A lot has been said about disruption in the insurance industry, but to date, there has been very little in the retail market. Editor in chief Rick Pullen recently sat down with Samir Shah, CEO of Ledger Investing, to talk about the future disruptive nature of insurance-linked securities.
How did you get here in the first place?
It wasn’t a straight line. I became a pension actuary by accident and then moved into general management consulting. I’m probably one of the few people who worked across HR and risk management, across pension, life and p-c insurance and banking. I became chief risk officer for Scottish Re, at the time a global life reinsurer, and it was an interesting business model. They essentially wanted to be the Fannie Mae of the life insurance industry, buying huge blocks of life insurance and then securitizing them. Unfortunately, they were doing this right in the middle of the crisis and a lot of their securitization structures had collateral that was with Lehman Brothers. So they got caught up in the subprime crisis just as I was joining them in 2008. I learned a lot.
Then, I became chief risk officer for Validus, a global property cat reinsurer soon to be part of AIG. I left Validus and joined AIG in 2010, initially as the first chief risk officer of their global p-c business. I built the risk management function at AIG p-c and did the first allocation of the cost of capital. I then became chief reinsurance officer and the head of insurance capital markets. I started observing a lot of things that were happening in the capital markets, which led to a lot of the ideas underpinning what we’re doing now with Ledger Investing in leveraging capital markets.
Why Ledger Investing?
When I took the reinsurance role in late 2011, there was a fundamental shift in the investor base in the insurance-linked securities (ILS) market—from sort of opportunistic investors, like hedge funds, who were seeking reinsurer type returns to more stable long-term investors, like pension plans and endowments, who were seeking some relative value compared to other fixed-income investments. Interest rates had been low, and they were expected to stay there. So they were seeking some yield.
As they started coming into the market, they observed the performance of the securities was not correlated with the rest of the markets. In distressed markets, everything becomes correlated. And yet, even during the crisis—you can see statistics on this thing—insurance-linked securities were not correlated.
We started tapping that market and noticed its cost to capital was much lower than reinsurance. I started examining why. I wanted to know whether that was simply opportunistic, cyclical or something fundamental. I realized what was happening was that ILS investors were pricing in the diversification benefits of risks, which are completely uncorrelated to the traditional asset classes. So they were OK with lower returns instead of the returns insurance stock investors needed.
For example, insurers’ cost of capital is around 10%, but ILS investors don’t need 10%—instead needing only 3% to 6% to create relative value compared to their fixed-income investments.
Especially when interest rates hover around 1%.
Exactly. Insurance risk should naturally have a much lower cost of capital because of diversification, but when you put it on an insurance company balance sheet, two things happen: first, it is mixed with all the other risks and becomes opaque. As a chief risk officer, it was hard enough for me to understand all the risk in the insurance company much less being able to convey that to the management team and then to the board. Investors are so far removed they really have no clue. That opacity has a cost.
Second, for large p-c companies, you know, it’s not unusual to have maybe 30% to 50% of capital there simply for investment risk that by definition is fully correlated to the capital markets. So if you’re an investor and you buy insurance company stock, you need a 10%+ return. But if you invest directly in cat risk, you may need only a 5%-6% return.
Also, when you hold risks on a balance sheet, even though the entire insurance industry or reinsurance industry in total has $2 trillion, let’s say, it’s still concentrated, and there’s a cost of that concentration. When you place this concentration into the deep pool of the capital markets, it diversifies, and the cost goes down. So that’s another reason why the cost of capital on an insurer’s balance sheet is artificially high.
In 2012, I saw this as fundamental. It was not cyclical. This was an enduring trend. I thought, in fact, the prices should go down even more until they converged to equivalent rated risk in the traditional asset classes, like high yield, fixed income. And that’s what happened over the next four years. And every year in Monte Carlo the reinsurers would say, “Oh, I think it’s at its bottom and it’ll go back up.” They were treating it like a cycle because historically that’s what’s happened in the markets. Capital flows and prices have been cyclical.
