#354 | SVB, Liquidity & Life Insurance

Quick Take

The collapse of Silicon Valley Bank has shaken confidence in the financial system and raised questions about whether life insurers could find themselves in a similar situation. At first blush, the data is not comforting. SVB’s unrealized capital losses equated to its entire Tier 1 capital, meaning that the bank would be insolvent if all of the losses were realized. Of 35 insurers analyzed for this article, 21 would be in a predicament as bad or worse than SVB. If unrealized capital losses are a tumor, then liquidity restrictions determine whether the tumor is benign or malignant. For banks, the tumor is malignant. Liquidity is unrestricted and depositors can demand it immediately. For life insurers, however, the tumor is benign. They have layers upon layers of liquidity restrictions built into their products and contracts. As a result, the danger of an SVB-type situation is minimal. Complications may remain, but the fundamental prognosis is strong.

Full Article

In the wake of the stunningly quick collapse of Silicon Valley Bank last week, one question that keeps coming up – can this sort of thing happen to life insurers? The fundamental problem at SVB was unrealized capital losses on long-duration bonds bought when interest rates are low. According to JP Morgan, SVB was an outlier in that a full recognition of its unrealized losses would have completely wiped out its Tier 1 capital. Hence, the run on the bank.

But SVB obviously isn’t alone in grappling with this issue. The problem of unrealized capital losses sitting latent on bank balance sheets is universal. According to the FDIC, the banking industry was sitting on about $620 billion in unrealized capital losses at the end of 2022. How big is that number? Put into the context of Tier 1 capital – the banking equivalent of what insurers call surplus – unrealized capital losses equate to about 40% of the $1.7 trillion total. Immediate recognition of all of these losses would not destroy the entire banking industry, but it would destroy quite a few individual banks. Take a look at a JP Morgan chart comparing the relationship of Tier 1 Capital to unrealized capital losses at some of the nation’s largest banks:

Life insurers are in a not-so-dissimilar situation as banks. They invest in long-duration liabilities that have been devalued as a result of rising interest rates. They hold a certain amount of regulatory capital to cover situations that aren’t explicitly covered by product reserves, particularly investment losses. Certain aspects of regulatory capital are actually specifically designed to reflect the liquidity characteristics of products. A product with greater liquidity requires more capital and vice versa. Like banks*, insurers hold their fixed income investments at book value for statutory accounting purposes. Unrealized capital gains and losses therefore sit latently on insurer balance sheets and are activated only if traded. It’s the exact same dynamic as with banks – the losses are there but are only realized if people start asking for money. There is only a problem if both parts happen at the same time.

To get a picture of the current state of the life insurance and annuity industry in terms of unrealized capital losses, I pulled year end 2022 statutory statements for 35 large US life insurers with a total of a little over $2.1 trillion of book value invested fixed income assets. These insurers, I think, cover the vast majority of the life and annuity space. Below is a recreation of the JP Morgan chart, but this time with data from the 35 life insurers. Capital, for this chart, is pure statutory surplus without IMR and AVR**. I used fixed income assets as the denominator rather than total assets, so this is not a complete picture of capitalization. It’s focused entirely on the risk addressed in this article, which is unrealized capital losses in fixed income. Blue dots reflect the ratio of statutory capital to fixed income assets. Gold dots, like the JP Morgan graph, reflect statutory capital less unrealized capital losses as a ratio of fixed income assets.

This chart is not particularly confidence-inspiring. In aggregate, unrealized capital losses stand at a whopping $222 billion, or about 10.5% of the total book value of the invested fixed income assets. Unrealized capital losses currently equate to 90.7% of the $245 billion in aggregate surplus held by these 33 life insurers, but more than half of the insurers actually would have negative surplus if all of the unrealized capital losses were marked to market. At first blush, the life insurance industry looks like a whole lot of SVBs.

