Axos Financial, Inc. (NYSE:AX) Q3 2025 Earnings Call Transcript

Axos Financial, Inc. (NYSE:AX) Q3 2025 Earnings Call Transcript April 30, 2025

Axos Financial, Inc. beats earnings expectations. Reported EPS is $1.81, expectations were $1.74.

Operator: Greetings and welcome to the Axos Financial Third Quarter 2025 Earnings Call and Webcast. At this time, all participants are in a listen-only mode. [Operator Instructions] A question-and-answer session will follow the formal presentation. [Operator Instructions] As a reminder, this conference is being recorded. It’s now my pleasure to introduce your host, Johnny Lai, Senior Vice President, Corporate Development and Investor Relations. Please go ahead, Johnny.

Johnny Lai: Thank you, Kevin. Good afternoon everyone and thanks for joining us for today’s third quarter 2025 financial results conference call. Joining us today are the company’s President and Chief Executive Officer, Greg Garrabrants; and Executive Vice President and Chief Financial Officer, Derrick Walsh. Greg and Derrick will review and comment on the financial and operating results for the 3 and 9 months ended March 31, 2025 and we will be available to answer questions after the prepared remarks. Before I begin, I would like to remind listeners that prepared remarks made on this call may contain forward-looking statements that are subject to risks and uncertainties and that management may make additional forward-looking statements in response to your questions.

Please refer to the Safe Harbor statement found in today’s earnings press release and in our Investor Presentation for additional details. This call is being webcast and there will be an audio replay available in the Investor Relations section of the company’s website located at axosfinancial.com for 30 days. Details for this call were provided on the conference call announcement and in today’s earnings press release. Before I hand over the call to Greg, I’d like to remind listeners that in addition to the earnings press release, we also issued an earnings supplement and 8-K with additional information. All of these documents can be found on axosfinancial.com. With that, I’d like to turn the call over to you, Greg.

Gregory Garrabrants: Thank you, Johnny and good afternoon, everyone and thank you for joining us. I’d like to welcome everyone to Axos Financial’s conference call for the third quarter of fiscal 2025 ended March 31, 2025. I thank you for your interest in Axos Financial. We delivered solid results this quarter, generating over $700 million of net loan growth linked quarter, stable net interest margins and a 19% year-over-year increase in book value per share. We continue to generate high returns as evidenced by the 16% return on average common equity and a 1.8% return on average assets in the 3 months ended March 31, 2025. We deployed some of our excess capital to repurchase approximately $28 million of common stock in the quarter ended March 31, 2025 and an additional 517,000 shares of common stock for $30.3 million from April 1 to April 30 after the quarter end.

Other highlights in the quarter include net interest income was $275 million for the 3 months ended March 31, 2025, up 5.3% from the $262 million in the prior year period. Net interest margin was 4.78% for the quarter ended March 31, 2025, down 5 basis points from the 4.83% in the quarter ended March 31, 2024. We continue to benefit from a best-in-class net interest margin with and without the benefit of the accretion from loans purchased from the FDIC. Total on-balance sheet deposits increased 5.4% year-over-year to $20.1 million. Our diverse and granular deposit base across consumer and commercial banking and our securities businesses continue to support our organic loan growth. We managed our operating expenses well this quarter, with total noninterest expense for the quarter ended March 31, 2025, up by only 0.6% from the prior quarter.

Excluding the seasonal increase in FICA expenses and legal accrual reversals, noninterest expense increased slightly quarter-over-quarter. Net annualized charge-offs to average loans were 9 basis points in the 3 months ended March 31, compared to 7 basis points in the corresponding period last year. Excluding the auto loans covered by insurance, net annualized charge-offs to average loans were 8 basis points in our fiscal third quarter of 2025. We remain well reserved relative to our low current and historic net credit losses. Total nonaccrual loans declined by $66.5 million linked quarter, resulting in our nonaccrual loans to total loan ratio improving from 1.26% in the quarter ended December 31, 2024 to 89 basis points in the quarter ended March 31, 2025.

