gains for these Norwegian pension investors would not have been taxed. A reduction in the withholding rate from 15% to 0% for Norwegian pension investors increased the likelihood that Norwegian pension investors would be attracted to Prologis and other US REITs.
In contrast, some countries enjoyed a 0% capital gains withholding tax rate prior to the Act as a result of country-specific tax treaties (again, as long as the investor owned less than 5% of the REIT shares outstanding). For example, Dutch pension funds held almost 4% of the outstanding shares of Vornado Realty Trust at the end of 2014. Any capital gains for these Dutch investors in 2014 would not have been subject to a withholding tax, treatment that remained unchanged under the PATH Act.
To test the sensitivity of foreign pension funds to the FIRPTA exemption applied to REIT capital gains, we compare the annual change in aggregate investment from pension funds in a particular country to a specific US REIT before and after the PATH Act. We examine changes in investment flows from pensions in eight countries to 129 US REITs. As predicted, we find an increased rate of foreign investment in the post-PATH Act period that came exclusively from those countries where the withholding rate was not already lowered to zero by tax treaty. In addition, our results suggest that pension funds in those countries responded almost immediately, increasing their relative investment by the end of 2015.
These results empirically support that taxes do in fact matter to foreign real estate investors. They also shed some light on the responsiveness of any real estate portfolio investor, foreign or domestic, to changes in capital gains tax rates and the magnitude and speed of that response. Specifically, to our knowledge, this is the first article to provide evidence that the 2015 PATH Act QFPF exemption increased foreign pension fund investment in publicly traded US REITs, with the increase occurring immediately after the passage of the Act and directly motivated by the drop in tax rate.8 These results should be informative to real estate investors, REITs, and US policymakers.
The rest of this article is organized as follows. The next section provides background on FIRPTA. We follow that with a discussion of expectations of how the PATH Act would and did affect QFPF investment and our method of testing whether those assertions are valid. We then present our findings. The final section provides our conclusions.
FIRPTA BACKGROUND
Prior to FIRPTA’s enactment in 1980, foreign investors were generally exempt from US capital gains tax on the sale of US real estate and thus held a competitive advantage over domestic investors. There was a growing perception in the 1970s that this discrepancy hurt US farmers because US farm prices were being bid up by foreign buyers. FIRPTA was enacted to rectify this, adding a number of sections to the Internal Revenue Code (IRC). Most importantly, IRC section 897 subjects a foreign investor to US income tax on income from the disposition of US real estate property interests (USRPI), which includes both direct investment in real estate and indirect investment through stock of a corporation whose primary assets consist of real estate. Therefore, FIRPTA applies to investors in US REITs.9 Furthermore, IRC section 6039C imposes federal tax reporting requirements on foreign investors that are subject to capital gains withholding.10
Given that REITs are structured as pass-through entities to avoid double taxation, each form of REIT distribution to shareholders retains its character and is taxed on the shareholder’s return based on the type of distribution. Therefore, the taxation of a REIT distribution depends on whether the distribution is attributable to ordinary operating profits or to gain from the sale of real estate. A REIT’s ordinary income dividend distribution to a foreign investor from rental or interest income is not covered by FIRPTA. The ordinary income withholding tax rate for foreign shareholders is 30%, or lower (between 0% and 30%) depending on tax treaties. When a REIT sells US real estate and passes that gain to foreign shareholders through a capital gains distribution, the distribution is considered a gain from a USRPI and therefore subject to a 35% FIRPTA withholding tax.11 This type of REIT distribution is not reduced by tax treaty provisions. An exemption was enacted in 2004, providing that REIT capital gain dividend distributions are treated as ordinary dividend income if (1) the REIT is traded on an established securities market in the United States and (2) the foreign shareholder owns 5% or less of the REIT. The PATH Act increased this ownership threshold to 10%. This means that the withholding tax rate for capital gains distributions to de minimis foreign shareholders in publicly traded REITS ranges from 0% to 30% depending on tax treaties.
