Good day, and welcome to the Essent Second Quarter 2022 Earnings Conference Call. Please note, today's conference is being recorded. All lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question-and-answer session. If you would like to ask a question during this time, simply press star followed by the number one on your telephone keypad. If you would like to withdraw your question, press star followed by the number one again. Thank you. At this time, I would like to turn the conference over to Philip Stefano, Vice President of Investor Relations. Mr. Stefano, you may begin your conference.
Thank you, Erica. Good morning, everyone, and welcome to our call. Joining me today are Mark Casale, Chairman and CEO, and David Weinstock, Interim Chief Financial Officer. Also on hand for the Q&A portion of the call is Christopher Curran, President of Essent Guaranty. Our press release, which contains Essent's financial results for the second quarter of 2022 was issued earlier today, and is available on our website at essentgroup.com. Prior to getting started, I would like to remind participants that today's discussions are being recorded, and will include the use of forward-looking statements. These statements are based on current expectations, estimates, projections, and assumptions that are subject to risks and uncertainties, which may cause actual results to differ materially.
For a discussion of these risks and uncertainties, please review the cautionary language regarding forward-looking statements in today's press release, the risk factors included in our Form 10-K filed with the SEC on February 16, 2022, and any other reports and registration statements filed with the SEC, which are also available on our website. Now let me turn the call over to Mark.
Thanks, Phil, and good morning, everyone. Today, we released our quarterly financial results, which continue to reflect the favorable credit performance of our in-force portfolio. For the second quarter of 2022, we reported net income of $232 million, as compared to $160 million a year ago. Similar to last quarter, our results benefited from the release of COVID reserves associated with defaults from the second, and third quarters of 2020. On a diluted per share basis, we earned $2.16 for the second quarter compared to $1.42 a year ago, and our annualized return on average equity was 22%. From a macro standpoint, our long-term structural outlook for the housing market is positive despite near-term headwinds.
Rising rates in response to inflation and home price appreciation have pressured affordability, resulting in a slowdown in housing activity and mortgage production. However, the undersupply of housing, and a strong labor market should continue to support home prices and credit performance in the short term. Longer term, we believe that demographic trends are favorable and should continue to bolster housing demand. At June 30th, our insurance in force was $216 billion, a 6% increase compared to $204 billion a year ago. The credit quality of our insurance in force remains strong, with a weighted average FICO of 746 and a weighted average original LTV of 92%. Strong home price appreciation in recent years has enabled the accumulation of embedded home equity for a material portion of our book.
While home price growth will likely moderate going forward, this embedded value should mitigate the risk of future claims on our in-force book. Our 12-month persistency at June 30 was 73%, while the 3-month annualized persistency was 82%. The weighted average note rate of our book is in the mid-3% range, while 81% of our insurance in force is comprised of 2020 and later vintages. Our in-force portfolio remains well-positioned after the recent rise in rates, with higher rates translating to higher persistency. We continue to execute upon our diversified and programmatic reinsurance strategy. In the second quarter, we closed an excess of loss forward transaction to cover an additional 20% of our 2022 NIW. Combined with our 20% quota share transaction from the first quarter, 40% of our current year business is covered with forward reinsurance protection.
As of June 30, approximately 98% of our portfolio was reinsured. Our reinsurance entity, Essent Re, continues to write profitable GSE business and support our MGA clients. As of June 30, annual run rate revenues are approximately $60 million, while our third-party risk in force was nearly $2 billion. We remain pleased with Essent Re's performance and its contribution to the profitability of our franchise. Cash and investments as of June 30 were nearly $5 billion, and the investment yield for the second quarter of 2022 was 2.5%, up from 2.1% in the first quarter. The recent rise in rates is providing some tailwinds for our investment portfolio, as yields in the second quarter on new money approximated 4%.
