The Stuff That Dreams Are Made Of – The Rise And Fall Of Malta Pension Plans

I am pleased to introduce a new guest blogger. Andrew Gradman alerted me to Malta Pension Plans (MPP) an idea that has come and gone without me noticing. What happened with MPP is an illustration of Reilly’s Third Law of Tax Planning – Any clever idea that pops into your head probably has (or will have) a corresponding rule that makes it not work. Before MPP Malta evoked three things for me. One is there was some sort of military religious order called the Knights of Malta, but the Templars get a lot more publicity. Then there is the strategic significance of Malta during World War II, which made the Luftwaffe bomb the hell out of it and everybody was so brave the King of England gave the whole island a medal or something like that. Most memorable of all, though, was the entirely fictitious Maltese Falcon, which like MPP is the stuff that dreams are made of. – PJR

By Andrew Gradman

——————————————————————————————————————————————————————————

You may have heard of Malta Pension Plans (MPPs). Until recently, these were touted as “a supercharged, cross-border Roth IRA.” This referred to the idea that—thanks to an oversight by US treaty negotiators, as well as the unique flexibility of Malta’s domestic pension laws—Malta pensions could accept unlimited contributions of non-cash property, even from US citizen-residents. Thus, you could use an MPP as a conduit to sell appreciated assets tax-free, without ever leaving your ranch house.

Later, you may have seen IRS’s announcement adding MPPs to their Dirty Dozen list, and its announcement that US and Malta authorities agreed that MPPs do not work as described. To the people who felt that MPPs were “too good to be true,” this was a vindication, and ended the story.

However, before we close this chapter of tax history, let’s take a moment to understand exactly what happened. Was there really a flaw in the treaty? How did the US treaty negotiators miss it? How did MPP promoters exaggerate its significance? How do we know this flaw won’t be resurrected?

In this guest post, I explain how MPPs were supposed to work, and why they didn’t. This will form the basis for a later commentary by Pete Reilly.

INTRODUCTION: PENSIONS (DOMESTIC)

When people retire, they often can’t afford their costs of living. To take pressure off the welfare system, governments try to encourage private saving. One way to do this is via tax break.

For a person who chooses to save (aka defer consumption), there are three moments when tax could be imposed:

(i) when one initially earns the Earnings;
(ii) as one achieves Investment Returns on those earnings, e.g. as dividends or sales, and
(iii) when one liquidates those investments in exchange for Liquidation Proceeds, for the purpose of consumption during retirement.

Typically, the Investment Returns at step (ii) are tax-free. For example, in the United States, IRAs, qualified plans, and similar arrangements are not taxed on their income, just like charities. (The same happens to be true in Malta.)

In addition, these arrangements are commonly not taxed at either on Earnings at step (i), or on Liquidation Proceeds at step (iii). A traditional IRA follows the pattern exempt-exempt-taxed (EET). A Roth IRA follows the pattern taxed-exempt-exempt (TEE)..

TREATY TAXATION OF TRADITIONAL IRAS

What if the worker emigrates, after he earns the salary but before he liquidates his investments? Unless I say otherwise, I will assume taxpayer “T” who works in work-state “W”, then retires to retirement-state “R”.

Before the creation of Roth IRAs in 1997, US tax treaties did not address the TEE model. They only addressed EET (traditional IRAs), under this general rule: If T moves from W to R, then his Liquidation Proceeds may not be taxed by W. Versions of this appear in Article 18(1) of the 1996 US Model Treaty; Article 17(1)(a) of the 2006 US Model Treaty; and Article 17(1)(a) of the 2016 US Model Treaty (which is almost identical to the 2006 model).

Taken out of context, this would suggest that if W is the USA and R is a foreign country, then T could escape US taxation on his traditional IRA by migrating to a low-tax R. However, one must always check the “Savings Clause,” which appears in Article 1(4) of each treaty. As a general matter, the Savings Clause can change a treaty result in two cases.

