Life insurance and your money, with Zac Townsend

This week, I'm joined by my buddy Zac Townsend. He is currently building an insurance company from the ground up. We talked extensively about the life insurance industry, giving me a chance to give my standard advice about it (which I broke into its own section below), and also letting us examine the tax, regulatory, and even international relations angles on what some people consider a bit of boring financial plumbing.

Sponsor: Mercury

This episode is brought to you by Mercury, the fintech trusted by 200K+ companies — from first milestones to running complex systems. Mercury offers banking that truly understands startups and scales with them. Start today at Mercury.com 

Mercury is a financial technology company, not a bank. Banking services provided by Choice Financial Group, Column N.A., and Evolve Bank & Trust; Members FDIC.

My annual exhortation about insurance coverage

Although it is covered in the conversation below, I’ll just frontload the Official Patio11 Buyer’s Guide to Long-Term Insurance, which is optimized for tech workers and business owners. (Insurance isn't rocket science, but there are parts of this recommendation I wouldn't make to literally everyone. If you are not in one of those groups, you can either talk to an insurance salesman or read general personal financial advice literature.)

Are you rich? If so, congratulations. You care about this far less and/or can ask the people who handle your money to do it for you. If you do not have liquid wealth in excess of 25X your family’s expenses for a year plus any extraordinary expenses (college, stock exercise expense, etc), you should get the following two policies.

You must get these policies in your own name and not connected to your employer. Many tech employers offer life and/or disability insurance as a benefit. It is fine to buy this (or receive it for free) but this should not be everything you buy. You will not reliably continue working to receive that benefit in all scenarios which result in you needing insurance coverage. For example, you might experience health deterioration for a few years, downstep from your current career to deal with it, and then lose your current benefits suite. At that point, it will be too late to get benefits in your own name, because you will not be able to pass medical underwriting. Sell this risk.

Your shopping list:

Term-life: You pay an insurance company $X0 a month in premiums (fees). They pay you $X million if you die in next Y years. That’s it, that’s the whole ballgame.

You want at least $2 million in coverage. Don’t know what term to pick? 20 years. There, done. This is an absolutely commodity product. It doesn’t particularly matter which insurer you use; whomever quotes the best rate is best. You should expect (given that you are young and healthy) that optimizing for a quote is probably not worth more than a few minutes of your time.

I got mine in the U.S. through Ethos, which doesn’t require a medical exam for lower value policies (around +/- $1 million). But again, get $2 million or more. 

You will not notice the month-to-month difference in spending power (an extra hamburger, whee, if you’re reading this you can buy as many burgers as you want), but (particularly post-pandemic) your heirs will really notice the difference between $1 million and $2.5 million.

Long-term disability: Pays a portion of your current income, on an ongoing basis, if and only if you are unable to work. You want “own-occupation” private long-term disability; that pays you if you can’t do your actual job. (The default type of long-term disability coverage will not pay if you are still capable of doing any job in the economy. There are events which are disabling for professionals which would still allow you to arguably work at McDonalds; you want to be paid in event of one of those. (Consider e.g. severe carpal tunnel syndrome.)

Probably pick the longest “elimination period” they offer. This is the time between a disabling event and the time benefits kick in, and the longer that time is, the less you pay in premiums. You will cover the period between losing your job and benefits kicking in out of savings, if you need to.

Long-term disability insurance will typically require a medical exam. Get it, set it, forget it. LTD is more expensive than term life insurance, and I consider it more optional than term life insurance. Ballpark ~2% of your income per year to insure you for ~60% of your income.

If you have a relatively uncomplicated health and life situation, your friendly local insurance professional can get this for you easily. If you have a complicated life situation but are still healthy, you might benefit from Petersen International Underwriters. They will basically not talk to you directly; you need a broker to sell you the policy they issue. I used Insubuy. They are particularly solicitous of business owners, highly compensated professionals, and people with complicated international living situations. (One downside: they have a total exclusion for basically any form of mental-related disability.)

If you are not in very good health, apologies: private LTD will likely not be available to you at any price in the U.S. And thus, for the rest of you: get it while you are still capable of doing so. The list of relatively minor mishaps that will make you essentially permanently uninsurable is very long.

Timestamps

(00:00) Intro
(00:23) Overview of Meanwhile Insurance
(00:56) How Bitcoin is being used in the insurance industry
(02:12) Understanding different types of insurance
(06:13) Term life insurance explained
(07:26) Life insurance tips for tech professionals
(12:47) Permanent life insurance and annuities
(18:57) Sponsor: Mercury
(20:11) Insurance regulatory insights
(25:46) Principal-agent problems in insurance
(37:09) Tax considerations in life insurance
(42:16) Leveraging life insurance for estate planning
(44:55) Premium financing explained
(47:39) Wealth management and market segmentation
(52:09) Regulatory challenges and mispricing in insurance
(01:03:32) Reinsurance and the role of bermuda
(01:10:27) Private equity's interest in insurance
(01:14:24) Building a crypto life insurance company
(01:20:02) Wrap

Transcript

Patrick: Hi everybody. My name is Patrick McKenzie, better known as Patio11 on the Internet, and I'm here with my buddy Zac Townsend, who is currently the CEO of Meanwhile. So what does Meanwhile do?

Zac: Meanwhile is a life insurance company. We are a fully regulated licensed life insurance company in Bermuda. Bermuda is somewhat the insurance capital of the world. We can talk about why that is. 

We are full-fledged. We have an independent board, a chief risk officer, and chief compliance officer. We've built a policy admin system and do all the things of a life insurance company… from scratch.

How Bitcoin is being used in the insurance industry

What's unique about the company is that we have a go-to-market around Bitcoin. So right now, this Bermuda Insurance Company is entirely in Bitcoin. It takes premium in Bitcoin, pays claims in Bitcoin, does our audited financials, our solvency calculations, all our regulatory filings in Bitcoin.

You and I might have a whole discussion about whether Bitcoin is gonna be worth anything, but I think basically what we do is we are using that both as a way to provide a critical financial service to everyone who does think that Bitcoin will be a long-term store of value. And we also got to build a fully functioning life insurance company with policy holders, which is a pretty hard thing to do in dollars. So it was a really fun way to learn and build in this ancient, boring industry.

Patrick: So the ancient boring industry is a sub-collection of industries. And I think both of us are saying this in a somewhat jocular, joking fashion. I think the life insurance industry in particular has created one of the best things in the history of the financial industry, which is that term life insurance is a policy, it's an upgrade in civilization.

[Patrick notes: Term life insurance is designed for a young, healthy person to buy a fixed-price fixed-term (10, 20, or 30 years are common) policy which pays out if and only if they die. It is extremely inexpensive when used in this fashion. Very few healthy young people pass away, but some do, and historically—and in the present day, for people who do not do 1-2 hours of admin every 20 years—this results in their heirs needing support from either charity or government.

Term life moves this from non-deterministic tapping of others’ largesse in the future to an extremely deterministic, trivially budgetable expense.]

Understanding different types of insurance

But just to scope our discussion somewhere, we were talking before the call on how there is a vast difference between life insurers, property casualty insurers, and health insurers. And for the benefit of people who haven't been in the industry, can you just sketch out why they are so different from each other? And then we'll focus most of our time on life insurance.

Zac: Well, I think actually the first thing that people probably know about or think about is property and casualty. So, P&C is things like auto insurance, homeowner's insurance, slip and fall insurance, professional liability insurance, director and officer insurance - basically all those liabilities that get created in the world related to property or casualty.

The distinction that is made in Bermuda, I actually find really useful, which is, they don't call it property and casualty and life, they call it “general business” and long-term. And that distinction is that most P&C contracts are for one year. So you buy auto insurance, you're buying it for a one year term, and then you either renew or you shop around, redo whatever. And then life insurance is long-term.

