Docoh
Loading...

ESNT Essent

Participants
Chris Curran SVP, IR
Mark Casale Chairman and CEO
Larry McAlee CFO
Mark DeVries Barclays
Rick Shane JP Morgan
Mihir Bhatia Bank of America
Bose George KBW
Douglas Harter Credit Suisse
Ryan Gilbert BTIG
Call transcript
Operator

Good day and thank you for standing by. Welcome to the Essent Group Limited Second Quarter Earnings Call. [Operator Instructions] And please be advised that today’s conference is being recorded. [Operator Instructions] I would now like to hand the conference over to your first speaker today, Chris Curran, Senior Vice President of Investor Relations.

You may begin, sir.

Chris Curran

Thank you, Brian. Good morning, everyone. And welcome to our call.

Joining me today are Mark Casale, Chairman and CEO; and Larry McAlee, Chief Financial Officer.

Our press release, which contains Essent’s financial results for the second quarter of 2021 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 26, 2021, 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.

Mark Casale

Thanks, Chris. And good morning, everyone. Today we are pleased to report our second quarter earnings, which exhibited both strong performance and capital generation.

Our results for the quarter reflect a favorable operating environment as credit continues to normalize and housing demand remains elevated. The economic engine of our business remains firmly in place as our high quality earnings and cash flow to the quarter demonstrate the strengths of our buy, manage and distribute operating model.

Our outlook on our business remains positive as the underlying fundamentals of housing are strong. And we continue to make solid progress on the next generation of our EssentEDGE technology.

On the housing front, strong millennial demand and historically low rates continue to provide positive underpinnings.

As for EssentEDGE, we continue to enhance its utility by combining increased amounts of data with the use of artificial-intelligence-based models. We believe this capability will be a long-term advantage in pricing and managing credit risk.

Now let me touch on our results.

For the second quarter we reported net income of $160 million as compared to $136 million last quarter. On a diluted per share basis we earned $1.42 for the second quarter compared to $1.21 last quarter. And our annualized return on average equity for the second quarter was 16%. At June 30 our insurance in force was $204 billion, a 17% increase compared to $175 billion as of the second quarter a year ago. The credit quality of our second quarter NII was strong, with a weighted average FICO of 747 and loan to value ratio of 92%. Also we continue to be pleased with credit performance as our default rate at June 30 was 2.96% compared to 3.7% last quarter and 5.19% at the end of the second quarter a year ago.

On the business front, we continue to focus on optimizing our unit economics.

While the more tangible aspects of this include using ILNs to minimize loss volatility and ceding more business to Essent Re, we also continue to invest in technology-related intensives. A platform like ours is technologically intensive as we need to seamlessly deliver pricing and services to thousands of customers located throughout the U.S.

For example, we are nearing completion of migrating our platform to the cloud which provides more data storage, processing and computing power. This enables us to deliver our EDGE technology more efficiently, given the need to quickly analyze large amounts of data from a variety of sources and combine with machine learning techniques. We believe that this will benefit our unit economics over time and optimizing premium levels and credit costs.

In fact, we also believe that customer efficiencies of using EDGE will deliver our best price to borrowers and challenge the industry practice of using the negotiated rate cards. At June 30 our balance sheet and capital were strong. With over $4 billion in GAAP equity, access to $2.4 billion in excess of loss reinsurance and over $800 million of available liquidity at the holding company, we are well positioned.

Our most recent ILN transaction was our largest to date where we obtained $558 million of reinsurance through the capital markets. Also Essent Guaranty remains the highest rated monoline in our industry at A by Invest, and AAA and BBB+ by Moody’s and S&P respectively. We remain pleased with our base business, which is an earnings and cash flow engine during positive economic environments.

For example, for the first half of the year our operating margin was 73%. And we generated $340 million in operating cash flow.

Given the use of reinsurance, which enhances the sustainability of our business, there’s more certainty in the earnings power and capital generation of our franchise.

While our practice has been to retain cash and invest, the strength of our business model enables a measured deployment of excess capital amongst the business strategic investments and shareholders.