FIRPTA also imposes a 15% withholding tax on gain from a foreign shareholder’s sale of US REIT shares (10% prior to the PATH Act).12 This contrasts with the sale of other US securities, for which foreign shareholders are not taxed. Foreign shareholders can avoid this tax on the sale in two ways. First, gain from the sale of a US REIT by a foreign shareholder is not subject to FIRPTA if (1) the REIT is regularly traded on an established securities market and (2) the shares are sold by a foreign investor that owns 10% or less of the REIT (expanded from 5% by the PATH Act). Second, gain from the sale by a foreign shareholder is not taxed if the REIT owned is domestically controlled.13
In the 1970s, foreign investment in US real estate was perceived as harmful to the country, but since the financial crisis, there has been a growing opinion that foreign investment in US real estate could be beneficial. The Obama administration viewed foreign capital as a way to finance US infrastructure improvements and actively supported FIRPTA reform, finding that “foreign investors including large foreign pension funds regularly cite FIRPTA as an impediment to their investment in US infrastructure and real estate assets” (White House 2013). After several years of reform efforts, the PATH Act was enacted, making changes to FIRPTA that became effective on December 18, 2015. The stated purpose of these changes was to encourage additional foreign investment in US real estate and, in particular, to help provide foreign capital to fund much-needed investment in US infrastructure.14
The PATH Act included changes to FIRPTA that benefited all foreign investors in US REITs; relevant pre- and post-PATH Act tax rules are set out at Exhibit 1. The most significant change made by the PATH Act specifically targeted QFPFs by providing them with a complete exemption from FIRPTA. To be qualified, a foreign pension fund needs to meet certain requirements, including being part of a plan established to provide tax-free or deferred pension or retirement benefits to a broad-based class of current or former employees and satisfying certain reporting rules. Once qualified, a foreign pension fund is exempt from FIRPTA tax on gains from any type of USRPI, including gains from sale of stock of and distributions from public and private REITs.15 Foreign pension funds were previously able to avoid tax on sale of shares if they were a de minimis owner in a publicly traded REIT or if the REIT was domestically controlled. Now pension funds pay no tax on sale of shares regardless of the percentage of shares owned or the nature of the REIT. Foreign pension funds were not able to previously avoid withholding tax on capital gains distributions from REITs, although the tax rate could be improved from 35% to the country-specific ordinary income tax rate of between 0% and 30% for de minimis investors in publicly trade REITs. Now the withholding tax on REIT capital gain distributions for all pension funds from all countries is 0%.
Relevant FIRPTA Provisions
PATH ACT AND QUALIFIED FOREIGN PENSION FUND INVESTMENT
Since 2005, when REITs distribute capital gains from the sale of appreciated real estate to foreign pension funds, the capital gains withholding rate is equal to the ordinary income tax rate if the REIT is publicly traded and the foreign investor owns no more than 5% of the REIT. This rate varies by country tax treaty from 0% to 30%. However, as discussed in the previous section, the post-PATH Act capital gains withholding rate is now 0% for all QFPFs. Therefore, the effect of the PATH Act was to lower the withholding tax for some country foreign pension funds from as high as 30% to 0% (nonzero countries). For other country foreign pension funds, the withholding tax rate was already 0% so the pension funds from those countries received no rate decrease from the PATH Act (zero countries). In other words, rates varied for foreign pension funds in the pre-PATH Act period, but the rate is now zero for QFPFs in the post-PATH Act period. This leads to our hypothesis: Post-PATH Act foreign pension fund investment in US REITs increased more from nonzero countries than from zero countries, ceteris paribus.
We generally model our tests after Howard, Pancak, and Shackelford (2016) and begin by focusing on foreign pension fund investment as our main dependent variable. We calculate each pension fund’s investment in a given REIT by multiplying shares held by share price.16 We then sum investment on a country–REIT basis to obtain all foreign pension fund investment from a given country in a given US REIT for a given year. We then measure the annual change in investment by taking current year investment less prior year investment, scaled by prior year investment. Therefore, we capture the rate of growth in foreign investment year over year. We do this rather than simply measuring the level of investment to provide additional assurance that our results are not simply capturing an ongoing trend of increased foreign pension fund investment in US REITs over time, as documented by Hentov, Petrov, and Odedra (2018). Similar to Howard, Pancak, and Shackelford (2016), we also control for REIT sales and home-country gross domestic product (GDP), although we calculate those variables as annual changes instead of year-end levels to mirror how we calculate investment growth. Our sales variable helps control for differences in investment driven by differences in the REITs themselves. Similarly, our GDP variable helps control for differences in investment driven by differences in home country economic growth and investment opportunities.
