Shifting Gears: The Forces Driving Insurance Risk in Commercial Real Estate

What if the insurance program you thought was protecting your deal is actually what's putting it at risk? In this episode of our 4 Real podcast, hosts Jon Gaynor, Sam Gilbert and Kate Mylod sit down with Emily Rasmussen, President and COO of Resilience Insurance Analytics, to break down what lenders, borrowers and attorneys need to know about property insurance in commercial real estate today. The hosts also explore the Basel III endgame re-proposal, and take a turn down memory lane to discuss childhood vehicles and names of their cars.

Transcript

Jon Gaynor: 0:10
Hello, and welcome back to the Dechert 4 Real podcast, where we discuss current issues and trends in commercial real estate finance. We aim to bring market commentary about developments, updates you can use, and maybe a little bit of banter along the way. I'm Jon Gaynor, a partner based in Dechert's Philadelphia office.

 Sam Gilbert: 0:23
I'm Sam Gilbert, a partner based up in the Boston office.

 Kate Mylod: 0:27
And I'm Kate Mylod, a partner based in the New York office.

 Jon Gaynor: 0:30
In this episode, Sam, Kate and I are joined by Emily Rasmussen from Resilience. We're going to have a broad conversation covering Emily's career and the issues shaping insurance and commercial real estate finance today.

 Sam Gilbert: 0:41
But before that, in our 411 segment, we'll talk about the Basel III Endgame Re-proposal and what the new capital rules mean for CRE lending and securitization.

 Kate Mylod: 0:50
But first, let's get for real with the hosts. So I have an interesting question for everybody, and it has to do with whether or not you name your car. Apparently, 40 to 50% of the general population name their cars for all sorts of reasons. One, nostalgia. Two, it's just fun. So I'm curious if anybody in this group names their cars. Sam?

 Sam Gilbert: 1:16
So I think I'm in the 50 to 60% that don't. Although my daughter has a name for my car, I have a Rivian, and she's come up with a very unoriginal name of "Rivi" for the car. But I don't think I've named a car previously, so I'm in that 50 to 60% majority.

 Jon Gaynor: 1:35
Well, an electric car in the current environment must feel really smart. So good for you.

 Sam Gilbert: 1:43
Well, my wife has a Yukon, so we offset.

 Jon Gaynor: 1:45
Oh, you balance. On average, you're not doing so well. Sorry to hear. We just got a hybrid, actually, and it has this, like moon roof. And so my wife has taken to calling it "Celine" because of like the color and like the moonish element to it. So it's her car, not mine. I wouldn't name my...well, I wouldn't name my car Celine, but she she named the car Celine. So that's mine. What about you, Kate?

 Kate Mylod: 2:11
You might want to make sure your wife doesn't listen to the first portion of our segment here, Jon. But unlike Sam, I come from the 40 to 50% who do name their cars. My family always named our cars growing up. Now I can't remember any of those names, but I will remember the name of my first car, a 1986 brown two-door Toyota Celica named Paul, but she was a girl. Don't ask me how I knew that. And then my current car, which is in the SUV family, is Lulu, and they actually, at the car dealership when I take her in or wherever I go to for her servicing, she's in the file as Lulu.

 Jon Gaynor: 2:46
Are your cars all usually ladies? It sounds like they're like ships in that sense, I guess.

 Kate Mylod: 2:51
No. Like, like my husband doesn't really name his car. But most of them have had genders, but the two that I can remember have been women that have been in my side of the garage.

 Jon Gaynor: 3:03
All right, I think that counts as getting 4 Real. Let's move on to our 411 segment. On March 19, the Fed, the OCC and the FDIC jointly issued proposed rules that would overhaul bank capital requirements. This is the long awaited do-over of the original Basel III Endgame Proposal from July 2023, which would have raised Common Equity Tier 1 capital for the largest banks by roughly 16 to 19%. That proposal generated enormous opposition — over 97% of the comment letters raised significant concerns — and it was never finalized. Michael Barr, who was then Vice Chair for Supervision, and the architect of that proposal, stepped down from the supervision role in early 2025. Michelle Bowman replaced him, and she had actually dissented from the original 2023 proposal. So the re-proposal reflects a very different perspective. The Fed voted six to one in favor, with Barr now the sole dissenter, and the direction in the new proposal is completely reversed. Instead of a capital increase, the agencies estimate that these proposals would actually decrease the aggregate CET 1 requirements, that's the Common Equity Tier 1 requirements, for the largest banks by somewhere in the range of 2-5% and that's a remarkable swing. Comments on the re-proposal are due by June 18.

 Sam Gilbert: 4:30
On CRE lending, the proposals introduce risk weights that are tied to LTV values and whether the property cash flows the debt. You know, that's a genuine improvement over the flat charge approach under the current rules. But there's an odd result embedded in the calibration. A CRE loan, where debt service depends on property cash flows, gets 110% risk weight when the LTV exceeds 80%. And compare that to an unsecured corporate loan which maxes out at 100% or goes as low as 65% of the borrower's investment grade. So you can end up in a situation where a fully secured real estate loan carries a higher capital charge than an unsecured corporate credit. That's the kind of outcome that's going to generate those comment letters.

 Kate Mylod: 5:10
And for anyone in the subordinate debt business, there's a definitional problem to qualify for the proposal's favorable CRE risk weights. A bank needs a first priority security interest in completed real property. That's a direct lien on the dirt. Mezzanine loans don't satisfy that test. They're secured by equity interests in the borrower SPE, but not by a mortgage on the property itself. So they get categorized as "other real estate exposures" at 150% even though, economically, a mezz loan is performing the same function as the first lien position in the capital stack. This issue was raised in comments on the 2023 proposal, and the re-proposal brings it back unchanged. We think this one deserves real attention in this comment cycle.

