#312 | The New Money Tipping Point

grayscale photo of person popping a water balloon

With rates making a decisive and sustained move upwards for the first time since 2019, we have to ask an uncomfortable question – at what point will life insurers begin to wall off in-force portfolios and offer new portfolios that are invested in current, higher yielding assets? It’s a matter of when, not if.

The industry has been here before, but it’s been a while. The last time that new money rates were consistently higher than portfolio yields was the mid-1980s, and it had been that way for decades stretching all the way back to the early 1950s. Here’s the relationship between new money yields, modeled with a 50/50 blend of the Moody’s Aaa and Baa composites versus a rolling 10-year average of the same blend, which is a simplified proxy for a life insurer portfolio yield:

Using the simple rolling average model, the portfolio yield does exactly what you’d expect it to do – lag. It lags when interest rates increase and it lags when they decrease. How closely does this simple math match up to actual portfolio yield histories? The closest proxy we have for portfolio yields are the various dividend interest rates paid by major Whole Life writers. Here are Northwestern Mutual, New York Life and Mass Mutual’s declared dividend interest rates relative to the simple rolling average modeled portfolio yield:

It is amazing, really, how well a simple rolling average model for portfolio yields mirrors what life insurers actually did with their dividend interest rates over time – with the notable exception of the rising rate period from 1955 to 1984. During that time, actual DIRs significantly lagged not only market interest rates but even the basic portfolio yield model. Why did that happen? I called a friend of mine who was the head of product for a major mutual company during that period of time and he gave me, in broad strokes, the story of what was going on.

There were three things at work – tax law, policy loan structure and competitive dynamics. Prior to 1982, it appears that life insurers had to pay income tax on the difference between the guaranteed interest rate in the policy and their net investment rate. To make matters worse, rising interest rates also increased the marginal tax rate, potentially all the way to 100% above a certain threshold. Hence, a reason why DIRs lagged behind the simple moving average portfolio by an even greater margin as interest rates increased.

But the tax treatment of life insurance had another quirk at the time. As unbelievable as this may sound, policyholders were allowed to deduct interest on their policy loans as long as they had paid 4 out of 7 premiums out of pocket. This was a huge incentive to take policy loans and consequently many life insurers had a significant portion of their balance sheet tied up in policy loans. These loans were generally structured as non-direct recognition with a fixed loan rate of 5-6%. As interest rates increased, policyholders began to take out policy loans to invest in fixed income assets with much higher yields, forcing DIRs down as low-yielding policy loans dragged down the portfolio yield.

The solution was twofold – either switch to a policy loan rate indexed to fixed income yields (such as the Moody’s Aaa composite) or to direct recognition with a fixed loan interest rate. Some companies chose the former, some companies chose the latter, some companies offered both. For companies that went the direct recognition route, they could immediately post a huge increase in the DIR for non-loaned rates, which is exactly what Guardian did in 1983 when its DIR jumped from 7.65% to 12.25%.

Finally, competitive dynamics were such that prior to the advent of Universal Life in 1981, the small cadre of major mutual companies were operating more or less under a fragile peace and shared anguish. But when EF Hutton, Executive Life and Transamerica rolled out new money Universal Life products sporting 14% interest rates as early as 1980, the mutual companies had no choice but to respond.

For most of the major mutual companies, the response was to modify their policy loan structures, take advantage of the new tax structure that reduced their tax bill and ride it out while their portfolio yields increased to catch up to market yields. But for other companies – most notably Principal and New York Life – the response was also to switch to a new money structure for their Whole Life block. I couldn’t find a historical dividend report for NYL, but I was told recently that the company created a new portfolio in 1982 with a significantly higher DIR that reflected the market conditions of the time. This spurred me to do a bit of digging in the dusty corners of the internet.

Turns out, there was actually a class action lawsuit (Banks vs. New York Life) for NYL policyholders from 1982-1994 and one of the claims in the lawsuit was that “beginning in the early 1980’s New York Life began to base the method of projecting interest and dividend rates of their life insurance policies on a higher rate of return based upon the market of the past few years rather than a rate previously used.” In other words, a new money portfolio. I also stumbled on a couple of old earnings releases that paint the rest of the picture. In 1981, New York Life posted a 2.2% growth in individual life insurance sales. But in the 1983 release, New York Life reported a sales increase of 43.6% for individual life insurance sales. Little wonder why Guardian, MassMutual and Northwestern Mutual all massively increased their dividend interest rates in 1984. New York Life’s gambit with new money rates to compete with Universal Life worked and the other companies were forced to compete, even if they stuck to their guns by using portfolio crediting rates.

The irony, of course, is that 40 years after the fact, the New York Life standard portfolio and the 1982 portfolio series now have, from what I understand, exactly the same dividend interest rate. Why? Because the effect of new money yields necessarily burns off quickly because of investment rollover and recurring premiums. New money is only new for so long. Then it just becomes a portfolio yield like any other. Take a look at how the simple rolling average portfolio yield would have looked if a life insurer had, for example, started a brand-new portfolio at the absolute height of interest rates in the mid-1980s:

The problem, though, is that even though the ultimate fate of both portfolios is identical, they don’t illustrate that way. At the inception of the new portfolio, the illustrated DIR for the life of the contract would be around 16%, well north of the illustrated DIR for the old portfolio of around 10%. This is a temporary disconnect, but it is illustrated as if it is permanent.