As crazy as it sounds, I would argue that the capital situation is actually worse than it appears for some companies. Many of the firms in the chart above – Pruco***, Ohio National, Brighthouse, Lincoln, John Hancock, Penn Mutual and Accordia, to name a few – rely heavily on captive insurance companies to bolster their statutory surplus. These captives effectively allow life insurers to swap hard capital for soft capital, which means that actually liquidating the capital to cover losses will be more difficult in those structures than if all capital was held as hard capital.

But that’s only part of the problem. The captives themselves own huge portfolios of long-duration fixed income securities that also have presumably seen steep declines in valuation. As a result, companies that are heavily reliant on captives face a potential double threat. Not only do they arguably overstate their aggregate capital position because of “soft” capital from the captives, but they also dramatically understate their exposure to unrealized capital losses because those losses sit latent on the balance sheet of the captives.

Soft capital isn’t only a phenomenon at companies that are heavily reliant on captives. I would also argue that mutual life insurers also have soft capital but in the forms of surplus notes and unrealized capital gains. Virtually every mutual life insurer has made ample use of surplus notes over the years, particularly when rates were low. Surplus notes are theoretically liquid, but there are contingencies and counterparty risk. More importantly, some mutual insurers have also allowed unrealized capital gains in non-fixed income assets such as equity positions in subsidiaries, limited partnerships and joint ventures to become a large portion of surplus. MassMutual, for example, sports a $28 billion surplus, but nearly $20 billion of that is unrealized capital gains. Those gains are real – but they may not be nearly as liquid as Guardian’s $10.8 billion in surplus with effectively zero unrealized capital gains.

In terms of long-term solvency and the ability to pay claims, I would argue that the distinction between hard capital, surplus notes and unrealized capital gains isn’t necessarily meaningful. All three are valid forms of capital. But in a situation where liquidity is at a premium, as in a run a la SVB, the difference will start to show itself. Even more so for “soft” capital deployed by companies reliant on captives because the capital may be encumbered by parental guarantees, supported by trust agreements or any other myriad of strings. I recall one reinsurer coming to see us at MetLife and pitching us on their financial reinsurance services. Their story was essentially this – “We’ll help you minimize reserves by offloading liabilities to us but we aren’t in the business of actually taking financial risk. We just want it to look like we’re taking risk so you can get it off of your books.” How many captive deals are structured like that? It’s really hard to tell. And for the record, we passed on the deal.

Purely based on the capital position of the industry compared to its mammoth unrealized capital losses, the picture is bleak. Very bleak. But remember – what happened at SVB requires both the presence of unrealized capital losses and bulk demands for immediate and unconstrained liquidity. To use an analogy that hits close to home for my family, unrealized capital losses are like a tumor. The presence of a tumor does not mean cancer. What makes a tumor malignant, in this analogy, are demands for liquidity.

Fortunately for the life insurance industry, our tumor is benign – mostly. We have layers of protection against demands from policyholders for unconstrained liquidity. Is a run impossible? No, it’s not, but it is orders of magnitude less likely than with a bank. And as a result, I would argue that the life insurance industry is far more insulated than banks from the specific type of event that triggered the collapse of SVB. Let’s go through it, layer by layer.

State Guarantee Associations

The analogy to the FDIC in the insurance world are state guarantee associations. The website for their trade group, the National Organization of Life and Health Insurance Guaranty Associations (NOLHGA), offers an almost excruciating level of detail about the operations of state guaranty associations. Long story short, the protections offered by state guaranty associations are relatively strong – $100,000 of life cash value, $250,000 of present value of annuity benefits and $300,000 of death benefit. These guarantees are funded with claims on the assets of the failed insurer and by assessing other insurers who are doing business in the state for the difference. This is similar to how the FDIC operates, but the FDIC has an additional backstop through the ability to borrow from the federal government.

