The world is increasingly marked by the unpredictability of extreme climate events, the urgent need to build resilient communities, and the biggest question of all—how to finance these projects.
I reached out to Dr. Shalini Vajjhala whose work stands at an intriguing confluence of innovation, policy, and finance.
With a background as eclectic as it is profound—spanning architecture, international development, engineering, and public policy—Dr. Vajjhala has navigated a path that is as unconventional as it is impactful.
She holds a PhD in Engineering & Public Policy and a Bachelor in Architecture from Carnegie Mellon University. She was a visiting professor at Johns Hopkins University School of Advanced International Studies and is a nonresident fellow with The Brookings Institution.
At the heart of Vajhalla’s work is re:focus, a firm that creates resilient infrastructure solutions and private-public partnerships. Here, the resilience bond—a financial instrument designed to fortify urban and coastal areas against the vicissitudes of climate change—marks a pivotal leap forward in how we conceptualize and fund resilience.
This novel approach, crafted in partnership with titans like Goldman Sachs, RMS, Swiss Re, and the Rockefeller Foundation, embodies a profound shift towards valuing and financing preventive measures.
Dr. Shalini Vajjhala has challenged conventional wisdom and charted a course toward a resilient future. Her solutions unite rather than divide people. They uncover money where originally there was no money to be found.
As we delve into this conversation, we explore the intricacies and insights of the resilience bond. Following is the long-form transcript of our conversation.
Dominique: Please tell me a bit about your background and the impetus for co-creating this concept of a resilience bond.
Vajjhala: My background is fairly unusual. I am trained as an architect, and I worked in international development for a small while before getting pulled into a degree in engineering and public policy. All of this adds up quite nicely, in hindsight, to following problems where communities need infrastructure and where it is lacking.
For my Ph.D., I went to an economics think tank and then got pulled into the Obama Administration. I may have the distinction of being the only person to hang up on a presidential transition team that called about a position. I thought it was my younger brother joking with me, so I actually hung up.
Dominique: [laughter] How did this resilience bond with Goldman Sachs, RMS, Swiss Re, and the Rockefeller Foundation come about?
Vajjhala: You know, it was an invention that was born out of necessity. Our team was working with cities across the U.S. and around the world, designing major infrastructure projects, and we realized that a number of the projects with things like flood protection and coastal protection were actually creating insurance benefits. They were making it less likely that property owners and insurers would lose their money in the event of a major disaster. But, there was no real way to capture that value.
Think of it a little bit like calling your insurance company and saying, “I’ve joined a gym, so you should give me a discount on my health insurance.” There isn’t the data there to make the connection to what the value of prevention is. We found that we needed to develop a financial instrument to capture that value, so that is how we were pushed into creating resilience bonds.
Dominique: When you say, “We,” are you talking about Re:Bound and re:focus, correct?
Vajjhala: I am. I am talking about my colleague Jamie Rhodes.
Dominique: How did Swiss Re, Rockefeller, and RMS all come in?
Vajjhala: With re:focus we have an interesting business model: We work directly with cities, with government agencies across the world, and we work with large companies as well as on different types of projects.
And when we hit a problem like what I described—infrastructure creating benefits that can’t be captured—we generally approach a foundation and propose a plain-sight, philanthropic, charitable purpose initiative. And we pull partners from our public sector partner list and our private sector partner list.
So we put together the RE:bound program to solve this problem in a way that it didn’t disappear into any single insurance company but it became a solution in the public interest.
Dominique: Okay, so what is a resilience bond?
Vajjhala: A resilience bond is a variation on a catastrophe bond, which incorporates an infrastructure project, or any type of solution designed to reduce the risk that property owners face. And that property owner can be a homeowner, and it can be a large industrial port facility.
What’s nice about catastrophe bonds is they are very transparent. You do all of this modeling to say, “If a hurricane of wind speed X hits my community, then here is how much I would get paid.” What that means is that you can use the models in reverse to say, “If I reduce my risk, here is how much less likely investors are to lose their money.”
And so that is the basis for everything we did on resilience bonds. We took this existing insurance instrument that has about $30 billion in the market and we created a tweak to help capture value for proactive and protective projects.
Dominique: I see that there has been huge interest in catastrophe bonds and some were even oversubscribed. What kind of an investor has an appetite for this? And can investors lose all of their if there is a trigger event?