Net income was approximately $105.2 million in the quarter ended March 31, compared to $104.7 million in the December quarter. Diluted EPS was $1.81 for the quarter ended March 31, 2025, compared to $1.80 in the prior quarter. Net growth in nonpurchased loans for investment was $700 million for the month ended — the quarter ended March 31, an increase of 3.6% linked quarter or 14.5% annualized. Fund finance, equipment leasing and lender finance had strong originations and net loan growth this quarter. Headwinds from high levels of repayment of the jumbo single-family and multifamily mortgage business improved significantly with net declines of only $36 million in those 2 loan categories combined in this quarter, compared with a $384 million decline in the December quarter.

While the interest rate and competitive environment remained unstable, we feel good about keeping our jumbo single-family and multifamily loan balances flat to down $100 million per quarter versus the prior $200 million to $400 million quarterly headwind we experienced since the Fed started raising rates in 2023. Average loan yields for the 3 months ended March 31, 2025, was 7.99%, down from 8.37% in the prior quarter. Average loan yields for nonpurchased loans was 7.66% and average yields for purchased loans were 14.32% which includes the accretion of our purchase price discount. The FDIC purchased loans continue to perform well and all loans in that portfolio remain current. New loan interest rates were the following: SFR mortgages, 7.5%; multifamily, 7.3%; C&I, 7.6%; and auto 8.5%.

Ending deposit balances were $20.1 billion or up 1% linked quarter and up 5.4% year-over-year. Demand money market and savings accounts represent 96% of total deposits at December 31, 2024, increasing by 6.9% year-over-year. We have a diverse mix of funding across a variety of business verticals with consumer and small business representing 58% of total deposits, commercial cash, treasury management and institutional representing 23%, commercial specialty representing 9%, Axos Fiduciary Services representing 6% and Axos Securities which is our custody and clearing business combined, representing 4%. Noninterest-bearing deposits were approximately $3 billion at the end of the quarter, roughly the same as the prior quarter. Client cash sorting deposit balances have been volatile, increasing to over $1.2 billion during the peak of the market sell-off in March 2025 before ending the quarter around $900 million as advisers made tactical changes throughout the quarter in a turbulent market.

We’re focused on adding net new assets from existing and new advisers to grow our assets under custody and cash balances. In addition to our Axos Securities deposits on our balance sheet, we had approximately $450 million of deposits off balance sheet at partner banks. Our consolidated net interest margin was 4.78% for the quarter ended March 31, 2025, compared to 4.83% in the quarter ended December 31, 2024. Even though we deployed some of our excess liquidity and organic loan growth this quarter, we still have more deposits than we typically carry on our balance sheet. The excess liquidity was a 13 basis point drag on our net interest margin in the quarter ended March 31, 2025, down from 18 basis points last quarter. Our net interest margin remains above the high end of our target with and without the benefits of the FDIC loan purchases, largely because we’ve been able to offset the gradual decline in our earning asset yields with corresponding decreases in our funding cost.

Total interest-bearing demand and savings deposit costs were 3.59% for the quarter ended March 31, 2025, down 36 basis points from the prior quarter. We’re seeing strong growth in account and balances from our Axos ONE consumer bundled deposit product which includes a checking and a savings account. Growth in Axos ONE and other deposit businesses has allowed us to reduce our high — our cost consumer high-yield savings and wholesale funding. We continue to grow our lower-cost deposits in our commercial cash, treasury management and specialty businesses. We’re also making good progress cross-selling deposits across selected lending businesses such as fund finance and multifamily lending. Continued strong net new asset growth and normalizing in cash sorting will be a tailwind in our ability to grow lower cost deposit balances going forward.

We expect our consolidated net interest margin ex FDIC loan purchase accretion to stay at the high end of our 4.25% to 4.35% range we have targeted over the past year. Despite increased competition from banks and nonbanks driving new loan yields lower in many lending categories we compete in, we continue to win our share of new lending opportunities. Our loan pipelines have improved meaningfully in our auto and multifamily lending businesses over the past few quarters as a result of strategic actions we have taken. Better execution and expanding our distribution channels across certain commercial lending categories, including equipment leasing, have contributed to improved loan growth and pipelines. We expect loan growth to come in somewhere between the high single-digit and low teens range on an annual basis that we have targeted for the past several years.

