Flagstar Bank, National Association Q3 FY2020 Earnings Call

Good day, and welcome to the Flagstar Bank Third Quarter 2020 Earnings Call. Today’s conference is being recorded. At this time, I would like to turn the conference over to Mr. Ken Schellenberg. Please go ahead, sir.

Thank you, Sherry, and good morning. Welcome to the Flagstar third quarter 2020 earnings call. Before we begin, I would like to mention that our third quarter earnings release and presentation are available on our website at flagstar.com. I would also like to remind you that any forward-looking statements made during today’s call are subject to risks and uncertainties. Factors that could materially change our current forward-looking assumptions are described on Slide 2 of today’s presentation, in our press release, and in our 2019 Form 10-K and subsequent reports on file with the SEC. We are also discussing GAAP and non-GAAP financial measures, which are described in our earnings release and in the presentation we made available for this earnings call. You should also refer to these documents as part of this call. With that, I’d like to now turn the call over to Sandro DiNello, our President and Chief Executive Officer.

Thanks, Ken, and good morning to everyone listening in. I hope all of you and your loved ones have been able to stay safe and healthy. I'm joined this morning by Jim Ciroli, our Chief Financial Officer; Lee Smith, our President of Mortgage, and Reggie Davis, our new President of Banking. As you may know, we had a bit of a change in our leadership team during the quarter. Reggie joined us on August the 3rd, and I'm thrilled to bring a person of his stature in the banking industry to Flagstar. He could have gone to a lot of banks, but chose to join us. Further, Kristy Fercho, our President of Mortgage, left Flagstar on that same date. While we hated to see Kristy go, fortunately, we had a great internal solution. Lee Smith, who has been with me from day one of Flagstar's turnaround, was already overseeing mortgage fulfillment as part of his COO responsibilities. Plus he was also overseeing mortgage servicing. So it made all the sense in the world to place him in the role of President of Mortgage, allowing us to have originations, fulfillment, and servicing under one highly respected leader. Given these leadership changes, Jim will expand his remarks a bit today to cover some of Lee’s former areas. Reggie will follow with an update on the Community Bank, and then Lee will handle the mortgage segment, including servicing. As usual, we'll then open the line for questions. A few more comments about Reggie. He's a talented, versatile, seasoned banker, with a proven track record of success. As many of you are aware, diversifying our earnings has been an ongoing key strategy, and growing our Community Bank is a pillar of that strategy. And we've made a lot of progress on that strategy, and therefore feel very fortunate to have Reggie onboard to move us forward, not just in Community Banking, but in the company as a whole. Also during the quarter, we announced that David Hollis joined our team as our new Chief Human Resources Officer, replacing Cindy Myers, who retired. While we all hated to see Kristy leave, David is doing a great job of filling her shoes. His strong background in both financial services and other industries has demonstrated results in leveraging HR to improve organizational performance, and the depth and breadth of his experience in all aspects of HR, including diversity, equity, and inclusion, will help ensure Flagstar stays in the forefront of this important field. As for our performance for the third quarter, it was simply unprecedented, exceeding our second quarter performance, which at the time I said, was the best in Flagstar's history. Not that long ago, I would have been thrilled to earn $3.88 per share in an entire year, let alone in a quarter. This was an all-hands on deck performance, led by mortgage, followed by a phenomenal warehouse business, and supported by complimentary servicing business. In fact, the first nine months of this year were also unprecedented, with tangible book value per share growing by $7.03, a stunning 25%. There's a story here that we don't want overshadowed in our results, and that's how successful we were at growing net interest margin, excluding the impact of loans with government guarantees. We were pretty much alone amongst our peers in expanding them, and I think you can put that at the feet of our unique business model and the way we've managed it. We had the foresight to protect the yields in our warehouse portfolio with floor rates, and these were yields already enhanced by significant revenue from drop fees. The result was an overall increase in adjusted net interest margin from 2.88% to 2.94% quarter over quarter. And you'll recall that credit losses in the warehouse portfolio are de minimis, less than $5 million over the last 12 years. The other important story from the quarter is the aggressive way we continued to manage deposit costs, with our average cost of deposits dropping again in the third quarter, a trend that has continued throughout 2020. And because we carry a large amount of wholesale funding where costs are tied to LIBOR, these costs have dropped in tandem with Fed actions. Importantly, we think our margin will remain resilient. I’m also proud of the success that the retail team has achieved with respect to deposits. Average Community Banking deposits, which excludes custodial accounts and broker deposits, increased by $0.5 billion during the quarter. On to mortgage. What can I say? $346 million in gain on sale revenue, $11 million more than all of last year. Pretty amazing. And then quietly, the servicing business just keeps chugging, supporting our mortgage business, providing efficient funding, adding fee income, and even growing the number of sub-service loans in the face of historically high payoffs related to the current refinance boom. Finally, let me address credit. So far, everything is holding up. In fact, commercial loan deferrals now total only $47 million. We're encouraged by that, but cautious and staying close to borrowers, especially those who are now on tap to start making payments again. As always, our watch words are to be careful and conservative. That's why we increased our coverage ratio again this quarter, not because we see any weakness in our books. In fact, just the opposite, but your guess is as good as mine as to what the future holds for the economy. So we choose to be conservative. I don't know how I could feel much more satisfied with where we stand. We've moved into Q4 with a stable interest margin, outstanding power to generate non-interest income, and a fortress balance sheet. All things considered, we're in the best possible position. As we hope for the best, we are prepared for the worst. Let me now turn it over to Jim.