We also create an indicator variable, Post, which equals 1 for observations occurring as of December 2015.17 Our hypothesis predicts that the response in investment depends on whether the foreign pension already enjoyed a 0% withholding rate prior to the PATH Act. Therefore we create a second indicator variable, Nonzero, that equals 1 for observations in which investor country pension funds were subject to a positive (nonzero) withholding tax rate prior to the PATH Act. We then interact that variable with Post to capture the investment from foreign pension funds in nonzero countries after the passage of the PATH Act. This gives us our main model:
Our coefficient of interest is β3. Based on our hypothesis, we expect β3 to be positive and significant, indicating that in the post-PATH Act era foreign pension funds from nonzero countries increased their investment in US REITs at a faster rate than foreign pension funds that already enjoyed a zero withholding tax rate prior to the PATH Act.
DATA
We assemble our data from a variety of sources. We obtain the investor name, country of residence, type (e.g., pension fund), and position (number of shares held for a given US REIT) from S&P Global.18 At the time the PATH Act was passed, there was some confusion about exactly which entities would qualify as a QFPF (e.g., Deloitte 2016). When gathering our data, we were aware that not all pension fund providers would necessarily be considered QFPFs. Therefore, we only included entities in our sample that seemed to be entirely focused on pensions so that our sample would be composed of entities that we could be relatively sure would be considered qualified (and likely considered themselves qualified). We eliminated investors from our larger sample that provided pension services, but in combination with nonpension services, because we believed they would not meet the qualification requirements. As a result, our sample is composed of 22 entities that focus solely on pension services.
S&P collects its position data from a variety of sources, some of which are only provided on an annual basis. Therefore, we structure our data on an annual basis. Compustat provides data on REIT total sales. The Center for Research in Security Prices provides data on REIT year-end stock price and number of shares outstanding. We obtain data on REIT property types and US withholding rates on foreign investment in US REITs from the National Association of Real Estate Investment Trusts. The World Bank provides data on annual country GDP. We restrict our sample to the years 2006 through 2015 because the American Jobs Creation Act of 2004 included exceptions to FIRPTA that were effective beginning in 2005, as documented by Howard, Pancak, and Shackelford (2016).
Exhibit 2 displays some descriptive statistics on the REITs in our sample. As mentioned earlier, our sample includes 129 unique REITs for a total of 881 unique REIT-years. Even though our sample focuses on publicly traded REITs, there is still variety in the sizes of those REITs, with total assets ranging from $216 million to over $33 billion. On average in our sample, foreign pension funds invest approximately $160 million in a given REIT and own just over 2% of outstanding REIT shares. Exhibit 3 shows that our sample includes a wide variety of types of REITs, with retail, office, healthcare, and lodging/resorts being the most popular.
REIT Descriptive Statistics
Notes: A total of 881 unique REIT–year observations for all variables. Sales, Total Assets, and Total Foreign Pension Fund Investment all in millions of dollars.
REIT Property Type
Exhibit 4 presents the number of observations for each country in our sample, divided between zero and nonzero countries. Zero countries make up a slightly larger portion of our sample. Although we have eight countries in our sample, Exhibit 4 shows that five of those countries (Canada, France, Netherlands, Norway, and Sweden) each account for over 15% of the sample and in total account for 95% of the sample.
Investor Countries
RESULTS
Exhibit 5 includes the results of our main tests. We first run our test without segregating our sample between zero and nonzero countries (column 1). The coefficient on Post is insignificant, indicating that in general foreign pension funds did not increase the rate of growth of their investment in US REITs after the passage of the PATH Act, all else equal. In column 2, we use our main model to parse out the response of nonzero countries. We see that now the coefficient on Post*Nonzero is positive and significant, providing support for our hypothesis and indicating that only foreign pension funds that did not already enjoy a zero withholding tax rate prior to the PATH Act increased their rate of investment in publicly traded US REITs in response to the Act. Given that we do not see an increased rate of investment from pension funds in zero countries (based on the coefficient for Post), these results also suggest that the existing tax treaties were an effective means of encouraging investment in publicly traded US REITs.
Changes in Foreign Investment from Nonzero Countries in the Post-PATH Act Period
Notes: The sample includes 2,972 investor country–REIT–year observations from 2006 through 2015. In columns 1 and 2, the dependent variable is ForInvtGrowth. In columns 3 and 4, the dependent variable is Country%OwnGrowth. The t-statistics, in parentheses, are based on standard errors clustered by investor country–REIT pair.