We may have more variance from quarter-to-quarter due to uncertainty regarding the pace and timing of payoffs and the potential impact of tariffs and interest rates on loan demand. The credit quality of our loan book continues to be solid and our historical and current net charge-offs remain low. Total nonperforming assets declined by $63.3 million linked quarter, representing 79 basis points of total assets compared to 1.06% in the quarter ended December 31. The sequential decrease in nonaccrual loans was broad-based, declining by $26 million in our single-family mortgage and warehouse businesses by $15 million in our multifamily and commercial mortgage business and by $25.7 million in our commercial real estate lending business. We do not anticipate a material loss from loans currently classified as nonperforming in our single-family, multifamily or commercial real estate loan portfolios.

Our commercial real estate specialty portfolio continues to perform very well and in line with our expectations. Nonaccrual loan balances in our C&I lending portfolio were roughly flat linked quarter at $71.2 million. All C&I loans classified as nonaccrual at March 31, 2025 but 3 totaling $12.2 million continue to make contractual interest, principal and curtailment payments. We continue to monitor the credit trends across all loan portfolios and have not seen any broad-based deterioration in any individual lending category. We don’t have significant exposure to any specific industry that is expected to have an outsized negative impact from proposed or enacted tariffs. Axos Clearing which includes our corresponding clearing and RIA custody businesses had a good quarter.

Total deposits at Axos Clearing were $1.34 billion [ph] at the end of the quarter, roughly consistent with where they were in the prior quarter. Of the $1.34 billion [ph] of deposits from Axos Clearing, approximately $900 million was on our balance sheet and $450 million were held at partner banks. Client margin balances grew by 2.9%, up from $274.5 million at December 31 to $282.4 million at March 31, 2025. Net new assets for our custody business were $289 million in the March quarter, extending the positive net asset momentum we had experienced in the past several quarters. Despite a turbulent first few months of 2025, many of our legacy and new RIA clients have increased their AUM. The pipeline for new custody clients remains healthy and we expect continued organic net new asset growth in Axos Advisory Services.

Pre-tax income for the Securities business segment increased by 23.6% year-over-year to $9.1 million due primarily to better operating expense control. From a product perspective, we continue to identify ways to generate incremental fee and partners — and partner with third parties to offer additional services such as access to new asset classes and investment strategies. We are consolidating certain back office and servicing functions in our clearing and custody business to leverage the processes and systems we have to more efficiently serve broker-dealers and advisory clients. Once completed, we will have a more competitive and stable cost structure in order to expand the types of custody and clearing clients we can serve profitably. One important strategic initiative in the securities business that is the development of Axos professional workstation, our proprietary client service platform that will replace the current third-party workstations used by our clearing clients and allow better integration of banking and lending to those clients.

We leverage low-cost — low-code development to reduce the time, cost and resources required to complete our Axos professional workstation build-out. We’re also actively using artificial intelligence in our software development and across a wider set of workflows to enhance development and operating efficiency which should result in better operating leverage over time. Additionally, we’re modernizing core components of the technology infrastructure for Axos Invest, our direct-to-consumer securities trading and digital wealth management business. The primary objectives are to make the platform more flexible so we can add new products and services faster and cheaper as well as improve the customer experience by eliminating frictions caused by a reliance on third-party integrations.

We see Axos Invest as a channel for low-cost consumer acquisition and cross-sell to existing Axos clients as well as a white label offering to institutional clients such as RIAs and IBDs. Now, I’ll turn the call over to Derrick to share other further details.

Derrick Walsh: Thanks, Greg. A quick reminder that in addition to our press release, an 8-K with supplemental schedules and our 10-Q were filed with the SEC today and are available online through EDGAR or through our website at axosfinancial.com. I’ll provide some brief comments on a few topics. Please refer to our press release and our SEC filings for additional details. Noninterest expenses were approximately $146 million for the 3 months ended March 31, 2025, up by about $900,000 from the 3 months ended December 31, 2024. Salaries and benefit expenses were $74.6 million, up by $0.6 million compared to the 3 months ended December 31, 2024. Excluding the seasonal increases in FICA expenses in the March 31 quarter, salaries and benefit expenses were down by $0.8 million on a linked-quarter basis.