First, it provides that nothing in the treaty causes a contracting state to relinquish its right to impose taxes based on citizenship (rather than residency). This matters because the US is one of the few countries which taxes its citizens regardless of residence. Thus, any treaty provision which seems to limit taxation to the country of residence turns out to be a dead letter.
Second, it provides that nothing in the treaty causes a contracting state to relinquish its right to impose taxes based on residence within that state. Thus, if some language in the treaty purported to give one of the contracting states the exclusive right to tax a person (for any reason at all), that exclusive right would always collapse if the person took up resident in the other state.

In short, the “Savings Clause” has the potential to dilute the treaty. You then need to check the next paragraph to see if the treaty gets its powers back: Article 1(5) will contain exceptions from the Savings Clause. Thus, if a country appears to make a concession in some substantive provision of a treaty, it might turn out to be illusory unless you see that provision mentioned in Article 1(5).

In the case of traditional IRAs, none of the three Models incorporate the above-mentioned pension language into Article 1(5). This means that these traditional IRA provisions are diluted; the US continues to assert its right to tax these distributions to its citizens, even if they reside abroad.

TREATY TAXATION OF ROTH IRAS

Under the above-mentioned general rule, T’s pension could only be taxed in R, the state where he resides. However, the 2006 Model Treaty provides an exception to this general rule, relevant to Roth IRAs. The key language is in Article 17(1)(b). Similar language first appeared in the 2001 US-UK treaty.

Under this new exception, even R cannot tax the pension, if the amount arose in W and would have been tax-exempt had T resided there. In other words: Suppose you worked in the USA; put after-tax money into a Roth; and then moved to the UK. Had you stayed in the US, your withdrawals could not have been taxed in the US. Thus, this provision is satisfied: The UK will not be able to tax your Roth IRA withdrawals.

In both the 2006 and 2016 Model Treaties, this provision is excepted from the Savings Clause (and thus, cannot be defeated by the taxpayer’s citizenship or residency).

TREATY TAXATION OF THE MPP

Finally, we turn to the Malta treaty. To illustrate, I’ll compare the MPP to the Roth IRA, and I’ll ask: Why can’t a resident-citizen of the UK use a Roth IRA in the same way as an MPP, to the detriment of the UK?

Contributions of Appreciated Property

First, there are limits on the IRA’s capacity to accept Contributions of Appreciated Property. This is a function of domestic law, in both jurisdictions.

In the United States, IRAs and Roth IRAs can only accept contributions of cash. A noncash contribution would constitute an “excess contribution” under IRC 4973, triggering a 6% excise tax. By contrast, MPPs do accept noncash contributions.

Furthermore, to analyze the Roth IRA, one should ask whether UK law would treat the transfer as a deemed sale. (I have no idea.) Turning to the MPP, a transfer from a US person to a foreign pension would be a nonrecognition event if the foreign pension were a foreign grantor trust. The proponents of MPPs argue that this is true. To show this, they argue that (i) the MPP is a trust, (ii) which is a foreign trust, i.e. not subject to US jurisdiction or not managed by US persons, and (iii) which is a foreign grantor trust, i.e. with US beneficiaries.

Contribution Limits

Second, Roth IRAs have strict contributions. They only accept a small amount of contributions. And as AGI increases, this amount quickly phases out to zero. IRC 408A(c)(2), (3). By contrast, MPPs do not have contribution limits.

Tax-Free Investments

As mentioned, most pensions do not seek to tax Investment Returns. This is honored in Article 18(1) of the Malta and 2006 Model Treaties, in identical language. To paraphrase: If T has retired in R, but is a participant or beneficiary of a Pension Fund which resides in W, the income of the Pension Fund may not be taxed before it has been distributed to him.

Notice:

In both treaties, this clause is carved out of the savings clause (and thus, cannot be defeated by the taxpayer’s citizenship or residency).
The treaty does not literally require that T actually worked in W.
Together, these two facts open the door to MPPs, where US persons become participants without ever setting foot in Malta.
This requirement refers to a defined term, “Pension Fund.” This is important; I will discuss it later.

