I can't speak for everybody involved in the markets and I shouldn't speak in generalities about large groups of people but for the most part I think it's safe to assume that most market participants start each day trying to make money. Now, whether by means of trading debt or equity instruments, purchasing debt or equity instruments for the long-term, or using speculative contracts, most often money is made by spotting inefficiencies in pricing caused by any of the moving pieces that create the market.
This same desire to find the greatest inefficiency and exploit it to the greatest financial advantage is largely what caused the last financial crisis. For those of you beginning to roll your eyes, this article is not about the next financial crisis and the core of this articles conversation couldn't cause a financial crisis (not yet at least) in a worst case scenario - just significant impairments.
What I'm talking about is an accounting inefficiency that insurance companies have been using to generate larger and larger profits since early 2009 and have continued to use to present day - although it should be noted that greed and hypocrisy can only go so far and that recently the use of this gap in regulation has been slowed.
The same products that will compose the bulk of this article with slight variations have been used at publically traded companies like Lincoln National Corporation (NYSE:LNC), MetLife (NYSE:MET), Prudential (NYSE:PRU), American Insurance Group, Inc. (NYSE:AIG) and many others. I have personal familiarity with products used at the above listed firms with AIG being the company I am most familiar with. I have also seen quite a few Jackson National Life variations. The purpose of this article to get persons long insurance stocks to take a granular look at each company's financial filings (although the risk won't be shown there as it is not mandated) and decide a risk adjusted ownership.
As a final note, I love the insurance business, I think it is incredibly profitable, I will continue to own insurance companies as a broader investments and risk management strategy and I can appreciate the insurance industry in general.
It started in 2009
In early 2009, at the peak of fear, something innovative was happening at insurance companies in an effort to asset grab. Insurance companies, who offer annuities as vehicles for safety and income, saw an opportunity to acquire assets on the backs of the greatest free advertisement blitz seen since the 20's - fear. They figured they could acquire assets cheaply while offering what seemed like minimal benefits and use these assets increase their fee base, cross-sell, and eventually bring in more assets as the provider who brought the customer piece of mind.
Now, even after getting smoked most folks realized that they had to stay invested (depending on age) if they ever wanted to see their account balances close to the pre-crisis highs but they also wanted the opportunity to asset allocate to a safe fund or two in the meantime while the markets decreased in volatility. The Variable Annuity being sold to them by "that insurance guy" seemed to be able to do it all. The VA could offer 7% interest for any money kept in short term Treasuries for the first 12 months and allowed for investments in a variety of equity based funds if the opportunity presented itself. It also guaranteed payout of at least what was invested at inception and would step up that payout from the baseline figure, regardless of the actual account value, for any contributions made for the first five years - a death benefit. Finally, the VA was willing to step up the "income base" from which payments would be calculated by 8% a year every year or the actual account values gains, whichever were greater (on a specified date known as the Anniversary Date). Boy, what a dream investment.
Ok, let's summarize where we are at this point. The VA's being offered beginning in 2009 offered the following features and carried the following REAL risks to insurers (please keep in mind there are hundreds of variations of this example product:
- Bonus Rate in Treasury based fund for at least the first 12 months, sometimes up to 36 months: the risk to insurers was that Treasuries were paying all-time lows at the time and they took immediate losses on the spread minus any fees charged to the overall VA
- Guaranteed Death Benefit: this is the baseline risk that the insurance company would suffer should the client die and the actual account value be worth less than the GDB. I'll explain how this is not the risk that is required to be marked to market on books of insurance firms.
- Guaranteed Minimum Withdrawal Benefit: this rider (addition to the original contract) essentially created two "account balances" for the client. The first, the client's actual fair market account value inclusive of gains and losses of the investments chosen by the client. The second, a pie in the sky account that grew by 6-8% per year (depending on the rider and the product) or the account value (the higher of the two) of which the clients end payment options would be chosen. This rider came with an additional fee that is a front-end benefit of the insurance company.