The NOLHGA website keeps a running list of all insurers that have gone into receivership and also a year-by-year tally of assessments for each state and in each line of insurance. The numbers are pretty interesting. Across the board, state guaranty associations assessed large amounts to member insurers in the early 1990s related to life insurance – and I would surmise that a fair bit of that is due to the failure of several insurers around that time, most notably Executive Life. The assessed amounts are not trivial. California insurers were assessed $229 million from 1991 to 1997. The failure of Executive Life of NY in 2012 caused a $565 million assessment to insurers in New York over the subsequent two years. In total, from 1991 to 2017 (when the NOHLGA data stops), insurer assessments for all 50 states have tallied just shy of $3.3 billion. These guaranty associations really work and really do provide a level of protection for policyholders.

Lack of Bulk Demands for Liquidity

Although SVB was one of the largest banks in the United States, it was essentially a mega-community bank that was highly exposed to the conditions in its local town and industry. It just so happens that its local town is one of the most prosperous in the world and its local industry is the arguably the most well-funded in the world. SVB didn’t have a diversified depositor base – and it considered that one of its core strengths until it turned into its fatal weakness. Life insurers don’t have this problem. Outside of a few companies that seem to have more of a geographic or industry concentration, the vast majority of major insurers write business across a huge footprint and have an incredibly diverse client base. No client is big enough to make of break the fortunes of any insurer.

However, there are pockets of concentration for some insurers in some product lines. Bank-Owned Life Insurance comes to mind. If banks universally need liquidity to meet demands, then they may start pulling on their BOLI lines, which in turn would force life insurers to liquidate the assets supporting those blocks. Concentration risk is higher in BOLI than in individual life, without question. Arguably the same might apply to very large COLI transactions. And you might even argue that the same even applies for companies that have taken huge amounts of program-based premium financing transactions that may go sour, forcing policy surrenders and asset liquidations. But none of these actions, I think, would be fatal. Traditional retail policies dominate at every insurer, even those with the largest COLI, BOLI and premium financing blocks. There might be damage, but it will be limited.

Loss of Benefits

At the end of the day, life insurance and annuity products are intended to be used primarily to provide longevity, mortality and/or morbidity benefits. Every liquid dollar that is housed in our products is there because it funds a benefit. Period. That is not the case for bank deposits, where the benefit is the liquidity. As a result, accessing cash values in life insurance and annuity products always has consequences to the sustainability of the long-term benefit. For that reason, many clients will hesitate to access the values in their policy, if they’re even aware that they can.

For some clients, the benefits are not the reason why they bought the product. Instead, they were focused on the cash value growth within the policy. But even those folks have a reason not to access their cash – taxes. The whole point of using life insurance and annuity products purely for accumulation is long-term performance and tax savings. Surrendering the policy would eliminate those benefits.

The only way to preserve those benefits is to do a 1035 exchange, which must go into another insurance policy. There is unquestionably a risk of increased 1035 exchange behavior, a potential flight from danger to safety if one insurer is in a worse position than another. That could potentially force asset liquidations at the first company while sending flows elsewhere. It’s the equivalent of moving money out of SVB and into Bank of America, which has seen an influx of deposits over the past few days. But the problem with this strategy is that 1035 exchanges incur charges associated with switching to a new policy – and, of course, potentially charges for going out of the old policy. The costs may be worth far more than any perceived benefits. Switching from one bank to another is financially frictionless. Switching from one insurance policy to another is usually not.

Surrender Charges & Market Value Adjustments

That’s intentional. Virtually all life insurance and annuity contracts begin with liquid values that are less than what the client has contributed thanks to explicit (or implicit) Surrender Charges. In life insurance products, Surrender Charges are typically seen as a necessary evil in order to provide for heaped commissions. But Surrender Charges have a secondary effect of extending the expected duration of the product by constraining liquidity. In other words, higher and longer Surrender Charges mean that the client is likely to stick around until the charges burn off, which means that the carrier can invest in longer duration assets. It goes without saying that bank deposits don’t have surrender charges. There are no natural deterrents to withdrawing from a liquid bank savings account.