We continue to operate from a position of strength, with $4.3 billion in GAAP equity, access to $2.6 billion in excess of loss reinsurance, and approximately $1 billion of available liquidity. With a trailing 12-month operating cash flow of $613 million, our franchise remains well-positioned from an earnings, cash flow, and balance sheet perspective. As of June 13, our book value per share was $39.67, an increase of 9% from $36.32 a year ago. We remain committed to managing capital for the long term, taking a measured approach to maintain strength in our balance sheet.
Given our financial performance during the second quarter, I'm pleased to announce that our board has approved a $0.01 per share increase in our dividend to $0.22. We continue to believe that dividends are a meaningful demonstration of the confidence we have in the stability of our earnings, and cash flow as a result of our buy, manage, and distribute operating model. Now let me turn the call over to Dave.
Thanks, Mark. Good morning, everyone. Let me review our results for the quarter in a little more detail. For the second quarter, we earned $2.16 per diluted share, compared to $2.52 last quarter and $1.42 in the second quarter a year ago. Net premium earned for the second quarter of 2022 was $212 million, and included $13.1 million of premiums earned by Essent Re on our third-party business. The net average premium rate for the U.S. mortgage insurance business in the second quarter was 38 basis points, a decrease of one basis point from the first quarter.
Net investment income increased $4.7 million, or 19% in the second quarter of 2022 compared to last quarter due to higher yields on new investments and floating rate securities resetting to higher rates. Other income in the second quarter was $1.6 million, which included a $5.5 million loss due to a decrease in the fair value of embedded derivatives in certain of our third-party reinsurance agreements. This compares to $7.2 million last quarter, which included a $4.4 million gain due to the increase in the fair value of these embedded derivatives.
The provision for losses and loss adjustment expense was a benefit of $76.2 million in the second quarter of 2022, compared to a benefit of $106.9 million in the first quarter and a provision of $9.7 million in the second quarter a year ago. As a reminder, last quarter, as defaults in the second and third quarters of 2020 continued to cure at elevated levels, we revised our estimate of the ultimate claim rate for these defaults from 7% to 4%. During the second quarter of 2022, these defaults continued to cure at elevated levels, and we revised our estimate of the ultimate claim rate for these defaults from 4% to 2%.
As a result, the provision for losses this quarter includes a benefit of $62.9 million from the second, and third quarter 2020 defaults. Other underwriting and operating expenses in the second quarter were $42 million, an increase of $1 million from the first quarter. The expense ratio was 20% this quarter, a slight increase from 19% in the first quarter of 2022 and the full year 2021. We estimate that other underwriting, and operating expenses will be approximately $170 million for the full year 2022. During the second quarter, Essent Group paid a cash dividend totaling $22.4 million to shareholders. As a reminder, Essent h as a credit facility with a committed capacity of $825 million.
Borrowings under the credit facility accrue interest at a floating rate tied to a short-term index. As of June 30, we had $425 million of term loan outstanding with a weighted average interest rate of 2.92%, up from 1.99% at March 31. Our credit facility also has $400 million of undrawn revolver capacity that provides additional sources of liquidity for the company. At June 30, our debt-to-capital ratio was 9%. During the second quarter, Essent Guaranty paid a dividend of $100 million to its U.S. holding company. The U.S. mortgage i nsurance companies can pay additional ordinary dividends of $303 million in 2022.
As of quarter end, the combined U.S. mortgage insurance business had statutory capital of $3.1 billion, with a risk-to-capital ratio of 10.2 to 1. Note that statutory capital includes $2 billion of contingency reserves at June 30, 2022. Over the last 12 months, the U.S. mortgage insurance business has grown statutory capital by $253 million, while at the same time paying $347 million of dividends to our U.S. holding company. Now let me turn the call back over to Mark.