*, **, and *** indicate significance at the 0.10, 0.05, and 0.01 levels, respectively.
For the tests in columns 1 and 2 of Exhibit 5, we calculate foreign investment by multiplying number of shares held by year-end stock price. Therefore, an increase in investment growth could be driven by an increase in the number of shares held and/or an increase in share price. To determine whether the results in those initial tests are simply caused by rising REIT share prices, we replace our investment variable with percentage of outstanding shares owned, Country%OwnGrowth. We calculate country percentage ownership of a given REIT by taking the total number of shares in a given REIT owned by pension funds in a given country in a given year and scaling that number by the total number of outstanding shares for that REIT in that year. We then take the current year ownership percentage, subtract the prior year ownership percentage, and scale that difference by the prior year ownership percentage. The results are presented in columns 3 and 4 of Exhibit 5. There is a significant positive coefficient on Path*Nonzero in column 4, mirroring the results in column 2. This indicates that the increased investment from nonzero countries immediately after the passage of the PATH Act was not just the result of rising share prices but that foreign pension funds from nonzero countries actively acquired additional shares of US REITs.
In both columns 2 and 4, there is a significant negative coefficient on Nonzero. This indicates that nonzero country investors’ annual growth in investment, whether measured by dollars of investment or percentage of shares owned, was significantly lower compared to zero country investors in the period before the passage of the PATH Act. As mentioned earlier, the positive significant coefficients on Post*Nonzero in both columns indicate that immediately following the passage of the PATH Act nonzero country investors had significantly larger annual increases in their investment. We conduct tests using the Nonzero and Post*Nonzero coefficients in each regression to compare annual nonzero investor investment growth in the pre- versus post-PATH Act period. In both cases, we are unable to reject that the absolute value of the coefficients are equal. Therefore, although nonzero country investors exhibit significant annual investment growth following the PATH Act, it appears that growth was compensating for the relatively lower growth in the pre-PATH Act period.
We performed three additional tests to help confirm that our results are actually driven by the QFPF provision of the PATH Act. First, we reran our tests using a sample that excludes the financial crisis period. Our results remained qualitatively similar to those in Exhibit 5. Second, we reran our tests using data from non–pension fund foreign investors (primarily foreign investment banks). Although there were provisions of the PATH Act that affected non–pension fund foreign investors, those provisions did not vary according to our nonzero variable. Therefore, we would not expect a similar increase in rate of investment from these non–pension fund investors in nonzero countries. As expected, the coefficient on Post*Nonzero was insignificant in all specifications. Finally, we believe that it is reasonable that sovereign wealth funds (SWFs) set up as a government pension fund qualify as a QFPF. However, it is also possible that REIT capital gain distributions to SWFs with de minimis REIT ownership already qualified for exemption from taxation before the PATH Act per IRC provisions related to sales of US real property and taxation of foreign governments. To the extent this is true, including SWFs in our sample would only bias against our finding results because the QFPF provisions of the PATH Act would not provide any incremental tax benefit to these entities. However, to ensure SWFs are not driving our results, we reran our tests after removing potential SWFs from our sample. Our results remain qualitatively similar. The results of all of these tests further support the conclusion that the PATH Act drove additional investment in publicly traded US REITs specifically from QFPFs in nonzero countries.
CONCLUSION
Real estate is in large part a tax-driven industry. Consequently, changes in US tax policy have an impact on the relative attractiveness of real estate as an investment. However, very little evidence exists regarding the effect of taxes on real estate investment. This study expands the real estate literature by looking at the magnitude and speed of the response of real estate investors to changes in taxes. Although we feel there is much to be learned by exploring this one specific tax policy change, our hope is that this research serves as an impetus toward the goal of more fully understanding the impact of taxes on real estate investment.
Specifically, we looked at the impact of the 2015 PATH Act provision that exempts QFPFs from FIRPTA capital gains withholding taxes. This change had the effect of reducing capital gains withholding taxes by as much as 35% depending on investor position and home country tax treaties (0%–30% for de minimis investors). We find evidence consistent with foreign pension fund investors in publicly traded US REITs responding to the 2015 reduction in capital gains withholding taxes, with the motivating factor being country-specific tax rate decreases.
