BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

IRS Noose Starting To Tighten On Sham Risk Pools

Following
This article is more than 10 years old.

A captive insurance company ("captive") is an insurance company that is created by the parent organization to underwrite the insurance needs of its operating affiliates. An oil company might, for example, form a captive insurance company to provide environmental cover to its various exploration, refining, and transportation subsidiaries.

The goal of a captive is to reduce insurance costs, by having the captive retain the underwriting profits that might ordinarily be lost to the third-party commercial insurance carriers from whom the business previously bought its insurance.

But of course, one might suggest, the business doesn't need an insurance company to do that -- it could simply self-insure against these risks internally by setting money aside for claims. The fallacy of that argument is that risk and liabilities often last for many years, but the business can only lawfully deduct in a given year the amount of claims that it pays out.

The benefit of a captive is that the business is allowed to take a current-year deduction now for risks that might not materialize or be paid until some years in the future. That, in a nutshell, is why businesses form captive insurance companies as opposed to simply self-insuring themselves by setting aside internal reserves.

For a captive insurance arrangement to be valid, as opposed to a disguised self-insurance arrangement, the captive must engage in the spreading of risks among many insureds, what is known as "risk distribution". Here, the IRS has given us two safe harbors to determine if risk distribution has been met.

In Revenue Ruling 2002-90, the IRS said that if the captive was insuring at least 12 insureds, with each insured having between 5% and 15% of the total risk (not premiums) in the captive, then risk distribution exists even if all the insureds are wholly-owned by the same common parent as the captive. Nearly all of the largest American companies now have captive insurance companies, and nearly all of these meet the requirement of risk distribution by this method. This safe harbor is known as the "12 insured" safe harbor.

But large companies usually have dozens, and sometimes hundreds, of subsidiaries that they can throw into the captive to meet risk distribution; smaller organization might not have the required number of entities in some cases -- how can they benefit from a captive?

In Revenue Ruling 2002-89, the IRS gave us the second safe harbor, known as the "50% third party insurance" safe harbor. Widely used (and abused, as discussed below) by smaller businesses, this test posits that if the captive derives at least 50% of its premiums (not risk) from unrelated third-party insureds, then risk distribution has been met.

The problem here is that a captive insurance company has a limited insurance license, which basically restricts its underwriting to either companies affiliated with the common owner of the captive, or to other insurance companies. A captive cannot, for example, open an agency on the street corner and start selling insurance to whoever walks in the door. However, a captive can sell insurance to another insurance company -- which arrangements are known as "reinsurance".

Captive managers are businesses that actually operate captives for their owners, at least when the captive is too small to have its own full-time staff, and executives approved by the local insurance commissioner. In other words, the owner of a captive will outsource the operations of their captive to a captive manager, who will draft policies and do the things necessary to keep the company in tax and regulatory compliance.

To assist their clients, and of course to sell more captives, many captive managers have put together arrangements among their clients who cannot meet the 12 entity test, so that these other captives can meet the 50% third-party insurance test -- these arrangements are known in the industry as "risk pools", and they are themselves organized as insurance companies.

The way a risk pool works is this: Each captive owner has their underlying operating business purchase insurance from the risk pool. At the same time, each captive reinsures the risk pool for roughly the same percentage of premiums that the operating business paid into the pool, compared to all the other businesses in the pool. Once the policies have expired, and all the claims against the risk pool have been paid, then the pool releases the premium moneys to the captive. Voila! The captive now has third-party insurance, which presumably has been calculated in advance to be in the neighborhood of 50% of the total premiums received by the captive.

To illustrate, let's say that Widget Maker Company's owner sets up Widget Insurance Company as is captive. Widget Insurance will buy insurance from its captive manager's risk pool and pay a $500,000 premium. Let's say that there are 20 other unrelated businesses in the risk pool, such that the premiums paid by Widget Maker constitute 5% of the total premiums of the risk pool.

At the same time, Widget Insurance will issue a policy of reinsurance to the risk pool for 5% of the total liability of the risk pool, in exchange for a $500,000 premium. This means that Widget Insurance now has a $500,000 premium received in that year from a third-party source.

Next, Widget Maker will make a direct premium payment to Widget Insurance for $500,000 so that Widget Insurance has now received a total of $1 million in insurance premiums, of which 50% or $500,000 was the third-party insurance obtained from the risk pool. If I've lost you, the following graphic shows how all this ends up.

Under this arrangement, the 50% third-party insurance test of Revenue Ruling 2002-89 has now been met -- in theory.

Astute readers will probably be saying: "Whoa! Hold on there! What about claims? Don't worry, I'll get to that touchy issue in a second. Suffice it to say that for now, just focusing on premiums, the 50% third-party insurance test has been met.