Now, sure the insurance companies aren't silly and they understood as a group that there was quite a bit of risk involved in doing this. There was the risk that the client could die and the clients account balance would be under the guaranteed death benefit - the insurance company is out the difference. There was the front-end risk minus fees collected of paying the client a massive spread for investments in treasuries that it wasn't able to get - the insurance company was out the difference. There was the largest risk of all, that the clients pie in the sky balance would greatly outpace the actual balance of the account at the end of the accumulation phase and the insurance company would be out a substantial difference - but everybody knew the market would certainly not go to the moon anytime soon after the crisis of 2008 and if it did it wouldn't hold long enough for the riders contractual "anniversary date" to pass, the date that the variable annuity takes a "snapshot" of the account value and measures if it's higher than the annual step up provider and locks in the higher of the two. Surely, this would never happen. Well guess what it did.
So, you have 3 ½ years' worth (sales of VA's with the full suite of bells and whistles slowed dramatically in mid-2012) of these products floating around with the selling of these particular risk enhancers still happening but on a smaller scale.
For clarity, the scale isn't all that small and the concentration between the top-20 sellers is significant (SOURCE: LifeHeahlthPro.com). The risk enhancing doesn't stop there.
I'm of the opinion that based on risk already on many insurance providers who also sell VA's books that the newer VA was created to offset this historical risk, which didn't look like risk until the last five years. For clarity I'll give the following example.
Many of the VA's sold to large institutions historically (last 20 years or so) have completely separate but just as large risk issues. One company in particular has over $300 million in "older version" VA's at the University of Texas alone, $800 million at Duke, so on and so on, and this particular variety has permanent short-term treasury fund (12 months or less expiration) guaranteed interest payments of 3.0%-4.5% - think about the deficit that creates. This is an example of the total risk these VA's are adding to books historically. This particular example is prior to the more recent version of the VA and has been sold for much longer durations. I'll give a more granular example.
I came across a professor at the University of Texas, and I have seen this example hundreds of times at different institutions and across several companies' products, who had roughly $450k in her VA, the particular company is irrelevant. She had the maximum allowed under her VA's allocation parameters (100%) invested in "Fund 001" or short term treasuries. Her particular UT special rate guarantee for this fund was 4.5%. What motivation would she ever have to move a single dime of this to any other bucket of asset allocation being in her late 50's and otherwise well prepared for retirement? The fun doesn't stop there. She had no caps on outside money she could roll into her VA but any further contributions wouldn't count towards her death benefit should the account decrease in value. Decrease in value? From short-term treasuries? With a guarantee of 4.5% and a fee just under 1% all inclusive from the insurance company? Oh, the insurance company meant if she ever wanted to utilize the variable portion of her VA. I see. Shortly after realizing what she had and further realizing that she was completely out of surrender period (a surrender period is a period of time in which an owner of a VA would pay a penalty for taking their money elsewhere) she flowed into her VA 100% of external dollars that were in fixed instruments yielding less than 3.5% fee inclusive and risk adjusted. She is now an even larger liability of the insurance firm.
It's important to remember than historically interest rate environments have never mirrored ones that we're currently experiencing. The rick of these absurd guarantees didn't appear to be all that high until recently.
I think, as stated before the example that the newer VA was created for three reasons:
- The risk of the above client earning the 4.5% (and billions of dollars of others like her) has to be accounted for as a liability according to GAAP and FASB. This loss hurts the financials and can be offset by the fees taken in by the newer contracts
- To gather large amounts of assets under the assumption that the market wouldn't immediately move much higher and that if it did it wouldn't hold and if it did the economy would be much better and would permanently hold. Clarity will be provided shortly.
- The actual risk of the death benefit and the pie in the sky account value does not need to be marked to market according to GAAP and FASB because it does not qualify as a derivative.
The Raging Bull
Ok, now it's time to justify you reading the article. I'll break down this section according to the three reasons listed above.