Thanks, Sandro. Turning to Slide 6, net income this quarter was $222 million or $3.88 per share. This performance compared to the $116 million or $2.03 per share last quarter. The increase on a linked quarter basis was largely due to stronger mortgage results, and a nice increase in net interest income, the result of higher warehouse lending, and a lower credit provision score. Diving deeper into this quarter's performance, our pre-tax pre-provision earnings were $327 million, compared to $250 million last quarter. Net interest income increased $12 million or 7%, while average earning assets grew $2.0 billion, and the reported net interest margin decreased by 8 basis points. These were impacted by loans with government guarantees that have not been repurchased, and are only the result of an accounting gross-up. Excluding these assets, the net interest margin actually increased six basis points. This performance was primarily driven by the strength of our warehouse business that has rate floors in place to protect from margin compression, and our core deposits, which benefited from higher custodial balances, and also from the maturity of higher cost CDs, and the repricing of promotional rates on savings accounts. We'll review these numbers in a couple of slides. Mortgage revenues were $358 million, an increase of $63 million compared to the very strong number we reported last quarter. During the quarter, we saw gain on sale margins increase. Asset quality remained strong. Net charge-offs were only five basis points, and early stage delinquencies stayed low. Non-performing loans ticked up slightly, as we had one commercial credit of $10 million that we put on non-accrual status. Despite all of this, we increased the allowance for credit losses to $280 million, up from $250 million at the end of the second quarter. This reflects our views on the continued uncertainties within the economy. As a reminder, the allowance for credit losses or ACL includes the reserve for unfunded loan commitments. We'll provide more details when we get to the asset quality slide, and take a deeper dive into CECL. Capital also remained solid. Total risk-based capital was 11.3% at September 30. Our CET1 ratio was 9.2%, and our tier one leverage ratio was 8.0%. We allowed these capital ratios to come in at slightly lower levels than where they otherwise would be, to support higher levels of warehouse loans. We don't expect any losses coming from these assets, so there is a lower need to support them with capital. We'll go into more details on our capital ratios later. Finally, we continued to demonstrate significant capital generation, with growth in our tangible book value per share to $35.60 at quarter end, up $3.86 from June 30, and $7.98 from one year ago, a 29% increase. So let's turn to Slide 7, and dive deeper into the income statement. Net interest income increased $12 million to $180 million this quarter, up 7% from last quarter. Average earning assets grew 9%, led by warehouse lending and a full quarter’s balance of loans with government guarantees that are reconsolidated due solely to forbearance. Deposit costs came down 15 basis points, while average deposit balances increased $2 billion. We'll dive deeper into net interest income and our interest rate position on the next slide. Non-interest income increased $74 million to $452 million, due to higher mortgage revenues and servicing related fee revenue. Non-interest expense was $305 million, $9 million from the prior quarter. Finally, you'll notice that the tax rate tipped up this quarter. This is a function of a higher level of pre-tax income, which is much more than the tax benefits we have in our run rate earnings. With our higher level of earnings, we decided to delay certain tax planning strategies, as we expect that they will provide more value should corporate tax rates be higher in the future. Considering this, it would be appropriate to use our year-to-date effective tax rate of 23% for future periods. On Slide 8, beginning of this quarter, we've combined our average balance sheet with details on our interest rate risk position at the end of the quarter. Average earning assets increased $2 billion from last quarter. This resulted from a $1.9 billion increase in warehouse loans, and a $1.3 billion increase in the loans with government guarantees that have not been repurchased. Declines in securities of $0.6 billion and in mortgage loans held for investment of $0.2 billion were due to faster prepayments and partially offset the balance sheet growth. C&I balances also declined by $0.4 billion, with the sale of the PPP loan portfolio at the end of July. We expect loans with government guarantees to peak this quarter, and we will start to see balances gradually decline through the rest of 2020 and through 2021, where we own an MSR for Ginnie Mae loans. We have the option to repurchase these loans after the loans have gone three months without a payment, due either to delinquency or forbearance. At September 30, we had $1.8 billion of such loans we had not repurchased, which is consistent with what we indicated on the second quarter call, but which the accounting rules made us