 Jon Gaynor: 5:58
So on securitization, and I want to be specific about what I'm talking about here. First is the capital treatment a bank gets when it uses a securitization structure to transfer credit risk off its balance sheet. And the second is when a bank achieves securitization capital treatment through its warehouse and repurchase facility lending. The biggest item is the P factor. So the P factor is the calibration parameter in the formula that regulators use to calculate capital charges on securitization exposures. It functions as kind of a flat surcharge on the entire framework. The 2023 proposal would have doubled this P factor from 0.5 to 1.0, which would have significantly increased the amount of first loss credit enhancement a bank needs to achieve a meaningful risk-weighted asset relief from a securitization. In some cases, it would have made the trades uneconomical. The re-proposal keeps the P factor at 0.5, and that is a direct result of industry advocacy. CREFC, SFA and other groups flagged this in their 2024 comment letters, and the agencies listened. The risk weight floor also drops from 20% to 15%, and there's a new look through option that lets a bank cap its exposure risk at the weighted average of the underlying pool. Both of those improve the capital efficiency of balance sheet securitization transactions, or, you know, the normal day-to-day warehouse lending that we often see in commercial real estate finance.

 Kate Mylod: 7:22
On the concern side, and there are real concerns, the proposals add a sole dependence test, which could impact recourse guarantees. A transaction only qualifies for securitization capital treatment if the performance of the securitization exposures depends solely on the underlying pool. If there's a limited payment guarantee from a sponsor or originator, and you often see a 25% payment guarantee and repurchase facility specifically, the transaction may not qualify. That's a real problem for mortgage warehouse facilities and loan-on-loan CRE financing structures.

 Sam Gilbert: 7:55
And the proposals also expand the definition of commitment to capture uncommitted lines, which is how most CRE warehouse facilities are actually structured. And the credit conversion factor for these uncommitted facilities goes from zero to 10% for committed facilities. It's a flat 40% regardless of maturity, replacing the current approach that gave short term commitments a lower factor. The net effect here could be a meaningful cost increase for the warehouse lending business.

 Jon Gaynor: 8:22
Bottom line, this is a materially better proposal than what we saw in 2023 for banks, at least, and the overall direction seems welcome, but the mezz treatment, the sole dependence test and the commitment definition changes are substantive issues that are going to be addressed in comment letters, I'm sure. The deadline, again, is June 18, and we expect heavy engagement from the trade organizations like CREFC, the SFA and the MBA. And if these issues touch your business, now is the time to get involved.

 Sam Gilbert: 8:52
All right. Well, now on to today's guest. Emily Rasmussen is the president and COO of Resilience Insurance Analytics, which is the leading insurance risk advisor to the nation's top investment banks, equity funds and commercial lenders. Under her leadership, the firm's team of 130+ professionals reviews over 5,000 commercial real estate transactions each year for 33 of the top 35 U.S. CRE lenders. Emily was a 2024 CREFC "20 under 40" honoree — that's hard to say — and is a frequent speaker at MBA and at the CREFC conferences on topics ranging from the property insurance crisis to alternative risk transfer. Emily, welcome to the 4 Real podcast.

 Emily Rasmussen: 9:29
Hi guys.

 Kate Mylod: 9:30
Welcome, Emily. And it's our tradition around here to get 4 Real with our guests as well. So we have to ask you, do you name your car?

 Emily Rasmussen: 9:39
You know, I don't know that I have ever named a car personally, but I can attest that my current vehicle was named by my team, against my will. Really it was my own mistake, so I'll take accountability for that. But I really, really wanted a white car, and I wanted it so bad that I didn't ask about the interior color. Like, I was, like, just give me the white car, and I ended up with a white car that has this, like, a really weird, like, orangey camel interior, and so it looks like peanut butter and marshmallow, essentially. So my team calls it the Fluffernutter.

 Sam Gilbert: 10:19
Okay, I'm gonna have to come up with a name for my car now I think. I'm jealous. I'll find one for the next episode.

 Jon Gaynor: 10:25
I think Rivi is good. So Emily, it's good to have you with us on the podcast. Can you take us back to the beginning? What's your background? How did you end up at the intersection of insurance and commercial real estate analytics? Was this a path you planned, or did it find you?

 Emily Rasmussen: 10:40
So ever since I was a little girl, I wanted to be an insurance consultant. I'm just kidding. No, obviously, I think it found me. Once I graduated from law school, I came to New York, and I was doing securities compliance for Academy Securities, which, after that, took me to a commercial real estate fund. And I had never done commercial real estate at that point. I was, you know, coming from Philly, so I had, you know, very much a finance and securities background, but I took like, four property classes in law school. That was the, you know, that was the extent of it. So I had a really steep learning curve. I was really, really lucky at the fund where I landed that I had tons of really good mentors. And it was also, you know, definitely, like a lean fund, like, take what you can grab kind of a thing. And so I kind of learned on my feet. I joined the in-house counsel team. I was their, our closer at the CMBS Conduit Fund, and I stayed there for five years. I learned everything that I needed to know about the front end of things. And what I discovered was, especially because of the character of the work that we were doing, we were taking on, you know, boutique stuff, which really is code for hairy deals, right? And a lot of times that means that there's an insurance problem, or at least there's a property level problem, right? And so we're looking to insurance to see if there's a solution available. Costs are usually tight, you know. So there was a lot of trying to understand how underwriting, legal and insurance interacted to cover the risk, and over time, you know, as somebody who was really within the transaction as a facilitator, you know, to just find whatever solution was the right solution, I found myself leaning heavily into insurance, because no one else really seemed to lean in that direction. And I think, you know, maybe that's for good reasons. That's where I really kind of found a home, I think, was being that leg of the of the tripod. And so when I, when we wound down the fund, I ended up, you know, moving to the insurance consulting space. I've been there for the last 10 years, and really, you know, my goal now is to be that same person that I was then today, just on a larger scale.

 Jon Gaynor: 12:51
I think that's so cool that you have experience doing kind of the closer role. Because I think it really helps you in your role now, because you speak the language of the people originating the loans, and can kind of bridge gaps between the insurance constituency and the actual like, you know, lenders and borrowers out there. So that's, that's such a cool background.

