Good afternoon, ladies and gentlemen, and welcome to the NMI Holdings Inc. Third Quarter 2019 Earnings Conference Call. At this time all participants are in a listen-only mode, later we will conduct a question-and-answer session and instructions will follow at that time. [Operator Instructions] I would now like to turn the conference over to your host today, Mr. John Swenson. Please go ahead.

Thank you, Operator. Good afternoon, and welcome to the 2019 third quarter conference call for National MI. I'm John Swenson, Vice President of Investor Relations and Treasury. Joining us on the call today are Brad Shuster, Executive Chairman; Claudia Merkle, CEO; Adam Pollitzer, our Chief Financial Officer; and Julie Norberg, our Controller. Financial results for the quarter were released after the close of the market today. The press release may be accessed on NMI's website located at www.nationalmi.com under the Investors tab. During the course of this call, we may make comments about our expectations for the future. Actual results could differ materially from those contained in these forward-looking statements. Additional information about the factors that could cause actual results or trends to differ materially from those discussed on the call can be found on our website or through our regulatory filings with the SEC. If and to the extent the company makes forward-looking statements, we do not undertake any obligation to update those statements in the future in light of subsequent developments. Further, no one should rely on the fact that the guidance included in such statements is current at any time other than the time of this call. Also note that on this call we refer to certain non-GAAP measures. In today's press release and on our website, we provided a reconciliation of these measures to the most comparable measures under GAAP. Now I'll turn the call over to Brad.

Thank you, John, and good afternoon, everyone. I'm pleased to report that in the third quarter National MI again delivered record financial performance and strong momentum in customer development and portfolio growth. Broad market conditions were made constructive in the quarter in terms of origination, volume and demand for our product as well as the strength of the underwriting environment. The interest rate environment continued to spur increased, purchased and refinancing volume and helped drive incremental home price appreciation. Lower rates are particularly beneficial for first time home buyers, a key constituency who have seen an immediate improvement in affordability and access to the market. Looking forward, we are optimistic about the macro economic environment and the outlook for the mortgage insurance market. We see continued strength in our insured borrowers driven by record, low unemployment, strong wage growth and sustained home price appreciation. In fact, this is the first time in more than 50 years that both national unemployment and mortgage rates are below 4%, a dynamic that sets up well for both new business flow and the credit performance of our in-force portfolio going forward. Although our outlook is positive, we continue to take a long-term view of the credit cycle and have invested to develop the tools that we believe will allow us to outperform across all markets. Our view has always been that we need to establish broad credit risk management strategies when times are good in order to fully benefit when the market turns, and this is exactly what we are doing. We have developed and deployed the most comprehensive credit risk management framework in the industry spanning Rate GPS, individual risk underwriting, and our innovative reinsurance program. We believe that our ability to evaluate and price risk on a granular basis, the unique insight…

Thank you, Brad. In the third quarter we delivered record performance once again, expanding our customer franchise, achieving strong growth in NIW and insurance in force and delivering record financial results all while maintaining our disciplined approach to managing risk and return. GAAP net income for the quarter was $49.8 million or $0.69 per diluted share and adjusted net income was $49.9 million or $0.71 per diluted share, up 57% from the third quarter of 2018 GAAP return on equity was 23.6% for the quarter and adjusted ROE was a record 23.7%. Overall private mortgage insurance industry volume continued to benefit from increased origination activity tied to low rates and a strong national employment environment. We generated record NIW of $14.1 billion in the third quarter up 16% compared to the second quarter of 2019 and 92% compared to the third quarter of 2018. Monthly NIW was $13 billion, up 17% from the second quarter and 95% compared to the third quarter of last year. Primary insurance in force was $89.7 billion at the quarter-end up 10% compared to the second quarter and 41% compared to the third quarter of 2018. We continue to achieve the fastest rate of growth in insurance in-force in the industry. In the third quarter we activated 23 new lenders. We are now doing business with a broadly diverse group of over a thousand high quality originators. Equally as important, we continue to grow with our existing lenders, leveraging our value proposition of certainty and service and our consultative approach to customer engagement to further strengthen our relationships quarter-over-quarter. Rate GPS continues to be a standout success with our customers. More than 95% of our lenders are using the platform and over 90% of our third quarter NIW volume was delivered through the engine. Our…