Thanks, Dave. During the second quarter, our business continued generating strong earnings and robust returns. Our balance sheet, capital, and liquidity remain strong, providing optionality in managing our business both offensively and defensively. We believe that our measured approach of deploying excess capital, is in the best long-term interest of our franchise and stakeholders. Looking forward, we remain confident in the strength of our operating model, and view Essent as well-positioned in supporting affordable and sustainable homeownership. Now let's get to your questions. Operator?
At this time, I'd like to remind everyone, in order to ask a question, press star then the number one on your telephone keypad. We'll pause for just a moment to compile the Q&A roster. Your first question comes from the line of Mark Devries with Barclays.
Yeah. Thanks. Mark, I was hoping to get your updated thoughts on how you're thinking about deploying excess capital here, and what might you need to see before you get a little bit more aggressive repurchasing the shares?
Yeah. Mark, good morning. I think good question, and I would look at it a couple ways. I think in terms of returning excess capital to shareholders, you know, we increased the dividend. I would say we still believe, you know, that's the best demonstration of the confidence we have in the sustainability of the cash flows. Keep in mind, sustainability is in good times and bad times. You have to think about that, given where we are in the current environment. In terms of repurchases, you know, just to remind you, when we authorized the $250 million repurchase plan in May of 2021, that was for 18 months through the end of this year. We completed that in April of this year.
We reinstituted or reauthorized another $250 million, really obviously to give ourselves some flexibility. However, I think it was our intention when we said it was to have it run through this year. Again, we're going to pause. We paused this quarter. We may pause another quarter or two. A lot of it's just going to be the visibility. It's twofold, Mark. It's on a defensive side. You know, what's the visibility that we see in the economy? A lot of different data points, a lot of different opinions on where the economy is going. You know, clearly, you can sum it up for us in being levered to unemployment. You know, right now, credit is strong. The low-end consumer, as you know, has started to weaken, but it's something we have our eye on.
Eventually, you know, that could work its way up to our customer. Hasn't yet. Haven't seen any real clues of it yet. Again, you have to be prepared for that. On the offensive side, you know, as we've stated in the past, we are going to look to invest in or acquire other earning assets. You know, we continue to do the work on it. It's not a quarter-by-quarter assessment. It's not burning a hole in our pocket. We have a long-term view on it. You know, you can't judge these things in quarters. You know, when you sum it all up, you know, capital begets opportunity. We like our capital position and where we are now, but it's a quarter-by-quarter assessment really, depending on how things unfold in the economy.
Okay, g ot it. Thank you.
Your next question comes from the line of Rick Shane with JP Morgan.
Thanks, everybody for taking my questions this morning. Hey, Mark, you know, there's an interesting dynamic that's emerging right now, which is that there is obviously concern about affordability related to new originations. When we think about your portfolio, in the industry more broadly, there's a significant concentration in vintages with low rates, a lot of embedded home price appreciation. Obviously, the third factor for affordability is income, which is going to fluctuate. W hen you really look at the risk here, is it sentiment-driven if there's home price depreciation, borrowers' willingness to pay declines? Because it doesn't seem like when you think about the big vintages, that there's really much that's going to distort the existing policy's affordability.
Well, good morning, Rick. It's a good question. I think you're zeroing in on I think probably not a well-recognized strength of the portfolio. We've always said, you know, insurance in force is really how we judge the business because that's where it generates premiums. That 2020 and 2021 vintage is really strong, unusually strong, probably historically, just when you think about the two things that have happened to it, you know, simultaneously. We have a book that's, you know, rates at 3.5%. J ust from a stickiness or persistency standpoint, you know, that's going to be around a lot longer than we would have even modeled out when we first did it, right?
You know, if mortgage rates are hovering around 5%, where they are now, I know that's going to continue to generate cash, you know, for quite a long period. On the risk side, we have the embedded HPA. The mark-to-market is well below where we priced the business originally. For that to be harmed in any way, you're going to need a significant kind of HPD decline, almost to the point where unemployment would be at levels we haven't really seen for quite some time. We feel pretty good about that portfolio and the cash that it continues to generate. I would say more on the risk standpoint, Rick, it's on the newer originations, right? It's the business you're booking now at higher HPA, at higher rates.