 Emily Rasmussen: 13:13
Well, yeah, I mean, I think that's really kind of what our fundamental offering is at Resilience. We are not insurance brokers. We're not risk managers. Obviously, we have a lot of brokers and risk managers on our staff, just like we have lawyers. We've got engineers former underwriters as well. But our goal is to understand insurance as it relates to commercial real estate lending, and we also do some corporate finance loans, and certainly on the surfacing side as well for CRE so that's a fundamentally different position than somebody who is selling insurance or someone who is evaluating risk on behalf of an insured. I think often the frustration that people feel with insurance is that they have this feeling of going to the mechanic and somebody's telling them, "You need this, this and this," and they have no context as to whether that's actually true, and no ability to know you know whether this person is giving them good advice. All they know is that it's expensive, but they're being told that the thing might fall apart if they don't do it. And so our job is really to be that person who's like, "Let me explain to you what's going on here," and to translate the language between what you're hearing on the brokerage side or what you're hearing from your borrower and what you're hearing from the rating agency, or, you know, your underwriting requirements, et cetera.

 Sam Gilbert: 14:29
That sounds kind of like the elevator pitch for Resilience, right? And I know we work with you guys on basically every deal we do, and see that that role you have, kind of as a team behind the curtain advising the lenders. Is that what's sort of the day to day?

 Emily Rasmussen: 14:43
Well, yeah. I mean, I love that you said that, because that's the goal, right? I mean, we want to be seen as little as possible in the transaction. I think the less people see of insurance and the deal, the happier they are. We have no ego about that. We understand that that's our mission, right? And I think coming from the closing side, that helped me to understand, you know, this is really supposed to be something that's a backstop for what people consider to be more primary objectives which lead into the underwriting. And you know, how we think about, you know, the credit quality of the deal. And this is really just meant to, sort of, to make sure that stays true. And so I think whenever you are operating in the background like that, it becomes particularly frustrating if people feel like you're taking center stage. And so we really try to tailor our process in the way that we communicate with everybody, to say, listen, we're trying to be the least intrusive version of ourselves here. We might need some information from you, But, you know, we'll try to just roll with this.

 Kate Mylod: 15:42
Well, thanks for that overview and origin story, Emily, it certainly sets the stage. Let's shift and talk specifically about the insurance market and commercial real estate, right? So over the last few years, we've all seen property insurance premiums spike dramatically. Our clients, our lending clients have seen that impact their underwriting of certain deals and certain geographical areas have have been more prone to these types of spikes than others. From where you sit at this point in time, do you see any of these markets stabilizing? Or do you think we're still going to be in this hard market for a while, and if so, what's continuing to drive that?

 Emily Rasmussen: 16:20
I wish I could say that I think everything in insurance is going to get easier. Unfortunately, that doesn't always stay true, even when it becomes true, so I want to caveat this by saying I have some good news here, but I don't want to fool anyone into thinking that this isn't a cyclical process, right? So I think that there's absolutely a stabilization going on in a lot of geographies within the United States, obviously, areas that are more prone to cat risk are going to remain less stable. But I think that that looks a little different than how people are accustomed to seeing it.

 Jon Gaynor: 16:53
Just because my mom listens to the podcast, when you say cat risk, what do you mean?

 Emily Rasmussen: 16:58
Not the cute kind. Catastrophic risk, right?

 Jon Gaynor: 17:03
Very, not the cute kind.

 Emily Rasmussen: 17:04
Yeah, very not the cute kind. So that's named windstorm, at times convective. So that's tornado, flood, wildfire and earthquake.

 Jon Gaynor: 17:12
Gotcha. So this might help listeners understand also, kind of like what you do. But you know, we understand that the reinsurance market has been a major driver of primary insurance market conditions. Can you explain that market from like a 50,000 foot perspective for our listeners, and give a sense of where the reinsurance market sits today?

 Emily Rasmussen: 17:32
Yeah, of course. So I mean, the easiest way to think about reinsurance is that they are the insurer to insurers, right? So anytime that you can make something more complicated in insurance, you absolutely do it. So you know that's, that's the first piece here. But also, you know, if you can buy insurance on your own insurance, again, very insurance-coded. So essentially, reinsurers operate as a backstop to make sure that carriers don't get hit with single events that make them go insolvent, right? So this happens through contracts, reinsurance contracts, where they agree to back, usually in a layered program across reinsurers, a certain percentage of huge loss. Sometimes that might be 100%, sometimes it might be some portion of that. And that's really where the changes in stability live, right? So a reinsurer might be willing to back 100% of a particular loss that they view as low risk, or alternatively, something that they can predict very easily, for something that is unpredictable, whether that's because it doesn't occur very often, like an earthquake, or something that has a lot of volatility, like a named windstorm/hurricane, they're typically less likely to cover 100%. So what's going to happen is you're going to get into one of these layered programs, but also the rates for those contracts are higher. And so what this does is it drives the primary insurance carrier to layer out those programs. And Sam, you've seen some of this, like really get complex in some of the programs we've seen recently. So what happens is what used to be a traditional all-risk package where everything is just falling as a sub-peril underneath the classic all-risk program. Now we're really parceling these out into secondary perils, and they have their own sub limit because they're paying a different rate. Okay, so and those sort of create problems for us on the transaction side, and they tend to be kind of buried in the details. And what that means, too, is that borrowers often don't know that they exist. And anytime that a borrower is not aware of what's in their program, and then you tell them that this is there, they love it. I just want you to know, like, they absolutely love that. So, so you know, it does, it does create some agita for, you know, the deals that we work on, but it is really important to look at, because it does fundamentally change what coverage is there.

 Sam Gilbert: 19:50
Yeah, I think the borrowers and the lenders, right? They want the answer to be, "do we have insurance?" "Yes," right? Like, that's all they want to hear. Going back to the sort of team behind the curtain thing. And it doesn't cost too much. Right. Thank you. "What does it cost? Do we have it? Can we check the box?"