So there's life insurance, there's annuities, there's long-term disability insurance, there's all this different stuff. That's the distinction. And then actually when you get licensed, whether it's in a state like Illinois or California, or in a country like Bermuda, they tend to be entirely different licenses and actually the actuarial math and the actuarial training is different.

So, even if you're a member of the Society of Actuaries, which is the US Professional Association around actuaries, you can either be an expert in and take a bunch of tests around P&C or you can be an expert and take a bunch of tests around life insurance. And then health is completely off on its own.

Patrick: Health, I presume or must be more of a sort of embedded payment system than it is a true insurance offering. Right?

Zac: I think that's right. I mean, you know, in theory where all these things have in common is that we all, each of us or every business, we have this little thing we're walking around with, which is like a negative number on our personal balance sheets, and we just don't think about it. But that's like the liability that some risk is going to happen.

And when you think of it that way, the expected value of that liability might be pretty small, but the absolute value of that unlikely event might be very high. Right? So, the reason you bought insurance from a personal perspective is to cover off this little liability you have on your personal balance sheet.

And then how that works in the aggregate—say for life insurance—is we are all going to die, but we don't know when. And it could be tomorrow. There's a very small probability that it's gonna be tomorrow. But it could be tomorrow or it could be next month. 

Life insurance smooths variability between individuals through aggregation, basically through the law of large numbers. You get to have a pretty strong statistical sense of how many people in a population are gonna die. So, although I don't know when I'm gonna die, if you have thousands of similarly situated people like me, you could say, "Okay, well I don't know, two of these thousand people are likely to die, or 20 of 10,000." And then you can sort of pool that risk together.

The same thing is true of homeowners insurance. The same thing is true of auto. Risk pooling against catastrophic but low probability events is important civilizational infrastructure.

[Patrick notes: I only rarely make edits to the flow of the conversation, but Zac and I didn’t define term vs whole life until late in the conversation on the day of recording, and so I lifted that part of the conversation here for the benefit of listeners who might not have an insurance background.]

Term life insurance explained

Zac: The differences between term life and whole life are that term life lasts a certain term like 10 years or 20 years or 30 years. And particularly if you're young, you're very unlikely to die in the term of 10 years or 20 years or 30 years. So the ratio between the amount of money you might pay and the payout you might get is very high.

So, I forget offhand, I think I pay maybe like a thousand dollars a year and have two or $3 million of coverage on my life because I'm getting less young every day, but I'm relatively young, relatively healthy, and I'm very unlikely to die in the next 20 years.

Patrick: And just as a finger to the wind for people - if you're in your 20s or 30s, you can get a million dollars in coverage in term life for low tens of dollars a month in the United States.

Zac: That's right.

Patrick: It's a wonderful product. You should have more of it than you think you need.

Zac: Well, you should have more of it than you think you need, although you can't have an unlimited amount. Life insurance companies don't like it when you suspiciously want to buy a huge amount of life insurance on your life, vastly disproportionate to your wealth or your potential income.

Life Insurance tips for tech professionals

Patrick: Can I give my quick user guide to life insurance for professionals in the tech industry?

One, you definitely wanna have term life, regardless of other financial decisions you're making, unless you're so rich that no possible expiration timeline for yourself would matter to your family or causes that you care about.

Two, it's very possible that your employer offers insurance, either life insurance or professional disability insurance through work. That might or might not be a thing that you wanna buy, but definitely have policies outside of your employer. The reason being that you can lose your employment as your health declines in a variety of scenarios, but where you don't simply get to buy a car.

And then also the pricing tends to be better outside. And so, to avoid both of those things, get a policy in your own name which the insurance company will have a responsibility to maintain, rather than the company you might no longer be working for in six years, when you're thinking of a 20 or 30 year policy term.

How long do you have it for in the tech industry? If you're thinking about this question and can't come up with a number, 20 years is a fine number to think of, but you probably should optimize less than that and just make sure you have something versus nothing.

And similarly, how much should it be for? General rule of thumb is like 25 times expenses for what you think your family will need. And every time that you have a major increase in your outgoings over the course of the next couple of decades of your life, when you have a new addition to the family or you otherwise become liable for something, you can get a new policy. Or if you know what your plans are for the next couple of years, just assume that your expenses are going to increase and then map it out from there.

Then my final micro tip, which is specific to the tech industry: due to that factor about underwriters really not wanting to give an excessive amount of life insurance coverage to prevent people from making poor decisions by incentivizing those poor decisions, you might need to sort of argue through your agent to an underwriter that you should be allowed more insurance coverage than their standard rule would allow you to get.

Two magic phrases if you are in the industry for this are: one, "I work in the tech industry and might have stock options." This exercise will help get an underwriter to saying yes to approving you for more coverage than your current income would necessarily support. And the other one is "I have a closely held business," which for similar reasons gets them comfortable with: "Okay, the business might have outgoings or need to pay out a partner or similar." And then, "Alright, I can see a way to approving this person for a few more millions of dollars of coverage than I would ordinarily."

So that is my quick user's guide. None of the following was insurance advice, which of course I'm legally prohibited from giving because the law says that only people who have clicked through the requisite number of screens in their annual training are allowed to sell insurance. [Patrick notes: It’s a Compliance box ticking exercise but, as Parker Conrad can tell you, if you fail to treat the box ticking like it is important your friendly neighborhood insurance regulator will be very cross. (That situation was also something of a boardroom coup using failures-at-box-ticking as an excuse.)]

Zac: The only micro addition I'll have is, actually, the first time I bought term life was around the time that my wife and I got a mortgage. And our theory around that is we have this new 30-year liability that's dependent on our incomes. And if one of us died, that would be quite catastrophic for our ability to pay those mortgage payments. And conveniently, a 30-year term life aligns perfectly with our 30-year fixed-income mortgage.

[Patrick notes: I didn’t want to interrupt Zac, but for this particular use case, you can get “mortgage protection insurance.” It can be cheaper than having a 30 year term policy. Why? Because the benefit decreases over time as you pay down the principal of your mortgage. The most expensive part of that 30 year window actuarially is the last few years. Since most of your mortgage will be repaid by that time, you’re not paying premiums on that part of the risk anymore.

But, on flip side: it’s very unlikely statistically that you still live in your home in 30 years, and if you sell it in 8-10, MPI doesn’t really help you that much versus just getting a term life plan.]

Zac: But in general, it's quite inexpensive if you're young. It is providing a real service. One of the things that's tough, I think, about the mental model for term life is the most likely scenario is that you don't die, that you live, and that you just never get a payout. So people can feel like, "Well, what did I pay for?"

And we're just not, I think, well equipped to think about the protection. You're protecting yourself against this small probability, very high terribleness thing that could happen. And it is definitely worth tens of dollars a year, hundreds of dollars a year.

Patrick: The reason I give my annual exhortation to people who follow me on Twitter to make sure that they have term life in force is I've been around the industry for 20 years now and am a social individual, although a bit of an introvert. And so I've had conversations with a lot of people who work in tech and ambiently know a lot of people, and then actuarial math: 20 years times knowing a lot of people means I know some people who passed away with young families and in some cases ended up in a "pass the hat" situation.  

"Pass the hat" for people who don't understand the idiom: Before this risk was moved into the financial industry, it was often dealt with in a semi-socialized fashion via churches or something similar, where if someone passed away, a community leader would pass a hat around the room and everyone would put in money to support the family of the person that had passed away. And that's where the "pass the hat" idiom comes from. These days, it's probably GoFundMe or something.

But regardless of that, you probably don't want to be in a circumstance where, with a very low probability, your family will in the future depend on the kindness of strangers and/or on tapping your extended social network to do a GoFundMe. And so, life insurance can formalize that and make it much more predictable for you. And in the vast majority of futures, you will lose your bet with the insurance company that you die early and be very happy because you will still be living and your family will still see you at the dinner table every day.