It'll turn over faster, right, you know, given where rates are and where they're expected to go . You know, if you think of where the 10-year Treasury is today in historical spread to that, you know, you could see rates, you know, fluctuate above and below five. T he elevated HPA is something that we definitely are keeping our eye on, and I think the whole industry is. Yo u know, this is where the pricing engines really do benefit the mortgage insurers, because we can kind of go in and out of markets a lot faster, right? Just because of , you know, the way we filed and filing algorithms versus you know, actual rates, we can move a lot faster.
I think the industry proved that in COVID. Then clearly here, you can see with HPA, not so much, you know, certain markets have risen faster than others, but then you have to have a more forward-looking view in terms of, you know, supply and so forth, which were very, you know, instrumental to that. Really, at the end of the day, Rick, it comes down to unemployment, right? We're so levered to unemployment. Even at a super strong credit book, 745, if, you know, our borrower does lose their job, you know, they're going to default. I mean, we saw that in COVID. When unemployment spiked up, so did our default. Again, I think we're well built for it. I think from a capital position, we're strong.
W I think longer term, Rick, we would expect defaults to normalize. I mean, we've had such a benign credit cycle the last, geez, five, eight, 10 years. I think your colleague called it a super credit cycle back four or five years ago, and he was spot on. Longer term, you know, are we really going to see claims below 1%? That's a little hard to envision. E ven in a normalized claim environment, you know, you're still looking at a business that has 40%-45% combined ratio. Still just solid, right? Still operating margins in excess of 50%.. We're real positive on the industry going forward. Just even outside of your portfolio question, it's really looking forward.
Y ou know, we continue to see housing to be relatively strong, just given that intrinsic demand of the millennial buyer. E veryone has seen the stats. If you look at just, you know, the population that's in their 20s today, it's 45 million individuals. The average age of the first-time homeowner is in their early 30s. You're going to see 4 million-5 million new homeowners come on board for a while. Again, I always take the bigger picture, and what's the context of the macro environment that we play in? I f you don't like housing, it's going to be hard to like Essent. We continue to like housing, and I think when you break down kind of just, you know, the mechanics of the economics of the business, given where we are in the economic environment, I think we're still in pretty good shape.
Okay. Hey, Mark. Thank you so much.
You're welcome.
Your next question comes from the line of Doug Harter with Credit Suisse.
Thanks. You know, I guess as you look out over the next 12 months or so, you know, I guess how do you expect, you know, kind of the purchase market to trend and therefore, you know, what that might mean for your insurance growth?
I'm sorry. You broke up a little bit, Doug. Could you repeat that?
Sure. I guess as you look out over the next 12 months, you know, how do you expect the purchase market to trend, and what that might mean for your insurance in force?
Okay, s till broke up. I got the question, though. Yeah, I think, you know, purchases, we believe will be relatively strong the next 12 months. You know, maybe , you know, clearly not at the level they were the last 12 months. I think part of our volume is hidden a little bit by the larger loan balances. The new units is a little bit lower. Again, I get back to that intrinsic demand. You know, so even at a 5% rate, you know, a lot of decisions around purchases are life events, you know, versus, you know, a family of four staring at the 10-year. You know, we continue to think the demographics will have purchases strong.
I think, is it going to grow a lot over the next 12 months? You know, I'm not sure about that, Doug, just given the overall environment, you know, with inflation and where rates are, and just still some of the uncertainty around the employment picture.
Great. Thank you, Mark.
Your next question comes from the line of Bose George with KBW.
Hey, good morning. Actually, looks like your market share rebounded a bit. You know, I know some of that is quarter-over-quarter, you know, noise, but just curious if there's any read-through to pricing, whether you're more comfortable with pricing, just given some of the price hardening that your peers have talked about.