Professional service expenses were $8.2 million compared to the $9.1 million in fiscal Q2 ’25. General and administrative expenses were down to $6.8 million for March 2025 compared to $9.3 million for December 2024. We had a payment of a legal judgment that was previously accrued and resulted in a reduction to general and administrative expenses by approximately $2 million in the quarter ended March 31, 2025. We remain focused on managing our expenses and investments in a controlled manner in order to maintain and improve our operating efficiency ratio. Next, our income tax rate was 29% for the 3 months ended March 31, 2025, compared to 28.8% in the corresponding year ago period. We still expect our corporate tax rate to be approximately 29% to 30% with one caveat.

The California budget proposal currently includes a provision that would change the taxation of financial institutions. For tax years beginning on or after January 1, 2025, the provision, if passed, would require financial institutions to use a single sales factor for apportioning multistate income to California. Financial institutions are currently required to use a 3-factor apportionment formula which includes a corporate property factor, a payroll factor in addition to a sales factor. If the provision changing this tax apportionment from the 3-factor test to a single sales factor is enacted, the change would require the company to remeasure its deferred tax assets. Management estimates Axos’ deferred tax asset would decrease by approximately $6 million to $7 million as a result of the change.

The impact of the remeasurement will be a noncash charge recognized through continuing operations in the period which the law is enacted. If enacted, management expects the effective tax rate for the fiscal year ended June 30, 2026 and beyond, would be reduced by approximately 3% or approximately $5 million per quarter compared to the current effective tax rate. I’ll wrap up with our loan pipeline which remains healthy with $2.1 billion of total loans in the pipeline as of April 25, 2025, consisting of $576 million of single-family residential jumbo mortgage, $57 million of single-family gain on sale mortgage, $346 million of multifamily and small balance commercial, $63 million of auto and consumer and $1.1 billion of commercial loans. As Greg noted, we believe that we will be able to grow loan balances organically by high single digits to low teens year-over-year over the next 12 months, excluding the impact of the loan portfolio purchased from the FDIC or any other potential loan or asset acquisitions.

Our pipelines are up across several lending businesses and we expect the headwinds we faced from single-family and multifamily mortgages to subside. Due to elevated levels of uncertainty regarding interest rates, the economy and the shape of the yield curve, we may see more volatility in our net loan growth over the next few quarters. With that, I’ll turn the call back to Johnny.

Johnny Lai: Thanks, Derrick. Kevin, we’re ready to take questions.

Q&A Session

Operator: [Operator Instructions] Our first question is coming from Kyle Peterson from Needham Company.

Kyle Peterson: Nice results. I wanted to start off a little bit on loan growth and kind of what you guys are seeing. Obviously, there’s been a lot more volatility and uncertainty. I guess, are there any areas either that you guys are being a little more cautious in? Or on the flip side, like are there areas you guys are seeing competitors maybe be a little more cautious or shy where you guys think might be opportunities to go and take share?

Gregory Garrabrants: Yes. I think the logistics — I think there are certain C&I segments that in anticipation of the administration change or the potential administration change that we’ve been cautious about for a while. So we’ve been shying away from logistical sort of deals. Obviously, we got caught up in that one that’s still paying but is on nonaccrual. So we generally have shied away from those as far as adding a lot of exposure in that sector. And there’s been a few others that we’ve had views about from an economic perspective but those are either segments within C&I or specific companies that have unique exposure. As far as — it’s a little early to tell but I do think that we were seeing more spread compression and we’ve been able to push back against that a little bit as a result of volatility.

And I think that’s partially related to the fact that in certain instances, some of the takeouts or some of our lines or whatnot, the revolving lines are securitizations. And so to the extent you get any difference in that market that can flow through. So I think that’s positive there. But in general, the pipelines are pretty good. I would say the biggest impact it will have is just with respect to whether we can hit an 11% sort of loan growth in the year or 15%. Is this really going to be more related to prepays? And we had some periods during COVID that we were very cautious on any kind of project financing construction stuff. And so that means we sort of have a little gap where some of those loans kind of pay off and we may have a little gap there.