I observed this fundamental shift and realized it’s not just for cat risk. In fact, all insurance risk is uncorrelated. Over the long term of the business cycle, things can get highly correlated. But from an investor perspective, correlation has a different time frame. Investors are really focused on short-term volatility—daily, weekly, monthly, annual. And from that perspective, almost all insurance risks—loss ratios—don’t bounce up and down like stocks and bonds do.
I saw an opportunity to source capital for all the insurance risk and try to make the case to do that within AIG. But that is difficult to embrace in an industry that for a couple hundred years has thought of itself as a warehouser of risk, where it’s gone to market by saying, “I have the biggest balance sheet.”
Insurance companies used to leverage that, and to all of a sudden tell them “you shouldn’t be a warehouser of risk” is very difficult. But that’s the opportunity I wanted to chase. So I started recruiting big investors. There’s already a market of insurance-linked securities funds. There are maybe 50 or so in the market, and they’ve grown. But it’s still a niche asset class.
I went to big investors and said, “Why don’t you come into the market? We can issue securities.” And they replied, “We understand this whole premise. The thesis is sound. However, you guys understand your risk better than we do. So you could pick us off.”
So I focused on that. I saw how to make risk more transparent and reliable for investors and how to standardize the structures by focusing on using this for capital management rather than risk management. The industry still thinks of all this in terms of reinsurance. And reinsurance is thought of as a risk management tool.
But here’s the capital management perspective: you need 100 units of capital to take on 100 units of risk. You have to figure out what’s the cheapest capital. Is it your balance sheet, the reinsurer balance sheet, or the capital markets? And that’s the game. It has nothing to do with risk management, which is a separate objective; you need reinsurance for that.
You have to keep this independent of an insurer; otherwise, investors will say, “You’re going to keep the best risks.”
Exactly. So the risk analysis has to be reliable and transparent. What’s happening around the same time is the use of technology that has improved data, analytics and predictive modeling. I saw an opportunity to leverage that to improve transparency.
Ledger Investing creates a platform that allows insurers to securitize all of their risk classes—life insurance, health insurance, property-casualty insurance, short tail, long tail—by applying analytics that are much more transparent. Portfolio analytics are much more transparent and reliable. Their reliability can be gauged by investors without them having to become underwriters, and the structures are standardized for capital management purposes.
Every insurer, for reinsurance purposes, wants to negotiate their terms because what they’re saying is, “I want these risks, and I don’t want those risks.” But capital is fungible across all risks, so you don’t need the reinsurance kind of complexity. There’s an opportunity to standardize from a capital management perspective.
And the risk is spread among many investors.
Yes, the more reliable and transparent the risk analysis and the more standardized the structures, the greater the number of investors who would be interested in investing in ILS. This is key to commoditizing insurance capital to achieve lower costs while increasing capacity.
We’re not creating a brand-new product. We’re taking an existing product that’s a niche market that was focused on reinsurance and expanding it to all of insurance by standardizing, making things more transparent, and making it efficient and positioning it as capital management.Tweet
So instead of going to their balance sheet to cover risk, carriers would be going out to the markets.
Yeah. So, the opportunity is that, if an insurance company securitizes its risk, it has a bunch of different benefits. First, it has a smaller balance sheet. By the way, I’m not suggesting they securitize everything. I think 25% to 35% of the capital should be securitized.
Why not more?
Because there is a downside. Let’s say you have $1 billion of capital and 10 product lines, and let’s say $100 million in capital is attributed to each product line. And I take one of the product lines and I securitize that $100 million. Well, now, that $100 million is available for only that product. If I had kept it on the balance sheet, that $100 million would have been available for all products. So if you securitize too much, you lose fungibility—essentially the diversification benefit of your portfolio.
The good thing, however, is you have less on your balance sheet so you get a higher return on equity. The stuff you are writing that goes off balance sheet, you’re earning a spread on that. You’re earning some profit commensurate with the value you created originating and underwriting it and servicing that risk. You’re just not holding it.