Thank you, Claudia and good afternoon everyone. We had another strong quarter and achieved record results across a number of key financial metrics. We generated record NIW of $14.1 billion and continue the rapid growth of our high quality insured portfolio. This drove record net premiums earned of $92.4 million, record adjusted net income of $49.9 million or $0.71 per diluted share, and record adjusted return on equity of 23.7%. Now to the details. Primary insurance-in-force was $89.7 billion at quarter end, up 10% from $81.7 billion at the end of the second quarter and up 41% compared to the third quarter of 2018. 12-month persistency in the primary portfolio was 82.4%, down from 86% in the second quarter. Total NIW was $14.1 billion with monthly products contributing $13 billion or 92% of our total volume. Purchase originations represented 80% of our volume, compared to 88% in the second quarter. We're seeing an increased flow of refinancing volume driven by overall refi origination activity and an uptick in mortgage insurance penetration rates in the refinancing market. Net premiums earned in the third quarter were $92.4 million, up 11% from the second quarter and 41% compared to the third quarter of 2018. We earned $7.4 million from the cancellation of single premium policies in the quarter, compared to $4.5 million in the second quarter. Reported yield for the quarter was 43.1 basis points, compared to 43 basis points in the second quarter. Yield for the third quarter, reflects an increased contribution from cancellations largely offset by an increased impact from reinsurance tied to our most recent ILN transaction. We expect the net yield will trend between 41 to 42 basis points through the remainder of the year. Overall, we continue to capture business at rates that are supportive of our strong…

Thank you, Adam. The broader economic environment remains strong and private MI market conditions are healthy in terms of volume, industry competition, credit quality, and the regulatory environment. Against this backdrop, we are delivering record performance across our platform with continued momentum in terms of customer engagement, NIW volume, insured portfolio growth and bottom line financial results. We trace our strong performance in the quarter to the steady discipline manner in which we've built our business. Our goal from the start has been to build a franchise that is both durable and successful. We are executing on this vision and are well positioned to continue to win with customers, drive growth in our high quality insured portfolio, maintain the right risk return balance and deliver strong results for our shareholders. With that, I'll ask the operator to join so we can take your questions.

Certainly. [Operator Instructions] Your first question comes from Rick Shane from JPMorgan. Your line is open.

Hey guys, thanks for taking my question this afternoon. Not surprisingly given the path of rates over the last five years, the persistency declined the most among the – in the 2018 cohort. I'm curious if 2018 borrowers at this point have enough home price appreciation that they're able to refinance without taking PMI or as one of the things that's actually driving the strength of the book rotation from 2018 loans to 2019 loans. Are you basically recapturing existing paper?

Yes, Rick, it's Adam. I think that's exactly right. So we are observing an increased penetration of MI into the refi market and we trace it to – largely to the 2018 book year. What's unique this go-around is that we've had rates fall so quickly and buy so much without a corresponding macro credit event. And so that means that those borrowers who took out loans really in the back end of 2018 in a much higher note rate environment are now coming into the market and can benefit from a refinancing opportunity, but they haven't yet benefited from near as much HPA or also as much time to amortize down their principle balance through monthly payments. And so we do see increased penetration and expect that to continue to hold through at least the near-term.

Got it. And do you find that because the originators have incentive to go back to those same borrowers and you have strong relationships with those originators that your recapture rate tends to be pretty high on those transactions?