The effect of Surrender Charges is much more pronounced in annuities, where actuaries assume very low lapse rates until the end of the Surrender Charge period, allowing for a very clean match between the maturity of the investments and the expected duration of the contract. When a company sells an FIA with a 10 year Surrender Charge period, they can effectively buy a 10 year fixed income asset to match the product liability. On top of that, annuity products almost universally have Market Value Adjustment features that reduce the surrender value of the contract if interest rates increase, which provides a natural (if sometimes muted) hedge against capital losses. Asset/liability matching is pretty clean when it comes to accumulation-oriented annuities.

For income annuities and other long-duration assets like Hybrid LTC and Guaranteed UL, the effect is even more pronounced. Those products rarely get surrendered and, if they do, the client is usually leaving quite a bit of value on the table. For these reasons, I would argue that it makes sense for companies that predominately sell annuities, hybrid LTC and products with long-tail guarantees to have higher ratios of unrealized capital losses to surplus. The unrealized losses simply aren’t as dangerous because of the features and characteristics of the product. This is also arguably a reason why the unrealized capital losses sitting in captive insurers, which generally  house products with long-tail guarantees and very little liquidity, are less problematic than the ones held in assets that support more traditional and liquid products.

Take a look at the same chart as above, but this time color coded for annuity and hybrid LTC orientation. Yellow is heavy in either of those two lines, blue is heavy life concentration. That said, many of the companies coded as blue also have large blocks of Group annuities and other assets that have similar characteristics in terms of capital as annuities. As you’ll see, the majority of the companies that have the lowest ratios are heavy (if not exclusively) annuity sellers.

If annuities, hybrid LTC and products with long-tail indices have natural, built-in liquidity protections, what about mature life insurance policies with readily accessible cash value? These sorts of policies pose a challenge for insurers because they know simultaneously that they have the potential for liquidity demands and also that in order to keep policyholders around, they have to deliver strong returns courtesy of longer-duration investments. It’s the same sort of paradox that faces banks. But here, life insurers have another advantage over banks – awareness.

Lack of Liquidity Awareness

Unlike with bank deposits, where everyone who puts money on deposit knows exactly where there money is and how to get instantly, life insurance cash values are somewhat of an enigma for most policyholders. They’ve probably been told that they can borrow or withdraw from their policy, but also that if they do so, there may be unintended consequences. They’re not looking at their policies regularly to see what kind of cash value they have and they may also not be entirely clear about the actual process to get money out of it. Most clients don’t think of money in a life insurance policy like they think about money in a bank. It’s a different category of asset, even though it is theoretically nearly as liquid as money on deposit at a bank.

As a result, life insurance surrenders, withdrawals and loans are surprisingly low given the near-instant liquidity available within these policies. Take a look at the Big 4 Mutuals, the companies most likely to see cash value being accessed for liquidity:

The reality is that people who have policy cash value are not using it for liquidity, at least not in the same way that people use savings accounts. But let’s say they did. Let’s say that people suddenly woke up to the fact that they could drain their life insurance policies of cash value, either through withdrawals or loans, and reinvest the proceeds at higher rates. What would happen to life insurers then? Well, as it turns out, there are more lines of defense.

Deep Cash Holdings & Recurring Premiums

In aggregate, the 35 life insurers hold nearly $73 billion in cash and cash equivalents. Recognizing that life insurers would burn through cash before they get to surplus in the event of a run means that the net unrealized capital losses above surplus drops to 64%, not far off from the average for US banks. But as with everything else, the cash isn’t evenly distributed. Some hold a lot, some have very little. Below is the same chart used above, but this time with light blue signifying the ratio of unrealized capital losses to statutory surplus plus cash as a percentage of fixed income assets.

Overlaying cash onto capital, the ratios become stronger across the board. Without cash, only 14 insurers have surplus in excess of unrealized capital losses. With cash, the number jumps to 24. The effect is much more pronounced for some companies than others. Riversource, for example, holds a huge amount of cash. National Life does not. Neither does Securian. Cash holdings are all over the map.