You do need to get paid something, and that’s essentially what that spread represents. And so it’s essentially a fee business. Insurers have tried to grow their fee business. It’s extremely difficult. This is the easiest way to get fee business for a huge portion of your portfolio. That would give you a much higher return on equity.
Also earnings become more stable when you have less risk on your balance sheet. And most importantly you can write far beyond the limits of your balance sheet. This is one of the biggest problems in our industry. The industry says it has a lot of excess capital, but at the same time there’s a huge insurance gap—the difference between economic losses and insured losses. The difference is because of the cost of that capital.
People don’t talk about this. A couple of years ago, I was at a Bermuda conference where some very prominent CEOs were saying, “We have excess capital. You brokers should go out and write more business.” But they failed to touch on the fact their capital was very expensive.
There’s opportunity for the industry to organically grow instead of just stealing from each other by leveraging cheaper off-balance-sheet capital through securitization. The higher ROE, more stable earnings and ability to grow increase the valuation of the company.
You say carriers are reluctant to change their business model.
There is a significant cultural constraint there, even though there is a meaningful percentage that really believe in the capital markets as opportunities. I found brokers are one step ahead of carriers. They don’t have the same constraints, because they have always been looking for the cheapest capacity. They were never warehousers of risk.
Are you going to disintermediate insurers?
That’s not my objective. The insurance industry is huge. We’re a tiny little blip. We’re not going to disintermediate the insurance industry. I am suggesting that only 25% to 35% of capital should be securitized, because if you securitize too much, then you suffer the loss in fungibility of capital. So we will always need rated balance sheets that insurers provide.
However, sourcing capital and regulated entity/issuing paper will become a commodity. It’s just a question of how quickly. Capital has become a commodity in every other major industry, and the only reason it hasn’t in insurance is because risk is opaque. If we can break through that, capital becomes a commodity. It already is in cat. So it will happen.
Pension plan consultants are advising clients to put 1% to 3% of their assets into this class because of the lack correlation. Fixed-income investments in the world account for $80 trillion. So, that’s easily $1 trillion to $2 trillion in new capital for insurance. Compare that to all the capital in the insurance industry, now. Property-casualty insurance is less than $1.5 trillion, and the reinsurance industry has about $500 billion. So the potential is huge. I think of this as a tsunami of capital.
People talk about the disruption of technology, but the capital has already created new business models, whereas technology has not—yet. Technology has made things efficient. But all the major reinsurers are now asset managers investing for third-party capital for a fee. It’s clear this is a business that their shareholders had, and now third-party capital has it. So it’s already disrupted.
There are brokers who have arranged private deals directly into the capital markets. The NY MTA, Amtrak, Kaiser Permanente and many Caribbean countries have bought insurance directly from the capital markets. They skipped the entire value chain. So this is hugely disruptive. And we don’t have to create this. We just ride it and position it to channel to the insurance industry.
Then the question for all players is if you’re an insurance company, what are you going to do. There’s an opportunity for insurers to become what Scottish Re was, essentially the facilitator to the capital markets. But we can’t do 100% into the capital markets. There always needs to be risk on a balance sheet.
Since 1997, there have been cat bonds. So the regulatory and rating agency treatment already exists. Life insurance risks have been securitized, as well. So we’re not creating a brand-new product. We’re taking an existing product that’s a niche market that was focused on reinsurance and expanding it to all of insurance by standardizing, making things more transparent, and making it efficient and positioning it as capital management. In fact, by collateralizing more of the risk and making insurance more accessible and perhaps cheaper, regulators may look favorably at securitization.
How can brokers ride this wave, and what does that mean to carriers if they do?
You have to decide which part of the value chain you want to play on. Brokers need to invest in technology and compete on the basis of the connectivity with the customer. One thing they need to do that’s been missing is to gather data. They handle clients, but they’re missing the loss and exposure data. And when I talk to brokers, I ask, “Do you have any data?” They say, “No, we don’t. The insurance company has it.”
