Menu
Menu
A group of nonresident individuals want to pool their money to buy commercial real estate in the United States.
They are going to buy decent-sized properties — commercial and office properties in the $5 million to $50 million range. This is serious money but not crazy huge.
How should they set this up for optimum U.S. tax results?
The factors to consider are:
There are four categories of solutions:
I will add more on each of these structures below.
I will analyze the four choices one at a time. This discussion will not be an Encyclopedia of Real Estate Investment. Rather, I am going to identify a knockout punch — a fatal flaw — that will mean you should select a different option.
This leads to a “last man standing” result, in my opinion, that favors a structure using multiple trusts.
In summary:
| Structure | Knockout Punch |
|---|---|
| REIT | Very high legal, accounting costs |
| Partnership | Estate tax risk |
| Corporation | Higher capital gain tax |
| Trusts | Not perfect, but the last man standing |
Disclaimer. This is my opinion only. I am biased in favor of trusts because that’s what I know best. I am biased. 🙂
Real Estate Investment Trusts are a creature of tax law, designed for exactly the situation we are describing.
There are some potentially favorable U.S. tax results for nonresident investors who purchase shares of certain types of REITs. However, I will not talk about those benefits today, because I am looking for the knockout punch.
And the knockout punch for a REIT is complexity and cost.
Among other things, a REIT must have at least 100 shareholders. This, along with the standard legal and tax overhead of setting up and operating the REIT, is simply too much for a smaller investment pool.
Then consider the distraction cost. Setting up and operating a REIT will take a certain quantity of executive brainpower. Typically, an investment pool has someone who creates and manages it. That person’s time and attention is valuable, and should not be squandered lightly. I say that legal and tax structures are a waste of time for such an executive.
Yes, there are “REIT in a box” vendors who will do everything for you. Yes, you can drive costs down to a certain extent. But at some point, there is an irreducible financial and attention price that you must be willing to pay.
At some point the investment pool will be large enough to justify the expense. Postpone the change to REIT status until it is irresistible.
Limited partnerships have long been used by individual investors to pool capital and buy real estate. Limited liability companies — which provide the same tax results — are now widely used, due to familiarity and a slight reduction in complexity.1
Partnerships are unacceptable for nonresident individual investors. The reason: estate tax exposure.
A nonresident individual will not want to expose his or her U.S. real estate investment to a tax of 40% of the investment value on date of death.
The U.S. estate tax rules say that a nonresident will only face the 40% estate tax on “U.S. situs assets”. This is the way the law describes “assets located in the United States”.
The critical question we need to answer is “What did the nonresident investor own (as that term is understood in U.S. tax law) when he died? And the followup question is “Well, is that taxable for estate tax?”
U.S. real estate is located in the United States. Easy answer: taxable.
But a nonresident investor owns a partnership interest, not a direct piece of the U.S. real estate. This is impossible to answer, because no definitive answer exists.
Investors do not gamble on their interpretation of tax law.
The nonresident investor may treated as owning (for estate tax purposes) a proportionate share of the partnership’s real estate. (Hint: this is the “aggregate theory of partnership taxation”).
If the nonresident investor is treated as owning (for U.S. estate tax purposes) a proportionate share of the U.S. real estate, then we know it is game over.
At death, the nonresident investor has U.S. real estate that will be subjected to the estate tax, and 40% of the value of the investment will evaporate into tax. This is true because we know that U.S. real estate is a “U.S. situs” asset (located in the United States).
A nonresident investor can be treated — for estate tax purposes — as owning a partnership interest, not a proportionate share of the partnership’s real estate. (Hint: this is the “entity theory of partnership taxation”).
Problem: where that partnership interest located? If the partnership interest is located in the United States, then the value of the partnership interest will be subjected to U.S. estate tax. If the partnership interest is located outside the United States, then the value of the partnership interest will not be subjected to U.S. estate tax.
“Where is it?” is not an easy question to answer. There are, in fact, multiple competing theories.
So we come down to two theories that are fatal, and two theories that are uncertain. This is not the way a smart investor operates.
As you spend more time looking at real estate holding structures, you will come across ideas for multiple layers of partnerships. At the bottom is a U.S. partnership that owns U.S. real estate. All of the partners in that U.S. partnership are foreign partnerships. Those foreign partnerships might themselves be owned by more foreign partnerships.
The idea is that by adding layers you are adding abstraction. At some point you can argue that the individual nonresident partner can die without risk of U.S. estate tax, using the any one of the four theories of taxation:
I have seen no public evidence that the Internal Revenue Service has accepted a multi-level partnership structure as an accepted firewall against estate tax exposure. No Tax Court cases, no Revenue Rulings, no nothing.
Unless the nonresident investor is willing to accept estate tax uncertainty, I think the partnership structure should not be considered. The investor, I think, needs a firewall against the estate tax risk.