reconsolidate onto the balance sheet, with an offset to other liabilities. We do not believe there is significant downside to holding the loans, either by buying them or through this accounting gross-up. If we were to repurchase the loans, we can pledge the loans to the FHLB, and they are at 20% risk weighted assets. Further, if we do repurchase the loans, we could resell those loans at a later date, which is attractive for us, and they remain government guaranteed. Average deposits increased $1.8 billion last quarter. Custodial deposits drove $1.1 billion of this increase. We also saw a growth of $0.3 billion in non-interest bearing retail deposits, a 16% increase from last quarter, and a $0.3 billion increase in government deposits due to seasonal tax payments. We managed deposit costs lower by 15 basis points from the impact of pricing changes we've been making since March, when the Fed dropped short-term rates. Additionally, as we observed last quarter, deposits continued to reprice to the new curve environment, which provided support for our net interest margin expansion. While it's difficult to predict where rates might be in the future, we feel that our interest rate risk position is in a good place due to the actions we've taken in this lower interest rate environment. We feel that we can protect our net interest income and net interest margin, and believe that our net interest margin should be relatively flat to where it's been the last two quarters. The interest rate floors that we have in the commercial loan portfolio should protect us against further margin compression. We've completed a $2 billion program to lock in these low rates on our funding costs for the long term, using a combination of interest rate swaps and fixed cost purchase money, laddering that out for three to seven years. While this will make us more asset-sensitive in our banking business, our mortgage origination business is liability sensitive. So we believe this program will set us up for future success, regardless of where rates are. We continue to have a strong liquidity position, driven by the strength of our deposit base and access to multiple sources of liquidity, both on balance sheet, with our high quality securities portfolio, and off balance sheet with our undrawn FHLB facilities. At September 30, we had ready liquidity of $5.6 billion, not including the ample access we have to borrow with the Fed discount window. Let's now turn to Slide 9, which details our non-interest income and non-interest expenses. Non-interest income rose $74 million from the prior quarter on the strength of mortgage revenue. Our gain on sale revenue of $346 million represented an increase of $43 million, while our adjusted locks increased 9% to $15 billion. And the gain on sale margin was 231 basis points. While channel margins did come down slightly from last quarter, the overall margin expanded slightly, due largely to the gain we booked on our RMBS transaction this quarter, and other execution improvements. We also recognized a return of 12 million on our MSR. The current quarter represented a more normalized run rate as last quarter we made certain model changes that we believe are congruent with the economic forecast we use for CECL and for higher prepayments. These results were achieved even with the average value of the MSR falling two basis points from 85 basis points of UPB at the end of the third quarter. We would observe that the MSR market continues to show signs of improvement, as demonstrated with the small MSR sales that we executed in August. Loan admin income improved $5 million due to a decline in the LIBOR based credit that we provide our sub servicing customers for the custodial deposits that they control, and from a higher level of fees for loans in forbearance. Non-interest expense increased to $305 million for the third quarter, compared to $296 million last quarter, reflecting a $10 million increase in non-mortgage related expenses. This increase was primarily due to the capitalization of origination costs in the second quarter for the PPP loans, and the accelerated vesting of certain components of executive compensation that resulted from the recent secondary share offering. Despite increased volume, mortgage expenses were flat quarter over quarter, as the ratio of mortgage non-interest expense to closings for mortgage expense ratio actually declined. This improvement was due to certain expenses in the second quarter that did not recur this quarter, and are not expected to recur in the future, including certain performance-related incentives related to our Opes Advisors division. So let's turn to asset quality on Slide 10. Credit quality of the loan portfolio remains strong. Early stage delinquencies continued to be relatively low. Only $14 million of total loans were over 30 days delinquent, still accruing as of September 30, relatively flat from $15 million at June 30. Non-Performing loans ticked up slightly because we had one commercial credit at $10 million that we put on non-accrual status. The OCC completed its review of shared national credits recently, which resulted in no rate changes for our loans. Our allowances