So this is going to be a fight. Who’s going to be able to keep the data? That data is gold. Not only account-level data, but given the state of analytics today, there’s lots of third-party data that could be brought to bear. Brokers have the opportunity to maintain control and service the policy with claims. If they can do that, then it’s easy to connect up to somebody like us or anybody else who furnishes them the pipeline to capital.
Historically, tech has disrupted other industries by getting control of the customer. The number of startups that are focused on that connectivity in insurance is huge. And that means brokers are under attack from both sides. They’re squeezed by insurers and also getting competition from this new source.
They can defend themselves against insurers by pursuing this kind of a new, more efficient channel. Yet insurers have a decision to make as well. They can either continue what they’re doing and try to be that big balance sheet and hold on to some of the flow, or they can facilitate this flow into capital markets and make money because, as I said, not all the risk can go into capital markets.
Insurers who facilitate the flow into the capital markets rather than fighting it are in a position to control much more business than the size of their balance sheet currently can. So, whether you’re a $10 billion company or a $1 billion company, it’s conceivable you can have 30%, 40%, 50% of a market. That was previously impossible to fathom given the fragmentation of the industry. There’s nothing that stops you if you’re efficient at this value chain.
If you’re an insurer, if you buy into this proposition, you can control a huge portion of the market. Very much like how other tech companies have suddenly dominated their markets, there’s an opportunity for insurance companies to do the same.
Brokers can do this too by getting the connectivity of the customer and the data. Then, they can attach to this pipeline.
So what you’re telling me is that brokers could disintermediate 25% to 30% of insurers?
I’m talking about over a 10-year time frame. I mean, the pace is the only thing that’s uncertain. There’s no limitation, there.
Insurers who facilitate the flow into the capital markets rather than fighting it are in a position to control much more business than the size of their balance sheet currently can.Tweet
What does this do to premiums?
Pricing, naturally, is a market phenomenon. You go to a market, people buy and sell from each other, and you get a price. So price is something that you observe. It should not come out of a model, yet in the insurance industry, that’s what happens. That’s because those who actually hold this risk—the investors—don’t know anything about it. Because the risks are opaque, they generally say, “I want 10%+ return.” Yet, they can’t tell you how much return they want on workers comp, homeowners, commercial property or aviation. They don’t know the risks.
But if you securitize the major risk classes and create transparency, then investors, through their actions, price those securities—different prices for different risk classes based on their risk profile.
You see this in cat bonds. You see curves where the cat bond price will go up and down. An insurer can take those prices and transfer them to insurance buyers directly, so pricing no longer comes from a model but instead from interactions between investors on the one hand pricing securities and consumers on the other hand buying insurance based on those prices.
There’s a reason why when you want to get a good mortgage, every day the price changes, because that risk ends up in the capital markets where prices move every day. So why can’t we envision insurance pricing working the same way? Can’t you imagine in 10 to 20 years you go to some website—to the originator—and you get prices for a bunch of insurance policies based on capital market prices?
So if you’ve reduced pricing—
When cost of capital decreases, capacity increases significantly. Even now, there isn’t sufficient capacity. I saw commercial accounts where they wanted $500 million in peak cat coverage and they could only get $100 million. The industry was only willing to give them $100 million. There’s a reason why New York’s MTA [Mass Transit Authority] went to the capital markets. They couldn’t get capacity that they wanted at a reasonable price.
My point is that not only does the price go down to its naturally occurring and most efficient level but the capacity increases incredibly. If you have a limited amount of capital, you focus on risk selection; when you have unlimited capacity, you focus on pricing all risks.
Does this mean lower premiums and smaller commissions for brokers?
No, no. Rates go down, but the volume increases, which means bigger commissions. There’s a subtlety here that needs to be understood. I think everybody acknowledges there isn’t sufficient coverage. I mean, you see this all the time. Only a fraction of the economic loss is insured. Why is that we say that we have a lot of excess capital?