Vajjhala: We have to answer this question a lot for risk-averse public sector officials who are like, “What madman buys this kind of gamble?!” And the answer is, “It is not irrational.” Catastrophe bonds are attractive investments and they are oversubscribed because people are looking to diversify from more conventional investments.
A hurricane is not correlated with the stock market. It gives investors a chance to put their money in different places and the pricing of the bonds is done in a way that they are well compensated for taking the risk of potentially losing their principal invested. That’s why the returns on catastrophe bonds are also better than safer investments in stocks and bonds, for example.
Dominique: Beyond the potential of higher returns for investors, what is the key benefit of a resilience bond?
Vajjhala: The biggest benefit of a resilience bond is it helps you do the Holy Grail of resilience finance, which is to capture an avoided loss. If you think about resilience, when you are more resilient to a disaster it means that success is something that hasn’t happened, right? A storm hit, but your community wasn’t flooded.
It doesn’t actually change the event itself, it changes your exposure to it. What a resilience bond enables you to do is capture the value of being better off. In the same way that if you quit smoking your life insurance rates go down; if your city builds coastal protection, you should see that same kind of benefit.
Dominique: Would you say that there is a myth about resilience bonds then, that they are bonds instead of insurance instruments?
Vajjhala: A resilience bond is absolutely an insurance contract. It is nothing more. This is true of catastrophe bonds also. Especially when you work with the public sector it is very easy to think that these are like municipal bonds where you are borrowing money or taking on a loan or absorbing debt—and that’s not the case at all.
This is an insurance contract where you pay a premium and you receive a policy and that policy will pay out under certain conditions. We have to regularly clarify and explain this because these are two completely different sectors of public finance.
Dominique: Please run through a hypothetical and simplified scenario for, let’s say, a seawall with the timing of payments and savings. I have seen re:bound’s charts and know that this is a big question to ask you to put this in layman’s terms. Where do the payment streams come from? Who are the beneficiaries? Who receives and pays the money? Who takes on the risk?
Vajjhala: I don’t think I have ever done this out loud. The best example is the hypothetical city resilience program we created for a city at risk. Let me try to do that out loud and see if I sound at all intelligible.
If you take a city that is trying to build a seawall, and let’s just pick round numbers, say it’s $100 million for a small city, and you know what the risk of flooding is from hurricane-related surge, then what you would do is model the risk to the city and all the asset holders protected by this wall, with protections in place and without. Then you do a simple difference between those two things to see what is the financial value in insurance terms of this new sea wall.
And the same way that you could pay for a new toll road by saying, “I expect to have a certain number of drivers over the next 10 years that each pay this toll amount”, and then you can actually securitize that value—which simply means to take a loan upfront and then pay it back as the tolls come in.
You can do the same thing with your sea wall by saying, ”Here are my anticipated insurance savings, over this 10-year period, or 20-year period”, and bring them forward to fund or partially finance this sea wall.
That approach for a $100 million sea wall can look like, for example, $5 to $10 million in savings a year, depending on how much protection you are creating. And you bring that forward to pay off your project over the life of the loan, or the project itself.
And in some cases where you are only improving protection by a little bit, you may not cover the whole cost of the whole project. But in cases that are really exposed to devastating potential disasters, where you can increase protection by a lot, you can create an enormous amount of value.
Dominique: Therefore none of this cost is passed on to taxpayers because it is coming from savings, correct?
Vajjhala: It’s interesting, yeah, so just as we talked about how a resilience bond is not a debt instrument and it’s not a municipal bond—it’s an insurance contract. That means it actually helps make better use of taxpayer dollars.
What taxpayers would have been paying for in insurance and disaster recovery, now they can balance between prevention and protection and get more value for each dollar. But it doesn’t come directly out of tax dollars, you are correct.
Dominique: And their current insurance policy stays in place while you are doing this and there is room to pay for both?
Vajjhala: Correct, you would never completely displace the need for insurance. You need to have a balance between physical and financial protection. In the same way you would never cancel your health insurance because you have a life insurance policy. So that balance will always remain. And the key is how you optimize between the two.
And the healthcare industry is a perfect analogy for this because they have been struggling for many decades with how to fund prevention. The most cost-effective healthcare is the disease that doesn’t happen. This is taking that same approach to capturing savings and capturing the avoided loss in a way that actually benefits taxpayers.
Dominique: Who typically sits at the table when you are putting together a resilience bond? Who are some of the players?