We interpret these results as evidence that nonzero FIRPTA capital gains tax rates restrict the flow of foreign capital into publicly traded US REITs.19 We believe that this is the first empirical analysis of the PATH Act changes to FIRPTA and presume that it will be informative to investors, industry groups, and US policymakers when evaluating the implications of this legislation and in developing any additional future modifications to FIRPTA. The results from this investigation would imply that further exemptions from FIRPTA among other classes of foreign investors could also increase the flow of foreign capital to US real estate.
ADDITIONAL READING
A Global Perspective on Pension Fund Investments in Real Estate
Aleksandar Andonov, Nils Kok, and Piet Eichholtz
The Journal of Portfolio Management
https://jpm.pm-research.com/content/39/5/32
ABSTRACT: Real estate is the third-largest asset class for institutional investors, but research into the performance of pension fund investments in real estate is scant. This article employs a unique set of data, the CEM database, to study the real estate investment approach, cost, and performance for a global panel of almost a thousand pension funds. The results show that smaller pension funds invest primarily in direct real estate, through external managers and fund-offunds, and are likely to disregard listed property companies in their real estate allocation. Larger pension funds are more likely to invest in real estate internally, exhibit lower costs, and achieve higher net returns. Internal real estate investment management is associated with lower costs and disproportionally higher returns. Finally, we document that US pension plans’ real estate investments significantly underperform both their benchmarks and their foreign peers.
New Horizons and Familiar Landscapes: New Capital Sources Confront Shifting Real Estate Fundamentals
Jim Clayton, Frank J. Fabozzi, S. Michael Giliberto, Jacques N. Gordon, Youguo Liang, Greg MacKinnon, and Asieh Mansour
The Journal of Portfolio Management
https://jpm.pm-research.com/content/41/6/11
ABSTRACT: Real estate has continued to evolve as an asset class. In this introduction to the special issue, the authors examine new sources of capital that have entered real estate, some of which have already had a large impact and others that have the potential to bring major changes in the future. While the capital market foundations of the asset class are changing, the traditional cycle in property fundamentals remains a distinguishing feature, albeit changed from past cycles. As new sources of capital encounter real estate’s fundamentals cycle, in-depth understanding of the asset class is imperative. The authors review these trends and suggest how the other articles in this special issue can help investors navigate this confluence of new and old.
ENDNOTES
↵1 Certain high-income individuals may also be subject to an additional 3.8% net investment income tax on capital gains.
↵2 The term domestically controlled is defined to mean that less than 50% of the value of the REIT has been owned (directly or indirectly) by foreign shareholders during the five-year period ending on the date of stock sale.
↵3 This exemption is set out at 26 US Code section 897(l). This subsection provides that a QFPF is not considered a foreign investor for purposes of the FIRPTA law at 26 US Code section 897. To summarize: to qualify, an entity has to be established under the laws of a foreign country to provide employee pension benefits. No beneficiary can have a right to more than 5% of the fund’s assets or income, and the fund must be subject to regulation by the country’s tax authorities. Contributions to the fund must be tax deductible (or taxed at a reduced rate), or the fund income must be tax deferred (or taxed at a reduced rate).
↵4 In fact, the link between taxes and foreign investment in any asset type has not even been widely explored. Amiram and Frank (2016) found that relatively favorable tax policies on the dividend income earned by foreign investors are associated with larger amounts of foreign portfolio investment. Edwards and Shevlin (2011) reported on how changes in Canadian tax law affected the value of publicly traded income trusts to nonresident investors.
↵5 In addition, Ennis and Burik (1991) acknowledged that the absence of taxes was a shortcoming in their pension fund investment model, noting that “[r]eal estate investment is characterized by…tax advantages embedded in some property investments. In all likelihood, these factors tend to impede pension fund investment in real estate.” Andonov, Kok, and Eichholtz (2013) provided a look at global pension fund investment in real estate but did not comment on taxes.
↵6 Presumably, there are more than eight countries with pension funds investing in US REITs. However, as many industry experts and investment advisors noted (e.g., Deloitte 2016) at the time of the passage of the PATH Act the exact application of the QFPF requirements was still unclear. Therefore, in our sample, we only use data for “pure” pension funds (i.e., we do not include investment advisors who offer pensions as part of their portfolio of services) that most closely meet the QFPF requirements to try to ensure we are capturing those entities that were fairly certain of QFPF status in December 2015. See the “Data” section for additional details on our sample.
