PennyMac Financial Services, Inc. (PFSI) Earnings Call Transcript & Summary

Yes. So I think we'll get started here. So welcome, everyone, to the 22nd Annual Barclays Global Financial Services Conference. My name is Terry Ma, I'm the consumer finance analyst at Barclays. I'm very pleased to have PennyMac Financial Services here with us. We're joined by David Spector, CEO; and Dan Perotti, CFO. So welcome, gentlemen.

Thank you, Terry, and thank you for having us here, and thank you all for being in attendance. We appreciate it.

So with that short intro, let's just jump right into it. Maybe let's just start with mark-to-market on the third quarter across the industry, it seems like volumes are trending in the right direction. So can you maybe just talk about the current market environment thus far and what that means for margins. Obviously, you guys put out your August update today. So thank you for that.

Well, thank you. And look, I think you can see in the August update that we're seeing an increase in production, and we're seeing it -- really in August, we saw it in Broker Direct and Consumer Direct. Correspondent lagged a little bit, but I expect to see an increase in September and October. But look, I think there's a lot of good robust activity out there. We've been operating the company really for -- since we started to see rates increase that through our correspondent network, we wanted to continue to bring on high rate mortgages. For when rates do decline, we have the opportunity to seize on that opportunity to refinance those mortgages. And so as we sit here today, we have a very healthy portfolio of loans that are about 6% and 7%, about 20% of our portfolio. And so in our consumer direct channel, we're seeing increased activity there. And that's with the great story of the consumer direct increase in August is it's really a validation of a lot of the efficiency that we have been working to build into that channel. And so we're starting to now bring on additional capacity in our call centers as well as our fulfillment center to be able to continue to grow the amount of consumer direct origination. So I'm really encouraged by the increase of $1.3 billion in July to $1.9 billion in August. And as you can do the math, you can start to see a run rate that's going to be meaningful, a meaningful increase versus what we saw in Q2 and Q1. On the broker direct side there, we're continuing to grow share. And that's a great story of the work that we did there. We've spent a lot of time and again, creating technology efficiencies and tools. And we saw a lot of -- we saw a lot of market participants get out of broker when rates started to increase. And so we're the #3 broker direct originator, our market share is between 4% and 5%. And I'm really encouraged that we're going to continue to see growth there as brokers look for alternatives to the top two market participants. And in addition, we're seeing margin in broker really holding nicely. And obviously, as rates come down and capacity constraints begin to enter the equation, margins are going to continue to inch up a bit. And so I'm encouraged -- I'm really encouraged on the broker side. On correspondent, we're the leading correspondent aggregator. We're going to remain the leading correspondent aggregator. And while we have market participants coming in and out, all that really does is create a little bit of disruption on the margin front. And in this case, we have a market participant on the government side who's really pricing and participating in a really meaningful way. But look, I think that, that clearly, we've done a -- we're approaching $9 billion in corresponding originations, and that's going to continue to grow, and we're going to continue to get our share back up to where it was. And suffice it to say, you can see just the sheer quantum of new servicing we bring on every month, which really allows us to just get current [ near weight ] production. As rates decline, we'll have the refinance opportunity on that production as well.

Maybe just to take a step back, I think the big question is maybe what the 2025 origination market looks like. I think SME and MBA are at $1.7 trillion for this year and $2.1 trillion to 2025. What's PFSI's view?

Yes. So look, I think we share the $2.1 trillion view. That's how we're composing our strategic plans as we speak, and that's a good number to go in at. I tend to see that after the election, we'll see what happens with rates. But if they continue to bring them down, obviously, that will go up. One of the big stories in terms of market size. And I caution everyone to think about this is the fact that when you're looking at comparable periods, you have to take into account that the average balance has gone up considerably. And so in many cases, on a units basis, a more normalized mortgage market is closer to $2.3 trillion, $2.4 trillion and that's something that I see us getting back up to, if not in '25 and '26. But that's a more normalized way that we think about things.

Got it. And this is a good time to just pause for the first audience response question. Can you queue that up? Relative to standing MBA forecast with total originations of $1.7 trillion in 2024 and $2.1 trillion in 2025, what do you expect 2025 total mortgage originations to be? [voting] So 43% expect $2 trillion to $2.2 trillion fairly evenly split between the other three options. So everyone's kind of agreeing with your view. So you spoke a little bit on the last earnings call about how you have three classes of loan officers in queue already. So it seems like you're gearing up for rebound in refi. Can you maybe just tell us more about where you stand in terms of capacity and what you're seeing across the industry?