Yes, look, it's difficult to trace it to a specific recapture rate. One thing I would note is that broadly speaking, the interest rate environment that we're seeing as Brad mentioned, is really spurring a significant increase in overall origination activity. So that's a ripe environment for MI companies in general, ourselves included to really capture strong new business volume and that's what's happening. Beyond that, in terms of the business that's actually in rotation, we are a newer company with a smaller in-force portfolio. And if you look at it, we have about a 7% share of total industry IF. And we have a much higher share of new business production. And so simply the math of it, when business comes into rotation, there's a higher likelihood that we will be on the – I'll call it the winning end of that equation than actually having loss business, simply because we have less to lose and we're capturing a greater portion of new business flow.

And that's a great way to describe it. Thank you very much.

Your next question comes from the line of Sam Choe from Credit Suisse. Your line is open.

Hi guys, I'm on for Doug today. So just looking at the expense ratio, I mean, you guys have been making good progress on that front. And if you exclude the ILN, I mean, it's even better. So I'm just wondering how should we be viewing the progress on the expenses going forward?

Yes, it's a great question. And something we spend a lot of time obviously focused on. For us, operating leverage is a significant driver of embedded earnings growth. And that's certainly still the case. We would expect as we go forward that will continue to scale into our fixed expense base. And most importantly, right, it's not just about scaling into our fixed expense base, but balancing that with discipline around new investments that we might look to make in the platform and in our people. And so, what I would expect is that going forward we'll continue to see expense ratio improvement as to the pace and magnitude of expense ratio improvement in any given quarter. There's going to be a time where we naturally slow simply because we're coming off of lower and lower numbers on the expense ratio side so that the pace of improvement may slow. But we would expect that we'll continue to see improvements in our overall expense profile at our expense ratio quarter-on-quarter.

Great. Another one for me, there's revenues line was up. What caused that?

Yes. So other revenues for us is, it relates to our services subsidiary in MIS where we offer contract underwriting services for certain customers. And so we saw an uptick in the third quarter given really the broader growth in overall industry volume. And by extension the volume that lenders who engage with us through that services platform, we're looking for us to support them on. So in MIS we perform loan underwriting services on behalf of customers and that business just increased. What I would say is that the revenue that comes through, I’d note two items. One, it's a fee stream revenue, which is positive for us, right, its not underwriting income, it's fee revenue, but it's also generally offset by a corresponding expense that's embedded in our operating expense line item in the same period in which we recognize the revenue.

Got it. Thank you so much.

And your next question comes from Mackenzie Aron from Zelman & Associates. Your line is open.

Thanks. Good afternoon and congrats on another great quarter. My question is really around Rate GPS and the share of your customers that are utilizing it. I think it’s been relatively unchanged at 95% of customers and 90% of your volume. So I think this one is for you, Claudia. But are you expecting that 5% that's been holding out to eventually transition or are those customers comfortable with where they're at and you're not really hoping that they'll switch over to the engine?

Yes. Sure, Mackenzie. So we'd love to see all of our lenders over our Rate GPS. The lenders that aren't there just have usually a proprietary system that doesn't allow them to implement the Rate GPS. So we do provide cards to those lenders, but eventually, our focus has always been on all lenders getting the risk-based pricing and the more granular approach. So we'd like to see everyone on, but sometimes that's just not feasible for some of these large lenders.

Got it. And then maybe just an update on the competitive landscape. I know there was a lot of talk last quarter about the rate cards or the customized rate cards. Has there been any shift over the last three months?

No. I mean, what I'd say as far as the MI sector as it relates to risk-based pricing in general is. There are six MI companies with six distinct management teams, six different views on risk and that creates a very stable environment. What we see with the introduction of risk-based pricing is a rational approach to the market. So we're really thrilled with our results. We've had $14.1 billion of NIW and posted 24% return. So we're really proud of the team. But I’d say, the risk-based pricing introduced by all the MIs has really created a really granular approach that's good for the sector.

And Mackenzie, this relates to the rate cards and some of the discussions in the second quarter. We really haven't seen any shifts. The same lenders continue to engage in the same way. It doesn't seem to be expanding and nothing of note that's new in the third quarter.

Great. Well, thanks so much.

Your next question comes from Bose George from KBW. Your line is open.