↵7 The 5% refers to the threshold for treating the capital gain distribution to a foreign investor as an ordinary distribution. Depending on the foreign investor’s home county, there are then requirements that must be met for the ordinary distribution to qualify for the 0% rate. Those requirements are generally less restrictive (e.g., a 10% ownership threshold), so if a foreign investor qualifies for ordinary dividend treatment, that investor also qualifies for the 0% tax rate.
↵8 We note that our findings related to FIRPTA constraining foreign pension fund investment in publicly traded US REITs does not necessarily mean that the 2015 exemption increased overall foreign pension fund investment capital in US real estate. Foreign investors, including pension funds, are known to have participated in complex tax planning to address FIRPTA concerns. For example, blocker structures were used to reduce FIRPTA-related taxes and eliminate the FIRPTA-related requirement for a foreign investor to file a US income tax return. These structures were costly to implement and still required that the blocker pay US income taxes. After the tax law reform, foreign pension funds may have merely moved investments from some form of indirect ownership in US real estate assets to direct ownership in publicly traded US REITs.
↵9 REITs are entities that own, operate, and/or finance income-producing real estate. An entity must meet certain regulatory requirements to qualify as a REIT, including investing a certain percentage of its total assets in real estate and distributing a minimum percentage of income to shareholders in the form of dividends. Investors buy shares in REITs, which can be publicly traded or privately traded, providing a liquid, diversified, and professionally managed real estate investment.
↵10 As discussed later, a capital gains distribution is treated as an ordinary income dividend if certain requirements are met, in which case a foreign investor would not have an obligation to file a US tax return.
↵11 FIRPTA introduces additional complexities to this type of income. Specifically, capital gains are also treated as income that is “effectively connected with” the conduct of a US trade or business (effectively connected income [ECI]). Foreign investors that receive ECI have an obligation to file a US federal tax return and become subject to the subpoena powers of the US Internal Revenue Service with respect to all of its US investments. Foreign investors reportedly go to great lengths to avoid the requirement of filing a US tax return.
↵12 REITs may also have cash to make distributions greater than their taxable income because real estate depreciation is a significant noncash expense taken into account when calculating income. This type of distribution is deemed a return of a shareholder’s original investment and referred to as a return of capital. For US shareholders, it is not taxed as ordinary income but reduces the tax basis of shares by the amount distributed. In general, FIRPTA views income received from the return of capital similarly to the sale of REIT stock.
↵13 Our research has revealed that most publicly traded US REITs appear to be domestically controlled; however, it is not clear how a foreign investor could make such a determination, especially given that the definition refers to shares being held directly or indirectly. This likely created uncertainty for foreign investors as to whether they could rely on this exemption. The PATH Act provided investors with greater certainty in making this determination by creating a presumption that all shareholders holding less than 5% of a REIT are US persons unless the REIT has actual knowledge to the contrary.
↵14 For a more in-depth discussion of PATH Act changes, see Lipton et al. (2016).
↵15 Note that foreign investors would still be subject to an ordinary income withholding tax arising from rents and other sources of ordinary income.
↵16 In recognition of the 5% de minimis threshold pre-PATH Act, when analyzing shareholder data, we drop observations with foreign investors with ownership over 5% before calculating our total investment for a given country–REIT pair. In practice, we have very few holder–REIT–year observations (less than 1% of our sample) with foreign ownership above 5%. One reason we expect so few 5%+ observations is that foreign investors are aware of the tax consequences and therefore tried to remain below this 5% threshold. Including these observations in our country ownership totals does not affect our inferences or qualitatively change our results.
↵17 All data are measured at the end of a given month, so all December 2015 observations are as of December 31, 2015, after the enactment of the PATH Act on December 18, 2015.
↵18 Where S&P data on investor type are missing, we supplement with manually collected data.
↵19 We point out, however, that the results do not necessarily indicate that foreign pension funds increased net inbound foreign pension investment in US real estate; the change may have merely encouraged foreign pension funds to move investments from some form of indirect ownership in US real estate assets to direct ownership in publicly traded US REITs. If so, our results still suggest that US publicly traded REITs may be a more efficient means for at least some foreign investors to invest in US real estate.
- © 2019 Pageant Media Ltd
Don’t have access? Register today to begin unrestricted access to our database of research.
Log in using your username and password
In this issue