Permanent life insurance and annuities

Zac: Exactly. So, rewind - there's term life which everyone should buy. And basically the big tax benefit of the term life is there's a tax-free death benefit, or at least income tax-free generally.

And then the reason that people buy permanent life insurance policies like whole life - and they're called permanent life insurance because they're permanent. And since you will eventually die, there will eventually be a payout. So these are much more like intergenerational planning products.

So whole life products tend to just be very straight down the middle. It lasts your whole life. You put in a certain amount of money, maybe you're paying for 10 years, maybe you're paying forever. And then whenever you die, your beneficiaries get a certain amount of value.

Then there's more complicated variants on whole life, like universal life or indexed universal life, which are indexed to the value of the S&P 500 or something like that. And the two big reasons that people do that are - well, like, if you're running a relatively small policy, these tend to not be very good. You can probably just get better returns in the S&P 500 directly.

There was this whole movement in the sixties and seventies, which is "Don't buy permanent life insurance, buy term and invest the difference in the premiums." As you become wealthier, as you’ve maximized tax-advantaged investments—you've maxed out your 401k and you have done this tax vehicle and that tax vehicle—the reason that people end up buying permanent life insurance policies is the contents of permanent life insurance policies grow tax-free if they're structured properly.

There's a value inside the policy, and that value, sometimes called the cash value or the surrender value, is growing and accumulating tax free. You can withdraw from the policies in a tax-managed way if you want, or the big thing you do is you can take policy loans against your policy and those are tax free.

So what people will do, particularly if you have investment strategies—this is not true for most people in tech, but you have investment strategies where there's a lot of intermediate income tax. So you're doing high frequency trading, or you're the limited partner in a high frequency trading firm, or a crypto fund. At the end of the year, when you get your K-1, you've lost half your returns to income tax because of short-term capital gains tax on most of your income.

So what people will do is take that LP commitment and wrap it inside a life insurance policy, and then it's essentially accumulating tax free. And then when you need value from the policy - well, you could die and then your kids or your loved ones get the policy, or you can take a tax-free loan against the policy.

And then there's a lot of structuring that people do where maybe you front-load premiums into a policy and then you have 10 years where you're not paying a policy and then you can pull money out. But to go back to the very beginning, these tax rules were written in the case of life insurance because the government wanted people to be able to plan so that their kids wouldn't starve if they died.

And then what gets built around that is products that suit different levels of wealth all the way up to the wealthiest people in the world so that you can protect assets or investments inside of a life insurance policy.

And then similarly with annuities—what happens is almost no annuities in America actually annuitize. That means that most people don't opt to turn the value of their annuity into a stream of cash payments between now and when they die. Instead, like you Patrick or me, in our forties or something, you've maxed out your 401k and you've done your backdoor IRA or whatever.

And then what you can do is you can buy an annuity, and then that annuity is like a 401k, and then the contents of the annuity are compounding tax free over time. Maybe you have a wide range of different investment strategies that you can put in them. And then when you turn 65 or 75 or whatever, you don't annuitize them. Instead, you begin withdrawing them, which you can do after 65 in a tax-advantaged way. So it's like another investment account.

The challenge with them is that to go to the principal-agent problem, your agent wants to sell you something needlessly complicated that might be high margin for them. And what you want is something that is straight down the middle and not particularly complicated. But I would say as people's tax and estate planning gets more complicated, life insurance and annuities tends to be a part of that tax and estate planning.

[Patrick notes: Returning the conversation to the beginning of the recording here, now that we’ve discussed whole vs term.]

Patrick: My understanding is that there are some insurance lines where it is less risk pooling and more just enforced savings. For example, dental insurance and I believe auto insurance generally believe that they're going to pay the premiums back to the customer that paid the premiums rather than distributing some to the unlucky winner of the dental lottery.

Zac: I think in the case of auto, your loss ratio, that is the ratio of payouts—it's such a competitive market, it does approximate to a hundred percent. [Patrick notes: I’d ballpark auto at about 80. GEICO’s is about 80%.] And then you make money through the Warren Buffet style - the policy holders have paid you to hold their money for a certain amount of time, and that time itself is valuable and then you can go make investments with it. And then at enough scale, that float is obviously significant.

Insurance regulatory insights

Patrick: Do you mind if I geek out for a moment on numbers that people would know if they read more insurance regulatory filings?

So the loss ratio is the payouts in a year divided by premiums collected for the year - premiums being what you as the person buying the insurance pay the insurance company typically periodically.

And the combined ratio considers losses plus claims expenses. So, if they need to send an adjuster out to your house to figure out how much the hurricane repair actually costs, or if there is a back and forth with regards to selection of company to haul away debris, then the people doing that back and forth in the insurance company need to get paid. And so, that goes into the combined ratio.

One of the things that insurance regulators do as a primary activity is to manage those ratios. They don't want insurance to be too profitable. And so, particularly in certain lines, as the loss ratio or combined ratio gets farther away from a hundred percent, the regulator gets more concerned about whether the insurance company is pricing incorrectly.

[Patrick notes: And then we have title insurance, where the loss ratio is about 5%. Why so low? See my issue of Bits about Money on it.]

Patrick continues: And so they would like it to be enough of a margin such that the insurance company can absorb some variation, risk and get a bit of a profit, but not so much that they are not paying out claims. I think one of the things that people undervalue with insurance companies is that most believe that the job of an insurance company is to say no to all the claims all the time. Insurance companies never pay out for anything - and that is just incorrect. They are extremely regulated, extremely deterministic with regards to whether they will pay out claims. They're often annoying to deal with depending on the particular company. But somebody files a quarterly report with the insurance regulator that will say down to the penny what the total number of claims paid out for the quarter were.

Zac: That is all absolutely correct. And then I will say that all of those ratios and the way of thinking about that are much more, is very much a P&C view.

So let me just completely agree with that. And then how are Progressive or AAA or whatever thinking about it, and how are the regulators thinking about them - is definitely like loss ratios and combined ratios, and that's the financial and regulatory performance.

When you get into life insurance - and this is an endlessly nerdy topic - you're actually operating much more like a bank. But you're somewhat better than a bank in that your liabilities are much longer in duration than your assets.

The problem with banking, which I think you've written about a lot, is that you have really short term liabilities in your deposits and then you go invest those in mortgages. So you have this duration transformation in banking. In life insurance companies, it's some of the opposite - even with term life, you're making a promise for 30 years, you're getting paid premiums.

In the case of a whole life contract, it might be 40, 50 years depending on how old the person is. So you have these liabilities that are lasting a really long time, but you've also gotten paid the capital now, the premiums now, which you then put in as capital. Then you invest in a bunch of assets. But other than like 30-year treasuries and mortgage-backed securities, most of those assets do not have duration equal to your liabilities.

So you end up with a bunch of ratios that sound a lot like banking ratios - like a liquidity ratio, a risk-based capital ratio, and a surplus ratio. And all of the machinery of the regulation of life insurance companies is basically you do a lot of work to estimate what is the present value of all of your liabilities, and then what is the riskiness of all of your assets, and then making sure that you have enough capital to account for potential inaccuracies in your models of the world.

Patrick: So the intellectual/historical movement to have regulation for insurance generally, which actually predates to my understanding most regulation for banking, was that back in the day, the notion of "Hey, you pay me money every month for 30 years and then I will give you money back" has a bit of a failure mode in that someone can credibly make that promise for 29 months and 11 years to a lot of people and then not have the money.

And so that has been historically the single largest locus of regulatory concern with respect to life insurers. Although there are other ways to get in troubles as a life insurer.