Yes. So look, as we look at our organization and we look at our servicing portfolio, clearly recapture is a key component. We have this balanced business model and is really unique in the industry, where we invest. We invest in servicing, and we've been the beneficiaries of having a really solid servicing investment throughout our history. But really, over the last three years, if you look at the servicing performance, that's been a meaningful driver of earnings for the organization. Now that we're starting to see rates decline, we want to then really focus on the other part of the balance, and that is the production side. Clearly, correspondent feeds the servicing port, but really the recapture of the economics associated with refinance is something that we want to get the lion's share of. So that's -- there's a few different avenues that were going down, and that is, obviously, we want to be more efficient. And I'm really pleased that the increase in production that we saw early in July and August were the result of us being more efficient. And now that we're starting to bring on more loan officers, I expect us to get even more of the opportunity. And look, some of this is a little bit of timing on our part that obviously, the Fed, we believe, is going to decrease rates in September, whether it's -- I personally think it's a quarter, but whether it's a quarter or half they're going to come down, I think after the election, you may -- I tend to think you'll see additional Fed activity. But I think we need to have the capacity in place. And further to that point, I think on the hedge front, we all know the story in terms of we have a lot of current note rate servicing. That's a lot of volatility. The cost of hedging continues to remain high, albeit we're starting to see it get a little bit of relief with the more -- with the yield curve getting a little bit more normalized. But the cost to keep the additional capacity in place is, to me, more than justifiable versus the cost of the hedge and to integrate it into the hedge, hedge expense or how we think about hedging is something that we're doing organizationally. So whether it's the additional 200 call center people we're going to add or additional, call it, 200 to 400 loan fulfillment people we're going to add, I think it's going to be really important as we're -- as we see rates decline that we're able to get the lion's share of the recapture. The final piece I just want to highlight for people is that in a more normalized mortgage environment, rates come down, rates go up. It's not like '20, '21 and '22 they just went to 0, basically. And so with that volatility in rate and with that volatility and refinance opportunities, you have to have the capacity in place, and that's the final piece that's driving the decision to really focus on having even more capacity than we perhaps have historically had.

I think another piece to put in context too is just given where we've come from, right? We go back a couple of months when interest rates were near 7% basically, none of the portfolio or even the mortgage universe was refinanceable at all. So now we've moved down to 6.5%, a bit below 6.5%. The percentage of our portfolio as well as the overall mortgage universe that is refinanceable, still well below averages, but is multiples of what it was a few months ago. So adding this capacity on and we take down another leg of 25, 50 basis points. That percentage of our portfolio continues to multiply to actually be multiples of what it is today. So adding these loan officers in this capacity from where we've come from is the amount that's there to be recaptured is a meaningful distance from what we've been doing up to this point.

Got it. That's helpful. Can you maybe just provide more color on how you've constructed the portfolio? You're a top producer mortgage loans. You have a large portfolio with more current rate -- current note rate mortgage and loans between 6.5% and 7.5%. So maybe just talk about that strategy.

Sure. So as David mentioned, we're the largest correspondent aggregator and it's really been our thought and strategy over the past few years that we have a significant amount of our servicing portfolio already at low note rates, both have fairly stable prepayment speeds that are less likely to feed into our recapture and to our consumer direct channel. So really, we've been focused on adding loans at the higher end of mortgage rates through our correspondent channel. Our correspondent channel is generally going to be at current market rate loans that we're adding to the portfolio. And so as we've been doing that over the past few years, as David mentioned, the percentage of our portfolio that's above 6% is around 20% of our total portfolio with another 10% from 5% to 6%. The idea behind that is that both will generate current period income through our servicing portfolio, continue to add scale on the servicing platform. But then as interest rates decrease, it gives us a significant opportunity to be able to recapture those borrowers. We have all the information on the borrowers, we're able to reach out and market to them through our consumer direct channel, putting them back into the portfolio and gain the origination income through the consumer direct channel at a very -- in a very efficient manner given the leads that come in through that channel from that higher note rate section of the portfolio. And that really is what we viewed as our best deployment of capital, bringing those MSRs in through the correspondent channel over the past few quarters with a view to generating that origination come through the consumer direct channel as interest rates decline.

