#348 | The Next Front in the IUL Illustration War?

aerial photography of two airplanes flying

Quick Take

In each round of the past Indexed UL regulation, the seeds of the next round were already in the ground before the final language was even adopted. The same is true for the forthcoming revisions to AG 49-A, which I’m referring to as AG 49-B. The tactic that will still produce higher-than-intended illustrated returns is a specific loan account with a fixed interest bonus. Pushed to the limit, this simple tactic – which no life insurer is currently using – might produce an additional 10% of illustrated income with zero net benefit to the client. It’s purely an illustration tactic. But there is actually a much more potent variant already in market that is generating industry-leading income and may survive after AG 49-B is implemented. If it does, then it will almost assuredly start a new round of illustration warfare, this time focused entirely on esoteric bonused loan structures. And if that happens, then the fix likely won’t be another actuarial guideline – it will be an overhaul of the illustration model regulation.

Full Article

In all of the past Indexed UL illustration debates, the groundwork for the next round of illustration warfare was laid before the ink had even dried on the new regulation. Prior to AG 49, life insurers were already using buy-up caps (Equitable) and multipliers (Nationwide) to augment illustrated performance. There was tacit agreement amongst Indexed UL writers and non-Indexed UL writers alike that the loophole existed and should not be exploited.

That resolve lasted almost exactly two years until PacLife PDX was released in 2017 and quickly dominated – and consequently reshaped – the entire Indexed UL market. Prior to AG 49-A, AIG had been combining engineered indices with account-specific fixed interest bonuses to augment illustrated performance. The practice was identified during the AG 49-A debates but ultimately ignored, which now leads us to the forthcoming revisions to AG 49-A, which I’m calling AG 49-B.

I thought and hoped, perhaps foolishly, that AG 49-B did the trick. This time would be different. This time there would be no opportunities for pure illustration gamesmanship. The gamesmanship this time around, I thought, would not be in the product itself, but rather in the assets used to support the product. I’ve written as much in this newsletter. I’m already increasingly hearing about life insurers allocating certain assets to support their Indexed UL block in order to maximize illustrated rates at the expense of other in-force blocks of other types of policies. That is not surprising.

Indexed UL has demanded much of agents in order for them to discern what is real and what is pure illustration gamesmanship. Before AG 49, it demanded that they be statisticians. After AG 49, it demanded that they become options experts. After AG 49-A, it demanded that they understand complex engineered indices and philosophize about macroeconomics. Now, it demands that they become experts in general account portfolio construction and management under AG 49-B. Selling Indexed UL without being duped is a damn tall order.

And it’s even taller than I thought under AG 49-B. Ever since AG 49, there has been a side-show issue brewing in the obscure world of participating/indexed/variable policy loan illustration dynamics. Prior to AG 49, those loans could be illustrated with basically as much arbitrage as the life insurer thought was reasonable. A life insurer illustrating at 10.5% could illustrate 5.5% of arbitrage against a 5% loan cost without breaking a sweat. When I’d show up at regulator offices back in 2014 to show them the state of the Indexed UL market, I would show a Indexed UL illustration with $1,000,000 of total premium and around $150 million of illustrated distributions via policy loans. That was the one that made regulators’ eyes pop a little bit.

AG 49 rightfully clamped down on the amount of illustrated loan arbitrage, capping it at 1%. However, that opened up a side issue that I honestly had never thought about until a couple of weeks ago. The problem is basic math. Consider a situation where a life insurer could illustrate 6.2% for an S&P 500 indexed account with a 10% Cap, which costs about 5% these days. Let’s assume the loan charge rate is also 5%. According to AG 49, the maximum illustrated rate on loaned balances is 6%, which is 1% higher than the loan charge rate of 5%. That means policy loans are essentially leaving 0.2% on the table that they used to get to illustrate before AG 49.

However, there’s a workaround that – it must be said – no companies have directly used yet but I’m finding out that many of them are aware of. It’s pretty simple. The first step is to set up a separate account that credits interest to loan balances only. Once that has been done, the life insurer can “optimize” the relationship between the illustrated rate and the loan charge rate by increasing the loan rate from 5% to 5.2%, allowing the full 6.2% to be illustrated. But doing so also means that the life insurer now has an extra 0.2% of yield on the loan asset that it doesn’t need to support the declared Cap. So what does it do with the extra 0.2%? It credits it back to the loaned values in the form of a fixed interest bonus. Voila. The net effect is that the 1% restriction is circumvented and the full, pre-AG 49 illustrated arbitrage of 1.2% can be shown for competitive purposes. Here’s how it looks in a table:

Non-Loaned ValuesLoaned ValuesLoaned Values + Bonus
Account Cap10.0%10.0%8.7%
AG 49 Max Illustrated Rate6.25%5.50%5.50%
Hedge Budget5.00%5.00%4.40%
“Option Profits”25%25%25%
Loan Interest Rate5.00%5.00%
Loan Interest Bonus0.00%0.60%
Loan Credited Rate5.50%6.10%
Illustrated Loan Arbitrage0.50%1.10%

In the grand scheme of Indexed UL gimmickry, this is not a powerful tactic. It only affects loaned values, which take a long time to materialize. However, the tactic became more powerful as regulators further restricted illustrated loan arbitrage to 0.5% with AG 49-A. Now, the difference between the allowed loan arbitrage and the illustratable loan arbitrage isn’t 0.2% – it’s 0.7%. That’s enough to boost illustrated income by 10% or more. And if AG 49-B has effectively left this as the last remaining loophole, then that extra 10% could count for a lot.

It’s also worth pointing out that the actual impact to consumers is essentially zero. In my example above, raising the loan rate from 5% to 5.7% in order to capture the full illustrated benefit but then crediting the 0.7% back to the policyholder leaves them in the same position* as if the carrier had simply charged the original 5% loan rate. This sort of thing is illustration gimmickry, pure and simple. It has a single and sole purpose to boost the illustrated competitiveness of the product. That’s it.

How will you know if life insurers are playing this game? The tell-tale sign is that they have dedicated accounts for crediting interest for loans and those accounts have fixed interest bonuses that are different than what is found on the base contract. If both of those are present, then there’s a good chance that the life insurer is employing this tactic and agents need to handicap the illustrated performance of the product to account for it.

However, like the previous iterations of illustration gimmickry under the various Indexed UL illustration regimes, the question isn’t so much how it works but rather how far the limits can be pushed. The example above is pretty benign, but it doesn’t take much creativity to see how companies could make this strategy quite a bit more potent. In fact, we already have an example courtesy of North American’s Builder Plus IUL 3, already the most aggressively illustrated Indexed UL product in the market.

The story around Builder Plus 3 seems, at first blush, to be an amplified version of what other life insurers are doing by combining fixed interest bonuses with engineered indices. Builder Plus 3 has a Fidelity index with a 1.65% fixed interest bonus that pops to 2.65% in year 11 and stays there for the life of the contract. That eye-watering 2.65% is, I think, the quickest and easiest explanation for why Builder Plus 3 illustrates light-years better than its competitors and why North American is up more than 100% through Q3 of last year.

However, that’s not the complete explanation. In my review of Builder Plus 3 when it was launched, I alluded to the fact that something was up with the loan structure of the product. It illustrated far better income than it should, but I couldn’t quite figure out why. The fact that I couldn’t piece it together using just the materials I had available, including (obviously) Builder Plus 3 illustrations, is a problem worthy of a separate discussion. But now I have the rest of the pieces to the puzzle.

First, some clarifying points about Builder Plus 3. As I wrote in the original review, the product has a 1.25% asset-based charge on Account Value. Therefore, the net bonus actually credited to Account Values for the Fidelity engineered indexed account after year 11 isn’t 2.65%, it’s 1.4%. That puts Builder Plus 3 at the top end of the market for illustrated fixed interest bonuses, but it doesn’t blow the doors off of its competitors. And as I also wrote in the last article, the fixed interest bonus is also helped by the fact that North American used a hypothetical BIA which allows for a higher illustrated rate for the Fidelity account than the S&P 500 capped account.

Everything I wrote above applies only to unloaned Account Value. Once the policyholder takes a loan and a loaned value is created, a separate set of rules applies. The loaned value is not assessed the 1.25% asset-based charge. The illustrated rate drops to 0.5% above the declared rate, which is currently 5%, bringing the total illustrated rate to 5.5%. The loaned value no longer earns the bonus attributable to the indexed crediting selection, which is 1.65%/2.65% in the case of the Fidelity index and 1% for the other strategies in the product.

Instead, the loaned values – regardless of their index allocation – earn a 2% bonus which is configured as a part of the loan structure itself, not a part of the indexed crediting method. This brings the total illustrated loan arbitrage to 2.5%, by far the largest in the industry, and the key reason why Builder Plus 3 is such a potent competitor for income illustrations. Everyone, including me, was focused on the wrong thing. The real power of BP3 isn’t the 2.65%, which nets out to 1.4% anyway. The real power is the 2.5% illustrated loan arbitrage courtesy of the 2% loan-specific bonus.