for credit losses covered 1.7% of total HFI loans. Excluding warehouse loans from the denominator, given their relatively clean credit loss history, and considering that substantially all of these loans are collateralized with agency or government backed loans, our coverage ratio would stand at a very strong 3.1%. On Slide 11, we can see that we ended the quarter with $280 million of ACL, consisting of $255 million of allowance for loan losses, and $25 million in the reserve for unfunded loan commitments. In total, our ACL at quarter end increased by 12% over what we reported at the end of the second quarter. This quarter, we continued to use three different Moody's forecasts for the next two years to guide our allowance level: S1 growth forecast weighted at 30%, a baseline forecast weighted at 40%, and an S3 adverse forecast weighted at 30%. All forecasts used the September release. The resulting composite forecast for the third quarter was roughly equivalent to the scenario we used in the second quarter. Unemployment ends the year at 10%, and recovers only slightly in 2021. GDP recovers only slightly by the end of the year from current levels. It does not return to near pre-COVID levels until 2024. HPI drops about 2% from mid-2020 through 2021. While there are signs that point to a possible V-shaped recovery, we're going to be cautious in our confidence about the recovery until we see more success from the medical side of combating this pandemic. Accordingly, we have qualitative reserves of $63 million, primarily in our CRE and C&I portfolios, guided by the CECL model output using Moody's adverse scenarios to provide coverage for industries and customers that we believe could be more exposed to the stressful conditions in our forecast. We’ve provided a portfolio-by-portfolio breakdown of the resulting ACL coverage ratios in our appendix. On Slide 12, we've updated our exposure to those industries we believe are more likely to be most impacted. In total, we have $1 billion of outstanding loans in this category, representing 5.7% of our total loan portfolio. It's interesting to note, we have almost no loans in deferral in these portfolios today, down even from September 30. In our commercial and industrial loan portfolio, the COVID-impacted loans totaled $0.3 billion. You can see that the exposure here is relatively low, especially as deferrals in these portfolios total only $2 million. We have no oil and gas exposure. In our commercial real estate portfolio, we have $0.7 billion outstanding in the areas most likely to be impacted by COVID, including commercial real estate loans secured with hotels, retail properties, and senior housing. Of loans in this category, our average pre-COVID LTV was 55%, and our average pre-COVID debt service coverage ratio was 1.6 times. We still don't have any loans in these portfolios that we believe will default. In our hotel portfolio, we're seeing occupancy levels of 53%. Now, this is not quite the point of covering debt service, but at this level of occupancy, their cash burn slows considerably. While we believe that we will have losses, we continue to see strong borrower support across the portfolio. We feel good about our credit risk in this portfolio, as we're starting from a position of strength from our carefulness about who we lend to, to the disciplined underwriting of those credits, and the pre-COVID LTVs and debt service coverage ratios in the commercial real estate portfolio. Turning to Slide 13, our capital ratios remain solid and nicely above our stress buffers. Total risk-based capital was 11.3% at September 30, and our CET1 ratio was 9.2%, both relatively unchanged from the prior quarter, despite $2 billion of asset growth. As expected, our tier one leverage ratio of 8.0% increased 28 basis points this quarter. If we just weighted our warehouse loans at 50% — they're weighted at 100% under the current risk-based capital rules — you'd see that our capital ratios compare favorably to most other mid-sized banks. Now, this makes sense. The loans are fully collateralized by 50% risk weighted assets, and those assets remain under our custody, while the loans are on our lines. Further, there is even an outstanding proposal to make this distinction in the risk-based capital rules, a proposal that we wholeheartedly support. So, adjusting the risk weighting on the warehouse loans to 50%, our total risk-based capital would be 13.4%, over 200 basis points higher, which would put that ratio above the average for all mid-sized banks. Our CET1 ratio would be 10.9%. As we pointed out, with our warehouse loan portfolio of loans held for sale and loans with government guarantees, we have more than half our balance sheet dedicated to asset categories that have very little risk. Loans held for sale turn over every one to two months, and this portfolio was carried at fair value. The portfolio of loans with government guarantees has no real downside, and perhaps a modest upside. When you take all of this into consideration, we believe that we are operating at strong capital levels, given our low risk balance sheet composition. I’ll now turn it over to Reggie to cover Community Banking.