You know, I think well, just in terms of market share in general, I think our market share for the first half of the year was a little shy of 15%. Again, as you know, these things tend to normalize over time. I think in the quarter, though, Bose, I think we did sense some of the competitors backing up. That was clear. It's a little. It's an advantage for us in terms of EssentEDGE, right? Because EssentEDGE doesn't rely solely on FICO score. It's an Essent custom score. It can read things differently than our older models did.
We're a little closer to the market clearing price, I think we can tend to bring on a few more loans than we would when we're outside of the market clearing price. Clearly, by design, we were outside of the market clearing price in the, you know, kind of the third and fourth quarters. It was fairly evident. I know I sound like , you know, a broken record, but this whole industry is just price. It's flat out price. If you're tops in share, you have the lowest price. I think our view is, you know, we're given our focus on unit economics and longer- term growing book value per share, we're very comfortable being in the middle of the pack.
I think our view is, because that's where the market, you know, really is and I think we're trying to optimize our unit economics with our engine. If we can get 15%-16% share and just be a little bit better than the market on premium, be a little bit better than the market around expected losses, we feel like we have an expense advantage, and we clearly have an advantage around the tax rate. That's better ROEs. Again, you're talking about, this is a super competitive industry with very smart competitors. We're not going to outsmart them, so to speak. We're just going to try to out- execute people, you know, along those lines that I said. I think we've shown that over time, and I think the engine gives us just another way to squeak out a little bit better unit economics.
Okay, great. That's helpful. Thanks. Just actually on the tax rate, is there anything, you know, that you're following, like in terms of the global minimum tax or any of the other tax discussions in D.C. that could impact your tax rate?
Yeah, it's pretty early in that. I mean, everything we've seen there , t here's not an impact. You know, these things are always subject to change. I think we're good for now.
Okay, g reat. Thanks.
Sure.
Your next question comes from Mihir Bhatia with Bank of America.
Good morning, Mark, and thanks for taking my question. I wanted to maybe just go back to the price discussion from a second ago. Just wanted to ask about that, you know, I appreciate your comments about, you know, being closer to the market clearing price. What I was curious on is, how did that gap narrow? Was it more your competitors taking actions, getting closer to yours? Or did you also need to, like, tweak your prices because your view of the economy maybe got better or the risk coming in? Any additional color you can give there?
Yeah.
We're just trying to understand, you know, what's happening on competitive intensity?
No, it's a good question. I think we talked about the first quarter, that we're a little bit of a bellwether for that clearing price. I believe , you know, a few of the competitors kind of tweaked their pricing, and we could see it. Again, the way our engine works is, you know, we talk about diamonds. It's really good when we're close to the market clearing price. You know, if we're above it by a lot, i t's not as effective. You know, our view is, you know, the competitors or quite a few of them, you know, based on, you know, what you could see with the share, have started to tweak their pricing, which we think is a good thing.
I think it's a good sign, right? I think it's a sign that, you know, the pricing has really started to bottom out, which we think is positive. Because again, look at the environment. You know, it wouldn't surprise me. In fact, we actually raised pricing towards the end of the second quarter again, s o just the wholesale above, you know. We always have, you know, just, you know, a baseline raise. W e'll look potentially to do that in the back half of the year, depending on what others do. Again, our goal here is to kind of be in that middle of the pack and optimize unit economics.
You know, if we're at 20% share, our engine's not that good. Again, it's around the margins where it's incredibly effective. If we're at 20% share, then our base rate's lower than everyone else's. It's just relatively clear. I think just from an investor macro standpoint, again, I think it's positive where the industry's going and it makes a ton of sense, right? We're going into a period of uncertainty, you know, so everyone's expected claim rate or probability of default in this environment has to go up. It did during COVID. You know, I thought the industry raised very quickly.
