This blog was updated on March 8.
Fifteen years ago, a 23,000 square foot mansion in Washington D.C.’s swanky Embassy Row was bought for $15 million in cash. The purchase was made through a Delaware-registered shell company, a corporation that existed on paper, but had no real physical presence. Buying the mansion through a company kept the identity of the ultimate owner out of public records, with even those selling the estate remaining in the dark as to who they were. It was only a decade later that reporting by the Washington Post revealed that the true owner of the property was Russian oligarch Oleg Deripaska, a close ally of Vladimir Putin. Deripaska, who was once the richest person in Russia, also picked up $47 million worth of property in Manhattan through companies he controlled. He was subsequently slapped with sanctions by the U.S. Treasury in 2018 due to Russia’s ongoing instigation of violence in Eastern Ukraine.
Nonresident Fellow - Global Economy and Development
Florian M. Hollenbach
Associate Professor in Business & Politics - Copenhagen Business School
Assistant Professor of Political Science - George Washington University
Research Fellow - Higher School of Economics
It isn’t just Russian oligarchs who have shown an interest in luxury U.S. real estate. In 2011, a Manhattan penthouse once owned by Jay-Z and Beyoncé was bought for just over $30 million. Again, the purchase was made through a shell company, obscuring the fact that the true owner was Jho Low, a Malaysian national who was later found to have stolen around $4.5 billion from the 1MDB, the country’s sovereign wealth fund. The penthouse purchase was just one of many high-value properties across the U.S. that Low accumulated during the 2010s before his theft of Malaysian public funds was exposed and he became an international fugitive.
Stories like these have become all too common in recent years as real estate markets in countries like the U.S. and the U.K. have become prime destinations for illicit wealth. Real estate is an attractive investment for money launderers: It offers a safe return, the uncertainty around its true market value makes it useful for hiding illicit proceeds, and in many countries the sector is often subject to few, if any, anti-money laundering provisions. However, because of the secrecy shrouding this behavior, it is hard to know how much illicit money makes its way into real estate. In a recent study of known money laundering cases in the U.S., the think tank Global Financial Integrity found that at least $2.3 billion was laundered through U.S. real estate between 2015-2020, which they suggest is likely to be only the tip of the iceberg.
Real estate markets in countries like the U.S. and the U.K. have become prime destinations for illicit wealth.
As long as real estate markets in advanced economies remain vulnerable to abuse by corrupt actors and foreign kleptocrats, there will be little to dissuade powerful people ranging from Ukraine to the DRC from siphoning off public funds and stashing them abroad.
Enter FinCEN’s Geographic Targeting Order program
In January 2016, in response to the threat that money laundering poses to the real estate sector, the Financial Crimes Enforcement Network (FinCEN) announced it would introduce a real estate Geographic Targeting Order (GTO) program in both Miami-Dade and Manhattan counties. Beginning that March, the GTO program required that title insurance companies collect information on the ultimate beneficial owner of any company that bought a residential property in cash (i.e., without a mortgage). The program only applied to purchases above a set price threshold: $3 million in Manhattan and $1 million in Miami. It only targeted cash transactions, since those using a mortgage would already be subject to extensive scrutiny by banks, which are already required to perform due diligence on their clients. Later that same year and then again in 2018, FinCEN expanded the GTO program to 20 other counties, covering urban areas in nine different states and a large proportion of the U.S. housing market.
The program was initially seen as a success: When announcing its renewal in 2017, FinCEN reported that roughly 30 percent of reported real estate transactions involved beneficial owners who had also been subject to suspicious activity reports by banks, an indication the program was well targeted. The program also drew substantial support from Senator Marco Rubio, who wanted to expand the program nationwide, after an academic working paper indicated that the dollar volume of purchases by companies in cash declined by 66 percent, or $45 billion dollars, following the reform across the U.S.