 Kate Mylod: 20:04
So, Emily, that's really interesting about reinsurance, and the fact that so much of it is is behind the scenes. But the types of catastrophic events that reinsurance would be, you know, backstopping the insurance for are obviously, you know, huge, invisible. Can you help educate us about, like, what's an example of that where reinsurance would really come into play?

 Emily Rasmussen: 20:24
Sure. So, I mean, let's talk about, for example, hurricanes, right, named wind storm. We're doing a deal in Texas, Florida, somewhere on the Eastern Seaboard, that's going to come into play. And I think one of the most problematic things here transaction-wise is that you can look at one property that is very close by to another property, maybe a couple blocks away, and you are seeing a certain insurance rate being offered and a certain amount of coverage being available. And then you're seeing something totally different on a very nearby property. And obviously that creates a lot of frustration for borrowers. It makes people feel like they're at the mechanic. Again, they don't understand why this is happening. And the reality is that when it comes to cat risk, the secondary perils, like storm surge as an example, play a huge role in what the reinsurer is willing to come in for, or at what rate. Because if you have a program where storm surge is carved out, the likelihood that a reinsurer is going to have to back this loss is much higher, so they're going to charge a higher rate, and that's going to flow through to your premium. Your primary insurer is not going to eat that for you. That's why they've carved out storm surge, right? Because they don't want to take on that risk, and they're trying to keep the premium lower. So once you increase the capacity, and they need to access reinsurance, now they've got to pass through, you know, that increased rate. So it's really about saying, okay, there are certain geographies where, as a reinsurer, I feel confident that I'm not going to have to participate in a volatile way, right? Like my scale is not $50,000 to $50 million, whereas here I have no idea, because the category of a named windstorm doesn't even impact the storm surge. It is literally just about how this storm is turning when it hits the coast, how fast it's moving, how much moisture is in the air. It's not really something other than by using stochastic modeling that we can predict, so it's very difficult for a reinsurer to do anything other than assume the worst.

 Sam Gilbert: 22:26
That's a great example, but maybe let's switch to something a little more cheery than named windstorm. We obviously can't do a podcast without talking about data centers, and I am sort of looking forward to the day when Kalshi and PolyMarket start taking mentioned bets on our podcast, so I can start to profit a little bit on this. But obviously, as you know, data center lending has become one of the hottest sectors in CRE over the last few years, and there's a lot going on with it. But from an insurance perspective, how different is a data center than, say, just a regular industrial property? Assume it's fundamentally different?

 Emily Rasmussen: 23:01
Yeah, it is. It's a lot different. And that has to do with really kind of two things, I mean, but it lies really in the concentration risk that exists, and the fact that that level of concentration risk and the dependent factors that a data center relies on to maintain operating is more similar to an infrastructure project than it is to like a true industrial. And so that puts in a lot more for us to think about on the insurance side. And I think the insurance market is not particularly equipped, at least not yet, for these deals, and in some cases, is not willing to be equipped for these deals, at least in a traditional sense. So they are, yes, fundamentally different from an insurance perspective.

 Jon Gaynor: 23:46
So let's dig in. I think one key difference is that insurance on data centers work around a maximum foreseeable loss versus replacement cost. Can you just walk us through what that actually means and the practical impact?

 Emily Rasmussen: 24:02
Sure. So replacement cost is essentially based on the idea that if anything happens to building, it doesn't matter what the peril is. What causes it? If the thing falls down, how much does it cost to rebuild it with materials and labor? Okay, so that's replacement cost value. When we're talking about maximum foreseeable loss, we're talking about how much it costs to replace the building if it falls down based on a certain thing, okay? And it also doesn't necessarily mean all the way down, which is what replacement cost value covers. So there are assumptions that are being made by an engineer that's preparing an MFL, that there are certain things that are plausible, that will plausibly happen, and a replacement cost value doesn't go there, right? It just assumes that you've taken this thing all the way down, raised and cleared, and you're rebuilding it so you already have, in that sense, some of. Assumptions that are being made that may or may not occur. I mean, they're being made in an educated basis by an engineer. But the key is, the reason why this exists as a report is because there has not been enough lost data in the market for us to use modeling here. And so you can't use marginal and swift, you can't use RMS or AIR to do stochastic modeling. So even this engineer is really working up really limited data, and they're just making mathematical assumptions about what may or may not happen. And what they're doing is they're looking at whatever you designate as what you believe to be the highest risk events, and this is usually fire or convective storm in this, in the case of data centers, because they tend to be in tornado prone areas, yeah, which is, you know, fun for us, because there's also all of these sort of fundamental disconnects when it comes to data centers, because data centers are being built in areas where they're being built to code, but code usually translates to EF one tornado. Okay, so it's being built to withstand an EF one tornado. That's to code. But obviously that's not the only type of tornado that exists. So then, how do you predict losses when you essentially know that the building is not going to withstand this but you don't know to what degree? You can't necessarily predict a tornado. Some of these data centers are so large that it might only hit a part of the building right or the campus. But essentially, what an MFL is doing is it's trying to take the most plausible intense losses, catastrophic losses, and estimate what it believes is the maximum loss that makes sense in that scenario. So you're trying to size what you would usually access replacement costs for, but it does obviously have limitations.

 Jon Gaynor: 26:43
Right. The thing that's so interesting to me about data centers is that so little of the value is in the building and the land, and so much of the value is in access to power the actual equipment inside the building. And how do you think about packaging those various sets of risks? Because the risk of business interruption could be substantial. You know, there was an outage in a data center which took the entire derivatives market down for basically a few hours, right? Like, how do you think about those sorts of things? And how are you helping people kind of structure?