Thank you, Jim, and good morning. As this is my first opportunity to be in front of many of you, I'd like to comment on the potential I see in the Community Banking business at Flagstar. What excited me about coming to Flagstar was the tremendous potential of this franchise. Looking at this company from the outside, I've been impressed with the way that the company has been able to grow its loan book in a well-diversified manner, with a focus on building advisory relationships with our clients. Now, having been inside for nearly 90 days, I'm even more impressed with the energy and commitment of the Flagstar team. This team is on a very deliberate journey to build a best-in-class, client-focused organization. The two deposit acquisitions meaningfully transform the retail deposit base to where we can focus more on full relationships. This will continue to be our focus as we build out an omnichannel experience for our retail clients. In the near term, we want to focus on sustaining the strong performance we've seen from the warehouse lending team, and take this opportunity to improve our productivity across all businesses in Community Banking. Please turn to Slide 15. Quarterly operating highlights for the Community Banking segment include: average warehouse lending balances increased $1.9 billion or 51% to $7.6 billion in the quarter, due to the low interest rate environment driving strong refinance volume. Our relationship-based approach, and speed of execution, also enabled us to add new clients, as well as increased lines for existing customers during the quarter. We continue to maintain our disciplined underwriting in this business. Average commercial and industrial and commercial real estate loans decreased $450 million or 9%, with the decrease being driven predominantly by the sale of PPP loans that we closed in July. We continue to be thoughtful in terms of new facilities, and believe our strong credit policies and diversified portfolio will be a strength as the fallout from the pandemic becomes more apparent. Average consumer loans held for investment decreased $219 million or 5%, a result of increased payoffs in our first lien mortgage portfolio, partially offset by growth in other consumer loans, which is predominantly our indirect Marine and RV loan portfolio, which has performed rather nicely in this environment. I'm also proud of the success that the retail team has achieved. Average Community Banking deposits, which include custodial accounts and broker deposits, increased $0.5 billion or 5% over the last quarter to $10.3 billion. We saw nice growth in balances and governmental deposits due to seasonal tax collections, non-interest bearing DDA, and low cost savings accounts. We also saw CD balances contract $0.3 billion. The overall cost of these deposits declined by 22 basis points to 42 basis points, from 64 basis points last quarter. The retail team did a great job in retaining CDs where they're maturing, and redeploying these deposits into DDA and savings accounts. Turning to commercial lending on the next slide, we continue to manage our well-diversified commercial loan book. In the warehouse lending book, we've been using our core bank balance sheet to accommodate the needs of our customers, despite this having a direct impact on our period end capital ratios. Through October 19, we've averaged $7.4 billion, demonstrating these efforts to sustain the business growth are succeeding. In commercial real estate, we're in constant contact with our customer base. The home builder book has performed beyond our expectations, the result of strong management teams and the close working relationship that those teams have with our lenders here at Flagstar. The C&I book remains well diversified, and we're starting to see our customers get their business back on track. We're taking steps now to build our relationships in markets so that we can be in a position to fully serve these customers when the opportunities present themselves. I'll now turn things over to Lee.