Life insurers have another source of what might be termed recurring cash flow – premiums. As long as people are still paying premiums, then those premiums in cash can be repurposed to also pay claims and surrenders in cash. The near-term effect is that the company doesn’t have to liquidate assets and take losses in order to satisfy liquidity demands. The long-term effect, though, is that the company also isn’t getting the chance to reinvest those premiums in the new higher yielding assets. As a result, profitability will suffer and so will the rates for policies that remain.

But for the purposes of this article and the very specific situation of a potential run on life insurers, it doesn’t just take a run to cause a problem – there has to be a net run. Liquidity requests have to exceed recurring premium flow. Given that most companies have premiums outpacing typical surrenders several times over, that’s a pretty high bar to clear. But what if they did. What then? Well, there’s a final backstop that might be the most potent of all. And almost no one knows that it’s there.

The 6 Month Provision

Virtually every life insurance and annuity contract contains an NAIC-required provision that the life insurer can take up to 6 months to pay out. This provision was specifically designed to stop a bank run situation with life insurers. It is immensely effective. In the world of liquidity, 6 months is an eternity. It gives the life insurer plenty of time to selectively sell assets to mitigate damage. It lets policyholders cool off and relax because the insurer, in fact, isn’t going bankrupt.

The counter argument to the effectiveness of the 6 month measure is that if an insurer actually did this, it would freak everyone out and there would be even more of a panic and more demands for liquidity. Possibly, but I don’t think so. The life insurer wouldn’t announce that it is going to take 6 months to pay the benefits. It would just have, you know, processing delays that could take up to 6 months. They’d do it in a very orderly fashion. You’d hardly know it was happening. And, in the process, they’d slow everything down and mitigate the damage. By the time 6 months is over, there might not even be any damage to mitigate.

The Way Forward

For all of these reasons – and probably a few more that I didn’t think to include – there is no real possibility of a run on a life insurer. A situation like SVB playing out in the insurance space is virtually impossible to imagine. That’s intentional, not happenstance. Our products are designed primarily for their benefits, not for their liquidity. The life insurance industry is the beneficiary of a regulatory and capital regime that is specifically designed to prevent runs on life insurers. As a result, although life insurers are worse off than banks in terms of pure unrealized capital losses relative to surplus, the strength in the system is not in capital – it’s in restrictions on liquidity. The tumor is benign.

But even a benign tumor can cause challenges. Some life insurers routinely and opportunistically trade fixed income instruments for profit and loss. The mechanism for amortizing these realized capital gains and losses into statutory surplus is somewhat cryptically called the Interest Maintenance Reserve (IMR). As it currently stands, IMR can only be positive – in other words, it only applies to capital gains. This appears to be something of a historical oversight and Day 2 item that never got covered. A lot of carriers came into 2022 with large IMR balances from previous capital gains, but those balances have been obliterated as rates increased and carriers booked capital losses on trades.

The numbers for some insurers are staggering. Below at the IMR as a percentage of total assets for the 35 life insurers at the end of 2021 versus the end of 2022:

The impact of falling IMR balances is that life insurers have less cushion – or no cushion – before realized capital losses on bonds from trading flow directly to surplus rather than the being amortized over time, which is the whole intent of IMR to begin with. This is a real problem and does not match the intent of IMR. The NAIC is exploring the idea of allowing life insurers to carry a negative IMR balance that will allow them to amortize capital losses into surplus over time, but some insurers aren’t waiting around.

Northwestern Mutual, which burned through its $3.3 billion positive IMR in just one year, is currently holding a negative $212M IMR balance thanks to a permitted practice from Wisconsin. American Equity (Iowa) and New York Life & Annuity (NY) also have small negative IMR balances, although they don’t specifically reference permitted practices in their statements. The NAIC needs to act on a national scale before things get out of hand. IMR is a crucial instrument for financial flexibility for insurers to be able to opportunistically trade their book. Without it, life insurers may be stuck holding bonds they don’t want and can’t trade without taking an immediate hit to surplus.