 Emily Rasmussen: 27:19
Yeah, I mean, I think this is where I say we really kind of cross into the universe of infrastructure projects where it feels fundamentally different from a traditional BI structure on a CRE loan. Obviously, there's typically a lot of contracts written around replacement of the actual, you know, physical GPUs. A lot of places are now requiring their own power station to be involved. These are all good backstops, and certainly, depending on the size of the company that's using the data center, or if it's a dedicated center, they may have themselves as a company, certain standards that they've agreed to meet, you know, to their users. And so they're, you know, keeping parts on hand. They're, you know, they have planned ahead for outages. I mean, certainly, if you've got a warehouse next door and then a tornado comes through, that's not going to help you out that much. I don't want to discount the fact that there's a lot of innovation and interesting risk management tools that are being used by these data centers, especially, because a lot of them contain lithium batteries, and so there's a lot of like, really great science that's being done on fire prevention and all of that. But from a CRE underwriting perspective, the problem is that BI for insurance...

 Jon Gaynor: 28:34
Business interruption. But keep going.

 Emily Rasmussen: 28:37
Yeah, business interruption, is only triggered when there's a physical cause of loss, right? Okay, so like, if you have a power outage, or if someone does, you know a cyber attack that does not qualify, right? And so if your power station goes down and you lose revenue, or maybe even you lose a whole contract on that basis, your business interruption coverage is not going to cover that loss. Something has to actually take down or damage the building.

 Sam Gilbert: 29:06
And are there programs being developed to cover that risk? I mean, you mentioned earlier, the MFL is sort of specific to certain causes. Are there different policies you can get that will cover the power outage, like we had it, or hacking?

 Emily Rasmussen: 29:21
I mean, I think the insurance market, whenever there's money to be made, a product will be developed. They're not there yet, okay, but it will be probably a special product. We've seen this before with, like, zoning protector. There was a lease protection coverage that used to be available - gap, you know, this type of stuff where the insurance market will create a special product to come in and cover the risk for. Today, it doesn't really exist because it's very difficult to underwrite something like that. And it definitely is sort of in the universe of it's like the poly market of insurance, is how I would describe it. You know, it's, it's not necessarily tied. Anything other than just, like, will this happen or won't this happen?

 Kate Mylod: 29:59
It's like, totally random, like, underwriting for something that could be totally random, right? So, Emily, you know, thinking more about data centers, right? We all know, you know, one of the major players in our industry being the rating agencies. They have strong views on asset classes such as data centers. What are you seeing in your experience with rating agencies getting comfortable with data centers and the insurance coverage on them?

 Emily Rasmussen: 30:26
That's a great question. So, you know, we haven't really seen a lot of stuff securitized on the data center, at least the data center as we see it today, right? I mean, there have been, like, some of these more traditional like just racks of servers, type of warehouses, where those have obviously closed and have been closing, but I think that they have fewer of the issues that we're talking about here that are more attributable to these sort of hyper scale, campus level data centers that we're seeing being built. So we really haven't seen a lot of this stuff come to fruition, at least in the securitization space, like, you know, cm... Probably none of this would fit in CMBS conduit anyway. So we're really talking SASB, and it's, you know, it's not going to be on the the multifamily side, of course. So I think most of this stuff has lived in construction land, or it's in bridge land. And now the rating agencies are coming to us, they're coming to engineering firms, and they're saying, "How can we possibly view this risk when we know that it's fundamentally insured differently?" It doesn't look like what we have seen on other CRE deals. And I think, I mean, I don't want to speak for the rating agencies, because I love my friends at the rating agencies, but I will say this, they have a big task in front of them, and they certainly recognize that, I think they want to get these deals done. Obviously they want to participate in rating them, but there are fundamental differences between what you're getting in a traditional insurance policy and what you're seeing on a data center. And I think if I were making a recommendation to any one of my lenders who is approaching a rating agency, I think being transparent about that difference and talking about how it's been treated from an underwriting standpoint is going to set them up for the most success in that situation, because this is one of those perfect examples. Like at the very beginning of our conversation, we talked about the three prongs, right? Legal, underwriting, insurance. Insurance is not the solution for everything. Sometimes it doesn't exist, and when it doesn't, you know, that's when it's up to legal or underwriting to bridge the gap. And so I think if you can show how you have approached that, what attempts you've made, what's possible, I think it gives them a lot better shot at sizing the difference.

Jon Gaynor: 32:33
And picking up off of something you said earlier, you can look at the tenant, you can look at their policies, their plans, like if they have a certain redundancy built in, where interruption risk is lower, then maybe a deficiency in the rating agency's view and coverage is less fatal, because, you know, you've got this high quality tenant with a lower risk of actual loss happening along the way.

Emily Rasmussen: 32:56
Yeah, absolutely. And I will say one of the most common things that we see on data center deals is rent abatement for tenants. And that's not great from this standpoint. So you know, if you're looking at a data center and you have any control over that, I would say, you know, to the extent that you can push for, you know, excluding rent abatement, or at least eliminate limiting it to a certain degree, that will improve the credit quality from a rating agency standpoint.
 
 

Jon Gaynor: 33:20
 Great tip.
 
 

Sam Gilbert: 33:22
And maybe, let's continue on with the agencies. I mean, all the major ones obviously have their own criteria that that they need satisfied for a SASB deal. How consistent are those requirements across the different agencies and and where do you see the most meaningful differences?

 Emily Rasmussen: 33:37
You know, I think directionally, the rating agencies are very similar. Certainly we see differences between the approaches that they might take when things are not what they're supposed to be, right? But I think if you look at the core guides that have been issued by the rating agencies, they all align very closely. So the differences are really kind of what happens when you can't meet those things. And even more than that, I would say, you know, just in our experience, it has much more to do with like, what do they prefer to see in those instances than it does about differences in how they view the credit quality. All of the rating agencies are interested in papering, you know, whatever deficiency they're observing. Some rating agencies are more focused on the credit underwriting to come in and say, well, because we're comfortable with the credit quality. We think that even if there is a loss, the credit quality is there. We'll get it figured out. We'll put this thing in special servicing. We'll restructure it, you know, whatever. And then others are much more focused on the actual bridging of the gap, like, put the brick in the slot in the wall, right? So I think anytime that you can show what you've done to fill that slot. Certainly, I think Dan, who was recently on your show, would agree that he loves to see that.

 Sam Gilbert: 34:57
Okay, interesting.