The ostensible goal of the GTO program was not to deter people parking their dirty money in U.S. property, but instead to gather information that could help FinCEN better understand this kind of illicit behavior and then be shared with other law enforcement agencies. But a program of this scale should also produce deterrence effects: If people looking to launder money believed there was a chance that their information would be passed on to law enforcement, it should create a strong incentive to buy property anywhere the policy isn’t in place.
Deterrence effects of this kind aren’t just a pipe dream—they’ve been observed in other contexts where beneficial ownership reporting occurs.
Deterrence effects of this kind aren’t just a pipe dream—they’ve been observed in other contexts where beneficial ownership reporting occurs. When banks in tax havens were forced to gather information on the ultimate owners of their accounts, deposits fell between 11-22 percent as people looked for safer places to stash their cash. In the U.K., after Scottish Limited Partnerships were identified as a legal vehicle prone to abuse by money launderers, beneficial ownership reporting was introduced, and new registrations fell by more than 70 percent overnight. When the threat of reporting and follow-up is real, people relocate illicit cash to greener pastures. And it appeared—based on early research into the GTO program—this is what happened.
But this is not the whole story.
Measuring the impact of the GTO program on purchases normally shrouded in secrecy
In a new working paper, we ask whether—six years after its introduction—the GTO program led to deterrence effects in the 22 American counties where it was introduced. To do this, we obtained data from Zillow, the online real estate company, on all residential property purchases across the U.S. between 2015 and 2019 (just before the advent of the pandemic). We focused our main analysis on purchases by companies made without a mortgage (what we call corporate-cash transactions). Because these are the exact kinds of purchases being targeted by the GTO program, we would expect there to be declines in these types of transactions, particularly those above the price thresholds set by FinCEN. However, we’d expect little response in purchases that weren’t being targeted by the program, such as those involving individuals taking out a mortgage.
Using new methods from the difference-in-differences literature, we find that, in aggregate, there was a nonsignificant decline of corporate cash purchases by about 4.3 percent after the GTOs were announced, but that this decline was also matched by a nonsignificant decline in individual mortgage sales of approximately 5.6 percent (Figure 1). When we drill down on corporate cash purchases above the threshold, we actually find a positive (but also nonsignificant) effect of around 5.5 percent. We also fail to find any effect on aggregate price levels: Using Zillow’s Home Value Index, we fail to find any evidence that the price of top-tier homes shifted following the introduction of a GTO.
Figure 1. The Geographic Targeting Orders do not have a substantive impact on corporate cash sales after their announcement
Note: The above figure shows event-study estimates using Callaway and Sant’anna’s doubly robust difference-in-difference estimation. The method outcome is the estimated average treatment effect on the inverse hyperbolic sign transformation of the number of corporate cash purchases of residential properties in treated counties.
Our null findings stand in stark contrast to earlier work finding that the GTO program heralded a nationwide chilling effect on corporate cash purchases. In reconciling our results with that earlier paper, we discovered that its headline result was likely driven by the fact that not every company in the original Zillow data was being accounted. When we correct for these issues and identify every likely company in the data, we find no evidence of a general chilling effect on corporate cash transfers, nor one in the counties being targeted (Figure 2).
Figure 2. The impact of GTOs on corporate cash purchases becomes positive (but insignificant) when focusing on the high value transactions targeted by the program
Note: The figure shows the impact of the GTO announcement on corporate cash purchases when we consider all purchases (the first estimate) versus when we focus only on purchases at price levels that were actually targeted by the program (the second and third estimates).
Even though we found no aggregate evidence that the GTO program led to changes in corporate cash purchases, we investigated whether people started buying more properties using trusts, relied on mortgages from banks with a poor history of anti-money laundering compliance, or avoided using title companies all together. We also investigated whether people tried to avoid the reporting thresholds by buying properties just below them. In all these cases, we found no systematic evidence that people used these methods to circumvent the policy. We also examined the counties of Miami and Manhattan more closely. Since the policy was first introduced in these counties and many media outlets have reported on the substantial amount of dirty money flowing into luxury real estate, we expected a uniquely large effect in these markets. Again, we found nothing conclusive. In fact, we found no change at all in the percentage of total price volume from corporate all-cash purchases in reaction to the GTOs. Finally, we failed to find any evidence that the GTOs led to any net change in suspicious behavior, as measured by suspicious activity reports filed by banks.