There was a terrible time in United States history where it was common practice among many industries to act in dishonorable fashions with regards to customers who were black. And there were a bunch of insurers that sold small dollar life insurance policies to black Americans using a door-to-door salesforce, and then conveniently forgot those policies existed. And when insurance regulators found this, they said, "You have to open your books up for 80 years, and we are going to go through them line by line and find everybody."

[Patrick notes: Insurance companies are also primary targets of state escheatment regimes, where the state will require an unclaimed but earned payout to go into their safekeeping indefinitely until the actual heirs come forward. The states will ordinarily not bring the full competence the U.S. government is capable of to proactively notifying heirs, for a variety of reasons.

In most of the nation, you can find your state’s office in charge of doing this and do a search for free, though you’ll have to dodge many companies trying to insert themselves into the flow. For example, in Illinois, it’s run by the State Treasurer.

Actually getting paid requires quite a bit of hoop jumping; I’ve done it as an exercise for small dollar amounts and would not recommend it unless you have at least a few hundred dollars on the line.]

Patrick continues: So, insurance regulators - not one of the most exciting parts of the regulatory state, but one of the higher competence parts generally, I think.

Principal-agent problems in insurance

Zac: What's fascinating about that, actually, if you wanna think about the mental models, is there's a lot of principal-agent problems in insurance generally. To take that in life insurance, they sort of run in both directions.

So when you walk in the door and you say, "I would like to be insured for 30 years," the insurer is sitting there saying, "Okay, this is potentially an issue of adverse selection. This person knows a lot more about their health than I do as the life insurer. So I need to go through a lot of diligence to make sure that this person is in fact healthy because I don't want to write a life insurance policy on someone who's gonna drop dead the next day" - because there's this principal-agent problem.

And then the reverse is also true, which is that the policy holder is trusting that the insurance company will be there, trusting that the investments will be prudently invested and managed over time. So in that way, the insurance company is also an agent of the policy holder and their funds and sort of this mortality coverage.

Patrick: Can we speak briefly about how the insurance companies are a bit of quirky market structure relative to many other fields in that there's sort of a supply chain that gets you an insurance contract with a particular carrier, and the person that you're talking to as the unsophisticated retail user might assume really is called an agent?

Zac: Yeah. Okay. Well, let's take a step back to your first question, which is like, what is the supply chain of most contracts? So, the supply chain of most contracts is that you, Patrick - maybe you go online, but even that person or little company is an agent. So you have an agent, which is a named individual somewhere. It's like a broker or an RIA, and that agent usually can have one of three relationships in the world.

One is they can be independent agents or work for an independent agency. So this might just be like a person or 10 people who have an agency license. The other is they can be part of a big agency or what's called a managing general agent (there's a lot of different terms for it), but that human being is either independent or is part of a larger independent agency. Or they're captive, which means they work only for one company. So New York Life, Northwestern Mutual, or MassMutual - they have a lot of captive agents who really only sell their product.

And then agents can actually sell life insurance products and have not a fiduciary duty, but are supposed to have some duties to the user, to the policyholder to do the right thing.

The next level is there are carriers. So carriers tend to be the names I mentioned, like New York Life or MassMutual. Meanwhile is a carrier. Those are on the life insurance side. On the homeowners side, you've got Allstate and Progressive and things like that. So those are the companies that really manufacture the product and usually, to simplify, are making the underwriting decisions, deciding whether you really are gonna get the product, whether you qualify, what your pricing is gonna be. And then they also handle claims, they do all the investments.

And then there is another level of the business that is reinsurance. So these are insurance companies for insurance companies. And many of those are based in Bermuda. But they have big names like Munich Re and Swiss Re and RGA.

And then what's interesting is - oh, there's just so many things - on the agent side, one of the things in life insurance that's curious is there's a principal-agent problem there. So what do agents wanna do? They wanna sell you products that have high margins because then they get big commissions, and they get big commissions on products that have high margins.

And high margin products tend to be super complicated. Because the way you have high margins is - if Goldman Sachs calls you and is offering you a collar option strategy, you should be really suspicious. You know it's gotta be too good to be true. And that's basically what's true with a lot of embedded life or insurance/life insurance products - they are needlessly complicated. It's like, "You get three times the returns of the S&P 500, but it's floored, and blah, blah, blah." But that's just a way of complicated financial products packaged into annuities or life insurance.

So your agent doesn't necessarily have your best interests at heart, although frequently they do. And then the reverse is also true, which is the life insurance company really wants to sell life insurance products that they're manufacturing. But they don't have a right to sell those products out in the world - the agents do. So even your captive agency, you're beholden to the salesforce that you don't really control. And that can be saying whatever - who knows what they're saying about your product. So there's also a principal-agent problem there.

Patrick: There's also an angle of: where are you getting the truth from the agents/customers? That matters quite a bit. Insurance fraud is big business. And I don't think there's probably any industry anywhere that thinks that sales - that people make exactly the right decision every single time. If someone mis-sells, that typically ends up with a business purchasing software they didn't need and perhaps regretting that decision.

But if they don't tick the right boxes on helping you get through your questionnaire, then the insurance company could be out an awful lot of money in a few years. And a funny little nuance of life insurance contracts in particular, at least in the United States by regulation in most states, there is what's called the contestability period. This is a useful factoid to know for people who have life insurance policies.

Your life insurance policy has a tiny bit of risk embedded in it for the first two years. If you pass away for any reason in the first two years after you get a life insurance policy, your insurance company has the right, but not the obligation to say, "Oh, wait. We think there might have been a fib here on the policy documents - they misrepresented their health status" or that "We were adversely selected because someone took improper risks or attempted to terminate their own life."

After two years is up by law/regulation, the presumption is against the insurance company - the period under which the policy was contestable is done. And now like if they're dead, you must pay out the claim, period.

And so, why do I mention that? Well, one, the insurance company needs to make sure that its people give them correct information to underwrite because they have that two-year period during which they can put claims back on the customer, and then after that period, the claim must be paid out. And two, as a user, don't churn life insurance policies that frequently - you definitely want your policies to be past the two-year mark.

Zac: Yeah, that's absolutely right. And then the caveat I'll say is that that is definitely true in the United States and then not true in many other jurisdictions in the world. Maybe the best advice I can give you on this particular point is just tell the truth on your life insurance application. That's your best bet if you want to ensure that the policy is going to be paid out by your life insurer.

Patrick: Yep. So, obviously no one passes away at a time convenient for them to pass away. Many people are likely to not pass away at a time convenient for them to pass away. But life insurance bundles two very different risks, right? Or maybe bundles is not quite the right word. It addresses two very different risks. One is you could have a catastrophic outcome and pass away much earlier than you expected to, and the other is the opposite. Can you talk a little bit about that trade-off?

Zac: So in general, life insurance companies tend to be in the business of selling life insurance and annuities. And if we took a step back and said, okay, what are the risks - one of the risks that I have in my life is that I'm gonna die before I expect to. That's really what life insurance was originally designed to address.

So, we all get together - 10,000 of us, some number of us are gonna die. We don't know who. Let's all pool our capital together, pay our premiums, and then 30 people will die and get payouts over the next 30 years. So that's life insurance. And then that's often from an insurer's perspective, thought of as mortality risk. So I'm taking on the risk that these people are going to die.

And then an annuity - although we can sort of talk about how both of these products end up being a lot about taxes - an annuity in principle is: I don't know how long I'm going to live, and I could live a lot longer than I expect. So I'm gonna retire, or I'm near retirement, or I'm just thinking about retirement now. The mortality tables say I'm gonna live to 87. What's gonna happen to my poor family if I live to a hundred?

So you buy an annuity, and then that annuity pays you every month until you die. And if you live to a hundred, that's the life insurance company's problem. And there's risk flowing here as well. So you take 10,000 people, they're all supposed to die at 87 or whatever. And then you can get some statistical regularity on that. And that is called longevity risk from the perspective of the life insurance company. But these are just the flip sides of the same thing. You often wanna match your mortality and your longevity risk.

