A Conversation with Amy Gromowski
Natural disasters can have long-term effects on a community. As noted in our 2021 Hurricane Report, lower-income homeowners are left with few options when homes are destroyed. When affected homes are deemed unsuitable for habitation, displaced people may resort to living at motels and similar, creating added expenses. Natural catastrophes can result in financial hardship as people struggle with additional expenses.
In this episode, host Maiclaire Bolton Smith sits down with Amy Gromowski, senior leader of science and analytics at CoreLogic, to discuss financial implications as a result of natural disasters.
Maiclaire Bolton Smith:
Welcome back to Core Conversations, a CoreLogic podcast. I am your host Maiclaire Bolton Smith, and I’m the senior leader of research and content strategy with CoreLogic. In this podcast, we’ll have conversations with industry experts about key topics from housing affordabilities to the impacts of natural disasters on property. And today we’re really going to dive into that latter category, the impact of natural disasters on property. Natural disasters can really have some longitudinal effects on a community. As noted in our 2021 Hurricane Report, when homes are destroyed, homeowners, especially lower income ones, are left with few options. Damaged homes, often unsuitable for habitation, mean that people are displaced living in motel rooms and elsewhere spending money on newfound added expenses, money that they likely don’t have. Additionally, their place of work may be damaged, locking them out of earning an income.
They may have to pay some or all of the repairs out of pocket to restore their home. And in some really devastating situations, they may just abandon the community altogether, finding the cost and time to rebuild much too difficult. We saw a lot of this in the wake of the 2018 Camp Fire, which destroyed much of Paradise, California. Some people never returned. So in short, natural catastrophes can really result in financial catastrophes as people struggle to afford their loan, their insurance, their childcare, their groceries, rent, you name it. So to talk about these financial implications from natural hazards today, we welcome Amy Gromowski, senior leader of science and analytics. Amy, welcome to Core Conversations.
Amy Gromowski:
Hi, Maiclaire, thank you. Really, really excited to be here.
MBS:
All right. So to get us started today, why don’t you start by telling us a little bit about yourself and your role here at CoreLogic.
AG:
Sure, sounds good. So I grew up in the suburbs of Milwaukee, Wisconsin. I still live in that area today with my husband and my two kids. I went to Wisconsin state schools, which are great schools. I went to UW–La Crosse, which is on the Western side of the state, beautiful right on the Mississippi River. And there I studied math and psychology. Then I went to UWM, which very different, the other side of the state in downtown Milwaukee, where I studied statistics. So I started my career in insurance, both in health and then in property and casualty. I joined CoreLogic about eight years ago, and today I lead a team of data scientists who specialize in building AI, using machine learning for a lot of our different CoreLogic solutions.
AG:
So we build AI for mortgage underwriting solutions, for digital marketing and retention solutions, and for insurance underwriting and claims, and that’s around reconstruction cost, even run amount models. So this wide array of the different AI solutions that me and my team are building, it really allows me to have a pretty deep knowledge base on many of the key components all around the property ecosystem. So that just really affords me the opportunity to be able to be here today to talk about mortgage impairment, or said another way, the impact of natural disasters on the ability to pay mortgage.
MBS:
I love that. And that’s exactly why we’ve picked you to talk about this topic, Amy. So let’s just dive in. So we’ve talked about catastrophic events a fair bit on this podcast, and recently we discussed estimating losses. But what you do is something different, it looks at the impact of catastrophes on mortgage delinquencies. So before we get into this, can you just define for our listeners what we mean by mortgage delinquency?
AG:
Sure, I’d be happy to. So mortgage delinquency is essentially the borrower not paying their mortgage on time. So payment can be 30 days delinquent, 60 days delinquent 90 days, and even 120 plus, which then usually moves the loan into foreclosure. And foreclosure is when the lender attempts to recover the balance of the loan, so usually that’s by forcing the sale of the asset. It can be relatively common or it is relatively common to be 30 or 60 days late on a mortgage payment. But once a loan is 90 or more days delinquent, this is really considered a serious delinquency. It’s really difficult for a borrower to become and stay current on their loan typically after 90.