(1) This part is pretty self-explanatory. Although the funds being invested by the client are kept in a separate account at the insurance company and are a direct and wholly owned investment of the client, when the firm makes promises that cannot be achieved (giving guaranteed interest rates) by the investments in the legally specified vehicles, the firm is on the hook for the remaining balance and DOES have to recognize this as a loss as specified by GAAP and the FASB. This is according to FASB Statement 133: Accounting for Derivative Instruments and Hedging Activities. I would specify an implementation issue of Statement 133 but I don't think that is necessary. This is a portion, the fixed asset allocation at least, of risk that is being reflected on the balance sheet and income statements of the firms reporting.
(2) What the markets have shown historically is that when equity markets are doing well typically debt instruments offer higher interest rates and vice versa. It's very rare that you have an environment where the equity markets are doing poor and you have debt instruments offering high rates and it's equally rare to have environments where you have equity markets doing well and debt instruments offering low rates. The latter scenario is the scenario currently being offered by the markets in general. It is my opinion that the loophole to be described in section three was exploited because of this historical market spread between equity and debt instrument performance. I think that the plan was for the described scenario to happen:
- The hope was that new clients (new assets) would come into the newer contracts which indexed the guaranteed rate in the fixed portion of assets (making the firm take on zero added risk and fixes the above problem of needing to account for losses in this portion) and add the rider (increasing fees) that would step up the pie in the sky account balance. From there the insurance companies hoped the client would allocate just enough of the assets to variable funds to hit the historical market performance (depending on the firm this is justified as 6-9%) making the risk of the rider essentially nothing but not step up the balance so far that if the market had a sustained correction the firm would have a bunch of theoretical risk on its book (from the market falling and the firm still needing to base payout figures on the then much higher figure). For clarity, if the portfolio was stepping up the anniversary value 6-9% on its own the rider would remain out of the money (read: useless) and the firm would get to keep the fees for the rider without needing to provide a service. These added fees without an increase in risk would help offset the losses taking place in the example about the Professor. Now, assuming the market was growing enough to keep the rider "out of the money" at the time it would have also been a safe assumption that interest rates would have been much higher. The firm's losses on the Professor's fixed assets would slow as the spread of what the assets in the separate account (owned as investments of the professor) were earning and what the firm had promised would contract - offsetting the losses the firm had to recognize. Both situations would create a consistent profit driver for the firm and would eventually make up the financial drag created by the legacy book.
(3) By far the most important reason these types of VA contracts were created is because according to FASB Statement 133: Accounting for Derivative Instruments and Hedging Activities and FASB Statement 133: Implementation Issue No. B7 the mark to market liabilities DO NOT have to be accounted for on financial reports. Yes, that means what you think it means. In today's current environment, one where the worst case scenario for the issuers of VA's happened, one where we have a zooming market locking in the Anniversary values of all those higher balances that end payouts are based from, one where we have a sustained large spread between what the insurance companies have promised and what the sub account client investments are earning, the insurance companies DO NOT have to account for any of the future risk.
According to FASB Statement 133: Implementation Issue No. B7:
"The guidance in the second and third bullet points of paragraph 200 that a traditional variable annuity contract contains no embedded derivatives that warrant separate accounting under Statement 133 remains valid even though the insurer, rather than the policyholder, actually owns the assets. The following indicators provide the basis for concluding that a traditional variable annuity contract is not a hybrid instrument (meaning it could possess derivatives that need to be marked to market) to be accounted for under paragraph 12:
The variable annuity contract is established, approved, and regulated under special rules applicable to variable annuities (such as state insurance laws, securities laws, and tax laws);The assets underlying the contract are insulated from the general account liabilities of the insurance company; The policyholder's premium is invested in contract-approved separate accounts at the policyholder's direction; The insurer must invest in the assets on which the account values are based; The policyholder may redirect its investment among the contract-approved investment options; The account values are based entirely on the performance of those directed investments (the loophole);All investment returns are passed through to the policyholder (including dividends, interest, and gains/losses);The policyholder may redeem its interests at any time; however, it may be subject to surrender charges.