I guess there are two big picture things here that are worth thinking about. One is society has decided, and governments decide that having orphans and widows starve is bad - which I think we can all agree on. So what happens is most societies and governments that have taxes choose to give tax incentives to people to buy life insurance or to buy annuities.

But then what happens, like every complex system, is that the honorable reason for giving tax benefits to these products means that then there's an entire industry, and much of life insurance and annuities are built around actually those tax benefits. So whether or not you're actually trying to protect yourself against the mortality risk or the longevity risk, that's one aspect. We can explore that.

And another really interesting thing that's happening in the world is why every private equity firm in the world is buying a life insurance company. There's a whole world of that.

Patrick: Would love to get into the private equity question later, maybe as one of the B-rolls towards the bottom of the episode. But would love to address the tax discussion now if we could.

Tax considerations in life insurance

It's my general understanding that the life insurance decisions as they're actually made, both by customers and by companies, are heavily influenced by tax considerations. And I know this because they heavily influenced my own purchase of life insurance in Japan, but the tax regime there is very different. 

Can you explain to people what's the reason that there is a tax consideration advantage in life insurance, sort of as a matter of social choice, and also sketch out what does US practice look like on this?

Zac: As we've been talking about, risk pooling really protects people from either dying before they're supposed to - what happens to your family if you die before you're supposed to - and then what happens to your family or your loved ones if you live longer than you're supposed to.

And you know how this is sometimes expressed is like what happens to orphans and widows in the world. So there's just historical precedence for basically every government in the world that has taxes to make a decision to incentivize and offer tax advantages to life insurance and to annuities.

And like any complex system involving taxes, the original reason I think has a real societal purpose. And then over time it grows and becomes a benefit unto itself. So life insurance policies are often used as long-term savings and tax vehicles. And the same thing with annuities - they have tax benefits, and these are different in different countries.

And then those tax benefits can be quite profound. And just to give you some fun history here, in the sixties and seventies people were doing really crazy things with their life insurance policies where they would put yachts and family farms into their life insurance policy.

And then back in a time when Congress still could think about hard problems on some level and pass sensible laws, as part of the Reagan tax reforms, they wrote a bunch of rules about what is and isn't life insurance and what those tax benefits are. But in principle, on the life insurance side, the benefits of policies, whether that's term life or whole life, tend to be income tax free for your beneficiaries.

And that is obviously good because you've just died and your wife gets a million dollar payout on a term life policy. And it's good that there's no income tax.

Patrick: I understand from a tax management perspective that this is probably worth it for some people to put cycles into at some level of wealth and/or income, and probably not worth the time it takes to think about it for people below that level of income. Indicatively, finger to the wind - whereabouts would that break point be?

Zac: You know, my initial reaction is probably single-digit millions of dollars in net worth. I think it's interesting—I now get asked a lot of questions about life insurance for obvious reasons. And I would say among my tech friends, it really splits like—it's almost like how are you thinking about your money?

If your money is this collection of stuff that you're doing a lot of optimizations around, and you're the type of person who loves to optimize, and despite having $10 million in net worth, you're still on The Points Guy and doing credit card points, then 100% you should do this.

If you were early-ish at Stripe and you have a $10 million net worth and you're mostly sitting on Stripe stock, treasuries, and doing angel investments for fun, and the thrill of optimization is not what you live for—maybe you don't wanna do it. [Patrick notes: Zac presumably used Stripe as an indicative example of a high-flying tech company here. I once worked there but no longer do, am still an advisor there, things said by other people in my own spaces do not represent the views of the company, etc etc.]

I wouldn't say it's one of the first five things you should do in your tax and estate planning, but it is like one of the first 10 things.

Patrick: So, inheritance tax hasn't come up yet. And unfortunately I'm much more familiar with it on the other side of the Pacific than I am here. Does inheritance tax interact with the tax consequences here? Is it more about income tax optimization?

Zac: I'm not your tax lawyer or your tax advisor, nor is my insurance company, and I'm not aside you estate planning services or life insurance directly. What I'll say is that the way it generally works is that it's income tax free, but it's not estate tax free to get a death benefit.

So then there are like three optimizations that people engage in. The first is if you are wealthy enough that you have a trust for your kids or your loved ones outside your taxable estate - so you have an irrevocable trust, you have a dynastic trust.

Leveraging life insurance for estate planning

Often people will set up that trust, which will buy a life insurance policy on your life because it's a way to turn a little bit amount of money in a trust into a larger amount of money through a life insurance policy. So that's one thing that people do. And then there's this whole class of things called an Irrevocable Life Insurance Trust, which again, if you're in the business of optimizing your LP commitments to D.E. Shaw or some high frequency trading funds, then you might be in the business of getting an ILIT.

The second thing is actually some people deliberately put life insurance policies to cover an estate burden. So imagine that you have a hundred million dollar farm or company and you're gonna die and you're gonna leave that to your wonderful kids. Or even you're the founder of a tech company and you're gonna leave a bunch of stock to your kids.

You can end up in this unfortunate scenario where your $100 million business that your kids are gonna inherit - there's a $24 million estate tax exemption. So then you owe estate tax on the other $76 million. And what are you gonna do with this illiquid asset? Are you gonna sell it to cover the estate tax burden? This is a real problem.

So what people will do is they'll buy whole life insurance equal to the expected estate tax burden. And then you die, your estate gets the business and it gets the life insurance payout. You use the life insurance payout to pay the estate tax, and then you do not sell your business. So this is a thing that people do.

And then the third thing is the problem with taking a life insurance policy and putting it in an irrevocable trust is then the policy loan - that is, borrowing value out of your policy - ends up not being something that you can access, but something that the trust has to handle.

So sometimes people just deliberately buy permanent life insurance policies because they're planning to borrow against the policies later in their life in this tax-advantaged way. And they'll just accept the estate tax burden, but if the policy loan is why you're doing it, then you need to leave it in the estate.

Premium financing explained

Patrick: I think there's another form of lending which is relevant to insurance, although I'm not positive I have this mechanic right. For the example where someone has a closely held business which will pass to the next generation triggering estate tax, and they don't want to break the business up to pay the estate tax, someone might borrow money in the present day to put the loan proceeds against the insurance premium.

Zac: Yes, premium financing.

Patrick: Premium financing. Can you say a few words on that?

Zac: Yeah. So, they borrow money from a bank or similar in the present day, use the proceeds of the loan to buy an insurance policy. And then essentially they get to reverse-annuitize themselves and pay down the loan for the rest of their life rather than getting paid money by an insurance company for the rest of their life. And when they pass away, that will leave a lower net present cash burden to their heirs to deal with the inheritance taxes and similar.

Or you just do this because let's imagine a life insurance policy - you're indexing your life insurance policy to the S&P 500. You expect those returns are just higher, even net the fees and expenses that the life insurance company is taking, than your cost of borrowing against your business.

So obviously if the life insurance company will pay you net 8% returns and you can borrow against your business at 4% returns, then it's just a good deal. And the reason that the bank will do that is you actually have two pieces of collateral. You have the business or whatever that asset is, and you have the cash value of the insurance policy. So you end up over-collateralized in a way that can lower borrowing costs.

I've thought about this. This is like a whole business idea that someone should run with, or I'm happy to do an angel investment, which is: if you take employees of late-stage companies, they're somewhat cash poor, but they're asset rich. And this is another product where you don't always wanna lend against that, but maybe you do a dual-collateralized loan where you say, "I'll lend you a hundred thousand dollars against your million dollars of Stripe stock and the value of this policy." And then you just have to make these regular cash payments or just take out increasing loans.

No one's done that yet. But part of the challenge with some of those things is it's like a feature but not a product. It's also complicated, but there's a whole new world of agents that specialize in high-net-worth individuals.