This idea, you know, that it's a race to the bottom among MIs, I think it's mistaken. I think the MI Essent and the MI's in general price for their view on credit. If they think credit's really strong or it's going to be strong over a period of time of the policy, they're going to price appropriately given, you know, their capital structure and leverage, and all those sort of things. The moving up of it is really you know, I think it's expected. I think you could end up seeing, you know, potentially higher pricing, you know, over the next six to 12 months.
I know, t hat is interesting. Thanks. One other topic I did want to follow up on, you know, just some of the comments from this call and just in general, right? We have experienced, as you said, you know, a strong credit cycle for a few years. I appreciate that, you know, uncertainty is increasing on a forward basis, but a lot of your portfolio is kind of, you know, is what it is, and you've had a lot of embedded HPA in the portfolio. Consumers have had some wage growth.
You have longer life on the existing portfolio. I guess my question is just on the normalized delinquencies. I understand on new policies, you know, and I understand your underwriting has to account for that. Do you get to a normalized delinquency rate here in the next year, two years, realistically, just given all the other dynamics and how strong the portfolio is? Or is it more realistic that it takes longer to get there? Like, just to your normalized delinquency level.
I think Rick alluded to it earlier. It's really bifurcated, right? The core that 2021 book is isolated. We would expect, I think where the default goes is going to be where unemployment goes, Mihir. I mean, we're just levered to unemployment. Our view is unemployment could go up, our defaults could rise in that kind of 2021 cohort, but we don't see a high level of claims coming out of it because of the embedded HPA. I mean, y ou know, when , you know, we've talked to certain investors that are worried about a housing recession, of course i f you don't follow the industry, everyone's natural inclination is to go back to the Great Recession.
You know, the Great Recession, the portfolio that the MIs had versus today is literally night and day. I mean, you're talking about a 705 FICO layered risk. 30% of the product originated in the Great Recession is no longer even eligible for the GSEs, s o we have a much better kind of credit book. We have, again like I said, the embedded HPA, and obviously we have the reinsurance protection. You know, the MIs had that uncapped liability on their book, you know, back pre-crisis. You know, we don't have that. I mean, 98% of the book is covered, s o that whole mezz piece is being offloaded. We feel good about that.
A couple other things to think through just in terms of credit, just not of that portfolio, but just going forward. You know, we said there's a little bit more risk, you know, given the elevated HPA. Don't forget that the GSEs play a big role in this. I n fact, one of the GSEs has tightened their credit box recently around the tails, that we're a big beneficiary of that. As we said, I actually think we said this on our roadshow, one of the best thing that's happened to the MI industry, it was the adoption of the Qualified Mortgage rule. I mean, all of that non-QM that's being originated over the last few years would all be sitting in our portfolio and i t's not.
That's all what I call off-the-fairway type loans. There's a different execution to that through the PLS market. If the GSEs were taking on that risk, that would be sitting in our portfolio. I just think it's a much different dynamic, both in the core portfolio and obviously the risk, you know, on the newer business is probably absent HPA is still pretty good risk.
All right. Thank you. Thanks for taking my question.
Your next question comes from the line of Ryan Gilbert with BTIG.
Hi. Thanks. Good morning, everyone. Really great discussion around credit performance. Appreciate all the details. Just building on that, you know, given the quality of the in-force book and all the changes that have happened to the mortgage industry since the financial crisis, how are you sizing a realistic downside scenario in terms of mortgage loss rates, when, you know, you're stress testing the portfolio? Or how should we think about what a realistic downside scenario is over the next couple of years as, you know, presumably unemployment should tick up higher?
Yeah, I think it depends. I mean, we run, you know, a zillion different scenarios, but I would point t, we kind of point to the barbell of it, Ryan. Most recently, we were reaffirmed at Moody's for our A3 rating. As part of that, we ran through another Great Recession stress test on our portfolio. We volunteered to do that because w e do it all the time internally, but we really wanted to have some external eyes looking at it. If you go through that, you know, we do not lose money through a five-year period.