Partnerships are a useful business tool for real estate investments. But for nonresident individual investors, they should not be used alone. The nonresident investor’s estate tax risk can be eliminated by the addition of corporations or trusts as partners, or paid (in the event of death) by purchasing life insurance.
We come now to a structure that will provide a firewall against estate tax risk.
Remember that the estate tax risk is analyzed by looking at what the nonresident individual owned at the moment of death.
The idea here is that a group of nonresident individuals contribute capital to a non-U.S. corporation. That capital is then used to buy U.S. real estate.
Consider a nonresident who owns shares of a foreign corporation, which in turn owns U.S. real estate. (This is a highly simplified and not recommended strategy but I am using it to demonstrate the point of protection from estate tax).
When that person dies, he or she owns shares of that corporation. As a matter of U.S. tax law, we do not look through the foreign corporation to identify its assets. We do not treat the shareholder as owning a percentage of the corporation’s assets, even if those assets are all in the United States.
This means that a foreign corporation is an effective firewall against U.S. estate tax.
But this firewall comes at a cost. A capital gains tax cost.
Nonresident individuals who own an asset for more than one year will be tax on capital gains at 20% rather than the maximum 39.6% ordinary income tax rate. This is the “long term capital gain” tax rate.
By contrast, all corporations (foreign or domestic) will pay the same tax rate on regular income and capital gain. There is no special lower tax rate for assets held by corporations for more than one year. The tax rate varies by the amount of taxable profit, but let’s put the average tax rate somewhere in the mid-30% range.
In other words, the price for using a foreign corporation to eliminate U.S. estate tax is an extra $15 tax for every $100 of capital gain when you sell the property.
This, to me, is a fatal flaw. Especially for large investments, the extra capital gain tax on sale will be a large number. My guess is that the price you pay in extra capital gains tax will be greater than the value of the estate tax protection. Or let’s put it another way — you could probably buy life insurance for less.
We come to the Last Man Standing: using irrevocable trusts as an estate tax firewall, while taking advantage of the lower long term capital gain tax rates that apply to individuals.
The disadvantage? It is paperwork-heavy and expensive to operate.
The idea is simple. Each investor creates an irrevocable trust, naming family members (but not the investor himself) as a beneficiary. Each trust then becomes a partner in a U.S. limited partnership. The U.S. limited partnership buys commercial property.
When the property is sold, the partners receive their proportionate share of the capital gain, and are taxed at individual tax rates and therefore pay the 20% capital gain tax rate.
There is no estate tax when a nonresident investor dies. Indeed, the death of any beneficiary will also be without estate tax. A properly-constructed trust will protect its assets from the estate tax.
Like REITs, a “partnership owned by a bunch of irrevocable trusts” structure has the disadvantage of expense. Set-up costs are substantial, and operating costs are substantial as well.
These costs can be reduced. For a “big enough” group of trusts and highly standardized trust documents, it is possible to get economies of scale and reduced costs for trust administration by an independent trustee. Even better, use a beneficiary as a trustee, with suitable limits on that person’s power to make distributions to himself or herself.
Tax return costs are unavoidable, but again — if you do not permit customization of the trusts — economies of scale can be achieved.
The additional cost (and complexity) of this structure must be weighed against the expected benefit from its use: the 20% long term capital gain tax rate on sale. Let’s say for every $1,000,000 of capital gain there is a $150,000 tax reduction. For a bigger property that is expected to generate a lot of capital gain — let’s say $10,000,000 — some of the expected tax savings can be used to pay the overhead of operating the structure, with the extra tax savings going directly to the investors.
Setting up a trust requires an outright gift to the trust from the investor. This means the investor no longer has any claim on the money. This may create a tax event in the home country. It may also be unacceptable to the investor.
After all, Mr. Big did not become Mr. Big by giving away money and giving up control of his investments. He will want to control the investment and get his money back, as well as enjoy all of the tax benefits from giving up control and permanently giving away his money. More overhead and complexity is required.
Many countries simply do not have trusts within their legal or tax system. Maybe there is a clear treatment of a distribution from a trust to a resident of that country. Maybe there is no clear answer at all. Using a trust for an investor from such a country may require additional work or may not be workable.
There is no perfect solution. My suggestion is that wealthy nonresident individual investors seeking to build an investment pool to buy U.S. real estate:
At every stage, use an Excel model for the projected investment performance while owning the property — rental income and your best guess at capital gain.
You can compute an estimated after-tax and after overhead return on investment for the partners, for each structure, over the life cycle of the investment. Yes it will be a guess. But it should at least give you relative rankings for your different structure choices.
The structure that puts the most money in the investor’s pocket wins.
Again, my bias is to use trusts as partners in a partnership. But then, that’s what I know best, and individual (rather than institutional) investors are who I work with. For them, trusts are purpose-built for the desired result.
This is not legal advice, not tax advice, and not anything else either. Do things right and get someone smart to help you, please.
See you in a couple of weeks.
Phil.