That’s on the asset side. In terms of liabilities, some life insurers have dipped their toes into asset classes that potentially have concentrated liquidity risk as well such as GICs, pension buyouts, funding agreements and other products that don’t fall neatly into the traditional life and annuity buckets. Those sorts of things have taken down life insurers in the past. One example that comes to mind is General American, which had to be rescued by MetLife after a liquidity provision in an acquired GIC block was triggered as a result of a downgrade. It’s really hard to see some of these liabilities clearly simply by looking at life insurer statutory statements.

But, again, these things are relatively benign. IMR is a challenge that could put insurers in a tricky spot, but it’s not fatal. Some insurers probably have liquidity exposure laying around in COLI/BOLI and other institutional products, but few have blocks big enough to take down the enterprise. The fact is and remains that life insurers are built to be more resilient than banks precisely because they can control their liquidity exposure in ways that banks can’t.

This is our superpower. It’s the reason why life insurers persist through any economic storm. Since 2001, there have been 563 bank failures in the United States according to the FDIC. In that same period of time, according to the National Organization of Life & Health Insurance Guarantee Associations (NOLHGA), the organization for the state guarantee associations that backstop life insurers, there have been just 7 failures of companies that wrote primarily life insurance and annuities. There’s a reason for that. As long as life insurers invest in quality assets, write smart liabilities and hold appropriate capital, they are built to last. The business model is strong. And it’s times like these when we see its strength come through.

3/17/23 Update – I almost exclusively watch stat earnings for life insurers, but someone gave me a heads up that life insurers have the option to mark fixed income to market under GAAP accounting and the vast majority of their bonds are held that way. I pulled YE GAAP financials for several publicly traded insurers and the damage is clear. Prudential, for example, saw a $78 billion drop in the value of its fixed income portfolio and its book shareholder equity drop by $46 billion from YE 21 to YE 22. Lincoln’s fixed income portfolio took a $20B hit and equity dropped by $16B. For Brighthouse, the fixed income loss was $12B and equity was down by $10B. The carnage is real.

One final note – this article assumes that life insurers are not selling the bulk of their policies into systems or strategies that are contingent on unrestricted liquidity being accessed on a regular and recurring basis. These strategies have many names, but they’re generally clumped under the Infinite Banking umbrella, although IBC practitioners will tell you that a lot of agents selling under IBC aren’t following the principles. Regardless, these sorts of strategies can create liquidity problems at life insurers because of constant portfolio turnover, which is probably one reason why MassMutual recently sent out a memo condemning the strategy and forbidding its agents from selling that way. Agents who sell this way may find themselves met with more resistance at carriers for regular liquidity requests, lower returns on products and potentially termination. Carriers hold the cards. If this gets out of hand, they can exercise their right to wait to process withdrawal and loan requests for 6 months. Life insurers are not banks. Your life insurance policy is not a savings account. It’s a life insurance policy. And right now, that’s a really good thing.

*Banks can either hold fixed income assets as Available-For-Sale (AFS) portfolios or Hold-to-Maturity (HTM) portfolios. AFS portfolios are marked to market, which means that changes in valuation flow through to capital. HTM portfolios are held at book value. From what I understand (and I’m certainly not an expert), there’s a bit of fluidity and flexibility between the categories at non-SIFI banks, which allows the banks to essentially mark or warehouse capital gains and losses, depending on whether the asset is classified as AFS or HTM. Part of the issue with SVB is that they held a lot of their portfolio as HTM, which shielded their Tier 1 capital from the unrecognized capital losses that had been building up in their portfolio. Some of those unrealized losses were converted to realized losses when depositors started to ask for their money.

**The Asset Valuation Reserve is designed to reserve for credit-related losses within the portfolio. If this article was evaluating the overall capitalization of the industry against any threat, I think it’s appropriate to include AVR in statutory surplus. But because this article is exclusively focused on the effects of unrealized capital losses as a result of rising interest rates, I did not include AVR in the surplus figures.

***Pruco is the writing entity for the vast majority of Prudential’s life insurance and annuity business. Separated from Prudential and combined with its sister New York company, Pruco Life of New Jersey, it would be one of the largest life insurers in the United States.