How will rising interest rates affect the capital that’s going to flood into insurance?
That’s a good question. So, first of all, let’s break up interest rates generally into two parts. One is the risk-free rate. It’s Treasuries. And the other one is the spread on corporate fixed income.
When Treasurys fluctuate, it doesn’t affect insurance-linked securities’ pricing, because these notes are actually not fixed interest rate notes but variable. The capital that’s raised to cover the risk sits in a fund earning the risk-free rate. So if the risk-free rates go up, the investors get whatever the returns are in that risk-free account. Plus, they get the risk spread on the notes—that’s the reinsurance premium.
Risk-free interest rates going up doesn’t affect this. It’s when spreads of comparable risk, securities and other asset classes widen. Pricing will still go up and down, except it will go up and down with the capital markets, not with insurance markets.
We will no longer have the insurance cycle?
Exactly. Part of the logic in the insurance and reinsurance industry is if you have a loss, I’m going to raise your price. Not because I think there’s more risk next year. Not because exposure has changed but because you have a loss, so I’m entitled to charge you more. That’s not risk management. That is not really insurance. That’s like financing your loss.
You are starting to see in the market now that when you have big losses prices don’t go up much. That’s because the capital markets have come in and put pressure on this dynamic. So they will not follow the underwriting cycle, which, quite frankly, I don’t think was a real cycle.
When you have a lot of permanent capital, you have no choice but to deploy it. So if I’m an insurer, I have permanent capital, and I want a return of 10%. That’s what my investors want. So I price my policies to meet that goal.
For whatever reason, let’s say that you, a competing insurer, have a lower assessment of the risk and your price is lower than mine. So I’m going to lose some business to you. Now, if I lose business to you and if I can’t replace it, because it’s a zero-sum game in the industry, I have to lower my price. I can’t just sit and not change my price because otherwise I’ll have capital that will get zero return. And so I lower mine. And the next guy then lowers his. That’s what drives the soft market.
It’s not really a cycle. There’s a permanent dynamic that softens prices, and this is a huge problem. But if you had securitization, you could write a little bit less and securitize less. You can flex the capital, and you’re in a better position to protect against a soft market and manage to get the returns for your investors that they really want.
There’s also double taxation that securitization avoids. As you know, the insurance company gets taxed, and then the investor gets taxed on dividends. But if you securitize this risk and you give it directly to the investor, the profits were never in the insurance company. They go to that investor instead. It gets taxed once.
And the insurance company gets a fee for making it happen.
They get a fee for the value they create. They’ve originated the business, they service the business, they gather data, they do analysis and they make risk transparent. They’re entitled to make money on that. In fact, there’s an opportunity for insurance companies to be like tech companies. If you can imagine this going to the extreme, you have less and less capital. You have less and less risk. Most of your focus is on origination and gathering data and analyzing risk. That’s all tech activity.
You become more like a tech company. So you can get a tech multiple valuation rather than an insurance company valuation. Instead of being a highly regulated, very opaque, capital-intensive company, you’re now capital lite, very tech oriented, a fee-based business that makes multiples.
Does this mean you could finally get larger limits on something like cyber insurance?
No, not yet. Remember, the constraint here is that the risk must be transparent and the risk assessment has to be reliable.
Because we don’t understand cyber yet?
That’s right—from everything I’ve known in cyber, and I certainly understood the need for capacity. The issue with cyber is there are people who will say we know how to model cyber. But I don’t think there are investors who will buy that kind of modeling and say it’s objective, back-tested and reliable. Until you achieve that, I think it will be difficult. Remember, you need standardization, transparency and a reliable risk.
I can see our top brokers can easily step into this and work with somebody like you. Would it be more difficult for our regional and smaller brokers?
Well, if they don’t have the staff to do the data gathering and analytics.
They don’t have to do that. That’s my point. We have been spending a lot of time and effort on analytics. This is my background: modeling risk. We don’t need brokers to do the analysis.