We will continue to be thoughtful in deploying excess capital in doing that we believe is in the best long-term interest of the Essent franchise and our shareholders.

Finally, given our financial performance during the second quarter, I am pleased to announce that our Board has approved a $0.01 per share increase in our dividend to $0.18. Also in connection with our $250 million repurchase plan, we have bought back approximately 400,000 shares for a total of $18 million as of June 30. Similar to dividends, repurchasing shares is a tangible demonstration of the benefits of our model in generating capital. It also provides further balance in deploying excess capital between the businesses and redistribution to the shareholders.

Now let me turn the call over to Larry.

Larry McAlee

Thanks, Mark. And good morning, everyone. I will now discuss our results for the quarter in more detail.

For the second quarter we earned $1.42 per diluted share compared to $1.21 last quarter, and $0.15 in the second quarter a year ago.

Our updated estimate of the annualized effective tax rate for the full year 2021 is 16% before consideration of discreet tax items.

As a result, the tax rate for the second quarter was 16.1%. We ended the quarter with insurance in force of $204 billion, a 3% increase compared to $197 billion at March 31. And a 17% increase compared to $175 billion at June 30, 2020. Net earned premium for the second quarter of 2021 was $217 million and includes $13.3 million of premiums earned by Essent Re on our third-party business. The average net premium rate for just the U.S. mortgage insurance business in the second quarter was 41 basis points, down from 42 basis points in the first quarter. Persistency increased during the quarter to 58.3% at June 30, 2021, from 56.1% at March 31, 2021. The provision for losses and loss adjustment expenses in the second quarter was $10 million compared to $32 million last quarter. The provision for losses in the second quarter benefited from a decline in new notices of default and higher cure activity.

During the second quarter we received 4,934 new default notices, which is down 34%, compared to 7,422 defaults reported in the first quarter. At June 30 our default rate decreased to 2.96% from 3.7% at March 31. Consistent with the fourth quarter of 2020, and the first quarter of 2021, we have reserved for new defaults reported in the second quarter of 2021 using our pre-COVID-19 reserve methodology.

As a reminder, for new defaults reported in the second and third quarters of 2020 we provided reserves using a 7% claim rate assumption. This assumption was based on the expectation that programs such as the federal stimulus, foreclosure moratoriums and mortgage forbearance may extend traditional default to claim timelines and result in claim rates lower than our historical experience.

We have not adjusted these reserves previously recorded in the second and third quarters of 2020, which total $244 million as they continue to represent our best estimate of the ultimate losses associated with these defaults. Other underwriting and operating expenses in the second quarter we’re $41 million compared to $42 million in the first quarter.

We continue to estimate that other underwriting and operating expenses will be in the range of $170 million to $175 million for the full year 2021. Essent Group Limited paid a quarterly cash dividend totaling $19.1 million to shareholders in June and repurchased $18.4 million of stock.

Additionally, during the second quarter Essent Guaranty paid a $100 million dividend to Essent U.S. Holdings. On June 23 we closed the Radnor Re insurance-linked note transaction which provides $558 million of reinsurance protection on approximately $14 billion of risk in force. This transaction pertains to risk on our new insurance written from August 2020 through March 2021, including the portion of risk written from August 2020 through December 2020, which was not covered previously by our quota share reinsurance agreement. From a PMIERs perspective, after applying the 0.3 factor for COVID-19 defaults, Essent Guaranty’s PMIER sufficiency ratio is strong at 174% with $1.3 billion in excess available assets.

Excluding the 0.3 factors, our PMIER sufficiency ratio remains strong at 163% with $1.2 billion in excess available assets.

Now, let me turn the call back over to Mark.

Mark Casale

Thanks, Larry. In closing, we were pleased with our performance for the second quarter as we produced strong earnings and generated excess capital.

Our buy, manage and distribute model is operating on all cylinders and confidence in our economic engine is high. Combined with a strong housing environment, our outlook on our business is positive.

Finally, we are excited about the progress that we are making with the next generation of EssentEDGE and its potential to be a game changer in evaluating and pricing credit risk.