 Kate Mylod: 35:00
So and for the benefit of our listeners, that's Dan Rubock from Moody's, who has just referenced there.

 Emily Rasmussen: 35:05
One of the best.

 Kate Mylod: 35:07
One of the best, indeed.

 Jon Gaynor: 35:10
But if you could change one thing about how the rating agency insurance requirements were structured on these deals like something, you think it would make the framework genuinely more protective for lenders and investors. Rather than just a compliance exercise, what would it be?

 Emily Rasmussen: 35:25
This is going to be an unpopular answer.

 Jon Gaynor: 35:27
I love it. Hot takes. Let's do it.

 Emily Rasmussen: 35:30
I actually think that the way that the reporting is structured for insurance is not particularly protective, because we're talking here about how the insurance market has changed right from 2017 to 2020, to 2022, to 2026, all of that lives in the term of 110 year loan. Okay, so what we see is that, first of all, the reporting requirements are not particularly robust. They're typically one of the first things to go.

 Jon Gaynor: 36:03
So, as in, first thing that gets breached.

 Emily Rasmussen: 36:05
Well, first thing to get pushed back on.

 Jon Gaynor: 36:07
Oh okay. Got it.

 Emily Rasmussen: 36:09
In the docs, and then also, first thing to get breached, yeah, because there's this idea that, of course, goes back to like, insurance can't intrude on the loan, can't make the borrower mad, can't, you know, be a problem to facilitate. And so there's this, oh, there's such an administrative burden. And there's, oh, it's so difficult to provide this, you know? And it's, it's like, I think that if there was more consistency across enforcement of the reporting requirements, it would be easier for us to be more understanding on the front end, because if what you're saying is the market's bad today, and so I want to put forward a report that is insuring me at a market rate, which is a common, you know, insertion that people want to put in. It's very difficult to do that when you don't feel like you're going to get the benefit of that when the market improves. And so it's sort of like always just whatever the lowest amount of insurance coverage is, is what people are looking for. And so if we could be more dynamic and consistent in the reporting requirements for things, rather than forcing someone into special servicing before we actually ask them for their model that they're supposed to be submitting every single year, and also whenever they make a material change, we could probably be more comfortable facilitating exceptions on the front end.

 Kate Mylod: 37:29
So flood is one of the most misunderstood perils in commercial real estate, and I've actually been the victim of a flood so I can get how significant and trauma inducing it is, but a lot of borrowers in commercial real estate assume that their property policy just covers it, and in reality, it typically does not. Walk our listeners, Emily through how flood insurance is actually structured and what the difference is between the National Flood Insurance Program and the private flood insurance market.

 Emily Rasmussen: 38:00
Sure. So it is not typical for flood to be included within an all risk property policy, whether that's commercial or residential. NFIP is available to anyone, but in commercial real estate, it is required in circumstances when your property is located in a special flood hazard area. And that's important because, as the definition of flood has evolved from an insurance standpoint, a property doesn't necessarily have to be in a special flood hazard area to be involved in a flood, in a severe flood, even, which is, you know, something to be mindful of, especially when we're talking about flood combined with other catastrophic risk, but storm surge, yeah, like storm surge, exactly. But we'll kind of set that aside for a moment and just say so the NFIP is obviously a program that is funded by the government, and it does have a maximum available capacity for each building for commercial it's typically $500,000 sometimes $250,000 if we're talking about garden apartments, so it's not a lot, okay? And what is most common in commercial real estate is that NFIP is used as a deductible buy down. So essentially, you know, you can think about it like this, a deductible is a gatekeeping mechanism to access insurance proceeds, so you have to pay this much, you know, this from having health insurance and car insurance and everything else, right? So if you have a higher deductible, just like you know from health insurance, like your insurance, premiums are lower, if you have a private flood policy, you can decrease your premiums by increasing the deductible all the way up to $500,000 and then use an NFIP policy to buy down the deductible. So. So from an underwriting standpoint, in CRE we view that as like zero deductible. Or if you had $520,000 you know, then it would be $20,000 right? So this is very effective tool, especially for like your 20 to 80 borrowers, $80 million. So what that means, really, is that a lot of people are really worried about the NFIP going insolvent, and there's absolutely no doubt that that would be a huge problem for residential borrowers. It's not going to be a problem for commercial real estate borrowers if NFIP goes under, because there won't be flood insurance. There's already private flood insurance. Okay? It already exists. We require it on almost every large deal in such excess as lender may reasonably require, right? So this is very common. There's lots of different ways to calculate what's reasonable. But this market exists. The reg has actually been modified. I think it was in 2017 they modified the reg to accommodate and make sure that private flood uses the same definitions as NFIP. So this market exists already. It's ready to go. They are ready to step in. But the difference is the deductible. And so what that means is, if you no longer have access to NFIP, then as a lender, your decision is either to allow that $500,000 deductible, which is going to probably fundamentally go against your loan agreement, or it's going to be to require them to adhere to the deductible requirement, which could be anywhere from $20,000, $50,000 to 250,000 usually, sometimes it's $500,000 on the really big stuff, and that's going to change the cost by a lot. And so that's where this problem lies. It's not in whether the coverage exists, it's how much it costs and what decisions or trade offs you're willing to make as a lender.

 Jon Gaynor: 41:27
There are ways that we solve this. When we spoke in the last episode with Dan Rubock about terrorism insurance, you know, we had a cap built in. And if TRIA goes away ever, that's the Terrorism Risk Insurance Act, if TRIA goes away, it's capped at double the other insurance costs. And you know what it is. So if NFIP were ever to go away, it's a costs problem more than it is an availability problem.

 Emily Rasmussen: 41:55
Right, but of course, you know, I would say the difference between what is now TRIPRA, because it's reauthorized, right? Sorry.

 Jon Gaynor: 42:02
Ah right. Get that R in there too. Perfect, love it.

 Emily Rasmussen: 42:07

So the risk of a terrorism event is pretty remote. For 99.9% of CRE properties, the risk of flood is very high.