The fact that this temporary disconnect is illustrated as being permanent exposes a fundamental flaw in the illustration model regulation. The Reg does not take into account any future drift in assumptions, for positive or negative, even if those assumption changes are structural and guaranteed. We know that a portfolio created today and a portfolio created 5 years ago will, all else being equal, eventually converge to have the exact same performance assuming the same underlying investment strategy. Yet that is not how it illustrates.

Therefore, the incentives to switch to new money in a rising rate environment are huge. A life insurer can vault from the bottom of the heap in terms of illustrated performance to the top of the heap with no change to their profitability and no long-term impact to actual policy performance. It is purely an illustration game. And what have we seen every time life insurers are presented with a no-cost way to juice their illustrated performance? They take it, regardless of the consequences.

The time for new money products might be coming faster than you think. Take a look at the table below, which compares the 50/50 blend of the Moody’s Aaa and Baa composite yields against the 10-year rolling average of the blended rate which we’ve been using as a proxy for a life insurer portfolio yield:

Even though today’s corporate bond yields are similar to what we saw during the stretch from 2014 to 2019, the real issue is the relationship between new money yields and portfolio yields. During none of the previous spurts in interest rates, except perhaps a few months in 2019, did new money yields consistently pierce portfolio yields. Now, new money yields are decidedly in excess of portfolio yields – at least based on how I’m modeling it here – and it’s reasonable to assume that they’re going to stay that way as the Fed clamps down on inflation. This is the spark.

How long will it take to see a fire? That depends. The runup to the early 1980s is instructive. It took nearly 30 years of rising rates for any major mutual company to cave and create a new money portfolio. Even then, the bulk of Whole Life companies stuck to their guns and maintained portfolio crediting, preferring to restructure their financial and policy loan provisions to catch up more quickly. Those changes are still in effect today. There’s no reason to think that the mutuals will be as slow to keep up with rising rates in the future as they were 60 years ago.

Smaller mutual companies, however, may be quick to see an opportunity to scoop up sales by switching to a new money portfolio. It’s more likely that Penn Mutual, Ameritas, Lafayette or even OneAmerica decides to launch a new money series than one of the Big 4. These companies have much to gain and not nearly as much to lose – not unlike the much smaller and much less dominant New York Life of the 1970s. If history is any indication, it will only take one or two companies to lure the rest of the crew into it.

For Indexed UL, the incentives to switch to new money are even stronger than in Whole Life thanks to the illustrated leverage inherent in AG 49/AG 49-A. For Whole Life, switching to a new money portfolio to pick up an extra 1% in DIR will correspond (approximately) to an increase in illustrated performance of around 1%. Not so for Indexed UL, where a 1% increase in portfolio yield will buy something like an extra 3% in S&P 500 point-to-point cap (9% to 12%), which translates into an illustrated rate increase of 1.35%. That size of an increase would vault a product from the middle of the Indexed UL pack to the front by a country mile, excusing any sort of games being played with proprietary indices and fixed interest bonuses. New money is going to prove to be very, very tempting for Indexed UL writers.

And if rates keep going up, it will happen. Why? Because some companies are going to naturally benefit from rising rates more than others. Take, for example, two Indexed UL writers – Accordia and F&G. Accordia is a mature Indexed UL writer with a large in-force block of business that grew from $3.75B at the beginning of 2021 to $4.24B by the end of the year, an increase of about 13%. F&G, by contrast, has a much smaller block of business ($550M) that grew by a whopping 26% in 2021, double the rate of Accordia. Which company is going to benefit more in a rising rate environment? It stands to reason that F&G will, simply because their block is receiving a lot of new inflows that can be reinvested at higher rates. Discrepancies will develop and the only way for some companies to compete is by switching to new money. They’ll say that they didn’t have a choice.

This will ignite a new illustration war that revolves less around product designs and more around which carrier can fire up a new portfolio with a higher yield that produces higher illustrated performance. All of the rules would change. A life insurer might have two identical products, but one illustrates 50% more income than the other. A producer might tell a client to exchange an old portfolio Whole Life policy for new new money Whole Life policy simply on the basis of illustrated performance. Imagine a client paying the minimum premium to maintain coverage for life at a premium level less than his current Guaranteed UL premium simply because the company is using a new money crediting rate for the new policy.

In this world, it would be very difficult, if not impossible, for clients, producers and distributors to decipher between the sustainable and the ephemeral, the permanent and the temporary. They would all look the same. Clients would make real decisions and producers would make real recommendations on the basis of temporary dislocations being misunderstood to be permanent. Life insurers would cannibalize their own books of business on the false premise of new products outperforming old, forcing capital losses on surrenders and exchanges. In-force policyholders would be stranded after years of diluted yields to support new business.

In short, igniting an illustration war on the basis of new money yields would be a complete and unmitigated disaster for everyone involved. It would be a war of mutually assured destruction – and we should vigilantly guard against it at all costs.