Got it. Maybe just to switch gears. Servicing has been really strong, particularly last quarter with a margin at a 9.5 basis point UPB. Can you just talk about what this environment means for servicing profitability and whether or not there's any additional upside to margins longer term?

So as we look at servicing profitability, currently, it hasn't been as strong recently. A lot of that has been driven by our efforts to drive down the servicing costs over time. That's both a function of the scale of the servicing portfolio. As I mentioned, as we continue to grow the overall platform as well as our proprietary technology, SSE, which we've had in place since 2019. And that has given us further ability to scale and get more efficient and drive down overall costs. And we've driven down the cost of servicing the operational cost of servicing from 10 basis points to actually under 6 basis points in this most recent quarter. So a really significant decrease in our overall cost of service. We think that we have further ability to drive down those costs. There are certain areas of our servicing operation that we think that we can continue to gain efficiencies, at least in a stable delinquency market. And so we think that we can continue to drive that spread higher. Additionally, if you look over the past few quarters, a component of our sourcing income from early buyouts where we take loans out of the portfolio that are delinquent in Ginnie Mae securitizations, get them to reperform and then redeliver those into securitization, that's been a fairly small component of our of our income over the past few quarters with interest rates being higher. As interest rates decline, that opens up some additional opportunity there to be able to redeliver those loans. And so that could also be something that helps to drive that margin a bit wider as we go forward.

And look, I think -- I just want to add, this is one of the reasons that we are looking to get into subservicing. The ability to grow the platform and beginning and add capacity to that platform, our existing service team is going to be the beneficiary of that. And one of the biggest issue is sub-servicers have is replenishment like -- and so -- but we have that in place through our correspondent business and our other two direct lending channels. And so really, I think that we -- the market sees what we've seen as we've driven down cost 30% over the last five years in servicing, and that's a credit to the servicing team as well as the technology that we've created and that technology has been incredibly valuable to this story of driving down cost to make servicing more profitable. And to Dan's point, I expect that organization to continue to drive down the cost of servicing. And look, there are things that when you're a servicer, there are things you can't control like the macroeconomic environment. But even in that case, I look at the forbearance programs that FHA, VA, USDA and Fannie and Freddie have come out with and there's a real -- there's just a whole shelf of modification programs that are available to borrowers that are going to make the cost of servicing less if we get into a period of distress. And that's what makes us the service even more valuable even before you get into the discussions of the capacity issues or the lack of supply in the marketplace. And so I feel really confident about the servicing story, and I expect it to continue to shine even as we see rates inch back down here.

Got it. So you've spoken a fair amount in the past about PFSI's technology investments. Could you just expand more on the SSE platform and explain why you think it could be a competitive advantage?

Look, I think as I jumped the gun on your question, look, the SSE story is a great story for this company. It came with some unfortunate outcomes in terms of sort of the legal and the legal and arbitration issues that we face. But having said that, we have this really beautiful technology that's performing has performed great for us. You saw it during COVID when we implemented the CARES Act very quickly and over 90% of our borrowers were able to self-serve. If you had any doubt about the system, you can see that we've driven down the costs on the servicing side. So you've seen in just running the servicing platform what we've done. You then look at -- we just had an issue where the VA came out with the program that's really advantageous to servicers. It's called the BAS program. And we were able to implement it on August 1, while the rest of the industry struggled to get it implemented by the end of the year. And there's real economics for us as a servicer to be able to implement that. And what this all adds up to is the marketplace can see that we have something that's a real competitive advantage for us. And that's why we do want to now get into subservicing to be able to deliver that event that advantage while obviously to get the economics associated with it. And so we're going to be -- in Q4, we're looking to have a couple of subservicing clients on our platform that we look to grow and expand. It's going to be a co-branded product, but then sometime in mid-2025, we look to get a white label product out where we can service in the name of others and then we're going to see what we -- where else interest lies. But suffice it to say, we have something really, really valuable in the marketplace. And it's a marketplace that really lacks mortgage leadership running the technology, which is something that we have. We have the best servicing management team in the industry who understands how to service mortgages. They understand workflow, they understand technology, and they understand driving down costs. And so I'm really enthusiastic about what we're going to move forward with as we move into '25 and beyond.

That's helpful. Just wanted to touch on hedging for a moment. You touched on it already, but the hedging results have been a bit volatile in the last few quarters. Maybe just take a step back and talk about your approach to hedging and what you seek to accomplish? What -- and how you seek to accomplish it?