 Jon Gaynor: 42:14
Oh, yeah, yeah.

 Emily Rasmussen: 42:15
And so when we're talking about putting in place a cap, what we really mean is lowering the amount of coverage that we have. Because if you say the premiums can't exceed 200% of the original premium that was required at closing, then ultimately you are just going to reduce the amount of coverage that you would get. You're basing the coverage amount on the premium, not based on the actual risk. And so I think if you're putting in place a cap, the questions that you would have are, well, what does that mean for how much coverage I can actually access? Maybe it's a difference that you feel it like you can live with. Maybe you build something like a reserve or a sponsor guarantee or something to cover any potential shortfall, and that would be a solution there.

 Jon Gaynor: 42:59
Or you just have a guarantee or reserve to cover the deductible, or something like that, to keep the cost consistent, but address the lack of a deductible assistance in another way?

 Emily Rasmussen: 43:11
Right, and I will say that I think what is probably going to happen in this situation, if it occurs, or what would be a good solution, is a captive, right? So a captive is already commonly utilized for terrorism, and a captive is basically a private insurance company that you build for yourself. Okay? So it was, it was originally designed, I think, you know, at least, in my opinion, as a tax shelter. But basically it allows you to put your revenue in as insurance premiums into your own insurance company. You get evaluated by a state insurance agency to determine your solvency, and as long as they see you is in good standing, you can qualify for a CRE loan. Now, typically, you need to be reinsured with with, you know, paper on the back end, but that exists so you know, there are solutions available, certainly.

 Jon Gaynor: 44:01
Awesome.

 Sam Gilbert: 44:02
It all comes down to cost, it sounds like at the end of the day, right?

 Emily Rasmussen: 44:04
Well, it doesn't it always?

 Sam Gilbert: 44:05
Yes. I guess that's the ultimate question.

 Kate Mylod: 44:09
So the word modeling gets used in a lot of different ways in insurance and real estate finance. Emily, can you explain what catastrophic modeling actually is, sort of how it works at a basic level, and what it tries to tell us about properties or portfolios?

Emily Rasmussen: 44:29
Sure. So in the case of catastrophe modeling, we are talking about a Monte Carlo simulation. The. That is using available loss data from events that have occurred and reinterpreting them across, you know, hundreds of 1000s or millions of potential outcomes, and then using the curve of those outcomes to determine a probable maximum loss. And then those results are delivered based on a probability exceedance curve. So when we talk about the 100 501,000 10,000 year return period, that's what we're talking about. So I do think it's important to note, I think a lot of people sometimes misunderstand this about modeling. So if we say that this is a 1000 year return period. What we're not saying is that this will occur once in 1000 years, or even that this will exceed the amount once in 1000 years. We're saying it it has a point 1% chance of exceeding the PML in one year. Okay, so the return period is the exceedance threshold. So 1% is a 100 year return period. Point, 1% is 1,000.01% is 10,000 that's per year exceeding the PML chance. So if we're talking about, Are we okay with one return period or the other? We're talking about the difference between those two probabilities.

 Sam Gilbert: 45:54
People get confused when there's 100 year storm every year, and they wonder, what people are talking about, right?

 Emily Rasmussen: 45:59
Yeah, we had someone recently who was talking about the requirement for the 10,000 year, and they were like, this is like prehistoric times, like, this isn't gonna happen. I was like, there's a little bit of misunderstanding here.

 Sam Gilbert: 46:11
There are a handful of sort of dominant catastrophe modeling vendors. You got RMS and CoreLogic and others. How should lenders and their council be thinking about the differences between those platforms, and does it matter what model you're using? Not to have you plug one or the other.

 Emily Rasmussen: 46:27
Well, I mean, I think that there certainly are two main catastrophe modelers that we accept in commercial real estate. They are RMS and AIR. I think this is pretty well accepted core logic does great modeling for flood, wildfire, convective, but it's not necessarily the same style of modeling as what you're seeing from RMS and AIR, which is specifically developed for the insurance program, which is just a difference there. This has come up on deals before, like, can we use another modeling company? And it was difficult for us to answer that question for a while, and we really had to dig into understanding what it was that made RMS and AIR qualified versus some other vendor for modeling. But I think the reality is that it all comes down to scale. In any scenario where you are creating a model, the more data you have, the more access you have to better modeling results, and then the more scale you have in people working on the model and understanding how these different elements are changing these secondary characteristics, is what they're called, how they change outcomes, the better you will see the model perform. And performance is what has made RMS and AIR stand out. But both models are great, you know, and they really do perform very similarly, because ultimately, that's the point. But they have, you know, some different approaches, and there are obviously advantages to one or the other. Typically, that comes down to what your insurance brokerage is using.

 Jon Gaynor: 47:57
And I imagine picking a model the reinsurers have to be comfortable, because I imagine it's like one model that everybody is using to do the whole transaction, and then we're buying off slugs of risk inside of it, or whatever is happening.

 Emily Rasmussen: 48:11
Yeah. I mean, most of the time a reinsurer is going to be looking at the modeling on a portfolio level from the insurance carrier, right? So they're looking at a model the same way we would look at like a portfolio of assets. They're looking at it as a portfolio of insurance policies. So it's a little bit of a different layer, right? And they usually have their own internal models.

 Jon Gaynor: 48:33
They'll have a separate kind of evaluation of the risk. And it could be those models, it could be another.

 Emily Rasmussen: 48:38
Absolutely.

 Jon Gaynor: 48:38
Cool.

 Kate Mylod: 48:39
So being in the business of analytics, Emily, I bet that you're seeing in your space, artificial intelligence and machine learning start to play more and more of a role, right? We've heard about it all across our financial services industry. Specifically, what are you seeing in AI and insurance analytics and what is Resilience doing to take advantage of what AI has to offer?