Wealth management and market segmentation

Patrick: I think one of the reasons is that the wealth management industry dices up the population in the world into a few buckets. There's a group of people without investible assets. And then there's the largest group of people with investible assets, which are called the mass affluent in much of the industry in the United States. And there's high net worth individuals.

And the financial industry has gotten very good at solving the problems of the mass affluent over the last couple of years in that the prices of index funds have asymptotically approached zero. They're much better products to use than they used to be. The tax advantages are getting built into them as well over the course of the last couple of decades.

And as a result, if you're someone who has the typical life path at a typical income level, you probably get that through your brokerage or similar these days, or through 401k at work, and you know, click the obvious buttons and you're done.

And thus the products that we're talking about here get sort of crowded out of the market for the mass affluent that, forget what the percentage number is, it'll be plus or minus like 50% of the population, I think, up to like the 95th percentile of wealth in America, and then maybe 90th these days - good news, we're getting richer all the time.

But then the people who have good problems to have, but still problems, greater than that - they get innovation done on their behalf by the insurance companies and others.

Zac: Yeah, it's interesting. I agree with that, that basically there's people who have no wealth or very little wealth. There are people who - different people categorize these things different - but mass affluent is often like a net worth from $500,000 to $2 million. And then there's having a net worth of $2 million to $25 million, and there's $25 to $100 million, and there's $100 million plus.

And actually if you're in the $100 million plus or even $25 million plus, lots of people are holding your hands. You have a lawyer, you have accountants, you have a private wealth person who's not terrible.

And then if you are mass affluent, maybe $500,000 in net worth to $2 million, what you should do is buy index funds, you should do direct indexing, you should have a will, you should have a living trust, you should use some basic things. But they're very common there.

It is interesting to look at what is going to happen to people who have $2 to $25 million in net worth, because in my perception you're not getting the handholding that you get when you're richer than that. But also you do have some more space to do more sophisticated things, but you're not necessarily getting the sophisticated advisor.

And we'll see, because all the big firms - BlackRock, Blackstone, Fidelity, Apollo - they're all making this push to get folks in that bucket to buy more "alternatives," which is basically just code for private equity and private credit funds. And we'll see how wise that is.

But it is the case that there's something of a democratization of asset management strategies. And I think similarly with life insurance, one of our long-term theories of the case is that if you take these products and you actually make them simpler and somewhat standardized - even the complicated products for high net worth or ultra-high net worth individuals - and you package them up and make them relatively thin margin, there could be a larger population of people who want them.

Now my intentions, I think, and my belief in that are true and honorable, hoping that we can provide more access and tax structuring for folks. But that is always a double-edged sword - as you make a product more accessible, if it's a complicated product, it's a way to convince people that they are more sophisticated than they are, and they should engage in this risky activity to the benefit of the manager rather than the benefit of the individual.

Regulatory challenges and mispricing in insurance

Patrick: There have been a few regulatory settlements over the years with asset managers and others who have successfully sold a wealthy but unsophisticated individual on products which, under the totality of circumstances, would not support the actions that were taken. One case, which I'll link to in the show notes - someone sold someone an annuity and churned them out of the annuity every year. [Patrick notes: See generally FINRA’s settlement with one company and rulemaking elsewhere.]

And because of the way that the annuity was structured, this earned the salesman a 3% or whatever commission every year. But there's no way to pay an insurance salesman 3% commission every year and have that be a good thing for the person paying 3% every year. And eventually the math comes out, and of course they were not simply selling someone something that was not in their interest, but not being fully transparent as to what they were paying for under the hood.

And that's another factor here - as you mentioned, a term life product is very easy to understand. You can explain the mechanics, if not the exact actuarial math that goes into pricing, to a bright 10-year-old. But as you get to "it's index linked with a cap" and so on, you could be a professional in the financial industry and fail to diagram it given several hours to try.

[Patrick notes: There are many financial professionals who really grok complicated structured products with embedded options, but not everyone geeks out on this to the degree that e.g. Matt Levine would, even in the financial industry.]

Zac: Well, and actually, I mean, I could think of some cases for you to put in the show notes - there have been tons and tons of settlements, even with really big life insurers over the last 15 or 20 years on the issue of illustrations, which is basically how, when you're presenting the product to a user, you present their potential upsides and downsides on the product.

And this has become an issue of intense regulatory focus because if you illustrate them in certain ways, then it looks like, "Oh, there's no way I'm ever gonna lose my money and I'm gonna get 10% returns every year" - which they would never quite write down. But the diagram would strongly suggest that. The New York State regulator in particular has been really up in arms about that.

Patrick: Interestingly, and less for societal context than for an interesting poor decision made in capitalism that at least one insurer is still struggling with - there was once an insurance company, I believe in Europe, that had the bright idea... Policy innovation is happening all the time because insurers are in a competitive market with each other. There are rich people. You want to go out and sell them things, so you sell them things that sound good on paper.

And an insurance company said, "We will let you essentially buy mutual funds within this insurance policy. And we will let you essentially pick which fund you've bought after the markets are closed for the day." If you gave that offer to Goldman Sachs, Goldman Sachs would quickly say, "Oh, free money for me. You're conceding a lot of option value. That sounds awesome." But the bigwigs at the insurance company in Europe thought, "Eh, notionally, sure, but it's not like anyone is going to actually exploit this, right?"

And they have one policy holder who has for the last couple of decades kept them to the letter of that agreement and taken them for an absurd amount of money. Matt Levine has written about this better than I will.

[Patrick notes: I’m having difficulty finding the substantiating link, but here’s a similar case from the U.S.] 

Zac: Yeah, that really touches on something interesting, I guess on two sides. One is that, if we talk a little bit about reinsurance in the life reinsurance business, what you're doing is you're buying these big blocks of business if you're the reinsurer. So there are all these people sitting in Bermuda - Brookfield and Apollo and others - and they buy blocks of business.

And one of the things you're doing when you price a block of annuities or a block of life insurance is you are estimating how many people will optimally execute it or how many people will do the riders or this and that. And those estimates are always quite low, and they really impact the price of the blocks.

But it's just a case that a lot of contracts have some embedded optionality or a rider you can agree to or something. And very few people are optimally executing their contracts.

Patrick: We mentioned this in a previous episode about the mechanics of credit card rewards, where there are relatively few people who optimally execute their usage of, say, Chase Sapphire Reserve points. And so on a portfolio basis, the bank can model the math. And in credit cards, there are less people scattered around the right tail of having millions of dollars of potential to pull out of the bank if they execute optimally.

Even if someone executes optimally, it'll be the difference between making hundreds of dollars of money of margin on their account in a year, or potentially losing hundreds of dollars of margin on their account in the year. But in insurance, we're writing seven-figure, eight-figure insurance policies that could work out very badly if someone does their math wrong.

And so is that the reason why we have reinsurance in life insurance - primarily to insulate the insurers against mis-modeling risk? Or is it that there could be a COVID sort of scenario where they have vastly more claims among 30-year-olds in 2020 and 2021 than they expected?

Zac: Well, let me start by saying I think that most life insurance is quite similar to what you're saying in credit cards, which is the relative size of individual policies is small. If the entire portfolio is optimally managed by the users, that would be quite catastrophic. But any individual optimal management, although obviously when the policies get bigger, you have to think more carefully about these things.

It's often the case that life insurance companies don't fail very often. There are far fewer life insurance failures than there are bank failures. And of the big failures in the US, I think the biggest one was Executive Life in the nineties. [Patrick notes: AIG during the financial crisis of course threatened an even larger loss, but that was essentially bailed out.] They're almost never about liabilities. They're almost always about mismanaging assets.