Timing of Distributions

In the United States, there is a 10% tax on “early distributions,” i.e. distributions taken before age 59 1/2. IRC 72(t).

Under Maltese law, a 30% lump-sum distribution can be taken from an MPP as early as age 50 without penalty.

Taxation of Distributions

As mentioned, under Article 17(1)(b), country R can’t tax a distribution from a country W pension, if country W could not have done so.

With Roth IRAs, all distributions are free from US income tax. By contrast, with MPPs, only lump sum distributions are free from income tax. This represents 30% of the value of the pension. “Periodic” distributions are subject to tax.

CHALLENGES TO THE MPP

So far, we have a working loophole. However, commenters have raised many concerns with the MPP scheme described so far.

Query whether the MPP is really a foreign grantor trust. For example, one might ask whether the “trust” is more of a business trust, or a partnership. One might also ask whether it is a “foreign” trust; one would need to ask whether the US beneficiaries have enough control of the pension’s activities to cause them to be treated as its US administrators. In either case, the transfer from the individual to the MPP would be a recognition event.

Article 18 provides for a tax exemption on “income earned by the pension.” However, if the income in question arose from property which appreciated in the hands of the taxpayer, does it really meet this definition? A similar point can be made with reference to IRC 482. The IRS could argue that the treaty does not apply, because (under IRC 482) the “true” seller of the appreciated asset was the United States individual.

Article 18 requires a “Pension Fund.” As mentioned, this is a defined term. Under Article 3(1)(k) and Article 22(2)(e), this generally means an entity which (if established in Malta) is
1. a licensed fund or scheme,
2. of which more than 75% of the participants are residents of either Contracting State,
3. which is tax-exempt (except as to income from Malta real property), and
4. which is “operated principally … to administer or provide pension or retirement benefits.”

“OPERATED PRINCIPALLY … TO ADMINISTER OR PROVIDE PENSION OR RETIREMENT BENEFITS”

In my opinion, the final requirement—the Operated Principally requirement—is the true Achilles Heel of MPPs. I have never seen a convincing argument for why it is satisfied.

One difficulty: What are the limits on future pension plans? Suppose Malta amends its domestic law to allow tax-free distribution as to 100% of a pension (as opposed to 30% under current law). On what basis could the US defend itself against such malicious ingenuity?

Another difficulty: The treaty asks how the pension fund is “operated.” How does one test that? Many commenters have talked about the importance of managing one’s own personal account in a conservative way. But if the entire fund’s behavior is at stake, then even a squeaky-clean client account could get taken down if the IRS could show that the other accounts in the fund were not handled correctly.

Third: If this is true, then to satisfy one’s burden of proof on audit, one would need to present evidence regarding the entire fund, and its handling of other people’s accounts. It is unclear whether a Maltese pension fund would, or even could, cooperate with such a discovery request. Moreover, this would be an uphill battle; there is already a lot of literature in the public space, from MPPs and from opinion writers in the US, which have marketed MPPs as more of a tax-saving tool than a retirement vehicle.

Fourth: Who, ultimately, decides what these terms mean? This last point turned out to be the killer.

As it happens, under the plan language of the treaty, the IRS (as “competent authority” under the treaty) can get together with its Maltese counterpart for the purpose of defining terms in the treaty, using a competent authority arrangement (CAA), without going through a formal amendment.

Article 3(2) provides: “ny term not defined shall, unless the context otherwise requires, or the competent authorities agree to a common meaning pursuant to the provisions of Article 25 (Mutual Agreement Procedure), have the meaning which it has at that time under the law of that State for the purposes of the taxes to which the Convention applies …”
Article 25(3) provides: “he competent authorities of the Contracting States may agree … (c) to the settlement of conflicting application of the Convention, including conflicts regarding: … (iv) the meaning of any term used in the Convention …”

In December the IRS and Malta did just that. In the CAA, they clarified that a pension fund will not meet the “Operated Principally” test if it is allowed to accept contributions from a participant in a form other than cash, or if it does not limit contributions by reference to earned income from personal services. And they asserted that this represented the “original intent” of the parties.