 Emily Rasmussen: 49:07
Yeah. So, I mean, we're doing a lot of deals every year, right? And I think one of our strengths as a team has always been the ability to look across the market. You know, in every single type of CRE we are seeing outcomes. Outcomes, and we are constantly reinterpreting how we put forward our recommendations based on those outcomes. That's a great story for AI, right? So when you're talking about aggregating information and saying, Tell me what happened, that's a great use for AI, tell me what happened. Tell me why it happened, not a good use. And so I think for us, we are absolutely leveraging AI to help us organize information. So show me how many times this happened, or show me what characteristics existed that were commonly similar between these things like this is great for us, because we can say, well, usually, if these things are present, we have seen this happen, but it still requires you, one, to double check that and say, "Is this just correlation versus causation?" And then two, interpreting the why is not the job of AI, because it doesn't know why it can, but it will absolutely guess yes. And so I think AI gives us an opportunity to not only look at a larger set of conclusions and say, can you find trends that I'm I wasn't aware of? Can you tell me, you know, has this changed? Has this, what's the trend line over time? I can do it more quickly, right? But I think it is important for us to be careful in interpreting those conclusions, because this is the common example of a bar who says, but in this situation, this happened, and now I'm in this situation and I'm expecting the same outcome, because these things are alike, and it's like, well, but they're not alike in certain ways that matter, right? And obviously, you guys deal with that all the time. And so if you're using AI, you know, and I have seen other analyzes that have occurred of the insurance market generally, like, Oh, we're seeing, we're seeing overall, that premiums are decreasing for property insurance. That's true, but it's not true for everybody, right? And so if you're a par where who's in pre games have gone up, you're going, what the heck? But, like, that's not wrong for you. It's just that you fall outside the national trend, right? And so I do think it's important not to oversimplify.

Sam Gilbert: 51:41
Well, this has been awesome so far. I mean, we've talked about a lot of different things, data centers and windstorm and modeling, but if you could synthesize it and maybe get every commercial mortgage lender in the country to do one thing differently tomorrow, what would it be?

 Emily Rasmussen: 51:56
Gosh, that's a big question.

 Sam Gilbert: 51:57
Open ended question for you.

 Emily Rasmussen: 52:01
Don't be afraid of transparency. One of the things that is common for us as someone who's evaluating insurance risk is that borrowers tend to be afraid that the more information they give us, the worse the outcome is going to be for them. And I would say that about 95% of the time, that's not true, because the more information that we have, the more accurately we can understand their insurance program. Like insurance programs are different than lender requirements for insurance because your insurance broker or your risk manager has designed a program, at least in theory, that is based on what is best for your portfolio. That's not how lender insurance requirements work, but it doesn't mean that they can't work that way when it makes sense, but I have to have the information in order to make that call. And so if there's a disconnect between what's being required on the lender side and what's being offered. I need to have enough information to understand why that exists, and if I do, like, we can help you explain that. But I think people get really worried that we're asking them because we're trying to, like, trap them.

Jon Gaynor: 53:16
Yeah, or upsell them on premiums "Oh, looks like you're you need this additional rider. Sorry, but it's another ...

 Sam Gilbert: 53:22
The mechanic, and they get you, they sell you the brake pads that you don't need, right?

 Emily Rasmussen: 53:25
Yeah, which is why, you know, we aren't a broker, which is, you know, that helps us there, I think, to make people feel a little bit more comfortable. You know, we're not, we don't benefit at all. In fact, we generally do not benefit from telling people they need more insurance. But the reality is, insurance is a nuanced concept that is based on very specific characteristics, and lender insurance requirements are based on an aggregate concept that allows you to compare assets apples to apples, especially in securitization or just within your own internal underwriting. Those are fundamentally two different things, and so you can bridge the gap by giving us more information. If you don't give us more information, you're going to end up being the apple.

 Jon Gaynor: 54:14
Emily, really great to have you on the podcast. I feel like I learned a lot about insurance and about what you do. So thank you.

 Emily Rasmussen: 54:21
Thanks for having me.

 Jon Gaynor: 54:53
Our pleasure, and thank you for joining us for another episode of the Dechert 4 Real podcast. If you have any thoughts, please share them with us at our email inbox, realpodcast@dechert.com Also, if you like what you heard, give us a five star rating on whatever platform that you found us on. This episode was hosted by Sam Gilbert, Kate Mylod and me, Jon Gaynor. Stewart McQueen and Matt Armstrong produced it. Production support is by Kara Ray, Mallory Gorham, Alyssa Norton, Peggy Heffner, James Wortman and Jacob Kimmel. Our editor is Andy Robbins of Audio File Solutions. Thanks for listening, and we'll see you next time on the Dechert 4 Real podcast.

 Sam Gilbert: 55:00
All right, now, let's hear some quick dad jokes. We talked in the 4 Real about cars. So maybe if we have some car themed ones, can go with that.

 Emily Rasmussen: 55:09
Man. All right, let me access my my library here. I gotta sort for car jokes. Okay, why did the car go to therapy?

 Jon Gaynor: 55:18
Why?

 Emily Rasmussen: 55:19
It had too many breakdowns.

 Kate Mylod: 55:22
I actually do have a car joke at the at the ready here, and it's, how do you organize a car race?

 Jon Gaynor: 55:29
How do you?

 Kate Mylod: 55:31
Well, you start by steering the conversation.

 Sam Gilbert: 55:37
Well, mine's gonna double as a dad joke and a test to see if my mom is still listening. My mom, who is a librarian, so I'm gonna have a book themed joke. What is a car's favorite type of book?

 Jon Gaynor: 55:49
What is a car's favorite type of book?

 Sam Gilbert: 55:51
An autobiography.

 Emily Rasmussen: 55:53
Missed opportunity for Kelly.

 Sam Gilbert: 55:56
Yeah, it's true.

Jon Gaynor: 55:57
Okay, I'll do the last one, I guess. And this is to test to see if the kids are still listening. What do you get when dinosaurs crash their cars?

 Sam Gilbert: 56:06
What do you get?

 Jon Gaynor: 56:07
Tyrannosaurus Rex.

Sam Gilbert: 56:10
Four very bad ones. Good way to end.