Thanks, Reggie, and good morning, everyone. We couldn't be more pleased with how our mortgage and servicing businesses are performing right now, providing significant and valuable non-interest fee income in this low interest rate environment. The $346 million of gain on sale revenue generated during the third quarter was a record for Flagstar, as mortgage banking revenues increased an incredible $63 million or 21% quarter over quarter, as we continue to take advantage of the strong refinance market. Furthermore, we ended the quarter servicing or sub-servicing just over 1.1 million loans, an increase of 6% from the previous quarter, as we added over 100,000 non-Flagstar originated loans to our best-in-class servicing platform. Our unique one-stop shop mortgage business model allows us to originate mortgages across multiple TPOs and retail channels, provides optionality in how we sell or distribute the loans, and also enables us to sell the mortgage servicing rights created, and retain the sub-servicing on those loans. This is complemented by our warehouse and mortgage lending capabilities. And the custodial and escrow deposits generated by the servicing business also help us fund that balance sheet. This model allows us to build deep partnerships, remain nimble and flexible, and maximize earnings throughout the mortgage life cycle. I will now outline additional key operating metrics from our mortgage and servicing segments during the third quarter. Please turn to Slide 19. Quarterly operating highlights for the mortgage origination business include: full out adjusted lock volume increased 8% to $15 billion quarter over quarter, while the net gain on loan sale margin increased 12 basis points to 231 basis points. As a result, gain on sale revenues increased a significant $43 million to $346 million in the quarter. The majority of our lock volume growth was seen in our higher margin retail broker and non-delegated correspondent channels. Refinance activity accounted for 67% of our lock volume during the quarter, and retail accounted for 31% of lock volume. We continue to use margin as a lever to keep volume in check with capacity, and ensure continued exceptional service for our customers. Mortgage closings were $14.4 billion in the third quarter, a 19% increase from the previous quarter, as we continue to add underwriting and fulfillment capacity given the increased production volume as a result of the low interest rate environment. Our mortgage operations team continues to operate effectively in this work-from-home environment. We haven't seen any degradation in productivity during the pandemic, and we continue to hire and train new fulfillment staff, therefore building capacity throughout. We also maintain our disciplined approach to the types of products being originated, as effectuated at the outset of the pandemic, where we moved to stop originating higher-risk products, and tightened the credit box in certain areas to protect our position and minimize any future write-downs or losses. At period end, we have approximately $2.5 billion in Ginnie Mae early buyouts on our balance sheet. Of this, approximately $1.8 billion were a result of borrowers opting into forbearance as a result of the pandemic. The accounting consequence of owning the MSR is to show them as early buyouts, whether you buy them out or not. And therefore the biggest impact for Flagstar is against capital. These loans were all performing before the pandemic, and we believe a significant number will go back to making payments and get reinstated after the forbearance period expires, either on their own or through the partial claims process. Once the loan is reinstated, it's no longer categorized as an early buyout. We believe a small number will get modified outright, or modified in conjunction with a partial claim, at which point we will buy them out and re-securitize the loan, realizing the gain on sale benefit of doing so. Given the increase in home prices over the last few years and equity most owners have in their homes, we don't anticipate many borrowers going into foreclosure following the end of the forbearance period. The overall impact to Flagstar of this asset class is somewhat neutral. It does create an operational need to work through these loans. And as I mentioned, it impacts capital, given the accounting rules and recognition. Loans will only be bought out if we can modify the loan and re-securitize, therefore realizing the gain on sale benefit that would be generated. Finally, we expect volume and margin to decline in the fourth quarter because of the usual seasonality impacting the number of business days, and the winter months affecting the purchase market in particular, and we forecast gain on sale revenues to be approximately $200 million in Q4. We're very pleased with the performance of our record-setting mortgage business in the third quarter, and believe it will continue to be a meaningful contributor to the bank's earnings in future periods. Moving to servicing, quarterly operating highlights for the mortgage servicing segment on Slide 20 include: we ended the quarter servicing or sub-servicing approximately 1.1 million loans, of which almost 894,000 or 81% are sub-serviced for other MSR owners. Of the 1.1 million loans we service or sub-service, 93% are backed by Fannie Mae, Freddie Mac or Ginnie Mae. The number of loans serviced or sub-serviced increased slightly in the quarter, as we added in excess of 100,000 non-Flagstar originated loans. And despite the high levels of refinance activity, we're able to replace runoff with new loans from our mortgage origination business, another advantage of our business model. Today, we have the capacity to service or sub-service 2 million loans, as well as provide ancillary offerings, such as recapture services and financing solutions to MSR owners. If you look at Slide 38, you will see that we are generating $5 million to $7 million of operating profit before tax for every 100,000 loans we add to the platform. This is an increase from our previously reported $4 million to $6 million, as we continue to achieve economies of scale benefits in this business. As it relates to forbearance, through September 30, 112,427 borrowers, representing 10.6% of the first lien mortgage portfolio that we need to service or sub-service, have requested forbearance relief because of COVID-19. We’ve seen a significant decrease in new forbearance requests since the peak weeks at the outset of the COVID pandemic. Interestingly, 28% of those borrowers who have requested forbearance have continued to make their monthly payments through September 30, and have not taken advantage of the forbearance option. This effectively means that right now, 7.6% of the loan book we service or sub-service are actually using forbearance. As part of the forbearance period, we're also waiving certain fees and there will be no negative reporting to the credit bureaus. The peak number of loans in forbearance was 129,332. And as of September 30, that number has declined by approximately 17,000 or 13%, as borrowers who had initially opted in, have opted out, paid off their loan, reached out to say their hardship has been resolved and their loan is current, or had their loan modified. During the third quarter, we have onboarded approximately 10,000 non-Flagstar originated loans that were in forbearance. So comparing period over period forbearance activity isn't as meaningful. During the quarter, we sold $800 million in flow MSR deals. The market for MSR is certainly coming back after it dried up at the outset of the pandemic. And our MSR to CET1 ratio is currently 16%, significantly below the 25% threshold before it becomes capital punitive. Finally, custodial deposits averaged $7.3 billion in the third quarter, an 18% increase compared to the prior quarter. Again, this is just another benefit we get from our subservicing business, as it provides liquidity that helps fund our balance sheet. Our subservicing business continues to flourish and be successful. When combined with our mortgage origination capabilities, we believe the scale and quality of both operations give us one of the most valuable mortgage business models in the industry. This concludes our prepared remarks, and we will now open the call to questions from our listeners.