So with respect to hedging, hedging is a practice that we've had in place really since the beginning of the company, and we think it's really important from a risk management point of view. And to your point, philosophically, what are we trying to do with the hedging is really reduce the risk to the book value of the company, the equity of the company and changes that could result from that due to changes in the fair value of the MSR as well as to have a more stable stream of earnings over time. And so when we look at the risk of the mortgage servicing rights changes with respect to changes in interest rates, we don't necessarily seek to hedge out the entirety of the change in value of the MSR. At different points in time, we will have different effectively hedge ratios that we seek to achieve on the change in value of the MSR. And that's really driven by what we see as the change or the short-term upside from production. So we won't necessarily -- hedging has a cost, that's something that we focus on, on a daily basis around what is it that we're willing to spend in terms of the cost of the hedge for that protection from the change in rates and with respect to the change in the value of the MSR. What are we willing to spend there versus the risk that we're willing to take from a book value perspective on the MSR, any the upside that we see or downside in the case that interest rates increase in terms of our production earnings in the short term. So when interest rates were lower back prepandemic or during the pandemic, typically, we use the hedge ratio that was closer to 60% or 70% as interest rates increased in 2022 and through 2023 and a lot of this year, where there was not as much of a significant upside to interest rates decreasing. They're just that many loans that were close to being in the money. We were closing -- we were attempting to hedge close to 100% hedge ratio. We talked about on the last earnings call that we moved that down to 80% to 90% in the most recent quarter, and we'd probably expect that to decrease a little bit as we get more and more loans that are in the money where we see the upside from production as we decline or the incremental upside in the short term, more or less offsetting some of the loss that we might see from a change in value of the MSR. And so that's how we really think about it philosophically, to your point, around the last few quarters as we've been attempting to hedge around 100% as well as with hedge costs being fairly high, David mentioned with the yield curve being inverted, drives up hedge cost due to the fact that the carry on a lot of the instruments that we use that are not options is minimized or in some cases, even negative as well as with option volatility being very high and the market interest rates being uncertain. Overall cost of options has been elevated, and that's another component that we use in our hedging. And so with both of those being somewhat more difficult, there are certain exposures that we've opened up in order to maintain a somewhat stable level of hedging cost in the past couple of quarters that led to some of the volatility that you mentioned. As the yield curve begins to de-invert, that should drive our hedging costs down to more normalized levels, although short rate's still pretty high. That's driving -- continues to keep the hedge cost somewhat elevated in the current period. But as the yield curve overall and the short end of the yield curve comes down, we'd expect those hedging costs to become more normalized. And as we get further down in interest rates and see greater variation or expect greater variation in our production income as interest rates move up and down, we also expect our hedge ratio to come down somewhat compared to where we had been targeting in the past few quarters.

Got it. That's helpful color. Maybe just to tie everything together. ROE has been pretty strong, and it seems like you're positioned well to generate healthy returns in either a higher-for-longer environment or a more normalized origination market. You noted your operating ROE should remain in the mid-teens in the current environment and potentially expand to the high teens once the mortgage market grows. Do you feel -- do you still feel pretty good at that range and based on what you're seeing in the third quarter, what should investors expect?

The operating ROEs, I would still say are guidance or our view is very consistent with what we talked about in the second quarter -- or in second quarter earnings as interest rates have begun to decline. So as interest rates decline and a production volumes and income increases, that helps to drive up that operating ROE and versus where we have been in the beginning of the year, assuming interest rates stay at these levels and production similar to what we've seen in the past couple of months in our release today, we would expect that to have upward pressure on our ROEs probably to the higher end of the range that I talked about on the call. So really higher teens or operating ROEs as we're going through the rest of the year.

Helpful. We're going to switch gears and maybe just touch on PMT for a few questions. Can you just remind everyone about the significance of PMT in the PennyMac complex? And maybe just tell us a little bit more about the investment portfolio?