Our analysts can deliver account-level underwriting pricing models to brokers. Data science, quite frankly, is becoming commoditized. All the brokers we work with just need to have strong connectivity with the customer and get historical exposure and loss data.
You take the data and put it into your model and then price the securities?
On the investor side, we would do it as a portfolio. We don’t give investors detailed, account-level information. We give them portfolio analytics. We can back-test it. For example, we now have a client portfolio for which we have data for the last 10 years. We can back-test our risk models on this in many different ways. The investors will see the strengths and weaknesses of different models, just like they do now for cat, and they will price it.
On the other end, for the brokers, we take that portfolio price and convert into account-level prices so that they can price the individual policies. And what they essentially need to do is control the customer. By controlling the customer you should be able to get all the data.
So, predict the future. How are brokers and carriers going to react to this?
It’s soon going to become a free-for-all. In the very near term—I’ve already seen this. Until now, everybody respected their position in the value chain. You’re my client. I don’t go around you. Reinsurance doesn’t go around. But soon, everybody will be going around everybody.
I think capital and operating licenses are increasingly going to become commoditized. It’s a question of pace. I can imagine 10 years from now you will have some people who have large market shares in certain product lines. They will grow the market by writing new business, not just taking it away from others. So there will be natural growth.
The front end is entirely tech activity. All risks are priced rather than focusing on risk selection. The idea that I have to choose between higher risk and lower risk is a little bit frustrating for people outside the insurance industry. Let the investor price both. The higher risk actually may be more profitable because it also has the higher premium. Only when you have finite capital and you are a price taker, then you have to choose, and you would likely choose the lower risk.
It’s like investing in fixed investments. The risk and return of Treasurys versus corporate bonds.
It’s exactly the same thing. That’s what happens in credit, right? You have a credit score.
What about carriers? Sooner or later they’re going to have to do something about their business model. They can’t just ignore this.
I think the carriers are focused on predictive modeling on the origination side. They have not embraced the capital side. I think they’re going to be caught off guard because reinsurers were caught off-guard. They thought this was cyclical and then they all immediately turned. Within a few years, they all became asset managers. I think this will quickly become an issue. Because all these ILS funds are starting to go around insurance companies.
Are you talking about pension funds?
Pension funds invest through specialist ILS funds. Those funds are starting to have a lot of capital that must be deployed. So they have connected with fronting carriers that are essentially pipelines from MGAs directly into the capital markets. It’s small. It’s a niche. It’s going to grow exponentially. And the way it will grow exponentially is once they break into a new product line and set a precedent, the trajectories will be exponential, just like it has been over the last five years for cat.
The defining characteristic of the insurance industry is that it is not great at leading innovation but it is a very good follower.
Most technology is just making something more efficient. It’s not a disruptor. This idea, what you’re doing with the capital markets, it appears to actually be disruptive.
The business model hasn’t changed. This changes the business model from the carrier side or from the broker side. It defines the insurance 2.0 value chain. Insurance 1.0 has been very efficient for a couple hundred years. I agree with you that a lot of the technological improvements are simply things that were due. They should have happened 20 years ago when all other industries did this, and so now they’re due. All the hype about technology and insurtech, they have not actually changed business models. And the insurance company corporate venture capitalists—for the most part—who support this and finance this are doing so for ones that will make their current business model more efficient, not blow it up.
I think securitization is disruptive. It’s already disrupted reinsurers. Go and talk to reinsurers and ask how many of them are asset managers now. Three years ago, this did not happen, right? This is a tsunami, and it’s going to continue.
Disruption doesn’t come from within. That is one reason we went to Silicon Valley. We wanted to learn from them how they built marketplaces in other areas. We don’t think we can learn from the insurance industry.
So, I think the disruption will come from the outside and it’s a question of who will embrace it. It’s a free-for-all. These next five years are going to be interesting. And because it’s a free-for-all, it also, I think, creates a lot of opportunities in terms of business models, which could emerge in many different ways.