#223 | The Tangled Reinsurance Web – Part 4 – Coinsurance and Capital

As we’ve already seen, the power of coinsurance is that a life insurer can essentially borrow the economics of another life insurer within the defined confines of the reinsurance treaty. It’s not unlike what John and Jeremy, the protagonists of Wedding Crashers, would do when they’d show up to a wedding – take on a new identity to fit into the crowd for one night. That’s coinsurance. It’s a new identity for the life insurer, but only for the particular block being reinsured.

For most insurers, ceding to a third-party reinsurer isn’t much of a change in identity. Reinsurers are big, sometimes bigger even than their customers. Reinsurers and direct writers are generally licensed in nearly all states, which means they have to play by national disclosure and capital rules. Coinsurance with a major reinsurer is more about the particular risk appetite of the reinsurer than any structural difference between the companies. If all direct writers wear suits and ties, then all of the major third-party reinsurers also wear suits and ties – just different suits and different ties, maybe sometimes with some funky footwear or a fancy watch.

Captives, on the other hand, wear jeans and hoodies. It’s a major shift in identity. Captives are generally only licensed in the state or country where they are domiciled, making them an “unauthorized” reinsurer in other jurisdictions. They play by their domicile’s rules and certain domiciles are more, shall we say, accommodating of captive reinsurers than others. Vermont, Delaware, South Carolina and Arizona tend to come to the top of the list for life insurer-owned captives. There are a lot of ancillary benefits to a captive but for the purposes of this discussion, there’s one benefit to rule them all – capital relief.

When a block of business is reinsured, the ultimate liability remains at the life insurer but the assets and premium flow have now moved over to the reinsurer. In order to satisfy the liability, the life insurer books a reserve credit on the reinsured block so that, from a reserve standpoint, the transaction is net neutral. Theoretically, there should be no impact to the life insurer’s net capital position from using coinsurance. In fact, to highlight when reinsurance does anything other than produce net-neutral flows to capital, there’s actually a line item in the statutory filing that specifically serves to highlight if reinsurance has impacted surplus capital. For companies not using captives, that line is almost always zero.

But for companies using captives, that line is sometimes worth hundreds of millions of dollars. The strategy for using captives is quite simple. Captives are only required to follow the statutory rules of their domicile, which means that their capital and reserving regimes may deviate from the multi-state NAIC rules. For most insurance products, the reserve regime is very straightforward. The reserve for a standard Whole Life product, for example, is equal to its cash value. Same for an Indexed UL or a Universal Life product. There isn’t really any discretion, if you will, in determining the reserves for products like that. But the same doesn’t apply to products with long-term embedded guarantees. Determining the appropriate reserve for a product like that is necessarily dependent on the assumptions the carrier uses for future interest rates, lapses and mortality over the next 50+ years. Naturally, every company has a different view of what assumptions are appropriate. As a result, ten companies would likely come up with ten different economic reserves for the same block of Guaranteed UL policies. And the wiggle room is significant. Take a look at a schematic of how the formulaic reserve might compare to the economic reserve as calculated by the life insurer.

Economic reserves are discretionary by their nature, which is exactly why the NAIC previously used very conservative, formulaic reserves for products with embedded guarantees. Companies liked to call these reserves “redundant” as a way of making them seem unnecessary, but that’s not really accurate. The reserves are so high because the risks embedded in these products are so high. If 10 companies come up with 10 different valuations of the same block of business, what’s the probability that any one of them is correct? And more importantly, what happens when they’re wrong? The old formulaic reserves were arguably designed as a way to force life insurers to set aside capital in a tax-deductible way (as a reserve) in recognition of the fact that the actual liability of the block is unknown. Forcing carriers to back risky blocks with a lot of capital was a natural brake for carriers taking on too concentrated risk in one product category. A company could only write so much Guaranteed UL, for example, before it would start to seriously eat into its available surplus. At least, that was the way it was supposed to work.