Yes. So look, I think we have a really unique structure with PMT and a really advantageous structure that we're going to get to the follow-on question I suspect. But PMT is an investment vehicle that we've had since 2009, and it operates really with three different strategies. The first one is in the correspondent business. So PMT buys various amounts of conventional loans that it securitizes and retains the MSRs. And for -- and in buying those MSRs or buying those loans, it relies on PFSI for fulfillment capabilities, and PFSI uses its underwriting and due diligence capabilities to diligence all the loans before they get securitized. Then in the securitization process, what PMT teams up with is MSRs. And those are in the interest rate-sensitive strategies of PMT. And those -- and those are -- and in addition to MSRs, PMT also has Agency MBS that it holds for hedging the MSRs, but also as a requirement for being a REIT. And so that agency MBS position with MSRs that really composes the lion's share of the interest rate-sensitive strategies. The third area that PMT operates in is in its credit strategy. And that's where historically for the first x amount of years or call it, five years that PMT really operated and is really focused on distressed mortgage assets. And then in 2015, it started investing in lender credit risk transfers with the agencies, primarily Fannie Mae. In that process, it ended up with a CRT investment, that it still holds a good amount of equity against it today. But the remaining CRT that PMT has is very low LTVs around 50% LTV. It's really -- it's been a terrific investment for PMT. And in -- and it's going to be winding down over the next, call it, 1 to 5, 6 years. In the meantime, where we're PMT also credit in its credit-sensitive strategies, has been buying other CRT issued by Fannie Mae and Freddie Mac through its [ CASM ] stacker originations, our program -- and then from time to time sells those when it doesn't meet PMT's required return. And I think that's -- there's -- the way I think about PMT is there's organic creation, okay, that's its own CRT, its own MSRs. And then there's opportunistic investing that takes place when there's opportunities in the marketplace with credit spread widening or yields that exceed our required return to meet the -- what we think about a 10% net dividend PMT that will buy those securities and from time to time, sell those securities. But I think PMT is well positioned in the fact that between the two companies, there's $100 billion of originations every year, and the ability to seize on opportunities in the marketplace is really, really valuable.

Got it. And can you maybe just touch on the dividend. PMT's dividend was $0.40 a quarter with earnings power a little lower in the mid-30s. So how should investors think of the gap there?

Sure. So we've maintained the dividend at PMT at around $0.40. And part of our philosophy around the dividend of PMT is dividend stability. So to the extent that we see the opportunity for PMT's run rate earnings to increase back to or above that $0.40 level. We're not necessarily going to adjust it on a quarter-to-quarter basis. When we look out further, and this gets back to some of the dynamics that we had talked about recently, but in the interest rate-sensitive strategy, which is the strategy that has the largest amount of equity allocated to it currently, the earnings in that strategy have been somewhat contained over the past few quarters due to the inversion of the yield curve. So generally speaking, we finance with floating rate debt in that strategy secured by the assets. And then the assets are generally longer-term assets, mortgage servicing rights, mortgage-backed securities that have yields which we mark-to-market on a quarterly basis that are based off of a longer part of the curve. And so that the inversion of the curve has really driven sort of a spread compression in that strategy and contain the earnings from that strategy. As we expect the Fed to decrease rates as we're going forward through the next several quarters and for the yield curve to normalize, we expect the earnings capacity of that strategy to increase which should drive up the overall earnings potential of the entity back toward or above that dividend level. And that's why we feel comfortable maintaining and have felt comfortable maintaining the dividend at the $0.40 level despite the fact that our near-term expectation is a bit lower than that.

And I would just add in the medium and long term, I'm really encouraged and enthusiastic about the ability to securitize other products like investor loans and second homes that would otherwise be deliverable to the agencies and be able to take the subordinate bonds and be able to keep those as credit investments in PMT. And I think that there is a plausible scenario where guarantee fees continue to go up to drive additional production out of the agencies into the private label market. And PMT is going to be a leading securitizer in that market to retain portfolio investments in sub bonds. And that's just another -- if you think about it, it's not that dissimilar from the CRT work that we did between 2015 and 2020. There's also a potential opportunity to securitize closed-end seconds and jumbo loans. Our closed end second business in PFSI, we're doing about, call it, around $100 million, $125 million a month. And right now, we're looking at securitization that there could be a potential opportunity where PMT retains again, the credit investments as part of its credit-sensitive strategies. And on the jumbo loan front, look, we're doing a good amount of jumbo loan originations in our broker direct channel as well as increasing amounts in consumer direct and correspondent. And while those loans are be being sold on a whole loan basis, we're beginning to see a path that we can securitize those as well. So I think that our first securitization of investor in second homes will happen sometime and close sometime in Q4, and we're going to look to do another one. But I really am encouraged by the organization's responsiveness to a securitization program and doing increasing amounts of securitization is something that is a key strategic initiative for PMT in 2025 and beyond.