But I feel pretty good about where loan growth is. I mean if you look at the single-family pipeline, it’s a lot higher. It doesn’t mean every quarter it’s going to be perfect. But we’ve been really fighting against a pretty nasty tailwind to single-family and multifamily and that tailwind is over. So we may have areas like commercial specialty real estate that might have high payoff quarters just because of timing but I don’t think that’s going to be something that will occur every quarter. And so I feel okay about loan growth, actually.

Kyle Peterson: Okay, that’s very helpful. And then maybe just a follow-up, more of a housekeeping item but I noticed the fee income jumped up quite a bit this quarter, I guess. Could you just clarify whether there was anything like whether it’s seasonal or onetime or what we saw in the March quarter is a good run rate to you going forward?

Derrick Walsh: Yes, there were some — we had, had some mortgage banking impacts from the prior quarter. So that was where the prior quarter was somewhat depressed. This quarter is a little bit more representative of where we’re at. We had added BOLI in the December quarter. So that’s part of it. We did have a blip up in auto insurance recoveries which come through that line item and then there were some additional loan fees that came through unrelated to origination. So nothing necessarily of significant onetime that were in there this quarter. One item, actually — sorry, just remember, there is a fair value mark on our DTC stock that we have to hold for the clearing company, that was $750,000 that was — only happens once a year. And so that gave us a little bump up but that was only $750,000.

Operator: Your next question is coming from Andrew Liesch from Piper Sandler.

Andrew Liesch: Greg, you’ve highlighted some good investment opportunities to maybe streamline the company. I’m just curious, like what’s like the cadence of some of these investments? Is it — do you have enough revenue to keep that efficiency ratio with this 48% level? Or do you think it might step-up here in the near term?

Gregory Garrabrants: We’re going to really work extremely hard to keep it where it is. I don’t really think it’s acceptable for it to go up. And I think it’s the responsibility of me and my team to take advantage of all this amazing technology and operating efficiencies that we’ve been developing to ensure that we’re definitely keeping costs under control. So we’re targeting next year that personnel expenses will go up no more than 30% of the combination of net interest and noninterest income. I think we can do that. We want to make sure we’re getting operating leverage in the business. There’s a lot of AI improvements that can be made from an efficiency perspective. There’s — even in a company such as ours, there’s always an opportunity for people to step it up and do more.

And so I’m looking forward to cracking the whip on our organization to make all of us, including myself, run faster. So yes — so I think it’s going to be a very cost-efficient year going forward. And my team is smiling about how excited they are to make that happen.

Andrew Liesch: Got it. All right. Very helpful. And then just looking at net interest income going into the quarter, it looked like a lot of the loan growth may have come on later in the quarter. So maybe that may have been why NII was down sequentially. But if you’re looking into the — in your fourth quarter, maybe a little bit more margin compression if that continues but should we see NII step-up from here?

Gregory Garrabrants: Do you want to answer that, Derrick?

Derrick Walsh: Yes. Yes, we should. It was — to your point, from a point in time basis, loan growth was $700 million, it was closer to $100 million on an average basis during the quarter. So we expect that average this next quarter to go up and that will, in turn, increase the net interest income impact. But that’s why there was kind of a somewhat weak net interest income performance there. But as we look forward, it should return to an increasing value.

Andrew Liesch: Got it. Do you think the margin could step up here? Or there’s too much yield pressure where it could be hard to replicate 4.78%?

Gregory Garrabrants: So I asked the team to go and look at the difference between what was the net spread compression versus what was just lagging adjustments from interest rate declines. And it was about a 3 basis point net spread compression. So it’s — I would say that on average, the loans are coming in probably at a lower spread than in the past but we also still have — not that we have a lot, we still have hybrids adjusting. That will certainly happen going forward. There’s still a decent number of hybrid loans that have to adjust and things like that. So I feel pretty good about where margin is looking forward. I mean could it be a little bit down. But if we also use up some of the excess liquidity, that’s also going to move it the other way, too.

But I think that 3 basis points is an interesting number because that — if you have a flat rate environment, that was sort of — that’s the number that takes out the lagging decline just from index and floating rate loans.