Hey guys, this is Tommy McJoynt on for Bose. I wanted to ask about the 41 to 42 basis points of average premium yield that you guided to for the fourth quarter. Does that assume that the third quarter was the high point for earnings from cancellations?

Yes. Look, I think that the 41 to 42 reflects the possibility of a modest decrease in the contribution from cancellation earnings. And also some – what I call some movement in our core yield given sort of the runoff of what I'll call higher risk, but also some higher yielding business in prior years, as to whether or not the third quarter represents a high watermark in terms of dollars of cancellations. It's a difficult one to predict. But I would say, even if we have achieved the same dollars of cancellation earnings in the fourth quarter as we did in the third quarter, because the way the yield is calculated, which is it's in reference to our average IF – during the period, if we're growing IF, but still seeing the same dollars of cancellation earnings, you would see the yields come down. And so it's not necessarily to serve as an indicator of whether or not cancellation earnings have peaked or not. It's a very difficult one to forecast. It really depends on the interest rate environment and borrower activity and behavior around refinancings. But given that dynamic of even a constant dollar amount of cancellation earnings translates through a lower yield. That's how we get to 41 to 42 plus that in the core yield dynamic.

Okay. That makes sense. And I don't know if you guys get like October trends, but have you seen the kind of level of penetration on the refinance side for MI hold up in October, similar to it was in the third quarter?

Yes. So I won't speak to our trends specifically because we don't share results. We obviously have the ability to look into the portfolio. It's just not something we disclose. What I would observe is that as a general matter, what we would typically see is a seasonality pattern to refi volume, especially in the interest rate environment that we're in now. First and second quarters’ lenders really have a significant volume of purchase origination activity that they have to work through. And so their approach, pricing and eagerness around refi volume is somewhat scaled in the third and fourth quarters where there's not quite as much purchase origination activity coming through. It really lets lenders focus on refinancing activity. So as a general matter, what we would expect is for refinancing activity to stay roughly the same if not actually accelerate through the end of the year given that seasonal dynamic.

Okay. Yes, that makes sense. And then just last one from me, all of the growth that you've had, would you attribute more of that to new customers coming on board or ramping up wallet share with existing customers?

Yes. Our success really is driven by both. We continue to build new customers and then also drive wallet share with each of those customers. So I'd say both – I mean, we also have a lot of opportunity where we're bringing in larger and larger customers as we go on. Overall. I think also the introduction of risk-based pricing has helped us get to the market even faster. So I’d say both.

The interesting piece is a customer who we bring in today, how long do you want to define that as a new customer versus growth with an existing customer base. In the last three years we've activated over 300 new customer accounts and we are still achieving wallet share growth with some of those – many of those customers that we brought onto the platform and activated in 2017. There was no kind of same-store sales dynamic that we report. It doesn't – it's not applicable for the MI sector but so it really is both, but it's also both because those accounts that we activate in earlier periods don't just bring us to a level and leave us there and we can continue to grow with those relationships over time as well.

That's interesting. So after two to three years, there's – do you still see opportunity where your wallet share is increasing?

Yes, absolutely. We can see it in the numbers that we track. That absolutely happens.

That's great. All right, thanks guys.

And we have Mark Devries from Barclays. Your line is open.

Thanks. Most of my questions have been asked, but Adam, I have a question for you about execution around ILNs. I mean, clearly your last deal went quite well. Just wondering what risk, if any, you see the execution around Morningstar’s decision to kind of pull back from ratings or are they still kind of on the sidelines? And if so what do you think of the implications on the cost of future ILNs?