North American makes it really easy to see how big the impact is from the illustrated loan arbitrage. They also offer a Variable loan rate that does not have a fixed interest bonus but still illustrated with the standard allowable arbitrage of 50bps under AG 49-A. On a 45 year old Preferred Male funded with $1M of premium, the distributions using the loans with the bonus will illustrate $200k of income from years 21-40 using the S&P 500. But using the Variable loan rate with no bonus generates just $157k of income. That’s a 30% swing just because of a 2% bonus. In the world of AG 49-B, that’s a huge difference, enough to vault a company from middle of the pack to top of the pack.

How does North American pay for this bonus? I don’t know, so I’m going to speculate – in general – about how any life insurer might be able to do something like this. The key seems to be the assumption of the percentage of the Account Value (AV) that will end up being converted to Loaned Value (LV). This is a very important assumption for a structure like the one in Builder Plus 3 because two different sets of benefits apply to AV and LV.

For AV, Builder Plus 3 charges a 1.25% asset-charge to all accounts. From there, it credits 1% back and then applies additional interest for the Fidelity account, which means that the product probably nets about 25bps in asset-based fees from this strategy. So let’s keep it simple and imagine that all AV produces a 25bps net fee to North American. And let’s also imagine that North American keeps that 0.25% and invests it to “reserve” for the 2% bonus that it will have to pay out on loaned values. This is not so different than how companies fund persistency bonuses but, in this case, the persistency bonus is only applied to loaned values.

As in the case of a normal persistency bonus, surrenders and deaths will materially decrease the potential Loaned Values in the future while funding (at least for a time) the pool of “reserves” based on the 0.25% net fee. However, with normal persistency bonuses, companies can only push this logic so far while still passing actuarial testing for lapse-supported pricing, which requires that the company use a 0% lapse assumption for the test. There’s a limit.

But loans are a different animal because the company gets to make the assumption about what percentage of the future AV will be converted to LV through policy loans. Despite the fact that the vast majority of Indexed UL illustrations show policy loans, my understanding is that many insurance companies look to historical data on Universal Life, Whole Life and Variable UL as evidence that clients actually don’t take as many loans as are illustrated. Northwestern Mutual, for example, has just $17.2B of its $222B in policy reserves as policy loans (Q2 2022). If a company makes that sort of assumption, then the bonus that is payable on Loaned Value can be much, much larger than what they’d be able to do as a pure persistency bonus on Account Value for the simple reason that they expect fewer dollars will be eligible for it.

This is actuarial magic at work. A company can essentially take a small fee from Account Value to fund a massive bonus on Loaned Values that shows up in literally every single Indexed UL illustration while still arguably passing actuarial support testing because the company doesn’t actually expect very many policyholders to avail themselves of policy loans. If North American funds a 2% Loaned Value bonus with a 25bps Account Value charge, then what could another company do with 50bps or even 1%?

As it stands right now, the slam-dunk, bright-line approach for maximizing illustrated performance using loan-specific accounts will yield – I don’t know – 10-15% in income increase. That’s what the first part of this article was about. But if North American keeps this loan structure on Builder Plus 3 and continues to illustrate a 2% arbitrage on policy loans after AG 49-B, then they’ll probably pick up more like 25-35% more illustrated income than they should. That’s enough to turn heads. That’s enough to ignite a new illustration war. That is more than enough to get other life insurers to follow suit with their own series of Loaned Value bonuses.

A clarification to AG 49-B or even a new AG 49-C can eliminate the base issue with loan-specific accounts, but this issue with loan-specific bonuses potentially goes to the heart of the entire illustration framework under the illustration model regulation. Regulators have reluctantly begun discussions about opening the illustration model regulation. They know it’s a political minefield. But if strategies like this one become prevalent, then it might just force the issue.

We’ll see. What North American does in the wake of AG 49-B will be the tell. I expect the vast majority of companies to rework their rates and bonuses on engineered index accounts so that those strategies won’t be at an illustrative disadvantage when the new guideline hits in May. What that will mean in virtually every scenario is increasing participation rates and reduced bonuses. If North American follows suit and drops the 2% bonus on the loaned strategy, then we’ll know that it had more to do with the engineered index pricing than anything else. But if they keep it, then that means that the bonus was being funded by something other than the engineered index. And if that happens, get ready. There’s going to be a whole new kind of Indexed UL illustration war.

*That’s actually not quite true. The fact that the loan accrues at a higher rate means that the loan balance will be bigger than it otherwise would have been, although the increased loan balance will be offset by higher growth in the loaned collateral. The net effect will be similar, but not exactly the same.