Operator: [Operator Instructions] Our next question is coming from Gary Tenner from D.A. Davidson.

Gary Tenner: I wanted to ask about the sequential quarter improvement in special mention substandard loans. Can you talk to kind of where those — where that kind of improved credit rating was driven by and the color on there? I know NPAs were down quite a bit but kind of looking deeper, we’ll love some additional color.

Gregory Garrabrants: Yes. In a number of cases, some of the loans were — that were placed on substandard had payoffs that were coming that just happened to come at the end of the quarter or there was a loan being sold or refinanced that happened to hit at the end. So I think that in a lot of cases, like even if you look at the 2 large C&I loans that are on nonaccrual, they’re both still paying. They both still arguably have strong borrowing bases. So I think we’re always cautious about this stuff and we try to be conservative about how we look at it. But in a lot of cases, there really isn’t a lot of loss content there. And so frankly, a lot of those loans just paid off or — and then we did sell some of them, too.

Derrick Walsh: At par.

Gregory Garrabrants: At par. Yes. So we sold them at par. So yes, par plus accrued. So we got our interest and whatnot. So yes, it was a good quarter for that. I mean I think the reality of our real estate loans mostly with extremely small exceptions, if the borrower has something going on, the real estate is still worth so much more than what we’ve lent on it that someone wants it, right? And so we’re not really in the business of doing that, although sometimes I feel like we should be when I look at how much money people make on stuff that we sell to them but, in any event, that’s kind of what we did there.

Gary Tenner: Okay. And then another question, I guess, also around credit. In terms of the ACL build this quarter, just based on what’s in your supplemental deck, it looks like you lowered the reserve specific to multifamily and commercial mortgage by, call it, 30 basis points and increased the C&I reserve by a pretty similar amount. So if you talk about the moving parts there and the thoughts around those 2 categories?

Derrick Walsh: Yes. There were — I’ll start with C&I. The loan growth was primarily in C&I. And so that was obviously a major factor in that category. The other results were really coming from the quantitative model where — which incorporates the economic factors. So what were some of the key economic factors when you look at what was happening in kind of late February and into March was a lot of tariff fears and driving more negative outlooks on the economic factors. So when we run that through the model, what portfolio gets kind of hit a bit harder was the C&I portfolio. And so that’s what drove some of those — that combined with the growth were what drove some of those increases in that portfolio. And then the real estate portfolio still is continuing to perform.

And then some of the HPI indexes, housing price indexes, we’re continuing to remain positive or turn positive compared to where they were in the prior quarters. And so that’s what drove some of the benefits there.

Gregory Garrabrants: Yes. I mean these models are hooked up to Moody’s stuff. So if Moody’s gets in a mood, then that’s going to move the model around.

Gary Tenner: I’m sure we’ve got a lot of that to look forward to in the June quarter.

Gregory Garrabrants: Probably.

Operator: Our next question today is coming from Kelly Motta from KBW.

Kelly Motta: I guess starting off on capital. You noted you were active on the buyback here this quarter and continued into April. Wondering, especially given your outlook for what’s still pretty strong growth, how you’re viewing continuing the buyback here?

Gregory Garrabrants: Kelly, yes, look, I think we do have excess capital. Obviously, we look at where loan growth is. I think this is a nice opportunity to — where our stock is right now to be looking at buybacks. Obviously, we put our money where our mouth is this quarter and we’ll probably keep on looking at that. We don’t get opportunities like this all the time. And so we think it’s a good opportunity and we still think we can do right now good solid loan growth. We think our capital ratios are good. We’re not really seeing anything that we feel good about where nonperformers are and those kind of things. So things feel okay to be doing a little buyback. And it’s not really — we don’t ever go get too crazy and do something that’s too abrupt, right?

We kind of — we’re in there in the market and making incremental buybacks which are helpful. I mean we did — it was like almost a little bit more than 1% of the stock or about 1%, something like that for the quarter and including this month. So it’s a nice opportunity. If we see M&A type stuff floating around and maybe we won’t do as much. But yes, I think it’s a really nice opportunity for us right now. The IRR pencils really, really well when we look at our internal forecasts and stuff like that about where we think we’re going to be.