You know, we don't make a ton of money through some of that because it's a severe stress, but we don't deplete book value. I think that is really such a key point. You know, when you go to the Great Recession, the MIs depleted book value significantly. When you deplete book value, there's really no bottom for your stock. Whereas, you know, I think the way the MI industry will be in the next recession, whether it's mild, moderate, or severe, you know, take your pick. If we're not depleting book value in a severe stress scenario, that means the capital markets will be open, and that means we can raise capital and whether it's debt or equity, at a price, that's enough to keep it, you know, very similar to how the reinsurers do it.
I mean, you know, because again, if you're thinking, you know, severe recession, you're also gonna look in an environment where pricing is going to be significantly higher. The pricing would be significantly higher. The credit's going to be clean. We can just look at early days of Essent. We were fortunate enough to hit the market post-recession, and the book was squeaky clean. That'll happen again. We'll be around. Unfortunately, some of the MIs were not around to enjoy that period. We will be, and my expectation is so will all of our competitors because they're structured the same way.
I do think, so what's going to happen, if you think about, again, that points to the reinsurance community, when they have a cat event, generally pricing hardens, and they have the ability to raise capital off that and keep going. I think the same thing will play out for the mortgage insurance industry. I think over time, that'll be reflected on our multiples. Once investors understand kind of the sustainability and the strength of the model, and I give our competitors a lot of credit. Everyone's doing a lot of that, the same thing. The safety or the strength around the reinsurance will play itself out in a severe stress. I think we're pretty confident in that.
We're not confident we're going to make a ton of money through a severe stress. I never said that. I think it's a matter of the survivability. I think the old view of the MI industry is that we're going to go out of business in a severe stress. In fact, we were with an investor a month ago who asked us that question. We had to walk through this analysis. If you don't follow the industry at the level you guys do and the analysts do, and certainly our investors know it well, I mean, our top 25 investors make up 80% of our shares outstanding, and we don't have one hedge fund in it. We have a very educated investor base, and they understand kind of the dynamics of this.
The broader market, you know, again, it's a niche industry, so it's harder, you know, for others to pick up on it. I think over time, I think it'll be reflected. You know, we're going keep working through it, and do the things that we can control. If we think we're going into an uncertain environment, what are we going to do? We're going to raise pricing like we just talked about. We're going to look at different MSAs and adjust the pricing there. We're going to be a little bit more conservative at the HoldCo in terms of cash and distribution. We're going to continue to have a measured approach across both, again the business and what we do at the top of the house.
Okay, got it. Really helpful. Second question on purchase NIW. I get that buying a home is a life decision, and most families are not watching day-to-day fluctuations in the 10-year treasury. You know, mortgage rates are down, you know, maybe 100 basis points for the last month and a half. The question that we're getting is, you know, has that move down in mortgage rates generated a tick up or a lift in purchase demand? You know, based on the numbers that are rolling out for July, it doesn't seem like that's the case. I'm wondering if there's anything that you've seen in your business that would suggest that there's been any green shoots, or a lift in demand over the last maybe, you know, few weeks or month.
I wouldn't say anything in particular. I would say, again, we kind of have to express it in terms of kind of NIW for the industry, right? To me, that's a good proxy. You know, the first quarter was, you know, 105. Second quarter was 120-ish. You know, our view is it'll be lighter in the back half of the year. That being said, July applications , you know, they didn't fall. You know, they were down moderately, but they weren't down a ton. I think the third quarter, I don't know where it'll be relative to the second quarter. My guess is it's a little bit lighter and then y ou know, the fourth quarter, I think it's too early to tell, depending on what rates are.