We continue to believe that combining AI with large quantities of data is where the financial services industry is moving and we want Essent to be in the forefront of this.

Now let’s get to your questions. Operator?

Operator

[Operator Instructions] And the first question we have Mark DeVries with Barclays.

Mark DeVries

Mark, where are you seeing the most attractive places to deploy your excess capital today? And what do you need to see to want to get more aggressive repurchasing of stock?

Mark Casale

Hey, Mark.

I think it’s a number of things. Taking a step back, I think as we talked about last quarter and again in the transcript, I think it’s a measured approach around, let’s call it capital allocation.

So we are still investing capital into the business. We grew 17% year-over-year, which requires capital.

We’re making some strategic investments in the ventures part of the funds that we talked about, albeit a small amount. And we’ve increased the dividend, which is one way to return capital to shareholders. And we announced the buyback. We just announced it last May. The numbers were through June, I think in a transcript. And as of the other day we looked at it, we’ve repurchased $50 million.

So we’re well on our way there.

So I don’t think it’s anything particularly that we’re looking at, Mark.

I think it’s continued -- these businesses don’t get -- I know we come in front of you guys every 90 days, but we have a much longer-term perspective in terms of how we allocate the capital.

So again, I think we’re going to keep on this approach and continue to look for opportunities, again both within the business, outside of the business and obviously continuing to return capital to shareholders. It’s a little bit of the best of both worlds.

So we’re pretty excited about the optionality that we have around capital at this point.

Mark DeVries

Okay. It’s helpful. And next question, what percentage of the business is still coming from the rate card? And for those lenders that haven’t moved to the pricing engines, what’s behind that? Is it still like a technology systems issue, compliance, or -- and kind of where do you see that going longer term? Do you expect them -- to move to the engine?

Mark Casale

Yes, it’s a pretty good question. I would say right now it’s around, for us it’s 70-ish percent is coming through the engine.

I think for some -- and I mentioned this in the script. I do think some lenders enjoy the pricing power that they have with negotiated cards over the MIs. But we are in discussions now with 2 relatively large lenders around building APIs into their systems. And what that’s going to allow us to do is compete with the negotiated card.

So when they go to look for an MI price, it’s going to be our engine, which again, keep in mind, Mark, our engine is now 400 factors.

So what does that mean? It means when a lender goes for a price, we’re combining all the raw credit bureau information with all the mortgage information that we get and then obviously analyzing it via machine learning, which allows us just to run many iterations and obviously learns over time. And we’re delivering that price back to the lender in 3 seconds.

We’re competing -- we will be competing with a static rate card.

So we love our chances with that.

So I think the technology is moving faster than most people think. We think we’re at the forefront of it. And if we can get into those situations, we’re competing against a static card, I think we’re going to come out very favorable.

So again, we’re super, again, motivated to keep investing and looking and improving the engine because, again, that’s where credit’s going. We said this before. The MI industry is all going to be about credit selection. And it wasn’t before. All the pricing was the same. And you’ve seen it in other industries, you see it in the credit card industry, you see it in auto with Progressive and Geico. And I think we’re kind of applying those techniques to mortgage insurance. Again, it’s early. We just rolled it out at the beginning of this year. But again, if you sit in my seat and have more of a 3-, 5-, 10-year horizon, investing in that type of technology, and as I mentioned in the script, moving the platform to the cloud, which increases your computing power, this really set us up for future success down the road.

Operator

Next we have Rick Shane with JP Morgan.

Rick Shane

Mark, you pointed out that the persistency is trending higher quarter-over-quarter, and that’s obviously a favorable inflection. The reality is that CPRs are still elevated and burnout seems to be taking a little bit longer than folks anticipated. When we think about PMI, it actually increases the incentive for borrowers to refinance versus a borrower who doesn’t have PMI. On one hand that would suggest that burnout would take longer, but I’m actually wondering if that PMI refi incentive actually has pulled forward refinance in your book, and we should see burn out a merge more quickly than the overall market.

Mark Casale

I hope you’re right, Rick. I’m not sure about that.