Captives allowed carriers to remove the braking mechanism. Because captives in friendly states didn’t require the full NAIC reserve for the block of business, a life insurer could seed a captive with “hard capital” equal to its view of the economic reserves for a block of business plus an appropriate capital cushion. The remaining reserves that would have been required by the NAIC could then be satisfied with other financial instruments – “soft capital” – such as letters of credit or parental guarantees. Sometimes these financing packages are put together in a specialty trust which is then pledged to the reinsurer and used as collateral. The full reserve credit then goes back on the life insurer’s books despite the fact that the “hard” capital they’re using to back the reserve is less than what it would have been if the life insurer directly held the liability. Voila. Using captives, a life insurer can transform required NAIC formulaic reserves into deployable capital simply by exploiting the difference between the NAIC formulaic reserve and the “economic reserve,” whatever that means.

How much capital can these captives release? Usually that’s pretty hard to pin down because there’s so much going on within the life insurer’s books. Capital can come from a variety of different sources – operational earnings, unrecognized capital gains, you name it. Rarely do you get a chance to see a clear before-and-after picture of a company when they execute a reinsurance trade to a captive. Fortunately, we have something of a case study in PennMutual. First, some historical perspective. Take a look at PennMutual’s Total Adjusted Capital (TAC) going back to 2008 overlaid with the ratio of TAC to total general account liabilities.

PennMutual’s TAC tends to follow a pattern. From 2010 to 2013, TAC was floating around the $1.6B range but gradually declining. Then, suddenly, TAC jumps to $1.9B in 2014 and gradually declines until it pops in 2018 and again in 2019. But PennMutual is growing over this period of time – and fast. In 2007, PennMutual’s total general account liabilities were just $5.3B. By the end of 2019, they’re well north of $20B, nearly 4x growth over just 14 years. All of that growth means that more capital has to be tied up to support the growing pool of reserves. Hence, the reason why TAC as a ratio of general account liabilities has been dropping almost unabated from about 18% in 2008 to 10% in 2019. It almost looks like PennMutual is creating just enough capital to target a 10% ratio, although that might just be a coincidence. I’m not implying that PennMutual is any sort of capital pinch – a 10% TAC/Reserve ratio is about the same as Northwestern Mutual, significantly better than Lincoln (7.6%) or Symetra (6.8%) but well behind New York Life and Guardian (both 14.1%). That also puts an 18% ratio into perspective. I’m fairly certain that back in 2008 PennMutual would have been one of the most heavily capitalized life insurers in the country. So what do you do with all of that excess capital?

You put it to work. Growth and capital are always at odds with one another. If a company wants to grow, it has to spend capital to do it. Fortunately, for PennMutual, it had a treasure trove of capital to put to work. But as I’ve written before, manufacturing life insurance policies is like planting trees – it takes a lot of work to get it into the ground and takes a long time to grow. For most life insurance policies, the year where the product starts to kick off net positive capital is between 7 and 15. So if a life insurer puts a lot of business on the books, it takes a decade or more to see a net positive capital position from that sale. As a result, carriers have to balance both the capital strain of business already on the books and writing new business because of the time lag between when a policy issued and when it’s accretive to capital. You can see this playing out in PennMutual’s appetite for new sales. Take a look at the chart below that compares TAC to new life insurance sales at PennMutual.

Again, there’s a clear pattern. PennMutual appears to be targeting a certain TAC level and then writing new business that soaks up just enough operating earnings to maintain the desired TAC. As an outsider, I’d say that this looks like the strategy a company would employ to simultaneously grow and maintain its overall financial strength position. Rather than pouring those operating earnings back into capital, which it doesn’t need, PennMutual puts it back to work by writing new business. Whenever the pace of new sales gets a bit too fast, PennMutual applies the brakes. And then, suddenly, more capital hits the balance sheet and PennMutual starts to grow new sales again. This leads to an obvious question – where’s all of that extra capital coming from? You probably already have a guess by now and, if so, then you’re probably right.