Yes, Mark, that's a great question. It's something that we've been monitoring and are focused on. Difficult to say, Arch was out with a deal since there has been some movement in the Morningstar DBRS dynamic and they ultimately secure, went down the path of securing a rating from a different agency. We have to see where Morningstar comes out, right. What their model looks like and what it means for ratings of new deals as well as all the deals that have been issued to-date. For us, the positive is the timing. We got our last deal in July. We're not expecting to come back to the market until sometime in the first half of 2020. And by that point, we think that the – whatever is the dynamic that will ultimately develop will have been established, which is a positive for us to not – need to be in the market at this time. With respect to pricing, the key piece is the investors that we engage within the ILN market, all of them, every single investor that we engaged within the – in the course of our last deal, run their own credit models and aren't really relying on rating agency models as indicators of expected loss performance on the reference pools. And so really where the ratings become consequential is up the stack for those buyers whose capital requirements and also whose ability to finance some of those tranches are influenced by the ratings. At the end of the day, the credit performance of the underlying portfolios is assessed by each individual investor. And that's what sets the pricing, might pricing back up, I'll call it a touch because the movement with Morningstar and DBRS, that's possible. But for us, going from a deal where we had an all-in cost of capital on a pretax basis at roughly 265, even if that backed up by 10%, we're still looking at a cost of capital that's well below 3%. So it's something we're comfortable with but we're staying close to it.

Okay. That's helpful. Thanks.

We also have Phil Stefano from Deutsche Bank. Your line is open.

Yes, thanks. Question on underwriting and operating expenses. And some of your peers have started to point us towards looking at, this metric on a gross of the ceding commission basis. I guess, does that make sense and as we think about growth, should there be a materially different growth rate between the gross expenses and net expenses?

Yes, Phil like I think, ultimately the question is what's the volume that you're originating the top line business that you’re, the top line revenue that you're producing and how does that look relative to your expense profile? And so if you're looking at expenses gross of the ceding commission, you should probably be looking at gross premiums as well. From our vantage point, what ultimately matters is what's the mix of our underwriting margin? What is our underwriting margin and what does that allow us to drive from an ROE standpoint? And so I don't think, we would never look at our 23.7% ROE and say that should be scaled up or down because of our insurance profile, a reinsurance profile. It's simply is one of the items that drives the outcome that ultimately impacts the returns that we're delivering for shareholders. As to the trends in it, I think the biggest question around the trend is, well, do you have consistency in what your reinsurance profile looks like? So from, we've got a reinsurance decision to make for our 2020 renewal. And to the extent that we shift to a non-quota share structure that doesn't have a ceding commission that has a lesser ceding commission that may come into play. But that decision, if we go down that path, would likely be accompanied with an acceleration in our net premium earned because the contracts and the trees would have different profiles. So, I think at the end of the day, it's really a question for you and for investors from our vantage point, we care a great deal about ultimately what are our bottom line results and what's the ROE that we're delivering. All of that takes into account the costs that we pay for that coverage and the benefits that we receive in our expense ratio.

Got it. Okay. And second one I had, and look I'm making a mountain out of a molehill. I understand that, but there's a claims denied line in the default inventory role forward, it looks like it's the first time that it popped up. I guess I was just surprised to see anything there, given the extent of loan profiles that you review and certainty, ex-service, et cetera, is there anything notable in that or is this just things that are going to pop up in the normal course of business?

Yes, there's nothing notable in that. I mean, we certainly go through our due diligence process with claims based on the master policy requirements and we could have something that's denied but nothing, nothing to note.

And so the important piece is, so as it relates to our upfront underwrite, this is not a rescission, this is – so rescission means we are saying that the policy wasn't underwritten in accordance with our underwriting guidelines and we're terminating the coverage and we're sending it. We could be submitted, we could receive a submission for a claim payment for an item that isn't covered under the terms of our master policy. And so while the contract itself and the insurance coverage is valid, it's not a loss event that we are on the hook for. And so that's what's coming through. It's disconnected from the upfront underwrite that we do, which really speaks to levels of precision activity.

So is this something like a hurricane or tornado impacted house.

And one of those is in fact that.

Yes.

Got it. Okay. I understood. Most peers disclose the rescissions and denials in the same line. So, I'm pretty sure that's how I have it labeled. But that's the confusion. Thank you.

Your next question comes from Chris Gamaitoni from Compass Point. Your line is open.