The fly in the ointment is of course the claims, or at least the threat of claims. One of the primary reasons that business owners choose the captive route is that they don't want to mix their risk with anybody else -- they want to be in control of their own destiny. They sure don't want to walk in the office one morning to find an e-mail telling them that the loss experience in the risk pool was bad. In our example, the owner of Widget Maker doesn't want to hear that instead of getting a profit of $500,000 on his reinsurance agreement with the pool, Widget Insurance Company will now only get a $100,000 because of a lot of claims.

At the same time, the risk pool must actually have risk. The risk pool cannot just be an account where money goes in labeled as "insurance premiums" and routinely goes out in the same amount as "reinsurance premiums". If there is no substantial risk of loss in the risk pool, then there is no third-party insurance (or at least not enough to get Widget Insurance to the 50% of premiums mark), which means that the entire captive arrangement flops.

Here is where the deceit comes in to our picture. The captive manager wants to make an appearance to the IRS that the risk pool has real risk so that the 50% third-party insurance test is met. At the same time, the captive manager wants to tell the captive owner that there will not be losses in the pool (or else a substantial number of potential captive owners will discard the entire idea). So the captive manager sets up the risk pool to make it look like there is a mixing of substantial risk within the risk pool, while quietly (and never in writing) assuring the captive owner that these risks are not going to materialize.

To accomplish that deceit, captive managers engage in a number of subterfuges. One scheme is that if a business owner makes substantial claims against the risk pool, then the business owner must indemnify the risk pool for losses -- this protects the other captives in the risk pool from losses. Another scheme is to write the business owner's policies so that there are huge deductibles, meaning that there is risk shifting within the pool only in the event of really catastrophic claims. These and similar tactics of course negate any true risk-mixing in the risk pool, meaning that there really is no third-party insurance going on at all, but the risk pool is simply an account through which money passes.

It is here that I must digress and point out that the captive insurance business is relatively new, although captives have been around for some decades. The captive industry is "new" in the sense that after the IRS issued its 2002 guidance which legitimized captives, the industry literally exploded from the hundreds of mostly large, corporate captives preceding 2002, to the thousands of captives today.

As with any new industry, and new IRS guidance, bad practices tended to fly under the radar screen, and so it has been with risk pools. The IRS has not shown any interest in risk pools -- until lately. In September, 2012, I wrote in my article "IRS Filling The Pipeline With Captive Insurance Cases And Focusing On Dubious Practices" about Mr. John Glover of the IRS General Counsel's office commenting at the American Bar Association Tax Section's annual meeting, that:

The IRS has concerns about risk pooling arrangements and is beginning to focus on such arrangements, stated Mr. Glover, especially in cases where there is nominal laying and assuming of risks, but in the end reconciliation there really isn’t any risk-shift because the captive or business owner will reimburse the pool for significant losses. But there is also concern where the risk pool is simply an account where money comes in, and money goes out, and it is called “insurance” when it is really anything but.

In the months following Mr. Glover's presentation, the industry learned through various sources that the IRS now is conducting at least a half-dozen "promoter audits" of certain insurance managers, where the focus is on whether risks and premium levels, etc., have been dummied up, and captives sold not as any bona fide risk management tool, but rather as tax shelters where only lip-service and cosmetics has been given to any true insurance function of those captives. Nearly all these audits are focusing on some combination of the risk pools used by the captive manager (terrorism insurance seems to be a particularly bright red flag), and their marketing materials used to sell their clients into captives which proudly extoll the estate tax and income tax benefits of their captive arrangements with little mention of insurance risk.

All of which finally brings me to the point of this article.

Just months after Mr. Glover's presentation, in December, 2013, and in response to taxpayer requests, the IRS issued nine similar Private Letter Rulings ("PLRs") 201350008, 201350009, 201350010, 201350025-31, which were based on similar facts.

First, the captives in each were formed in jurisdictions outside the U.S., and as foreign insurance companies made the IRC § 953(d) election to be taxed as a domestic company.

Second, the captives each issued policies directly to two or more affiliated companies.

Third, the captives each reinsured risk pools with numerous unrelated parties, and the risk pool paid the captives a reinsurance premium equal to what it had received from the affiliated business ("ceded the premium" in insurance-speak), in exchange for the captives assuming a like amount of risk in the risk pool.

Fourth, the risk pools provided for experience refunds back to the operating business and expense loss carryforwards.

On these facts, the applying taxpayers sought rulings from the IRS that both the captive would be treated as an insurance company for tax purposes, and that all of the premiums paid by the operating businesses (both to the risk pool and directly to the captives) were deductible to the business.