AG:
So 30 to 60 days is often not considered eye-opening to a lender or servicer. But once that 90 days hit, they’re starting to pay attention and be thinking about what kind of action may need to be taken. Things that are contributing factors to delinquency very commonly understood are things like the loan to value balance, your FICO score, even the interest rates spread can be common contributing factors that a lender or servicer would look at to understand the likelihood, I guess, of delinquency, 90 days serious delinquency occurring. There’s other factors as well, but those are common ones.
MBS:
No, that’s really helpful. And we’ll probably get into this a little bit more too as we continue. But I know right after a disaster happens, because we talked about loss estimates recently on one of our podcasts, there’s always this demand for information immediately after the event happens. But something like this, you can’t give immediately after an event happens because you need that time of at least 90 days to understand where these delinquencies are. So this is not the kind of data that we can provide immediately following a catastrophe. So I think that was an important piece to start on.
AG:
That’s right. As we talk a little bit more about delinquencies and some of the work that we’ve done to understand how events impact or affects delinquencies, I probably won’t get into it much, but we can look at predicting that behavior based on the event, the damage to the home, and what we know about the characteristics of the loan. So there is a way to look at it in a future perspective.
MBS:
Well, that’s good to know. And we’ll dive into some of this now. I guess when we look at those events that have happened and where we do have some good data and some pretty staggering data from Hurricane Harvey in 2017 about the effect on mortgage delinquency, can you just talk a little bit about what we saw in Harvey after Harvey specifically?
AG:
Sure. And Harvey is a great example because we did study it pretty significantly. So let me talk just a little bit about some of the characteristics of Harvey just for context. So Harvey was a category four hurricane that hit in Houston area. It was the second costliest, at least at that time, since 1900. It had wind gusts as high as 132 miles per hour, it had hurricane force winds that went out 40 miles. So huge swath of land that it touched and lot of properties impacted, and 29 tornadoes spawned off of it. I think the most impactful thing when we think about the mortgage delinquency and the reason why we could study this is the 50 inches of rainfall that ended up hitting, and a lot of that inland.
AG:
So CoreLogic had estimated total insured and uninsured loss from Harvey was between $36 billion and $56 billion. Now, some of that was covered by NFIP and some by private insurance, but we estimated $18 billion to $27 billion was uninsured. So that’s really important to understand as we think about mortgage impairment and the impact of these events on the delinquency or ability to pay. So we had seen, when it comes to the 90 day delinquency, an increase in 90 plus day delinquency in FEMA designated counties. So where FEMA came in post events and said, “These are all damaged areas due to Hurricane Harvey and therefore NFIP.” And it being a presidential declared major disaster area, government funding, will take care of it. We saw in those areas, where we estimated damage, so we’re able to look at the footprint of the storm and we’re able to use our reconstruction cost values and we’re able to look at… Well, with the footprint and reconstruction cost values, we can really estimate damage at a property level.
AG:
So where we estimated damage, we did see a 90 plus days delinquency rate increase of over 200%. So it went from 2% to 6%. But what’s really interesting, Maiclaire, is that’s a staggering number. But where we don’t estimate damage in those same major disaster areas and in those same counties, we still saw a delinquency increase, and again, we’re talking about 90 day plus, of 167%. So delinquency rate went from 2% to 5%. Now, why might that be? That’s really, we hypothesized, that it’s really about closing of schools, blocked routes to work, damaged places of employment. So people just not being able to work and their employers not being able to stay open for business. So that all causes a disruption in income. So damage is critical, but even in those areas where homes may have not been damaged, there was still an impact and effect on their ability to pay their mortgage.
MBS:
Yeah, that’s really helpful and a great place for us to start today. And Hurricane Harvey really did show these unbelievable statistics that you’ve just mentioned, Amy. I do want to do a quick acronym check, you mentioned the NFIP a few times, that’s the National Flood Insurance Program, just to make sure that we make that clear. The other thing that I question too is how common is this? Do we see this a lot or was Harvey really one of those examples where things were really bad and it doesn’t happen that often?
AG:
I’d say both. So we have looked at other events. We saw a 90 plus delinquency increase for Hurricane Irma which hit Florida just a few weeks after Harvey. We saw it for Hurricane Michael, which was also in Florida in October of 2018. Hurricane Florence as well we looked at, and that hit the Carolinas in ’18 as well. So our research does suggest that there’s some effect on serious delinquencies for all natural catastrophes, even wildfire we’ve looked at. But the magnitude of the economic impact is highly correlated with the magnitude of the events and it’s damaged.