Kelly Motta: Got it. And on the M&A front, can you remind us what kinds of businesses would be top of mind in terms of being additive to Axos?

Gregory Garrabrants: Yes. We’d love to buy wealth and custody stuff but there’s not a lot out there and it tends to trade at multiples that are just so high that it makes it a little difficult. So we do look at stuff like that. And then obviously, we pay attention to what’s going on in banking but our model is somewhat unique. So we’re not the natural owner of a lot of banks, particularly if they’re heavy branch based. But there are specialty type of banks that we might look at, at different times if they really made a lot of sense. And then we look at spec thin companies if they’re the type of companies that would have bank-like credits and would be a benefit from our platform from the operational synergies we could bring and the tech and that kind of stuff.

So, we’ve — I mean, we looked at different — we looked at a premium finance business that a bank was selling. We lost the bid on it. We looked at kind of a leasing business that was a vendor leasing business, also lost that bit on price. We thought we had a good price but there’s — some people like to overpay for things. I do not. So — but we keep on looking at it. And I think when we do something, we do it in a way that has enough margin of safety in it.

Kelly Motta: Got it. That’s helpful. Last question for me, just changing gears. Greg, it sounds like you’re — you still feel good about growth even with kind of the noise and uncertainty about tariffs. I’m wondering if you could size your exposure directly or indirectly to construction and what implications rising input costs could have on maybe CRESL and other like construction elements of the portfolio. If you could just help frame that for me, that would be helpful.

Gregory Garrabrants: Yes. I mean the types of projects that we — I think the answer is not much. And the reason it’s not much isn’t because the input factors may not go up but we generally require that a substantial portion of the trades are bought out and that usually is a very high proportion. I mean it varies depending on the strength of the sponsor and our junior lender. But then you also have Subguard insurance that bonds those subs. So with respect to that, you don’t want to leave in a general instance, the idea that you’re going to have a project that could have a budget that blows out for whatever reason. And so theoretically, tariffs might be a reason but there’s plenty of things that happen in construction that have nothing to do with tariffs that blow out budgets.

And so that’s the thing you’ve got to get good at and you should make sure you don’t do it. I mean I think where it’s harder is, frankly which we don’t have a lot of is when you have a smaller mom-and-pop builder and they’re doing a $5 million multifamily project or something. And so having institutional level GCs with very strong balance sheets that are able to get all the subs bought out and then buy insurance for the subs and whatever, that’s a much different kind of lending that really doesn’t have anything to do with CRESL. So it’s a very different kind of risk that we bear with respect to that because of how we structure our deals.

Operator: [Operator Instructions] Our next question is coming from Edward Hemmelgarn from Shaker Investments.

Edward Hemmelgarn: Greg, I got a couple of questions. One is you seem to be a higher level of conservatism, both in your allowance for loan losses but also your — I think your equity. I’d like to see the share repurchase, that was good. But your equity as a percentage of the asset base was higher than I’ve ever seen it. Are you trying to — I mean, are you just being conservative? Or are you anticipating perhaps rougher times or what?

Gregory Garrabrants: Well, you know, I am a conservative guy as we’ve discussed many times. No, look, I think there’s a couple of things going on. One is that if you looked at our balance sheet when you first became enamored with us, we were much more of a 50% risk-weighted shop. And so that’s changed over time, right? So you sort of — you get — if you have a much higher percentage of 50% risk-weighted assets and you’ve got that leverage and that risk-weighted ratio tends to have a bigger disparity. And then as we’ve added fund finance and C&I and all these different categories and we’ve kind of shied away — single-family has not shied away, it has become a smaller portion of the portfolio, that ratio has sort of kind of moved together.

And so that’s an element that you’re seeing. That’s not all of it, though. I mean I think there is a recognition that we obviously want to continue to make sure that we have a very strong balance sheet. So that is a piece of that as well. So generally, in comparison with a longer period ago, we are targeting a higher equity ratio. But I feel like we’re good where we are. And as I said, credit performance looks to be — continue to be solid. So I think we are able to buy back stock and we had good growth this period. So I think we’re doing an appropriate job of balancing — taking appropriate risks and making sure that we’re really protecting the institution.