What the non-effect of the GTOs tells us about the program expansion and about beneficial ownership more generally
Why didn’t the GTO program lead to a reduction in corporate cash purchases or price levels in affected markets? There are two possibilities: 1) that the volume of illicit money moving through American property markets is too small to be detected by our analysis, or 2) that the policy was not implemented sufficiently well and illicit actors did not change their behavior.
The first possibility is that the volume of illicit money moving through American property markets is not as large as commonly believed. Analysis like ours that uses aggregate data to measure effects might not be precise enough to detect the small number of malicious actors that changed their behavior once beneficial ownership reporting was introduced. We think this is quite unlikely, particularly because we find little change in markets commonly accepted as rife with illicit wealth, such as Miami-Dade county. Roughly one-half of all real estate transactions in Miami continue to be made by corporations using only cash, a number that has not budged throughout the GTO program.
There are several reasons why we think imperfect implementation better explains the lack of behavior change. First, although title insurance companies were required to gather the information, there is evidence that their implementation of the program was less than comprehensive. FinCEN did not begin examining title company compliance until three years after the introduction of the program, and our own data suggest that, as of 2018, only 87 percent of eligible property purchases were reported through the program.
Even if the reporting was comprehensive, it isn’t clear whether the information being provided to FinCEN was accurate. Title insurance companies weren’t required to verify whether the reported beneficial owner was the true owner, making it relatively easy for a corporation to continue hiding the true owner by providing a proxy beneficial owner. This is a problem that plagues even public beneficial owner registers (when Luxembourg opened its register in 2019, a number of questionable entries appeared, including a major coal company that was owned by a 1-year-old). The verification of the data might fall on FinCEN itself if it weren’t already inundated with suspicious activity reports from other parts of the economy. Taken together, it is possible that the GTOs didn’t create a deterrent response because either people weren’t required to submit their information or could easily just provide another name.
However, deterrence was never the stated goal of the GTO program. Instead, FinCEN intended it to be an information-gathering exercise to help it better understand the markets targeted by illicit actors and to pass on credible intelligence to other law enforcement agencies. Thus, the surveillance value of the program may itself justify its existence. But for the U.S. Treasury to make a convincing case that the intel generated by the program is substantial, it needs to show how GTO data has assisted law enforcement in prosecuting known money laundering cases.
In the wake of Russia’s invasion of Ukraine, the need for high quality ownership information is greater than it has ever been before. For example, the U.S. Department of Justice recently announced a new task force—dubbed KleptoCapture—aimed at tracking down and seizing oligarch wealth. But the success of these initiatives will depend on, and will be limited by, the quality of the information already garnered by the GTO program. As FinCEN looks to expand the program nationwide, it might also consider moving to a permanent property register of ultimate ownership similar to what the U.K. government has committed to implementing. This would provide it with a permanent, updated database of ownership, rather than the snapshots provided by the GTO program.
Beneficial ownership regimes are a cornerstone in the fight against dirty money. But their introduction is not a panacea: Better implementation and evaluation will be crucial to their success.
The GTO program is now set to be expanded to the entirety of the U.S. property market, including both residential and commercial real estate. Given that our evidence suggests that GTOs have had no discernible effect on house prices, it doesn’t appear that the program has substantially affected the lives of either buyers or sellers. That means that while expanding the program nationwide will entail additional costs for FinCEN and for title insurance companies, it won’t be disruptive.
Beneficial ownership regimes are a cornerstone in the fight against dirty money. But their introduction is not a panacea: Better implementation and evaluation will be crucial to their success. Only then can programs like the GTO regime gather good, actionable intelligence and—eventually—dissuade corrupt and criminal actors from abusing our markets.