The IRS did not issue the rulings, but instead questioned:

(1) Whether the risk pool actually provided risk distribution and a true transfer of risk to the risk pool (as opposed to just passing it directly to the captives);

(2) Whether certain provisions of the arrangement negated the risk distribution and risk shifting;

(3) Whether the insurance being provided was "insurance" for tax purposes, as opposed to being a mere contract that covered a non-insurance investment or economic risk; and

(4) Whether the provisions of the risk pool arrangement reflected arm's length transactions between the parties involved.

The IRS further noted that even if the IRS were to issue a PLR, the PLR would only be good for the specific tax-year for which it was issued, meaning that the PLR would not protect the taxpayer in previous or subsequent years.

In other words, the IRS is not going to bless particular risk pools, without substantially more information about the risks they are underwriting, the premiums they are charging, whether there is true distribution (mixing) of risks, and similar factors.

ANALYSIS

Risk pools are a very attractive target for the IRS, for the simple fact that if the IRS can invalidate the risk pool as "insurance" for tax purposes, then all the participants in the pool will not have any third-party insurance, and their individual captives will fail to be considered insurance companies -- and the owner's operating businesses will not be entitled to any deduction for premiums paid to captive.

The IRS doesn't have to seek out abusive captives one by one though chance audits of individual captives, but rather can invalidate hundreds of them at the same time by simply invalidating the risk pool and using promoter audits to get the insurance manager's client lists. Going after risk pools is quite frankly one of the most potentially cost-beneficial strategies the IRS has adopted in some time. The IRS's potential return-on-investment if it can invalidate just a few pools will be astronomic, not to mention the natural deterrence that will arise as captive owners and their tax advisers balk at going into risk pools. The truth is that non-compliant risk pools are probably the norm, and not the exception. But even if the IRS simply challenges a risk pool, that by itself may be enough for many captive owners to decide that they do not want the tax risk of participation in a captive arrangement. Put another way, risk pools present the hope of "easy money" for the IRS, similar to how the IRS made easy money with promoter audits of 412(i) plans and 419A(f)(6) plans.

Worse, for "offshore" captives, the IRC § 953(d) election is only available to companies that qualify as "insurance companies" for tax purposes. If there is no risk distribution because the risk pool failed, then the § 953(d) election is probably invalid. That likely means what the owner then has is a Controlled Foreign Corporation ("CFC"), or some other type of onerous foreign company for tax purposes, which may cause all sorts of nasty tax and reporting requirements. This is one of the reasons that folks may want to think twice before they go with an offshore captive, as opposed to a captive domiciled in a U.S. jurisdiction that will never run the risk of being treated as some type of non-insurance foreign company

This is not to say that all risk pools are bad; indeed, it is quite possible to run a valid risk pool. Such a pool will not have the characteristics that the IRS mentioned in its PLR responses, but instead will have:

Reasonable insurance premiums. A good risk pool will have actuarially sound premium pricing, not too far from market pricing. While captive managers attempt to skirt this by reciting that "a captive is not necessarily required to follow market pricing based on the case law," that may be true for the captives themselves, but not necessarily for the risk pool which presumably exists on the front end so that the operating businesses can obtain insurance at reasonable prices. Thus, if one could purchase Directors & Officers insurance on the open market for a $2,500 annual premium for $1 million in coverage, but the risk pool charges $50,000 for that same $1 million in coverage, it is probably not going to stand muster.

Here, it should be noted that the IRS is very much aware that some actuaries are very much like some property appraisers, whose first question is "What number do you want?" and then try to back into the pricing through some combination of distorting industry figures and rubbing chick bones together at night in the graveyard. Premium pricing for risk pools should be close to market pricing (in theory, they should be better than market pricing), and not defy common sense.

Necessary insurance coverages. The other side of this coin is that to get to the higher premiums that business owners need to pay, and thus be reinsured into their own captive, captive managers will try to use coverages that are unlikely to ever have claims. In the past, terrorism insurance has fitted this bill; while there may be a real terrorism risk for a business in downtown New York, the odds of a business in Fayetteville, Arkansas, actually having a terrorism claim is astronomically small (and thus its premium should be priced accordingly small). A good risk pool will cover risks that are reasonably likely to occur -- though not certain to occur since that would not be insurance risk either.