Operator Instructions: At this time, we will open the floor for questions. Our first question will come from Daniel Tamayo from Raymond James.

Hey, good morning guys. Congratulations on a great quarter. I just wanted to talk a little bit about first, the sustainability. I asked this last quarter, but clearly holding up the sustainability of the warehouse yield — if you're still able or given kind of the competition there, but if you're still able to charge these higher rates for clients using the period end balance sheet, or if there's some other factor going on there and if you envision that changing at all going forward?

I do not envision a change, and we have not had to change any of the rates that we are charging. As you probably know, Danny, most, if not all of these warehouse lines, are at their floors. So I think the sustainability of the yield is high. In terms of the balances, as Reggie noted, through yesterday or so, we're at $7.4 billion average balance thus far in October. So that's hanging in there as well. I think growth will be measured against our capital growth, but the opportunity there going forward is pretty strong. Let me see if Reggie has anything he wants to add to that.

No, nothing to add. I mean, we've got a fairly disciplined approach. We like that business. We think we can get paid for it, and we're seeing more and more opportunities as our competitors struggle to execute.

All right, terrific. And then maybe just stepping back a little bit, you know, as the potential for a steeper yield curve comes into play here, could you talk about what levers you have to pull overall in order to mitigate that, how that might impact the business overall? And then you've talked a lot about capacity being outstripping demand on the volume side in mortgage for the last several quarters. How much of a delta is there that if you do see a steepening yield curve and lower demand before you would see a significant reduction in volumes on the mortgage side? Thanks.

Well, it's hard to look out that far into the future, but I’d say this: we've got some runway in the mortgage business, and we're going to continue to take advantage of that as long as we can. Should we see rates start going the other way and a steeper yield curve, other pieces of the business would kick in that might mitigate any reduction we'd see in the mortgage business. Let me see if Lee or anybody else wants to add anything to that.

I agree with those comments, Danny. You really have to look at the balanced model we have. And while that could crimp the mortgage business a little bit, think about how well the servicing business would do in that kind of an arrangement, and think about how the Community Banking business is going to perform in that scenario. We're not able to predict where rates are going to be, but we've positioned the company to do well regardless of where they're heading.