Vajjhala: It’s an amazing cast of characters. Typically there are 2 sides, which are the people who do big infrastructure projects in cities.This could be the public works department, capital planning, a water utility, a port, or a transit authority. And on the other side, you generally have the risk management team. This is the team that buys all the insurance, that manages everything, from worker’s comp and liability out to property insurance.
It’s pretty rare in cities that those 2 groups talk to each other. The risk managers don’t have a lot of visibility on projects, so they don’t actually know where their city is being made safer when they are going out for insurance coverage. On the other side, the project planners don’t know when they are creating insurance benefits.
A lot of our time in the early stages of doing this kind of insurance-linked finance is helping a city talk to itself or helping to pool together beneficiaries in areas where, for example, a city might be responsible for a sea wall, but it’s the utility or the transit authority that gets the most benefit.
Dominique: That’s interesting that they aren’t communicating. Can you elaborate on risk modeling, turning data into insights, and how these tie into resilience bonds?
Vajjhala: It is actually the core of everything we did on resilience bonds, and we were very fortunate to work with an amazing team at RMS out of London, Risk Management Solutions. They gave us an intense education in catastrophe modeling.
We were trying to answer one simple question, which is, “Can we model world scenarios with protection and without?” And that led to a series of other questions like, “What kinds of projects can be modeled well by the existing financial models and which ones are you less likely to see the benefits of?”
And the way that catastrophe models work—I will just describe one kind because these exist for all kinds of things that could keep you up at night: earthquakes, wildfires, hurricanes, you name it, floods, but hurricanes are sort of the easiest to wrap your head around—the way that the models work is that they are simulation-based.
They will effectively simulate 10,000 possible hurricanes hitting, for example, Florida. You will be able to see across the whole spectrum of possible hurricanes what the expected losses are in a given area.
So the models combine insurance industry data on assets, where people’s homes, powerlines, and ports are located and bring it together with this natural event simulation of hurricanes.
And what you get out of the models are different ways of slicing these expected losses and that’s your exposure for any given disaster.
What we were able to do was take a very basic approach and say, “Let’s run the models once in the baseline world where there is no protection, and then for a given area let’s look at the value of protections.
So if you take a slice of the Florida coastline and say, “We are now protecting this slice, run the models and show me where the water will come in and around this protected area”, then you can actually see from the model results what’s protected, what’s not, and for which types of events.
There are some things that you can build, like a sea wall, but it still won’t save you completely from those last few major hurricanes. So catastrophe models aggregate all of those possibilities into a single number, an average annual loss, and that’s what is used to price catastrophe bonds. And that’s what we were able to show—that if you build in protections you can actually bring down that average annual loss number and capture the difference.
Dominique: Is there anything that you want to add to that, that you think is important?
Vajjhala: The transparency of it is really, really important. The insurance industry can be a real black box for folks who may not have great data on exposure or their risks and this is especially true in smaller towns in the US and in developing countries.
You don’t have the same baseline that you do in a city like New York for example. Or even in Miami. And so what I like best about these models is they let public officials interrogate the information and say, “What is our exposure and why is this counted and why is this not?”
That helps us have a very frank conversation about the benefits of a project, but it also opens up an opportunity to speak about where could you enhance protection. So we found ourselves doing an unexpected set of projects, where we used catastrophe models in partnership with RMS, to help inform design standards to maximize protection.
Dominique: How has the program evolved since you first created it and what have you learned?
Vajjhala: We learned a ton and it’s been fascinating. My colleague Jamie and I put together, with the RE:bound team, the very first report on resilience bonds from the end of 2015 through the beginning of 2016.
So [it was] 3 years since we took what was really just a hypothesis and tested it with our partners at RMS and others. What we didn’t expect going in was how much this work would be a gateway to a much broader portfolio of simpler insurance-linked finance projects.
We, in effect, designed the most complicated mechanism for doing large-scale infrastructure projects, yet we found that most of our work has been much more basic where we can sit down with a public or private sector partner and look at their balance sheet and help their risk management side of the house better align with their capital planning side of the house.
And it has not required a resilience bond because oftentimes it can be an internal realignment of investment priorities.
It looks a lot more like a closed-loop set of decisions. For example, if you look at your life and health insurance, that analogy from earlier, and you decide to quit smoking, if you decide to go to the gym, which of these things creates value that you can capture and then fund the virtuous cycle of improvements and risk reduction?