True mixing of risk. Bad risk pools are arranged in a way that even if one owner has some really expensive claims, the rest of the pool participants are not significantly impact (after all, they want "their money" to go to their captive by way of the reinsurance premium). Side deals where the business owner with claims either indemnifies the pool for losses, or gets hit with a "retrospective premium adjustment" which serves the same purpose, means that there was no true sharing of risks among participants, and thus no true "insurance" for tax purposes. Likewise, some risk pool policies have such large deductibles, that the risk of claims is significantly lowered (but premiums are not correspondingly lowered). Good pools will operate as pools should operate, in the sense that if there are losses in the pool, everybody shares those losses and on a "first dollar" basis with only a small deductible.

History of claims payments. This one is more controversial, since some pools may never have real risk but never have a claim. The best example of this is earthquake insurance. Say that a pool insures nothing but earthquake risks for businesses in known earthquake zones, and all the participants share that risk. The premiums are priced on an actuarially sound basis. The pool operates for a decade, but no earthquake ever occurs; is this insurance? The answer is a firm "Yes!", since there was a real risk that an earthquake would occur and cause significant losses to the insured businesses, and claims to the pool. The problem here is that many risk pool participants don't want this exposure to losses, which might wipe out the risk pool if a bad earthquake hits -- again, they just want the money to go to their captive. Thus, the bad risk pools will not cover these types of exposures, but will instead cover things where losses are normally expected (a comparison of like commercial coverage will show that claims are routinely made and paid), but somehow no or few claims actually are made to the pool. Maybe in the first year, one can get away with this arrangement, but after a few years of no or few claims, it becomes obvious that the true risk to the pool are low and at the very least premium pricing should be modified downward to reflect this actual experience.

Since some risk pools are valid, and some risk pools are not valid, how can the average business owner protect herself from being caught in a bad pool? The answer is actually easy: Hire independent tax counsel familiar with captives (don't call me, I don't claim to be a tax professional) to review the risk pool arrangement and determine if it passes muster. Here, I do not mean tax counsel that is recommended by the captive manager (who very likely will just be a buddy who will blank-stamp everything), but truly independent tax counsel who investigates and asks a lot of questions about the risk pool, its structure, the soundness of its actuarial advice, its loss history, and similar concerns. If the captive manager balks at providing this information, then that is probably all you need to know and you should take your business elsewhere.

It should be noted that a similar analysis applies to so-called "cell captives" which are typically organized as Series LLCs or Protected Cell Companies, and which effectively offer "internal risk pools" where participants purportedly share risks. These arrangements are often offered to business owners who either do not have the insurance risks, or cannot afford the level of premium payments, to make their one standalone captive efficient.

Finally, it should be noted that -- completely aside from tax risks -- risk pools inherently carry the risk of fraud, which is that a buddy of the captive manager may participate in the pool and submit bogus claims (approved by the captive manager of course), as a way of siphoning money from the pool. While an individual and innocent participant will likely get to see the fact of the claims in annual reports, the contractual documents of the risk pool may prevent (ostensibly for privacy) that owner from looking behind the claim to see whether it is valid.

Some captive owners not having multiple subsidiaries may have to participate in risk pools if they want a captive; that is a fact. And, as stated above, not all risk pools are bad. But business owners had better go to some lengths to evaluate the arrangement beforehand if they don't want to face an unhappy situation later. My experience has been that it is often the case that many such business owners through some minor business re-organization can get to the number of entities that they need so that they don't have to participate in a risk pool, but can instead end up meeting the "12 separate entity" test of Revenue Ruling 2002-90.

Indeed, I have seen not just a few cases where particular businesses started out being able to meet the "12 separate entity" test, but were instead and quite unnecessarily slammed into a risk pool so that the captive manager could charge an extra fee for the participation in the risk pool. This usually occurs with those captive managers who offer very low formation fees (often below-cost, although one wonders about the quality of their actuarial studies) as an incentive to get more participants into their risk pool and earn their annual percentage of premiums. By contrast, the better insurance managers will charge their normal flat fee to form and manage the captive, but only a very small fee for participation in the risk pool (just enough to cover their costs), and a rare few insurance managers sometimes "comp" the risk pool fee entirely.

The worst captive managers will also bundle the sale of a cash value life insurance policy into the deal, which is how they really make their money on the life insurance commissions, but that is a story for another day -- suffice it to say that a captive arrangement involving life insurance seems to be another bright red flag to the IRS.

The bottom line is that a participant in a risk pool either has substantial risk of loss, or doesn't. If the latter, then the arrangement is not "insurance" for tax purposes, and the entire deal fails. But the business owner may not want any true risk of loss, but merely a conduit to move money into the captive.

Reconciling those two competing interests is where the sham arises, and you don't need to rub chicken bones together at midnight in the graveyard to know that.

This article at http://onforb.es/NpU14S and http://goo.gl/ntJVrC

Follow me on Twitter or LinkedInCheck out my website or some of my other work here