I think it’s a good thesis, but I would be cautious on that. Again, given where the 10-year, I mean, it bounced back a little bit this morning, but given how low it is I do think the refinancings will remain elevated. But take a step back, right, and look at it in context, Rick, from our standpoint, the purchase market remains strong. And we think the underpinnings of the purchase market will remain strong, right? I mean, when you think about the demand around millennials, the peak, I think the highest age group is still a cohort, it’s like right around 28 or 29.

The first-time homebuyer average is like 31, 32.

So we’re kind of getting into that sweet spot and I think it’s something like 5 -- close to 5 million new potential homeowners come into the market over the next few years. Rates will impact some of that, but rates remain low even if they go up. And I know there’s all talks about affordability. And that’s important.

I think the demand is going to really be there.

So over time, as rates go up, refinancings will start to dissipate. And I think the book will stick longer. It’s just -- it’s -- I wouldn’t expect it to happen quarter-over-quarter.

So we think it’s troughed. Maybe it gets into the mid-60s by the end of the year. But I think longer term, and that’s where we said the positive underpinnings of housing are relatively strong. And I do think that will help continue to allow us to grow the insurance in force. Does that make sense?

Rick Shane

Yes. Look, it totally does. And it’s interesting because when I originally framed the question I thought of it the opposite way in terms of it might extend burnout for you.

So it’s interesting to get your perspective. Look, the reality is that the fundamentals that are driving high prepayments are favorable for credit. And when you think about the 3 outcomes for a policy going to maturity, refinancing or a credit issue, the middle outcome is certainly a better outcome than credit problems.

Mark Casale

Agree, I agree. And remember, just -- and we’ve talked about this before, is we are -- when NIW is elevated, you have to expect that persistency is going to be low. There’s no free lunch.

So we’re talking about industry volumes at historic levels the last 2 years. It’s not surprising that persistency is at such low levels.

So I think we’re always a little bit more levered to higher rates, Rick. I’d rather see the book -- I’d rather see lower NIW and the book grow, persistency a little bit higher. And that’s just better from a cost basis for us. And slow and steady kind of wins the race, but this is the environment we’re in, and I think we’re adapting well.

Operator

Next we have Mihir Bhatia with Bank of America.

Mihir Bhatia

Maybe I’ll just start with just a follow-up on that last comment you made, Mark.

Just how much of I guess -- as persistency increases if you get back to your normalized levels, how much of a tailwind on like expense ratio is that? Is that -- are we talking like 1%? Or are we talking like 3%, 4%?

Mark Casale

Again, it’s hard to measure it an expense ratio, I mean, here, because that’s just a calculation based on the insurance.

So I think in terms of nominal cost, particularly it’s really around the underwriting cost, right? So we have a percentage of us, of our originations is a non-delegated.

So if you’re doing $50 billion a year versus $100 billion a year, there’s pretty good savings. And it’s just -- it’s the friction cost of doing all those things.

So you have turn time issues and all those things with customers because their -- this time last summer you couldn’t find an underwriter in the industry. I mean, there was such a shortage just because the volume was so high. And that just -- that’s hard on your employees, that’s hard on your customers’ employees. It’s not a sustainable environment.

So I think having -- and again, when I say levered, I do think the economics are better. Again, if we’re lower NIW, we could be significantly lower NIW and still end up growing the insurance in force because persistency is high. And I think that’s a little underappreciated part of our model.

Mihir Bhatia

Understood. And then just maybe broadening the discussion a little bit, just you’ve talked a couple of times on previous calls about things you’re looking at outside core MI and as you think about the business long term. Maybe give us an update on that? And maybe just talk about what you’re spending your time looking at outside of core MI? Are there particular segments or types of businesses that are most interesting?

Mark Casale

Well, I would say, first, a lot of the funds that were invested in, we spent most of our time tying that back to the core business.

So think about EssentEDGE. I mean, some of the -- and I mentioned this on one of the previous calls, one of the portfolio companies really helped us.

One of the companies helped us think through the machine learning technology and all the things in terms of multiple uses of data, emerging credit bureau with other information, that was learned through one of the portfolio company.

