#158 | Life Insurance in the Los Angeles Times
When the life insurance industry gets press, which is quite rare, it’s usually bad press. In this case, the bad press was more than justified. Take a look: https://www.latimes.com/business/hiltzik/la-fi-hiltzik-register-insurance-20190219-story.html. Obviously, there are many factors at play in the case of the demise of the O.C. Register than just a massive life insurance placement. But the article correctly points out that this was not a suitable sale and was based on some faulty economics and accounting. I wish I could say that this was the first time I’d heard of this scheme, but it’s not. Not by a long shot.
The basic idea behind the sale is simple. Insure employees of the company for, let’s say, $100M in total death benefit coverage on permanent life insurance policies. The pension plan pays for the premiums and the death benefits are paid to the pension plan, meaning that life insurance is essentially operating as an investment within the pension plan. At face value, it sounds somewhat reasonable. Life insurance can have favorable economics relative to other asset classes in the long run. Pension plans invest for the long run. Is it wrong, then, for a pension plan to invest in life insurance?
Excluding the issue of insurable interest, the real problem with this scheme isn’t the life insurance policies themselves. The real problem is the accounting. As the article mentions, the O.C. Register was advised by the actuaries and accountants hired by the insurance salesman to book the value of the policies at the net present value of the future benefits. This implies two major assumptions – mortality and the discount rate. I have no doubt that the “experts” employed by the insurance salesman made a compelling case for a steep mortality curve and a low discount rate. That’s the only way that, as the article notes, the purchase of life insurance would immediately serve to reduce (on paper) the funding gap in the pension plan. Spending $10M to buy $100M of death benefit valued at $30M, hypothetically, results in an immediate net $20M asset to the plan. It’s a pretty amazing little strategy – especially if the premiums are financed, as they likely were for the O.C. Register. As one promoter of a strategy like this one put to me, it’s “lightning in a bottle.”
Except that it’s complete garbage. The net present value of the death benefits paid by a life insurance policy are equal to its reserve which, in the case of most insurance policies, is equal to cash value. That’s the way valuations have always worked for life insurance and the way that the accounting for the pension should have worked, according to its auditors. Had the purchases been accounted for correctly, the pension plan’s funding shortfall would have increased as a result of purchasing the life insurance. Hardly a compelling case for a sale with a commission that was probably in the millions. That’s plenty of incentive to advocate for, shall we say, alternative accounting.
Unfortunately, the case I saw several years ago was even more egregious than this one. It involved hundreds of millions of dollars of VRDO-financed premium placed into Voya Indexed UL products illustrated at 8%. The scheme was supposed to close the shortfall in this particular pension plan by billions of dollars. Fortunately, the plan didn’t go through with the scheme and the promoter eventually went away.
Both of these cases serve as cautionary tales. We insurance folks sometimes get too clever. Right now, we think we’re very clever with our Indexed UL products and their 50% option profit assumptions, which are sold not just to shore up pension liabilities but also for premium financing, executive benefits and as the end-all, be-all retirement income solution. But being clever has never ended well for life insurance. Vanishing premium was clever. STOLI was clever. 409(a) was clever. Section 79 was clever. For once, can we just acknowledge that we’re really not so clever and just sell some life insurance for what it is, not for what we make it out to be?