U.S. Treasury
We last visited Treasury’s proposed regulations for captive insurance companies in my article U.S. Treasury Department Issues Proposed Regulations To Finally Eviscerate Microcaptive Tax Shelters (April 11, 2023). In less than a week, this has set off a considerable discussion within the captive insurance sector as to the effect, if any, the regulations may have on ordinary small captive insurance companies that have elected treatment under Internal Revenue Code section 831(b). This requires a deeper dive into the proposed regulations, and so today we will look with more precision at what the proposed regulations permit or allow.
For any of what follows to make sense, we first have to make a pit stop to look at the two new computation periods that are defined by the proposed regulations.
First, there is a financing computation period that is defined in § 1.6011-10(b)(2)(i), which looks at the most recent five taxable years of the captive, or all taxable years if the captive hasn’t existed for five taxable years. If a captive has only been in existence for one year, for instance, then the financing computation period will apply to that captive.
Second, there is a loss ratio computation period that is defined in § 1.6011-10(b)(2)(ii), which is only available for captives that have been in existence for at least ten taxable years, and which looks at the most recent ten taxable years. Basically, if the captive has not existed for at least ten taxable years, then one must use the financing computation period.
With these two periods in mind, let us now turn to the transaction that is described by the proposed regulations § 1.6011-10(c). This describes what Treasury considers to be an abusive tax shelter. The there are really two definitions of transaction, either of which can trigger the heightened tax shelter reporting requirements, and we’ll take these in turn.
The first definition of transaction is found in § 1.6011-10(c)(1), and basically describes any arrangement where the captive’s owner obtains any non-taxed benefit from the captive during the financing computation period (most recent five taxable years). The arrangement can either be director or indirect, and would include such things as “a guarantee, a loan, or other transfer of Captive’s capital, or made such financings or conveyances prior to the Financing Computation Period that remain outstanding or in effect at any point in the taxable year for which disclosure is required.”
In captive parlance, this definition prohibits the various forms of what are known as loan backs involving the captive. These loan backs usually involve the captive loaning money back to the insured business that just paid it a premium, but the loans can also be extended to the captive’s owner either directly or indirectly as through an irrevocable letter of credit (ILOC) or standby letter of credit (SLOC). This definition would also encompass loans made, directly or indirectly, to other businesses of the owner. A good name for this definition would accordingly be the loan back test.
The way this works is that you look to see whether the captive has made any such loans during the last five taxable years, and if it has then the captive must be reported as a tax shelter. What Treasury is basically doing here is saying that there is only one way that money comes out of a captive other than by way of paying claims, claims expenses, and the expenses of the captive, and that is by making taxable dividends to the owner. No more running the money out of the back door by creative schemes to avoid paying tax on that money.
Suffice it to say that this definition of transaction (“test” may be a better term) will substantially chill the sale of captives sold as tax shelters. Most of the small business owners who form captives as tax shelters, as opposed to bona fide risk financing vehicles, desperately want if not need to deploy the money being paid to the captive for their business or to fund their lifestyle, and so the money often comes back out of the captive about as quickly as it went in. The proposed regulations finally put an end to this practice: If you want or need your money from the captive, you’ll have to pay tax on it.
The second definition of transaction is found in § 1.6011-10(c)(2), and relates only to calculations under the loss ratio computation period, which it will be recalled is a full ten taxable years. In other words, this test only applies to captives that are at least ten years old.
Basically, this second definition of transaction is an excess premium test. The way the test works is to look at the captive’s losses, claims administration expenses, and dividends paid during the last ten years and determine if the aggregate of those things exceeds or is less than 65% of the total premiums received by the captive. If all those things are less than 65% of the total premiums paid by the captive, then the tax shelter reporting requirements come into play. If all those things are more than 65%, then the test is not met. Or, to look at it from the other side of the coin, if the captive’s non-taxable profits after losses and claims (plus taxable dividends) were greater than 35% over the last ten years, then the captive must file the tax shelter forms.
It is this part of the proposed regulations that has given rise to some of the most vociferous objections within the smallish captive community, since there are who believe that the 65% figure is too high (or the 35% non-taxed number is too low). If one were only to look at a single given year, those arguments would be more compelling. But here we are looking at the results over a decade, and after a decade it can become obvious that a captive has been charging too much in the way of premiums based on its actual loss history.
Furthermore, Treasury gives captive owners at least one easy and obvious escape hatch to avoid this test, which is simply to pay taxable dividends until the captive is under the 65% threshold. As captive tax attorney Hale Stewart pointed out to me, the proposed regulations even say this on page 36: “Any Captives that would be required to disclose as a result of the loss ratio factor may consider paying policyholder dividends to increase the loss ratio and eliminate the need to disclose.”
There is actually a second way that this can be accomplished, which would be for the captive to refund what amounts to unearned premiums until the captive is similarly below the 65% threshold. So this 65% test is really a giant nothingburger — unless the captive owner’s real motivation all along was to misuse the captive as a tax shelter by deliberately charging excess premiums and then not making any taxable distributions.
The combination of these two tests will help to chill the marketing of abusive tax shelter captives, since the loan back test will prevent the captive owner from touching the money unless it is paid as a taxable dividend, and the excess premium test will prevent the owner from misusing the captive as a tax deference mechanism by way of charging premiums substantially in excess of losses and claims. At the same time, real captives being used as true risk financing vehicles should not be troubled in the slightest by any of this.