Executive Life, I think, was one of the junk bond guys who bought a life insurance company and then pivoted the whole portfolio into sub-investment grade stuff. And it blew up. But it had nothing to do with their liability management.

The big mispricing that's happened in the last couple decades, which is sort of coming to a head right now, is people wrote a bunch - life insurers wrote a bunch of - long-term care insurance. So this is insurance for if you're gonna go to a nursing home. And all of the long-term care that was written and sold in the 2000s or the 1990s was completely mispriced to the explosive expense of nursing homes and other long-term care. So all of those books were just really mispriced.

And actually, if you pay attention to life insurance quarterly earnings, which is a passion perhaps of mine, but not many other people's, tons of billions and billions of dollars of additional capital has had to be posted by certain insurers over the last couple years to account for the mispricing of that book.

Patrick: A related but not quite life story was long-term disability insurance, which by the way, is something that you should definitely have if someone is dependent on your income and you are a professional. For a variety of complicated reasons, professionals are much more insurable for long-term disability insurance than other people who work in the economy.

Long-term disability insurance used to be absurdly cheap for professionals. And then there was a genuine change in risk around the time of, particularly, the AIDS epidemic, and a lot of insurance companies found that they had greatly underpriced policies which obligated them to pay professional wages for the rest of what is now, thankfully, a survivable set of circumstances, but for a very long time, and for a larger percentage of their insured base than they had modeled prior to that being a thing in the world.

And I think there was some concern during the COVID years with the experience of long COVID - is that going to be a second wave of long-term disability that insurers will be on the hook for?

Zac: It's one of the interesting things about life insurance for me. I guess two things briefly that you mentioned.

One is it rhymes with so many other parts of FinTech. The more I've learned about the business, the more it sort of rhymes with the things I knew about banking and payments and these complex systems. I guess all financial systems end up being quite similar.

And the second thing is, for me, life insurance feels a lot like payments did in 2013. There've been a few startups. They all, in my opinion, did interesting but slightly wrong things. But it is an industry where trillions of dollars are sloshing around globally. It's like 3% of global GDP, but there's really not a lot of software and a lot of technology. So it just feels quite ripe to think about in the way I think that you, Patrick, think about things, and then build technology around it.

Patrick: I'll have to drop a stat in the notes because I don't have one cached off the top of my head. But the portion of all financial assets in the United States which are ultimately owned by Japanese people through life insurance policies is very, very large.

Zac: Oh, yeah.

Patrick: Much larger than people's baseline expectation of it. And that is also because Japanese people buy a globally high amount of life insurance.

I have a fascinating anecdote for that, by the way. We talked about the principal-agent problem. The life insurance industry was kicked off in Japan by a notion that some executives at companies would need a post-retirement job. And if your best asset as an executive at a company is that there's a lot of people who know you and trust you and have been addressing you as "Yes sir. Right away, sir. I will do whatever you need, sir" for the last couple of years - and they no longer do that because they don't work for you anymore - you should go back into that office where everyone trusts you and does what you tell them to, and say, "You know what you should do? Buy life insurance. It'll help your family." And so a lot of the large Japanese insurers have a second career workforce that is essentially using their social capital.

And another funny Japan story - you're not a real company in Japan until you have your own insurance company. And so a policy that I still have in force - in no way compensated for the following - the Sony Corporation of Japan is like, "Why are we leaking margin to an insurance company when we have a hundred thousand employees? We should just have our own insurance company." Then they were like, "We have an insurance company. Why are we not making money from it when we have a consumer brand name that people love and trust? We'll sell Sony insurance policies all over the country." And so Sony insurance policy, it does what it says on the tin.

You should definitely get them if you need term life from the Japanese market. Or consult someone who is regulated to sell you Japanese insurance products. I'm not permitted to do that.

Zac: Actually, one of my investors is MS&AD, which is owned by Sumitomo. The Japanese proclivity towards conglomerates is something that my poor American mind can scarcely comprehend.

Reinsurance and the role of Bermuda

Patrick: So continuing on the international issues - why does so much of the industry, the reinsurance industry in particular, find itself clustering in Bermuda?

Zac: Well, first off, Bermuda is such an interesting place. It's like a thousand miles east of North Carolina. It's really in the middle of nowhere. It's not in the Caribbean, so it's not anywhere near The Bahamas or Jamaica or whatever. And it used to be the tourist destination if you were in New York because you were getting a boat and it was the closest island.

Then once there were airplanes, why wouldn't you go to someplace that's warmer than Bermuda? So in the fifties and sixties, basically a group of enterprising Bermudians just created the insurance business there out of whole cloth.

As an aside, Bermuda has a really fascinating history of how they've survived on this barren rock in the middle of the ocean. They've been a penal colony. They ran a ruthless salt monopoly in the Caribbean. They were bootleggers during the Civil War to the South. They're a very scrappy set of folks out there.

And they really started by creating the captive insurance business. So, to your point about large companies building insurers, often large companies will build captive insurers, which are owned insurance companies to handle their P&C risks. And that was really their business for a long time. And then they got into P&C reinsurance. And life reinsurance is really just starting to play a big role on the island in the last 10 years.

But it's a very interesting place to go and visit. And I highly recommend it to anyone who has an interest in these sorts of topics because it's 60,000 people on an island, and 30,000 of those people are actuaries, lawyers and accountants. It is not your mental model of an island nation. It is a very well-developed, not that many tourists, professional island of people who think about things like taxes and stuff.

What I'd like to say about Bermuda is they've really positioned themselves as being the upstanding, hard-nosed offshore regulator, particularly when compared to Cayman or Bahamas or BVI or Gibraltar or Channel Islands or whatever. But they're just a little bit more permissive on the two things that insurance companies care about.

So they do all the stuff right on the regulatory side. They're really respected by the National Association of Insurance Commissioners, which is the US body that sets up such matters. And they have Solvency II equivalents, which is this European standard. They're respected by the Europeans.

And basically, the two differences are: when you're calculating your liabilities, they let you be just a little more creative. And I wouldn't even say "creative" is the right word, but they, instead of just following the rules by the letter of the law, they have this concept of a "best estimate liability," which is basically that you really just have to think hard about your liabilities and you have to justify them to your auditor and your approved actuary and stuff.

So it's not super loosey-goosey, but you really can think about what are the present value of these liabilities. And that can be just a little bit more permissive than the US or European rules.

And then similarly on the asset management side, they're just a little bit more permissive in your ability to - not take the entire insurance company's balance sheet and bet it on black or put it into venture funds, which would be quite crazy for an insurance company - but instead of having 40% of your balance sheet in treasuries and 40% in investment grade bonds and 20% in private credit, maybe you can have 20% in treasuries and 50% in investment grade bonds, and then 30% in private credit.

But when in an insurance balance sheet - a bank is levered, right? You put up a hundred dollars in capital, you can have like a hundred or a thousand dollars in assets from all the premiums you write and stuff. That can make a big difference on your return on equity to be just a little bit more permissive.

So the reinsurance business in general - the original idea is risk spreading, it's diversification. Like, "I run an insurance company in Minnesota" - now that's Northwestern Mutual, but it used to be the Upper Midwest. So you actually did have some genuine concern about localized underwriting failures or pandemic mortality, or "I'm just not big enough." So the idea was, okay, you have this global, larger reinsurer.

And the other thing is that even if you are a relatively big carrier in the United States or wherever, you might want to reinsure because it's a more effective use of capital - it's capital relief or balance sheet optimization. If I have a hundred dollars and I could only be 8x levered or 10x levered, then once I am 10x levered on my hundred dollars of capital, I'm tapped out - I can't write anymore life insurance.

But if I sell a bunch of that to someone else and they're reinsuring it, maybe I'm getting a little bit of economic exposure, but I'm using a lot less of my capital for that exposure. So that's the original reason.