So we have been basically busy rebalancing physical and financial risk with various partners, and we are working on resilience bonds with about half a dozen public sector clients and partners. We have been genuinely surprised.
Dominique: Maybe your next project is called RE:balance.
Vajjhala: [Laughter] I love that. I have not thought of that. I will happily credit you for that.
Dominique: Is there some kind of certification standard that resilience bonds follow even though they are insurance instruments?
Vajjhala: This goes back to the fact that it is an insurance contract, not a bond. Catastrophe bonds have junk status because investors can lose their money—that is the definition of what makes a junk bond.
And so the ratings don’t factor much into the decision-making on how to pursue this as an insurance mechanism. It’s much more common to think about it as an alternative to property insurance or as a complement to property insurance that might be more cost-effective.
But no, I don’t expect that you are going to see any catastrophe bond ratings out there. The trust that the ratings provided in the bond market at large is what the modeling provides in a cat bond market.
Dominique: Would you say the biggest issue that you run into is fear? Risk aversion? If not, what is it?
Vajhalla: I would say it’s just the right hand not talking to the lefthand. We don’t encounter a lot of fear in the markets that we are working on, and it’s largely because a lot of folks are really interested in solving this problem.
The biggest barrier is you need to get very, very different types of people talking to one another to agree to move forward on this kind of physical and financial risk management, or this kind of integrative risk management. And that is the biggest barrier.
You have people who speak completely different languages, with very different priorities, existing in their bubbles. And kind of breaking that down to show where the mutual value is on both sides. By saying, “We can make insurance more available and affordable, and you can get your projects funded. Come to this table and talk.” That’s the biggest barrier.
Dominique: What are the most frequent questions that you hear from officials about resilience bonds?
Vajjhala: You got the primary one, which is, “How will this affect my debt limit or credit rating?” And that’s the confusion about whether this is a municipal bond or an insurance contract. So our answer to that is this is an insurance contract that will not have an impact on your debt limit or credit rating, if you are a city, for example.
The other questions we commonly get are about what types of projects resilience bonds can be used for. There is a real set of misconceptions about how private money can support public priorities and that’s been fed in large part by private money seeking investment opportunities in projects.
We get a lot of curiosity about things like, “Can a resilience bond be used to fund my fire station? And the answer is, “Probably not because it is too small scale and you can’t attribute a risk reduction to that tiny project.”
We also get asked, “How can you use this to fund wetlands or mangroves?” The answer is, “Although people are doing really cool work on this, right now the models are not set up to do that kind of simple difference between a world with protection and a world without. So we answer a lot of curiosity questions about what’s a good fit and what’s not.
Dominique: Okay, now what are the frequent objections that you hear about resilience bonds and what are your answers to those objections?
Vajjhala: One is the measurability of the risk reduction and knowing that it will be implemented and implemented well, is a major issue. This is something that is true for any large infrastructure project.
You have to be able to show what benefits it will deliver in order to get it funded. That is something that we feel like there is a solid basis in most engineering to be able to do. It gets more complicated when you are dealing with green infrastructure and natural assets, which is why it is great that organizations like the Nature Conservancy are building up the data and the investment case for those kinds of assets as well, but it is still going to be a little while.
The other kinds of objections are, for example, when you are protecting major assets from a disaster. Some people think the assets shouldn’t be there in the first place if they are exposed to these risks. That is a harder conversation to have. It’s a conversation about retreat and adaptation at large. It’s one that you need to be careful about because you need to manage transitions well.
You can’t simply go seamlessly from the world as it is now to the world as you wish it would be. We sense that resilience bonds can hopefully provide a physical bridge and financial protection while communities work out these bigger transitions. That can be along coastlines, along fault lines, and flood plains. This is a national and a global conversation that is going to take decades to play out—and we shouldn’t leave ourselves increasingly vulnerable while that happens.
Dominique: I see the benefit to taxpayers if projects get funded, and I see that they aren’t paying for this directly. What would be the downside, if any, or is there a way that this can disproportionately impact communities that can’t afford the cost? I am asking as a social justice issue.
Vajjhala: I appreciate that question a great deal. We spend a lot of time thinking about this because we try to be very aware of how any work we do can be used to further inequalities or inequities, and our sense of resilience bonds is they don’t create inequity directly, but they are built on a system that is unequal.