I wouldn't call it a green shoot. I think purchases actually held up pretty well. I think really, you know, the issue around originations to us has been, you know, refinancing. 98% of our book was purchased in the second quarter. I mean, we had certain days, Ryan, where it's like 99% purchase. Like, we've never seen that in the history of the company. I mean, I know w e've only been around 10 years, but still that's quite remarkable. Again, I think just the intrinsic demand of purchase will continue. There's fits and starts, right, j ust given where the rates are. I think it does take, you know, back to my earlier comment.
You know, not that people don't stare at it, but even if rates are higher, which they are, it takes folks a while to adjust, you know, to that new rate. I mean, purchases, again, longer term, just given the intrinsic demand of that, you know, the folks in their 20s right now, I think we're set up still pretty good longer term. It's tough to tell quarter by quarter.
Great. Thank you.
You're welcome.
Your next question comes from Geoffrey Dunn with Dowling & Partners.
Thanks. Good morning. Mark, I wanted to revisit the comment you made, you know, about, you know, defaults could affect or rise from the 2021 cohort, but you don't see a lot of claims coming out of that. That echoes, I think comments you made in the past where, you know, new notices drive reserving, and then it's HPA and employment, et c, that can affect the ultimate claim.
The problem is, with new notices going up, that drives incurred losses higher, affects your PMIERs assets. How do you approach the reserving for what is really a unique book of business? Do you attempt to speculate at lower ultimate claim rates and speculate at severity assumptions based on HPA? Or do you stick to a more conservative approach, which would imply the prospect for maybe more recurring favorable development down the road? How do you approach a book that you've never had precedent for reserving for in the past?
Yeah. It's a good question. I would say we'll clearly, you know, tend to be more conservative. I mean, it is an actuarial model that has gone back not just for our history, but we have all of the Triad data. We feel just more comfortable, you know., and i t's a model that, we review quarterly and, you know, we'll tweak it, you know, at least once a year for some items. I think in general, we'll stick with it. I mean, we've gone off the model before. We've gone off it twice in our history, once during the hurricanes.
We just thought the second and third quarter cohorts of, you know, 2020 were so different that we felt like the model wasn't quite built for that, and we kind of used our best estimate of the 7% to get around that. As you remember, we quickly went off of it. You know, we went off of it after two quarters, which did create a lot of noise. As we saw through 2021, you saw in this portfolio, you would see a borrower default, you know, declare forbearance, accumulate reserves over three, five , six-month period, then boom, they would cure. That's why you saw some of the, you know, kind of the large ins and outs around the reserve. That will happen again.
I mean, again, I think if folks get forbearance, they're going to tap into it. I'm not sure we'll be the beneficiary as well as we were last time, you know, just because jobs came back quickly. Yeah, I think in short though, Jeff, you know, t he actuarial model is based on history. There's obviously a forward view to it, but, you know, I don't think we feel comfortable enough to make those type of changes. I think from a PMIERs perspective, you know, I think the GSEs would probably rather see us be more conservative.
Okay. Just in terms of how you thought about HPA in your pricing and in your reserving, what levels were you assuming in the last couple of years relative to the actual performance? I mean, were you assuming a couple percent and we saw the double digits? Or can you just kind of frame up your rough approach on HPA?
Yeah. Hey, good morning, Jeffrey. It's Chris. With regards to HPA within our modeling and certainly our pricing, I would say we've taken a moderate view. It does not reflect certainly the HPA actuals that you've seen over the last couple of years. When we look at the unit economics, it's certainly more reflective of a, I'll call it a moderate view of HPA going forward.
Okay. As your reserves develop, do you do any kind of mark to market for HPA, or do you let it play out in your claim rate?
Hey, Geoffrey, it's Dave Weinstock. We do look at the HPA of the defaults, and we refresh that regularly. That does have an impact, and will play itself through our reserve numbers as we update them.
Okay. Thanks.
There are no further questions at this time. I'll turn the call back over to management for closing remarks.
Well, I'd like to thank everyone for joining today, and have a great weekend.
This concludes today's conference call. You may now disconnect.