So a lot of this is just how do we tie it back to improve the core business, right? Technology initiatives around lifting the business to the cloud.

In terms of new initiatives, that’s also in terms of newer investments outside the business, that’s kind of a nice to have. That’s not the necessarily the driver of it. But we clearly think that some of the skills that we have around capital management, buy, manage and distribute, understanding consumer credit, they’re applicable outside of mortgage insurance. And remember, mortgage insurance is, it’s only so big.

So if you think about the industry, it’s right around $1.3 trillion, $1.350 trillion at the end of this quarter.

We’re approximately 15 percent of that. Trees don’t grow to the sky. And I think you run the risk of, if you have a lot of capital, you just start kind of keeping in the core business, clearly there’s the return aspect to shareholders.

We have the ability to do both. And I think our ability to kind of apply capital to grow and grow outside of MI, I think it’s real while we’re still maintaining redistribution to the shareholders.

So again, we’re not in a hurry to do it per se, but we do think these skills are applicable. And I think long term -- and remember, I have a 3-, 5-, 10-year horizon.

I think Essent will be more valuable to shareholders in the long term by growing. And like just returning all the capital to -- all the capital to the shareholders, I think that looks good analytically.

I think it puts shareholders in long-term danger. I really do.

Just because you kind of shrink the company and you shrink the equity base. And remember, what have we always said, credit kills these businesses.

So bigger is better and capital begets opportunities.

So again, we have process around this on the investment side.

We’re very disciplined. And remember, we built this business from scratch. We built Essent Re from scratch.

So we understand as we look at other businesses outside of MI, kind of some of the fundamental things it takes to build businesses.

So again, these are -- this is a long-term perspective here, and I wanted to kind of get that out there as people talk about capital distribution and bringing capital back to the shareholders, which, by the way, is extremely important. It’s just I think our message is we have the ability to have a measured approach around capital allocation.

Mihir Bhatia

Right. No, no, I appreciate that. And look, you -- after you made all those comments, you still instituted the buyback when you thought it was appropriate.

So I think shareholders get it.

Okay. I will leave it there.

Operator

Next question we have Bose George with KBW.

Bose George

I don’t know if you said this in your prepared remarks.

So what was the default-to-claim rate this quarter?

Larry McAlee

Bose, this is Larry. The default rate at the end of the quarter, is that the number?

Bose George

No, the default-to-claim rate that you had used for new notices this quarter.

Larry McAlee

Okay. There was no change, Bose, from what we assumed prior quarter. The initial default-to-claim rate is still around that 9% level that we’ve historically experienced.

Bose George

Okay, great. And then can you just talk about the cadence for the margin over the next few quarters, just in terms of that, the base average premium number? Is that kind of the 1 basis point decline in a quarter level?

Mark Casale

Yes, Bose, I think we’re saying still at the end of the year kind of in that 40-ish range for the year. Could exit a little bit lower, but we are starting to see a little bit of a flattening out. We know there’s a lot of moving parts there. And I know everyone thinks through singles cancellation is one additive to it. But the reinsurance cost is obviously a big cost to that, and that has started to level out as most of the book is reinsured.

So again, 40-ish to the end of the year probably could exit a little bit lower, but I think the premium levels are relatively stable at this point.

Bose George

Okay. Actually one more just on the regulatory side with -- the changes of the FHFA, the risk-sharing market, do you think that that could get more traction?

Mark Casale

I’m sorry, risk-sharing in terms of the…

Bose George

The…

Mark Casale

…sharing?

Bose George

No, just the other business that you guys do, the ACIS and the -- where you invest in the…

Mark Casale

Yes, yes.

Okay. Yes, yes. I mean we’ve been pretty active. And Freddie Mac has been active the whole time in the market.

So we actually wrote the most business last year than we’ve ever. And it’s a pretty good start this year too.

So it’s mostly Freddie.

So if Fannie enters the market, certainly, that could expand it. But I think that remains to be seen. I haven’t heard much about that from Fannie Mae at this point.