In addition, although the liability to policyholders is not specifically required to be remeasured at fair value with changes reported in earnings under existing GAAP, current accounting practice for traditional variable annuity contracts is to record a liability equal to the summary total of the market value of the assets held in the separate account for the policyholders (loophole)."
Obviously, the insurance companies are following the letter of accounting law. They in fact are marking liabilities for the account values of the VA's but remember, they created these secondary pie in the sky accounts from which the payouts are based. Because technically the pie in the sky accounts won't ever be the actual account value of the contract they never need accounted for even though the endpoint liability will be substantially changed by the pie in the sky account value.
For clarity, if my account at the end of the accumulation period and at the beginning of the distribution period equals $100,000 the firm has to base my GAAP liability with the $100k being the minimum baseline risk (which is the actual account value - should I die early the remainder of the starting distribution value would be paid to my spouse over a particular period of time or to a non-spouse over a much shorter period of time). The $100k is the risk that will be indicated in financial reporting. That is my actual account value.
Now, let's say I was one of those that took the opportunity and added an 8% step-up and 8% annual payout rider (not at all unusual of VA's being sold between 2009-2013 in size and still sold but in less size) and that my pie in the sky account value is $185k from a combination of a few good years in the market and the rider itself growing my pie in the sky account. My annual payout instead of being $8,000 from the actual account balance would be $14,800 which substantially shortens the period time that I have to live to exceed the marked and accounted for risk that my insurance company has marked.
As scary as that seems, it gets worse. Consider this, with equity markets at all-time highs each calendar day that passes with markets at or near these ranges is another 1/365th of global VA contracts with once yearly Anniversary riders locking in the higher (assuming highest contract inception to date values as the markets have never been higher) values of the account value or the pie in the sky value. That means that each day that passes with markets at these levels the VA books at insurance companies have a materially understated amount of risk. This isn't considering that, for a small fee, some contracts were allowing buyers up to 8 "snapshot" Anniversary days a year - which means an even larger amount of outstanding contracts than the 1/365th per day are actually locking in.
One final piece of information to consider, assuming the actual GAAP account values of the contracts are fairly close to the pie in the sky account values the current risk situation is significantly less than it could be should the market ever experience a real pullback. Historically, the market witnesses a gut wrenching pullback at least one a decade. If you start to imagine spread between the GAAP account values and the pie in the sky payout values should a 20-30% pullback in the markets happen you can start to realize the full scope of this potential time bomb sitting on the books of so many firms. Once a client exits the accumulation phase and enters the payout phase the spread between GAAP account value annual payout and pie in the sky payout determines just how understated the risk is and after how many years the actual payout will be a direct loss of the firm.
For me, I'm certain at some point the market has to pullback. For the insurance companies, they better hope it happens sooner rather than later because each year it doesn't the clients get to lock in a higher and higher annual payout that the firm hasn't accounted for.
This type of VA on steroids has been greatly reduced in selling the last few years and for good reason. In one example, I knew a guy who had worked four months on closing a client who was going to move just under $1million into a Jackson National Life VA and had printed paperwork and had already sold the client on the idea. A few hours after the client appointment his Jackson rep called and said the firm was no longer offering that VA but would still offer a closely similar VA. The change in terms, the new VA no longer offered an immediate 6% bonus to the pie in the sky account on the first day of deposits, broke the deal and left my buddy out $60,000 of commissions that would have hit the payout grid. Ouch.
The fact is there still are billions of dollars of these VA's in circulation and because compliance officers at B/D's are so sensitive to risks of allowing somebody to surrender an annuity they probably won't be coming off the books anytime soon. The insurance companies selling these will eventually have to pay the piper and when they do it won't be pretty for the financials.
Again, I'm not here to pound the table on the next financial crisis or even saying these will lead to an insolvency but the unaccounted for risk matters and it's not something a lot of market participants are familiar with. I hope the next time you get on a conference call with management or decide to invest in an insurance company you take a long look at the VA book and if possible how much of it was sold from 2009-2013.
Good luck to all.
Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.