AG:
So as I mentioned, Harvey was really a great case study to look at this because of how big it was and how much rainfall came inland. So it really, from a data scientist’s perspective, it created this ability for us to study in more of a Petri dish, if you will, a good study design. But just when we look at what we were able to understand from Harvey and try and look at the behaviors and apply some of those same learnings in these other catastrophes, we do see the same phenomenon, it’s just not quite at the same level. So it’s just highly correlated with the level of damage and the extent of the catastrophe event.
MBS:
So the severity of the event really does impact how extreme the delinquency is that we are seeing, that’s really what the data is showing.
AG:
That’s right.
MBS:
One other thought that just triggered as you were talking about this is the behavioral side of this. You alluded to this a little bit, Amy. And I wonder is there also… I know a lot of it is deemed people can’t work, they don’t have money to pay their mortgage, but is there also, well, I can’t live in my home because my home is totally destroyed, so why would I pay for it if I can’t be living in it? Has there been any research or any studies done on that that you know about?
AG:
I think it’s a good question and a good point. A lot of what we’ve studied is partial damage, if you think about flooding or your roof being ripped off and structural damage. I did mention wildfires, we see some of that at a higher geographic level, I guess, around the delinquencies. But as far as people completely walking away from their mortgage and not rebuilding… And the reason why I bring up wildfire is because that’s a complete loss typically, it’s the entire home.
MBS:
Right, yeah.
AG:
So it’s a different dynamic, and we haven’t studied it to that level of, for a complete loss of a property, what’s the likelihood of them walking away completely from their mortgage rather than going delinquent? We’ve really kept our studies at that 90 day plus. So what’s happening after that, foreclosure, did they actually rebuild, that’s further down the line and we really haven’t studied that.
MBS:
I just think it’s nice little anecdotal. Because my brain went there as you were talking, I’m sure other people might’ve had the same thought.
AG:
You gave us a new research projects.
MBS:
I did. I did. So this has all been talking about delinquencies following catastrophes, but there’s different types of catastrophic events. We’ve been talking about natural catastrophes like hurricanes, fires. But when I think of things like the pandemic most recently or also the 2008 financial crisis, the whole news story there was about mortgage delinquencies foreclosures. How does that compare to what we were talking about today specifically related to natural catastrophes?
AG:
It’s a really great question. And I do think you’re right, Maiclaire, there are a lot of similarities like loss of job or just general inability to do the job. But instead of it being a blocked route to work or my place of employment closing, although we have that as well, it might be more like elements of, I’m home with my child and I can’t go to work because I need to be in a lab or in a hospital or in a restaurant serving or cooking or whatever that might be. One of the key differences, at least as far as what we’ve studied with mortgage delinquency, is the level of damage on a property is really a significant factor in that impairment. So with a pandemic, we don’t have damage to property typically that you would associate with the pandemic. So therefore that’s definitely a major difference in these two types of events.
AG:
So if you think about damage in what we’ve studied, the reason why that is so significant is if you think about it from a borrower’s perspective, if I have to pay $3,000 mortgage every month and now I have damage to my property and I may be not able to even work, that’s… And we know loan to value is a significant driver. So the closer my loan is to the actual value of my house, or maybe I’m even upside down, that additional damage, the damage that occurred from the event just further puts me closer to that value amount or even further upside down. So now I maybe owe 110% of the value of my home and I don’t even have the value of my home anymore. So that is what’s really significant in mortgage impairment, is that whole experience of the financial impacts to the borrower and the homeowner with the damage and what I have to pay because it’s not typically covered by insurance, especially if I don’t have NFIP.
MBS:
No, that’s really helpful. So this data, it’s fascinating really. And I guess the question is, who uses it? You’ve talked a lot about it’s mortgage delinquencies, I’m assuming it’s intended for mortgage companies. How is this helpful for mortgage companies after the fact of an event has happened?