Operator

Next question we have Douglas Harter with Credit Suisse.

Douglas Harter

I know it’s relatively early still, but any early read on kind of how EssentEDGE is performing in terms of picking credit quality and kind of delivering that better credit quality as you would expect?

Mark Casale

Yes, don’t think of in terms of better credit quality, Doug, think about in terms of optimizing premium levels.

So again, the example could be the 700 FICO, that our score will score them higher than 700.

So we’ll give them a slightly lower premium, but we would expect kind of a better credit loss.

So the unit economics of that loan would be better.

So I think longer term, I think the impact on the portfolio -- and again, it’s on the margin.

So it’s all relative to where the industry is pricing. And our view is, we’ve talked about it before, longer term, we’re 1 of 6, so our share is kind of in that 15 to 16.

So how do we optimize that premium over the long run. It’s a game of interest.

So if our average premium is 40 now, and we can get to 42ish or higher with slightly lower losses, then that’s going to roll down to the ROE.

So again, I think it also allows us just to select the credits we want. And I still think that’s hard for the industry to grasp as we see -- we see certain -- we’ve heard certain lenders talk to our competitors where our competitors are still pushing their rate cards over the engines. That tells me they don’t have the technology to compete. It really does.

So if you’re still looking and saying you’d rather price off of 2 or 4 factors versus 400-plus, that doesn’t seem like you’re moving in the right direction.

So again, I think for us -- and more is always better in the mortgage industry, and it’s just not the truth. Again, 1 of 6 there’s a certain price level for this industry that if you move below it then you’ll gain share, but then you’ll be quickly matched. There’s a quick competitor response.

So again, our view is let’s take our 15% or 16% that we can get, optimize it and move on.

Douglas Harter

And I guess as you’re thinking about optimizing unit economics, I guess, are you looking to kind of optimize the persistency for EssentEDGE as well? How do you think about that component?

Mark Casale

Yes, absolutely. There’s a prepayment component to that.

So we certainly [indiscernible] extending the life is a big thing.

So, yes, I think it’s -- so few things on just technology in terms of unit economics, right.

So we’ve talked a little about EDGE which has helped us optimize kind of a premium. And clearly around the credit losses, there is the potential to expand that, right? So EssentEDGE, when you’re ingesting that much credit bureau information along with the mortgage information, it’s essentially an underwriting engine at that point.

So I wouldn’t be surprised for us to continue to develop that to help automate some of the regular underwriting that we do today, which again is going to help cost down the road.

We have a number of initiatives underway to help our customers have a much more of a self-service experience, which again makes it more efficient, easier to use to the client. And again, I think down the road that helps from an expense standpoint. And we’ve talked about kind of the affiliate reinsurance around the tax rate.

So I think when we think about how we run the business day-to-day, we think in terms of unit economics.

So what steps can we take to improve premiums, credit losses, clearly reinsuring the book has helped a lot of around credit losses.

So we kind of have like a little plan around each one of them. And then day in and day out that’s how we go about trying to execute that.

Operator

And our last question will be from Ryan Gilbert with BTIG.

Ryan Gilbert

First question, Mark, just going back to your comments around overall industry market share.

Looking through that NIW data and comparing it to some of the estimates out there around 2Q ‘21 total originations, it looks like there was maybe a pickup in share for the overall industry in 2Q.

So PMI is taking a bigger piece of the overall pie.

On the one hand, it’s within the realm of normal quarterly fluctuations. But on the other hand, given really strong home buyers appreciation, maybe more first-time homebuyers, you might expect the PMI percentage of total originations to increase in the quarters and years ahead.

So I just love your thoughts on what you’re seeing on the market and if you think that 15% of the total pie can increase going forward?

Mark Casale

Yes, it’s a good question, Ryan.

I think you have to break it up out between purchase and refi.

So the penetration on refinances is always much lower on MI versus purchase. And our purchase share went from 62% in the first quarter to 82% in the second. And it’s been 90% -- I think last month it was 90% purchase.

So yes, I would -- which obviously is a higher penetration.