Hi everyone, most of my questions have been asked. Adam, what's the outstanding balance of ILN at the end of the quarter.

Great question. Chris. I don't have it off the top of my head. We could follow-up and give it to you. It'll be in our Q when we file it. And we could follow-up offline and give it to you.

Okay. Appreciate that. And this is a very small item for you. Just wondering, how we should think about the provision for a new default moving forward as the book has shifted, loan balances are probably higher than they were. Just kind of trends of expectations there?

Yes, look, it's obviously, we don't generally provide explicit guidance on losses. When I say that, the overall quality of the portfolio continues to be exceedingly high and we do expect that lifetime losses on the, on that portfolio will be, will be modest. I think for the next few years we've continued to expect that the quality of the portfolio and the vintage stacking dynamic that we've talked about in the past, will keep our loss ratio relatively low. There's also an interesting dynamic that's coming through because of the refinancing activity and the turnover in the in-force replenished by such a significant amount of new production. The way that you can think about that is it's somewhat, I'll call it resets the clock on our loss of merchants pattern. So as a general rule, we typically expect that loss incurrence on a book of business will peak between years three and six after origination. And now the fact that we're turning over older age books and importantly more than replacing them with newer production with fairly pristine credit characteristics is keeping our portfolio young, if you will for a longer period of time. And so that extends, the period over which we would expect to see a lower loss ratio because it basically is replenishing the portfolio in greater size with newer business.

It makes sense. Thank you so much.

We also have Mark Hughes from SunTrust. Your line is open.

Yes. Thank you. I was just curious whether Rate GPS is evolving much. If you think about how you are approaching the market and underwriting, you're obviously getting a lot more experience here than changing the way you are approaching the market. You seem to be doing obviously quite well with top line growth. The quality of the portfolio is improving. How much is the system evolving to help you grow or is it just, the market is moving more in your direction?

Yes, I think so – gas is never going to be static, and how we think about credit risk is never going to be static. I'd say from an evolution standpoint, the same risk variables that we have always consider a range of risk variables around the borrower, around the property, the loan structure, the lender itself where the home is located all of these things. We still consider what we'll change around Rate GPS. It happened yet in a meaningful way, but Rate GPS and how we express our risk appetite through the engine will change based on the risk environment that we see. But we still see real strength in the risk environment, and so our approach to the market through Rate GPS is really the same in the third quarter as it was in the first and second quarters, nothing significant to note. But the nice thing is it gives us an ability to react very quickly, if we see a need to because of risk developing in the market or alternatively to react, not just react, but to take action, if we see an emerging opportunity in the third quarter, things were relatively stable.

And just to comment Mark, Rate GPS is not an underwriting system. It's certainly very granular in defining what kind of risk we want. But we would still be underwriting with our underwriters and with our automation beyond that.

Okay. Understood. Thank you.

Thanks, Mark.

[Operator Instructions] Your next question comes from Geoffrey Dunn from Dowling & Partners. Your line is open.

Thanks, I've got just a couple. First, Adam, did you disclose the premium rate on new business this quarter?

No, Jeff, we did an, similar to the second quarter when we didn't disclose it. It's not something that, we put out there at this point in the – in a post rate card rate engine environment. It's a bit more information from a competitive standpoint than we want being out in the public domain. But what I will share is that, our pricing philosophy is the same as it's been all along, which is to target and most importantly achieve rates on new business that are supportive of that strong mid-teens return objective.

Gotcha. And then, Claudia, I was hoping you could talk a little bit about market share. Not what you did in the quarter, but in terms of how you're developing new business opportunities, is the incremental gain more on the small mid-market, adding new customers, or, and I guess this is more about future opportunity or is it more building the penetration of the top 10, top 15, top 20, whatever you want to think about? I want to get an idea for what's on the table, particularly for the large lenders and where you currently stand with them versus where you might be in the future.

