#11 | The NLG Trojan Horse

No Lapse Guarantee is supposed to be the most straightforward permanent life insurance product available. The story is simple – pay a premium, get a death benefit. But that simple sheen belies complex underlying pricing structures dubiously called “Shadow Accounts.” Shadow Accounts have some clear benefits. They allow NLG products to operate similarly to other flexible premium products without requiring liquid, accessible cash value in any given year. Shadow Accounts allow carriers to more precisely price the product across lots of different funding patterns. Shadow Accounts also gave carriers leeway to follow the letter, but not the spirit, of the Actuarial Guideline 38 reserving regime applied to NLG products as outlined in The NLG Reserving Debate. And, as it turns out, Shadow Accounts injected quite a bit of risk into NLG that virtually no one in the field, including myself, knew existed.

I’ll start with the story that changed the way I look at NLG. A policyholder was paying approximately $150,000 a year when, in the 5th year, he received a lapse-pending notice from the carrier. To his knowledge, he had paid each premium on time and in full. The carrier said that he had actually triggered a contingent Shadow Account by paying his first premium in two chunks and now owed $750,000 in order to avoid lapse. All of this on a guaranteed product. Stories engender more stories. I started talking about this issue and began to hear more and more stories about other policyholders in similar situations with other companies. What I thought was an isolated event probably isn’t.

The cause of the problem for this particular policyholder, the contingent Shadow Account, is also at the root of the NAIC’s current scrutiny of AG38 reserving. AG38 is a nine step calculation that hinges on a ratio. The numerator is the Shadow Account value and the denominator is the single premium required to keep the policy in force to maturity assuming the Shadow Account value is at zero. The primary Shadow Account pricing structure obviously determines the illustrated premium and the value in the Shadow Account at any given time.

But the trick was that carriers built provisions into policies that specified that another, much more expensive Shadow Account could be triggered contingent on a certain value in the Shadow Account. For example, if the value in the Shadow Account fell below $100 then the contingent Shadow Account would be activated and the client would be faced with ongoing costs to maintain the policy far in excess of any original estimate. For the purposes of AG38, the primary Shadow Account could be used for the numerator and the contingent Shadow Account for the denominator. The effect was to make the ratio very small, which means much smaller reserves. In short, using two Shadow Accounts instead of one allowed carriers to reserve less than they otherwise would with only one Shadow Account.

Guaranteeing a large denominator via the contingent Shadow Account obviously poses risk to the client because of the possibility that it could be triggered. The question, then, is how likely any client might be to trigger that contingent curve. Most early primary Shadow Account cost structures had COI curves roughly approximating to real mortality, so the costs would be low early and then become steep in the tail. This allowed for lots of accumulated value in the Shadow Account and, by extension, little chance that the contingent SA would be activated. But some of the later versions of very popular NLG products were designed specifically to limit buildup in the Shadow Account. This change had two effects. First, the numerator of the ratio was smaller (lower SA value) and, therefore, the ratio was lower holding all else constant. Second, the client was more likely to trigger the contingent curve by missing only a single premium.

Lincoln Financial’s old LifeGuarantee UL and LifeGuaranteeUL 09 provide insight into this shift. LifeGuarantee UL had a conventional SA cost curve and LifeGuarantee UL 09 had an almost flat SA cost curve. The impact is profound – after 5 $12,000 premiums, LifeGuarantee UL was sporting 46,000 in Shadow Account value while LifeGuarantee UL 09 had 8,000. After 10 years, LifeGuarantee was at 108,000 and LG09 had a paltry 14,000. LifeGuarantee offered ample cushion for mistakes. LifeGuarantee 09 offered almost none. It’s also little surprise that Lincoln is one of the companies (and there are several) at the center of the debate about AG38.

The primary argument made by carriers in defense of the dual Shadow Account methodology is that AG38 reserves are too onerous. I think that argument has some merit. The ratio outlined above is multiplied by statutory net single premium (based on guaranteed charges in the cash account and a specified crediting rate) and added on top of economic reserves. Why the statutory net single premium? I have no idea. Obviously the guaranteed charges in the cash account have very little impact on how an NLG product actually performs. The secondary argument, then, is that onerous reserves make life insurance too expensive and that carriers were actually looking out for consumers by selling cheaper products funded by holding smaller reserves. That’s also a fair point, assuming that AG38 reserves are actually too heavy. But the way that carriers eluded some of the reserves flies in the face of that argument. Yes, using two Shadow Accounts reduces reserves and provides cheaper insurance for the majority of policyholders. Yet at the same time, this method amplifies risk and imperils a small subset of policyholders who accidentally trigger the secondary curve.

One could also make the argument that carriers knew precisely what they were doing and did it to deter investors from buying the policies on the secondary market.  Settlement funding companies typically pay the minimum costs to maximize mortality arbitrage. Using minimal funding on NLG would result in very low Shadow Account values and likely trigger the contingent account. I’ve asked settlement funders about this and they’ve confirmed that it’s the primary reason why they don’t usually buy NLG policies. The trouble with deterring settlement funders with a contingent Shadow Account is a bit like tuna fishing and, in fact, snaring mostly dolphins. Settlement funders are savvy enough to stay out of the net – agents and policyholders aren’t.

So it turns out that NLG, the “riskless” product, is somewhat of a Trojan horse. NLG is only riskless on the static illustration that assumes no deviation from planned premium payments. In the real world, NLG can transform into a potentially toxic product with massive costs to maintain coverage – precisely the problem that NLG purported to fix.

Fortunately, the new AG38 regulations slated to come into effect in January will eliminate the benefit of using two Shadow Accounts in an NLG product. Prices will likely go up, but the risk that a client triggers a high, hidden cost curve will disappear. More on this in an upcoming post.