Edward Hemmelgarn: Yes. I was looking at the — just look at the loans balances and they seem to be getting better in terms of current payments and so forth as opposed to worse but I was a little surprised by the loan loss provision for the quarter. The other thing you…

Gregory Garrabrants: Yes. I mean what I would say on the loan loss — what I’d say on the loan loss provision, Ed, is you have to remember is that with CECL, what happens is there’s external inputs for us that are not — they’re marginally related to anything with respect to do to our loan portfolio. So if we have — so if Moody’s says the probability of recession is greater, that’s going to increase our loan loss. Now whether or not that ultimately ends up meaning anything for us, I mean, I can have my doubts but that’s — it’s a model-based part of it.

Derrick Walsh: And the one other thing is that certainly, it comes into effect a little bit a portion, what’s happened the last 3 months or what have you, right, from quarter-to-quarter. But the whole idea of CECL is the lifetime of the loan so that you’re going — you’re projecting forward multiple different waves of economic ups and downs, right? So you’re not trying to necessarily move the allowance massively based on, well, nonaccruals were up this quarter but they’re down next quarter, you’d have loan losses going kind of spiking up and down across the industry if you were solely using that as your one input, right? To Greg’s point, there’s a lot of different inputs that come into play to help inform the allowance and the provision each quarter.

Gregory Garrabrants: Look, I mean, that being said, we are — I think we’ve done — I think it’s a good thing that we can be as profitable as we are and continue to increase and make our loan loss balance where it is. I think that’s a positive. It is — I think it’s — I’ve talked to a lot of pretty smart people over the last couple of months about where they see things going. And I think the one commonality is everyone agrees that there’s just more volatility, right? There’s more potential for volatility and a higher sort of distribution of potential outcomes. So I think that’s reasonable that the model predicts that.

Edward Hemmelgarn: Yes. Okay. That’s fine. I mean I also appreciate the conservatives. The other thing is just about your spending I’ve noticed has really gone up for IT and data processing and so forth. I mean — and you briefly talked about that. You saw some real opportunities there. Can you talk about that a little bit more? I mean maybe some examples of where you’re seeing opportunities to really deploy AI and so forth?

Gregory Garrabrants: Yes. So we just released and we’re getting ready to do a transition for our clearing clients away from a couple of old workstations that are quite common in the industry but have been around for a long time and are sort of universally aided and replacing it with our own workstation. And we use the low-code platform to do that. And it, I believe, took, I think, about maybe 50%, 60% of the resources and about half the time to get that product out. So we do think that there’s a lot of really interesting AI opportunities happening in the software development life cycle. And we’ve seen some stuff that’s quite extraordinary. The ability to lift and shift old code that is in basic or in some sort of old language and be able to refactor it much more quickly or to be able to extract and document the code which would normally require somebody who is very skilled at reading the code and documenting it and all those things.

The documentation of code, for example, is becoming much more able to be done by artificial intelligence. The ability to take a plain language business requirements document and bring it into a set of stories and a set of documents that can be utilized by developers to code is a lot greater, too. So there’s a lot of that. We’re using software that can take a document that has unstructured data and appraisal or whatnot and let’s say, you had to pull 100 fields out of it, the AI can pull those fields out and put them into structured data, right? So those are just some examples. I give you a lot of them. I mean we’re really working hard on this. We have an AI task force. I think we have to hold ourselves accountable for actually seeing that in the results which to me means that you have for each dollar you earn, you don’t spend as much on people and technology, right?

And that’s the way I think you have to be able to eat AI. And I think it’s possible. It’s not always easy at every step of the way but there’s a lot of opportunity. I mean there really is. We have a good strategy there and we’re not fully at the Agentic AI sort of level yet but we’ll have to keep on pushing for that.

Operator: Thank you. We reached the end of our question-and-answer session. I’d like to turn the floor back over for any further or closing comments.

Gregory Garrabrants: Thank you, everyone. We’ll talk to you next quarter. I appreciate your interest.

Operator: Thank you. That does conclude today’s teleconference and webcast. You may disconnect your line at this time and have a wonderful day. We thank you for your participation today.