Great. I think we'll just queue up the last audience response question first before we go to questions. Over the next year, would you expect your position in PFSI to increase to decrease or to stay the same? [voting] So 43% increase, 36% stay the same and then 21% decrease. I'm pretty bullish. So I'll just turn it over to the audience for questions, if there are any. We have one up front here.

You made a comment earlier, the forbearance programs are very helpful. Now as. [Audio gap] Could you just simplify, like why is that so helpful for your servicing cost? How much of, say, the recession impact mitigated by these new clients?

So look, I think in the absence of forbearance programs, your outcomes get pretty binary. They get binary to the borrower gets -- finds new employment or new additional income to bring the loan current or you're marching down the road to foreclosure and property disposition. The work involved in taking a borrower through foreclosure, putting aside the borrower issues and tolls is really expensive. It starts with -- there's a series of requirements that you have to go through in terms of notifying not only the borrower, but also the agency, FHA, VA, USDA, and you have to meet certain requirements and time lines. And if you don't, there's a heavy cost there. Then in terms of the property disposition, there are -- the cost of service goes up multiples. And it's while I believe in our system, we've built in a tremendous amount of scale benefits for our default servicing, it still can't take away from the fact that, for example, on VA loans, if a property gets sold at a value that, that far exceeds the guarantee, the servicer takes the loss, otherwise known as VA no-bid risk. And so that's one of the benefits of this new VAS program is the fact that if you get the borrower modified, you can then sell that loan to the VA and for a borrower that's been through distress, obviously, the probability of redefault is probably a little bit higher than on a regular paying mortgage, you move that out of your servicing portfolio. And in addition, that VA no-bid risk was priced into your valuation, so you get the benefit pickup there as well. And so it's just -- it's a very labor-intensive business, primarily as a result of the fact you're having a deal of borrowers directly, and it's a very difficult business to drive to self-serve like you have on regular way servicing.

I think the other piece, too, is that a lot of these methods for modifications or forbearance have at this point from the financial crisis and the pandemic now been somewhat battle tested. So we -- as servicers and the industry has been able to see what works and what doesn't and a lot of the things that do work have been adopted. And so if you -- versus going back 10 or 15 years and coming out of the financial crisis, there are a lot more tools that have been battle tested around minimizing the disruption to the borrower, getting them back to current. And as David said, that's from a servicing cost point of view, sort of paramount in terms of keeping our overall servicing profitability high.

And look, when you talk to D.C., when I'm in DC and you're talking to legislators, there's consensus that keeping borrowers in their homes is the top priority. One of the -- I think as we look back to the Great Financial Crisis, there wasn't enough focus given to borrower programs and modifications like we have now. And if you look at the CARES Act in terms of the outcome there, in terms of giving borrowers modifications or forbearance by just asking for it and you look at the performance that took place the whole notion of strategic default of gaming the system, yes, on the extreme margins, there was cases of that, but not the systemic gaming of the system that some perhaps would have assumed took place. So suffice it to say, keeping borrowers in their homes is something that crosses party lines. And that's why I say that these modification programs are the most beneficial they've ever been. And look, if we get into a unique situation, suffice it to say, I got to imagine that D.C. regardless where there's sort of a republican or a democrat is going to react to keep borrowers in their homes.

We probably have time for one more question if there is one. One in the middle, yes?

Can you talk about what kind of recapture rates you're targeting for the borrowers in that 6 to 7 rack -- range in -- any evidence that points to?

Yes. So look, I think that my kind of flippant answer is as high as possible. But I think one of the things when I think back to the peak of refinance in '21 and '22. On the government side, we got above 50% recapture rates. And by the way, there's a lot of different recapture rates that are being bandied about in the industry. I'm kind of -- I'm kind of old school in how I think about it, my denominators, total payoffs, my numerators, loans that we originated. So that's my definition of recapture, and I would highly encourage you to ask people what theirs is when they throw the numbers out to you. But for government, we got about 50%. And I think that's a -- I declare that a really, really high recapture rate. Conventional, we got as high as 30%, and that's the highest that I've ever seen in my career. And I would -- but that I want higher, and that's why I tell our people every day. But that's -- those are kind of the numbers that are stuck in my head under my definition.

Okay. Great. I think we're out of time. Thank you.

Thank you, Terry.

Thank you.