So from that standpoint, yes, you could see the overall -- you certainly could see the overall penetration go up. And it gets back to our earlier, one of my earlier comments in response to one of the questions. And that does -- that just has to help persistency, right? I mean so as it gets to be more purchase and less refinance, our overall volume actually could be lower per se, but the book would end up growing a little bit -- because the book is a little bit stickier.

So it ends up growing a little bit more than in kind of a heavy refi environment.

Ryan Gilbert

Okay. Got it. And on premium rate, just going back to your response to a prior question, it sounds like with potentially premiums flattening out over the course of 2021, do you think we could be hitting the last year of premium rate compression? Or should we anticipate lower premiums for the foreseeable future?

Mark Casale

Again, some of it is just working -- it’s just math working its way through the book.

So if the premium levels on new business are lower than they were 4 years ago.

So clearly, and that’s kind of the pre-tax rate change.

So that’s clearly working its way through the books.

So I actually think you’ll continue to see premiums decrease, but not -- I think it’s at a much slower rate. They’re clearly getting to a trough period. I don’t know if it’s right now. But I do think it will start to trough at some point. But I do think, again don’t forget that part of this is because of the reinsurance that we’re spending, like the whole industry has -- I won’t say material in -- material reinsurance costs embedded in that net premium rate, which has really increased. The whole industry really only started reinsuring back in like 2018.

So it’s kind of working its way through premium rates now, Ryan.

And so the premium compression actually looks a lot worse than it is when you think about it.

So I don’t -- we don’t get too caught up in that. And again, look at the big picture, Ryan. This is a business that had operating margins of 73% for the first 6 months of the year, and we generated cash flow of $340 million.

You don’t measure that in premium basis points.

So I understand where people are coming from. But at some point, you got to take the -- again, that’s what’s allowed us to have this excess capital position and put us in such kind of the catbird seat in terms of our ability to allocate capital around potential new businesses, return capital to the shareholders with dividends and share repurchases.

So again, 41, 39 at a market like this, we’ll continue to enjoy these margins.

Ryan Gilbert

Yes. All good points.

Just one housekeeping question for Larry. Did I hear you correctly saying that you did not change your reserve assumptions for -- or you did not change your reserve levels for 2020?

Larry McAlee

That’s correct.

If you -- just to remind everybody, we assumed a 7% claim rate assumption on the new default notices we received in the second and third quarters of 2020, and recorded reserves of $244 million for those 2 quarters.

So we continue to hold those reserves and made no adjustments to them in the second quarter.

Ryan Gilbert

Okay. And it looks like there was a $15 million favorable development in the second quarter. Was that related to pre-2020 reserves?

Larry McAlee

It was related to the pre-COVID period, which would be Q1 of 2020 and prior. And then also in the fourth quarter of 2021 we had moved back to our historical methodology.

So it’s really the fourth quarter of 2021 and the first quarter ‘21 in prior periods.

Mark Casale

Yes, think of it -- Ryan, it’s Mark. Think of it in like 3 cohorts, right, you had the pre-COVID on a reserve model, you had the second and third quarter kind of special reserve where we froze it. And then, remember, for the fourth quarter of last year we went back to our normal reserve model.

So a lot of the changes are kind of at -- on the first and third part of that, where the second we’ve kept it constant. Really for the reason is that we just -- we haven’t seen -- we continue to see progress, but only 80% of the second quarter ‘20 cohort has cured, and we have 93% in the model.

So -- and part of that is forbearance is still around, the foreclosure moratorium is getting extended.

So we actually said back last May when we first did -- or August when we booked it, we said it would be 12 to 18 months before we start to really see kind of clarity, and it’s kind of playing out that way.

Operator

And there are no further questions. I’m turning the call back over to the speakers. Chris?

Mark Casale

Thanks, Brian. Thanks, everyone, for joining. I know Friday in August, exciting times to talk about MI. But thanks for your participation, and I hope everyone has a great weekend.

Operator

Thank you, everyone. Ladies and gentlemen, this concludes today’s conference call. Thank you for participating.

You may now disconnect.