But then in practice, what's happening now is lots of people, even big name brand companies, will reinsure offshore to get that slightly more permissive liabilities treatment and slightly more permissive asset management capabilities.

Patrick: It's interesting how many parts of the financial market end up being regulatory arbitrage.

Zac: I was trying to look for a different set of words. Historically, there's two ways to make money in financial services: bundling things and unbundling things. And then often helping things flow to the lowest permissible levels of riskiness or efficient points on the curve - that's some of the underlying economic impetus there.

Private equity's interest in insurance

Patrick: You mentioned private equity a couple of times. One, obviously it's a thing that an insurance company can potentially put a percentage of assets in. But as you've mentioned, private equity firms have recently decided having an insurance company is a wonderful business to be in. What's the justification there? 

Is it simply "We'd love to have a pool of permanent capital or semi-permanent capital"? [Patrick notes: Matt Levine has covered this several times.]

Zac: It is that we would love to have a pool of permanent capital. So, I say "private equity," but a lot of private equity firms are now also private credit firms. So Apollo now is almost more a private credit fund than a private equity fund.

And what these folks have realized is that if you can be in the business of writing long-dated liabilities, that is a set of permanent capital vehicles. Then you can invest - if you're in the business of doing various flavors of fixed income all the way from managing treasuries to 10% or 12% return private credit, that actually is exactly what an insurance company wants to invest in.

So the Apollos and Blackstones and Goldman Sachses of the world now have a huge business in being an asset manager for insurance companies. And then they realize, "Well, okay, we're in this business of managing the assets for an insurance company. Why don't we actually just own the insurance company, and then we'll in some ways be managing our own assets?"

So there's been a whole - and Apollo really originated the strategy on some level by building Athene. And Athene, I think, is the largest seller of five-year annuities in the United States. And then they reinsure all of that annuity business from their US affiliates into Bermuda, which allows them to be a little more permissive on what they can invest in. And they invest all of that into Apollo credit strategies.

And in fact, what Apollo does, I think now, is they say, "Oh, we have this credit opportunity," and then they allocate that credit opportunity to the Athene bucket and the fund bucket and the principal capital bucket.

And so this has led to KKR owning Global Atlantic, Blue Owl buying Qvar, and these are all the biggest providers of annuities in particular in the United States - all life insurers owned by these private credit funds.

Patrick: And again, the rationale for this is that there are some firms in the economy that would like to borrow money over a long time scale, which means someone needs to put up the money over a long time scale. We've been moving long duration assets out of banks for a while because funding those assets with deposits causes bank failures in uncertain times, as we most recently saw in 2023 with the collapse of a couple of US banks, including SVB, probably relevant to both of us in some fashion or another.

[Patrick notes: I realized a few seconds after I said this that Zac worked at SVB after selling his company to them. I had only meant “We were mutually professionally exposed because ~everyone in our subspecialties was professionally exposed.” Mentioning this because otherwise it sounds quite catty, and I do not do catty.] 

Patrick continues: And it is less likely that 80% of the book of business of an insurance company could decide to either cancel their policy or pass away in a period of 48 hours. And empirically it is possible to move 80% of a very large amount of money out of a bank in 48 hours under the right - or wrong - circumstances.

Zac: You do actually have to model this risk. It's called mass lapses. I think there's never been a mass lapse event, at least since the 19th century when insurance was somewhat the wild west. There were just tons of insurance companies that sprung up, and people would sell policies door to door. The 1800s were a little crazy. I joke that life insurance was the crypto of the 1800s.

Building a crypto life insurance company

But that's absolutely right - this is the original reason I got interested in life insurance. It's not actually why we started this business, but I went through Y Combinator with this company called Center Treasury. And Center Treasury was doing banking as a service - bank account as a service. Now there are like 10 people in this business, but we were early and eventually got bought by Silicon Valley Bank. Stripe Atlas - the original version of that was built on our software - the bank account at SVB that Stripe Atlas would give you was the API we built.

But in that company, we were thinking about starting a bank. And this was before Monzo had started. And so we were looking at the UK and starting a challenger bank there. And we looked at the US, and one of the big questions we had if we started a bank was: what do you actually do with the balance sheet?

Like, our intention was to do bank account as a service and payments - that's the world we came from. But what are you gonna do in the bank with the balance sheet? So I was thinking at the time a lot about lending as a service and what the equivalent of bank account as a service is, which no one has really ever done well or figured out to this day.

And in the course of that thinking, I was like, "Wow, the best bank is a life insurance company." The natural owner of mortgages is an insurer, not a bank, because they actually have these 30-year liabilities that they could match.

So after we sold the company to SVB, I went to Money 2020, and people were like, "What are you gonna do next?" I actually was telling people, "Oh, I'm gonna start a life insurance company." And the reason I didn't do that in 2015-2016 is I couldn't figure out how to get to market.

Because although when you're big, being great at technology is amazing because you're turning a 15% return on equity business into a 20% return on equity business. And actually, lots of the stuff we talked about here, like reinsurance or who should sell life insurance - I actually think it should be inside Wealthfront as an embedded API offering. And there's tons of technology opportunities, but that doesn't help you when you're small. It just doesn't really matter when you're small. It really matters when you're big.

So years later when we were sitting down, the way we came to this business is we were actually thinking about crypto, or Bitcoin in particular, and saying, "Okay, we think there's gonna be a Bitcoin economy and we are really good at regulatory BS. And we think that's gonna win," which was an unpopular position in January 2022, that the long-term winners are gonna be people who can think about and do regulatory stuff.

So we sat down and we wrote a list of, "Okay, every functioning economy in the world, they have a stock exchange and they have a payments company and they have a bank and they have this and they have that." And we actually didn't have life insurance on that list. And then I woke up in the middle of the night, two weeks later, and was like, "Oh, there should be a crypto life insurance company."

Which is in some ways not intuitive at all. And some people think we're crazy because we're doing this incredibly conservative, long-term, boring thing with Bitcoin. But to bring it just to life insurance, it has allowed us to build this company and have a product that people want to buy despite the fact that we're small and have some green technology.

So there is another life insurance company in Bermuda. They are older than us. They have raised more money than us. They have a much more experienced team than us, and they don't have a single policyholder because nobody wants to trust a brand new, under-capitalized life insurance company for their run-of-the-mill, commoditized, off-the-shelf dollar product.

But we have this product that no one else has, which is Bitcoin whole life insurance - which I'm not currently selling to you, but just describing - and that product no one else has. So we have policyholders and we have people who believe in us, and then we're able to build a policy admin system. We're able to build a regulated entity. And then the hope is over time to launch additional companies, maybe in stablecoins or additional companies maybe in dollars.

And that all starts from having a product no one else has, which we just realized by mistake after wanting to build an insurance company for the crypto economy for years. But anyway, that's how I've learned so much about insurance.

Patrick: I think that's the thing we see over and over again in FinTech - the math is very punishing and you have to get good distribution or you have no business. And then getting good at distribution often means getting good at the advertising machine that the incumbents are very, very good at and know exactly what they can pay for the marginal lead.

And if you don't achieve massive scale doing that, you have a lot of difficulty making headway as a FinTech business. And it is difficult to achieve massive scale simply by building a better mousetrap unless it's in a place that literally no one could follow you for whatever reason. And that can unfortunately be written as the epitaph of a number of FinTech businesses over the years. But that is what it is. 

I'll drop some perhaps more considered thoughts in the show notes on it. [Patrick notes: The margins of this transcript are too narrow for the piece I want to write, so I’ll have to do it for Bits about Money in the near future.]

So Zac, thanks very much for the interesting discussion today on the ins and outs of insurance. Where can people follow you on the internet?

Zac: On Twitter I am @ZTownsend. My company you can find at meanwhile.bm. 

Patrick: Well, thanks very much, and for the rest of you, I'll see you next week on Complex Systems.