AG:
Lenders and servicers do not want their borrowers to defaults, they don’t want them to go delinquent. More than that, they also want to be there for their borrowers in their time of need. So lenders and servicers are using this data to help with the outreach in the damaged communities. Some of the things that they can be doing with this are connecting the borrower to relief programs more rapidly, offering loan modification or other assistance. We did see in the pandemic a moratorium, things like that. So it definitely extending and really helping them stay above water and incurrent on their loan even if they’re just putting in place a moratorium. They want to provide the assistance and prevent delinquency progression. It’s really a win-win, it’s a win for the borrower obviously, but it also reduces costs to lenders and servicers. Because there is a cost to them as their borrowers are going delinquent. Especially if they start moving into foreclosure, that’s a very costly process for the lenders and servicers.
MBS:
No, that’s great. And it leads me then to the thought of, how can lenders contend with the reality of that damaging events, whether they’re natural or otherwise, can cause a homeowner to fall into the stages of delinquency? How do we help homeowners to stay on their feet during these hard times? And how can lenders contend with this risk to their portfolios with something that’s inevitable because we can’t stop mother nature?
AG:
For lenders and servicers, I think understanding the risk is really key so that they can financially be planning for that type of risk and understand the financial impact to them and the financial impact to their borrowers. It allows them to do the things like we just mentioned. The analytics that we talked about today, the different products that we have that we talked about today, this can really help them to do that type of planning and just understand the risk on their books. Very similar to insurance, you have to know what the risk is. And lenders and servicers should be thinking about that as well, and they’re understanding that more and more just to be able to be there in the time of need and stay financially solvent around it.
AG:
I think for homeowners, there’s a critical understanding here as well. Being prepared, both financially, understanding their options around insurance, am I in a FEMA flood zone? Do I have NFIP available to me? If not, what are my private insurance options? And if private insurance isn’t available, how do they plan ahead for the possibility of an event and the level of damage to them? So that’s what I would say. I think it’s all about knowing and understanding risk, both for the homeowner, for lenders, servicers, insurance companies. There’s a whole ecosystem here that’s impacted when natural catastrophes occur and when it affects their ability to pay mortgage and mortgage impairment.
MBS:
One our tag phrases here at CoreLogic that we’ve mentioned many times on this podcast is, know your risk to accelerate your recovery. And this, while it being in a very different context, that still applies. So that takes me back to something we talked about right at the very beginning, and I think it’s probably a good place for us to wrap up today, is as lenders are trying to understand their risk, do we do any forecasting of delinquencies for potential catastrophic events, either before they’re about to happen or just long-term, if X were to happen, this potentially could be the impact?
AG:
Yeah, that’s definitely where I would love to head. We don’t have anything off the shelf today that would have simulations, if you will, of level of damage and what its impact would be on delinquency. But we definitely have all of the data and the forecasting. So the data around the property ecosystem and the loan characteristics, as well as all of the science around what could happen. And those projections, if I think about our catastrophe risk simulations and stochastic event sets, we could take that and look at, okay, if we have this type of event, it would be this level of damage and it would have this type of effect on ability to pay mortgage given these types of loan characteristics.
AG:
It is something that is very loan and properties specific. So it becomes a little bit more difficult as far as how you get that information in a meaningful way to, let’s say, lenders and servicers prior to an event. But it’s definitely something I would like to explore more because like you, with this question, I think there would be a lot of value. It’s just about figuring out how it is that that information can be used ahead of time to plan and then getting the work done.
MBS:
Yeah, definitely. I think there’s not a perfect solution for anything with anything these days. And I think that’s something to aspire for as we help to try and help mortgage companies and insurance companies and others in the whole property ecosystem on how to prepare for bad things to happen when natural disasters do occur. So Amy, this has been so great. Thank you for joining me today on Core Conversations, a CoreLogic podcast.
AG:
Thank you for having me. It’s really nice to be able to talk about this topic and share some of the research we’ve done. I believe strongly that understanding the risks, preparing for it really just helps everybody, especially homeowners, borrowers, and us as society really. Because natural catastrophes and these types of events, especially around hurricanes, do not seem to be going away. So it’s great to be here and talk about this topic. And thank you.
MBS:
Yeah, definitely not going away and knowledge is power and know your risk, accelerate your recovery, that’s our tag phrase for a reason. Well, for more information on the property market and the housing economy, please visit us at corelogic.com/intelligence.