What Share of Institutional Real Estate Is Invested in the United States?

No single share covers all institutional real estate capital, because the answer depends on the investor type and on what is counted. Sovereign wealth funds put about half of their real estate deployment into the United States in one annual period, and nearly 40% of institutional direct purchases over a longer span were U.S. assets. Portfolio holdings, capital flows and deal counts are different ratios, so the figures describe different things and do not average into one allocation.

A U.S. share is a ratio, and its value moves with the numerator and the denominator. Holdings divide U.S. asset value by total asset value at a point in time. Deployment divides U.S. capital put to work by total capital put to work during a period. Transaction share divides the count of U.S. deals by the count of all deals. An investor-type cut, such as sovereign funds or pensions, changes who sits in the denominator.

Holdings are a stock and deployment is a flow. A portfolio built over many years carries the geography of old decisions, while one period of deployment reflects only current activity.

A transaction count gives a small deal and a large deal the same weight, so it can sit far from the dollar share.

“If we look at real estate in particular, as you can see the chart, the US represented about half of all the deployment into real estate.”
Park Madison Partners

One flow, one portfolio and one unlisted sleeve: three U.S. figures that coexist

Sovereign wealth funds deployed about half of their real estate investment capital in the United States in one annual period. That is a flow figure for one investor type.

A pension real estate portfolio can sit at about 80% U.S. and 15% to 20% outside the U.S., with only a third of it in core strategies and the rest non-core. That is a holdings figure for a single portfolio, and its strategy mix is part of what shapes the geography.

A global sovereign fund shows a third pattern. Its real estate was just over 3% of the fund at 208 billion kroner, and within the unlisted portion 46% was in the U.S. against 54% in Europe. The 46% is a share of unlisted real estate only, so it leaves out listed holdings, and it sits below half.

These figures do not conflict. About half, 80% and 46% measure a flow, a portfolio and an unlisted sleeve for different investors, and the gaps between them reflect those differences.

Transaction counts: nearly 40% of institutional direct purchases

Over its full measured span, nearly 40% of direct real estate transactions purchased by institutional investors were U.S. assets. That is a count of deals across a long window, so it blends many market cycles and carries no information about dollar size.

Family offices show a heavier domestic tilt in deal counts. In one dataset, North America took 69% of direct transactions and Europe took 25%. All remaining regions combined made up 6%.

In one half-year tally, 50 of 55 family office real estate transactions were in the United States.

Family offices are a separate population with a different denominator. Their counts sit beside the institutional figures and do not average with them.

North America is a wider label than the United States

In one annual allocations survey series, North America remained a major destination for institutional real estate capital flows, with market scale and liquidity cited as the reasons.

The region includes countries other than the United States, so a regional flow observation gives context without yielding a U.S. percentage.

A clean comparison matches the geographic category, whether that is the U.S. alone against the rest of the world or North America against other regions. It also matches the location basis, which is where the asset sits against where the investor is domiciled.

Why ticket size separates deal count from capital

Practitioner estimates put institutional check sizes at $25 to $100 million, with sovereign wealth funds needing at least $100 million per deal.

A fund with a $100 million floor cannot complete many small deals, so its deal count stays low and each deal carries heavy dollar weight. A count-based statistic gives a small purchase the same weight as a large one.

The gap between deal share and dollar share therefore widens as deal sizes within a population spread out. The worked example below shows the two measures pointing in opposite directions.

Worked example: one year of deployment counted by deals and by dollars

Take a hypothetical institution with $800 million of real estate, of which $480 million is in the U.S. The starting U.S. holdings share is 480 divided by 800 multiplied by 100, which is 60 percent. During the year it deploys $160 million across 8 deals. The figures are illustrative arithmetic and are not market estimates.

Base case: 5 of the 8 deals are in the U.S., and they total $60 million. The deal share is 5 divided by 8 multiplied by 100, or 62.5 percent. The capital share is 60 divided by 160 multiplied by 100, or 37.5 percent. After deployment, U.S. holdings are 480 plus 60, which is $540 million, and total holdings are 800 plus 160, which is $960 million. The holdings share becomes 540 divided by 960 multiplied by 100, or 56.25 percent. That is 60 minus 56.25, a decline of 3.75 points.

Alternative case: 3 of the 8 deals are in the U.S., and they total $96 million. The deal share is 3 divided by 8 multiplied by 100, or 37.5 percent. The capital share is 96 divided by 160 multiplied by 100, or 60 percent. U.S. holdings become 480 plus 96, which is $576 million, so the holdings share is 576 divided by 960 multiplied by 100, or 60 percent.

Both cases use the same $160 million, the same 8 deals and the same period. The base case reads 62.5 percent by count and 37.5 percent by dollars. The alternative reads 37.5 percent by count and 60 percent by dollars. The count ranks the two cases in the opposite order from the dollars, and only the dollar-weighted measure explains why holdings fell in one case and held steady in the other.

Treating a deal share as a capital share changes the unit. Nearly 40% of direct purchases says how many deals were U.S. assets. Converting that count into dollars needs transaction values, and the worked example shows a count of 62.5 percent beside a capital share of 37.5 percent.

Mixing a stock with a flow produces a ratio that matches neither. A holdings split such as 46% U.S. within unlisted real estate reflects decisions accumulated over many periods, while a deployment share of about half reflects one period of new capital.

Carrying a figure from one population to another changes the denominator. The sovereign deployment share covers sovereign wealth funds, and the 69% North America share covers family office transactions. Neither describes pensions or all institutions.

Reading North America as the United States overstates precision. A regional flow observation is wider than a U.S.-only number, and the regional label hides the split among its member countries.

Taking the unlisted share as the portfolio share leaves out the listed holdings. The 46% is a share of unlisted real estate, so the U.S. share of the full real estate allocation can differ.

Is the figure a holding, a flow or a count of deals? Each has its own numerator and denominator.

Which investor population does it cover? Sovereign funds, pensions, institutional direct buyers and family offices produce different pictures.

What period or span does it cover? A single annual period and a multi-year span are not interchangeable.

Does the geography mean the United States or North America? The wider region contains countries other than the U.S.

Does the figure cover listed holdings, unlisted holdings or both? A share of an unlisted sleeve describes only that sleeve.

Are deal values available alongside the deal count? Without them, a count cannot be converted into a capital share.

Do the check sizes in the population explain the gap between count and dollars? A high minimum ticket concentrates capital in fewer deals.