Not everybody has insurance and it’s the poorest who have the least protection. The insurance industry is better at protecting assets than people. If you look, for example, at a comparison of insurance costs from a tsunami in Indonesia versus a hurricane in Florida, the numbers are wildly different.
They are orders of magnitude higher for major real estate development. That is something that is fundamentally broken in how we think about valuing lives and valuing protection. So resilience bonds are built on the backbone of the data that supports the insurance industry as it is now.
That is something that we are very mindful of and very clear that if you are doing projects that create protections and that protection value is based on the assets, then you are more likely to protect high-value assets than the most vulnerable people.
And so in places where that is the case we have the conversation directly with the public interest party that we are working with. We will say that it is all good and well to protect this opera house or this downtown area, but what about XYZ, can this help you extend your capital to cover this extra cost or do a project phase that extends protections to the most vulnerable?
The other thing that we have been really excited about doing is working with organizations like DFID (the UK Foreign, Commonwealth & Development Office) and the World Bank to essentially do another level of resilience bond development to figure out how we can serve communities and countries that are at risk in places like the Caribbean where there aren’t great underlying insurance data.
It is also very much about livelihoods in places like Dominica where there isn’t a huge asset base—it’s the entire economy that’s at risk. It’s a really high priority for us to make sure that we use these things for good and not evil and to set a precedent for that in all of our work.
Dominique: You touched on transparency. What kind of reporting and independent verification is done?
Vajjhala: It’s a good question. Resilience bonds are relatively new. They have gone from an idea to a validated instrument; we still have yet to do the first instance. As far as we know, there isn’t an enormous amount of validation verification of the catastrophe bond market that is symmetrical to what you would see in the green bond market, which has an established credentialing system.
But I think for resilience bonds, what we expect, based upon the insurance-linked finance work that we have been doing over the past few years is that the modeling validates a reduced risk that is the basis for communicating the value of the bond itself.
So catastrophe modeling is the most important piece of this work. And I think that it is going to have to evolve over the next few years and get to the first issuance of resilience bonds and any variations that emerge.
But I am not sure yet and I don’t have an intuition for what that system should look like on what a good resilience bond is versus one that doesn’t create as much value as it could. Or one that leaves the most vulnerable behind.
Dominique: Do you think that people are waiting to see if others do these bonds first? Or proof of concept is already there with New York catastrophe bonds and that is enough?
Vajjhala: It’s interesting, I think that cities, utilities, and public sector issuers of bonds, generally always prefer a precedent, right? So the MTA and Amtrak are excellent precedents of public sector catastrophe bonds. We are seeing more and more public entities dip their toes into that market.
And so it’s not a wild leap from a catastrophe bond to a resilience bond. The extra effort is all about lining up the projects, doing the modeling, and understanding the value of the risk reduction. And that extends the timeline pretty dramatically. You go from doing, for example, a 3-month insurance policy design to now needing to design a sea wall which can take 3 years. And if you take that in mind, that is the reason why resilience bonds are slower and in some ways, more difficult than catastrophe bonds because they glue these two things together and they move at different speeds.
Dominique: If it’s, say, 3 years to get financing in place on some instruments, are government officials having proactive conversations before disasters hit?
Vajjhala: We have several active conversations and they are very much in that design space: What kinds of projects can we do? What’s the insurance value of them? How do we put these things together?
We like as much as possible to be able to say and show a community what the value of a resilience investment is because, otherwise, you find crazy things happening like the height of a sea wall being negotiated in a town meeting where it is completely disconnected from what the protection value is for the community at large, and it just narrows it down to a single issue.
We are finding much more interest and value in the process elements of resilience bonds than we ever expected. Like, can you use catastrophe models to inform design standards, or incorporate them into public-private partnerships? And that to me is incredibly encouraging that communities are not just waiting for the next disaster; that they are being proactive and also thinking about how to take incremental steps towards a full rebalancing.
Dominique: Is there a place that people can go to learn from other resilience bond projects?
Vajjhala: You can go to our website. We keep a collection of reports specifically geared toward public interest leaders, whether that is staff and local government or senior officials in government agencies, and we tend to pool resources through our social media of other people working in the field to highlight reports of interest or types of analysis that may be useful.
But this is still a very nascent field of work and there isn’t a huge amount out there. There aren’t a large number of firms working on this. We are part of a small ragtag band of firms that are trying to square the circle and invest in prevention and protection, the same way that we invest in doing disaster recovery.
Dominique: Thank you for your time today.