EVALUATING THE MERITS OF THE CAA

In my view, the CAA is ironclad. MPPs are dead. Still, it is worth considering possible objections.

One objection I have heard to the CAA relates to a doctrine of treaty interpretation: treaties should be interpreted in “good faith.” The implication is that this application of the power to define terms via CAA is not in good faith, at least when seen from the perspective of US taxpayers who relied on that term.

However, I don’t think that’s the right understanding of the phrase “good faith.” The phrase appears in the Vienna Convention on the Law of Treaties, with relevant language in Articles 31 and 32. Notice that the “good faith” principle includes looking to “any subsequent agreement between the parties.” Surely, that ought to include a Competent Authority Arrangement. Also, the principle takes into account whether a particular meaning “leads to a result which is manifestly absurd or unreasonable.” Since the marketers of MPPs enjoy replying to the objection that “it’s too good to be true” by saying “yes, it seems that way, yet it’s true,” I’d say we have a case of the “manifestly absurd or unreasonable.”

Turning to the US-Malta treaty in particular: the technical explanation is not very helpful. However, the analogous language in Article 25 of the OECD model treaty is more helpful. Section 5.1 of the OECD commentary to Article 25 seems to support the “good faith” argument:

The undertaking to resolve by mutual agreement cases of taxation not in accordance with the Convention … must be performed in good faith. In particular, … the competent authorities are obliged to seek to resolve the case in a principled, fair and objective manner, on its merits, in accordance with the terms of the Convention and applicable principles of international law on the interpretation of treaties.

However, it’s not clear to me that the “good faith” in this context should be owed to the taxpayers. Since it’s being used in the context of “undertaking to resolve by mutual agreement cases of taxation not in accordance with the Convention,” it makes more sense that the obligation of good faith is owed by each competent authority towards the other. In other words, this is simply a statement that each party to the treaty should take the mutual agreement procedure seriously. I believe that view is supported in discussion language which led to this 2017 text. The “good faith” doctrine is not concerned with the interests of the citizens of the treaty parties. It serves “to remind courts of … the need to show sensitivity for the views of our nation’s treaty partners.”

In any event, when you continue reading through the OECD commentary (particularly paragraphs 6.1 and 6.2), it turns out that this principal of “good faith” is completely consistent with what happened here. In other words, the “good faith” reading is simply the literal one: the treaty explicitly gives the competent authorities the power to define terms, and that’s exactly what they did.

CONCLUSION: REFLECTIONS FOR TAX PLANNERS

As tax planners, our job is to predict. Proponents of MPPs predicted that the arrangement would survive until the treaty was amended. Others were not so sure. Personally, I doubted it could pass the “Operated Principally” test. Others had other concerns.

But very few of us (including me) predicted that the whole thing would come crumbling down so quickly. I think it is fair to say that the swiftness of the CAA process caught most of us by surprise. It is amazing to think that the CAA process for defining terms is right there, on the face of the treaty, staring us in the face. So many smart people missed it. It is a humbling moment for all of us.

When the IRS dropped its guidance, it was like watching a chess game played in a sitcom, where one player is “beating” his opponent, only to be shocked when his opponent suddenly announces checkmate. There is no such thing as a mate-in-one in serious chess. I thought the same was true of tax planning. Apparently I was wrong.

Joe Kristan originated the idea of the “tax fairy”, the magical sprite that makes your taxes disappear. I am going to start a campaign to rename the “tax fairy” the Maltese Falcon. – PJR

Original YTMP content.

For great value continuing professional education. I recommend the Boston Tax Institute

You can register on-line or reach them by phone (561) 268 – 2269 or email vc@bostontaxinstitute.com. Mention Your Tax Matters Partner if you contact them.