Sure. So Jeff, the way we've approached the market, just when we think about, we all focus on share, but the way we've focused on the market is, is really always going after our top opportunities. So, we're very focused on top 100, top 200, top 300 and that's always been a focus for us. And then when we activate the customer, the answer to the – the other side of the question is we're always building share then within that customer. So the growth comes from both and we've got, although we have plenty of opportunity as well. But I believe we're at this point that we need, if we look at the top 200, there's another eight that we're working on with master policies. So, we'll continue to move down that path, with our top opportunities. The key is when you are doing business with them is that you're working on real consultative selling with them, building relationships and then building in that wallet share.

I guess on average is your market share of the top 10 or top 20 less than your market share of the top 50, top 100?

Yes. The customers ramp, and follow the many different ways. We don't really disclose any particular share within the customers, but we can tell you that we're doing, we do very, very well once we activate a customer and build wallet share.

Okay, thanks.

Sure, Geoff.

Your next question comes from Soham Bhonsle from SIG. Your line is open.

Hey, good evening everyone. So with the PMIERs excess of, 300 million and it gives you significant runway going forward. So, how do you think about, the uses of capital going forward outside of just the core business and what does that mean for feature ILN issuances?

Yeah. So I'll, take it. but I think, what we like about ILN as well as the traditional reinsurance market is both the capital efficiency but also the risk protection that we get. And so by extension, what it means from both a cost of funding standpoint but also from a potential volatility of results standpoint. Right? There's real benefits to both. We do have a significant amount of capital runway with the excess position that we have. We are, we wrote 14.1 billion of NIW though in the third quarter, and we see strength in our NIW production as we go forward. So we're deploying that access, every day and so as we look forward, we do expect to be active in the ILN market, really with consistency in the future as a risk management matter and also as a capital efficiency matter. In the near term that's going to be from a capital standpoint, we'll support, necessary capital to support growth. And at points in the future it'll be about capital efficiency. But for right now we will be active not withstanding the runway that we have, I expect at some point in the first half of 2020, we'll be looking to execute our next transaction.

Got it. Okay. And then just wanted to get your thoughts on some recent servicer reports that, we've been reading where they flagged, slight pickup in early stage delinquencies amongst first time borrowers. I know, this is still a small fraction of what it was before, but one, have you guys seen any of that on the margins? And if you have, has there, has it been prevalent in, either the non-banks or the bank side at all?

So yes, so we had not seen that, no change in EPDs in our portfolio, which is something we monitor closely. Right. For us that is, that's an important credit metric and performance metric. We just haven't seen that come through. I can't tell you. We just haven't also been abreast of what others are seeing from a servicer standpoint because it's not coming through in our portfolio. It may be different risk cohorts, different pockets of business that they're, that’s ultimately speaking to.

Fair enough.

I think it probably bodes to our servicers we're underwriting and making sure that we're looking through layered risk as we mentioned in our scripted remarks, so some of that could be a factor as well, that on those person homebuyers is there any important for us to make sure that we're sustaining those bars with ability to repay? So that could also be another dynamic.

Fair enough. And then just one more on expenses for the fourth quarter, Adam, what should we sort of expect from here, slight uptick from here or sort of flat.

Yes, I'd expect that we'll see expenses grow modestly from our GAAP base in Q3 and that's just going to reflect a certain variable expenses. Like we've got premium taxes and so as the portfolio grows, those that are teed off of the size of the enforced portfolio. And then, we'll have a couple of other items around the timing of certain technology and other projects, some additions to our headcount. We picked up four full time employees in the third quarter. So there full sort of, I'll call it run rate, salary basis will come through in the fourth quarter. So we'll see. I'll call it modest growth in expenses in the fourth quarter.

Got it. Thank you.

I'm showing no further questions at this time. I would now like to turn the conference back to the management for closing remarks.

I want to thank you for joining us on the call today. Awesome to note, we will be hosting our Annual Investor Day on Thursday, November 21, in New York, and we look forward to seeing you there. Thank you.

Ladies and gentlemen, that concludes today's conference call. Thank you all for joining. You may all disconnect.