Yes. And Danny, this is Lee. I want to emphasize that the diversified model — we've always said we can be successful whatever happens to interest rates. On the mortgage side, 70% to 75% of our expenses are variable or semi-variable. As volume goes down, you'll have natural flex in the expense base as well.

That's terrific. I appreciate all the color. Maybe just the last one on that — on the expenses. As you mentioned, lower than they've been running in the quarter related to mortgage. You mentioned the Opes Advisors incentives have come out. Is that just the driver there and do you think that percentage is going to be lower than what we've seen prior to this quarter going forward?

Yes. The main driver for that, Danny, is the Opes earn-out coming to an end. You won't see that going forward. In terms of where you’re seeing the run rate on mortgages, we have a chart — I think it's Slide 42 — that gives a sense of where you can expect expenses as a percentage of closings.

Terrific. Thank you for all the color.

Our next question will be from Scott Siefers from Piper Sandler.

Good morning. This is actually Jamie Benjamin on for Scott. Thanks for taking my questions. First, gain on sale margins held up very well, with some benefit from the securitization transaction this quarter. I was wondering if you could clarify what that benefit was this quarter.

How much was the benefit from the RMBS?

We haven’t actually disclosed the exact amount. It was sizable, but it was only maybe half of the benefit. The other half was really just what I call other execution improvements that we made on the secondary side.

I'll just add that half of it was the RMBS. We got better execution around the hedge, and we did see an increase in our non-delegated correspondent channel quarter over quarter from a margin point of view. Those were the three drivers.

Perfect. Thank you. And then one follow-up. Appreciate the clarity around the decrease in deposit costs. Could you tell us a little bit about what additional levers you could flex on the net interest margin? Whether that's balance sheet movements, more deposit cost leverage, et cetera. Thanks.

We feel pretty confident that the yields are going to hold because of the influence of the warehouse side. On deposits, there's still some runoff from higher rate CDs that are going to come into play and provide additional repricing opportunities.

One of the things we've been pleased with this year is that as CDs have matured, we've been able to move that into lower cost products. Because of the maturity schedule, we've got additional opportunities for the balance of the year, so we would expect that trend to continue.

Perfect. Thank you very much.

Our next question will be from Henry Coffey from Wedbush.

Yes. Good morning, everyone, and thanks for taking my questions. Two specific areas. One, you talked a little bit about the Ginnie Mae EBO opportunity. Exactly how does that work, especially both with the servicing you may own, or with the servicing you may be subservicing? I know it's a capital impact, et cetera, but I'm thinking about it more as an opportunity. What is the governor behind converting that $1.8 billion of loans in forbearance into EBO and then EBO into resecuritization gains?

Let me take this. In terms of the $2.5 billion in Ginnie Mae early buyouts we referenced and the $1.8 billion that are on the balance sheet because borrowers are in forbearance, that is all Flagstar-owned Ginnie MSR. So they are Flagstar-owned EBOs. The accounting consequence of owning the MSR is to show them as early buyouts, whether you buy them out or not. And so the biggest impact for us is against capital. These loans were all performing prior to the pandemic, and we think a big portion will go back to making payments and get reinstated after the forbearance period expires, either on their own or through the partial claim process. Once they're reinstated, they're no longer categorized as an EBO. A small number will get modified outright, or they'll get modified in conjunction with a partial claim. If that's the case, we will then buy them out and realize the gain on sale benefit of re-securitizing the modified loans.

Do you own them or are you just putting them on your balance sheet?

We’re putting them on the balance sheet. We have not bought them out. The accounting consequence of owning the MSR is to show them as an early buyout whether you buy them out or not. So the biggest impact for us is on capital. We're working through the options. The first six-month forbearance period under the CARES Act is now expiring for many borrowers, so they can extend for a second six months, opt out because they continued to pay, or we can correct through a partial claim or a modification. We're working through those loans, but we cannot yet quantify how many will fall into each bucket.

If you look at the $1.8 billion in forbearance in Ginnie Mae land, those are counted as delinquent, right? And so if they go 180 days or longer delinquent, isn't there an opportunity to buy the loan out? I'm thinking about it as an opportunity rather than a problem. When forbearance is over, don't many of these loans have to go through some modification or partial claim process to get current again?

No, not all of them. Some borrowers have continued to pay and will just opt out and remain current. Some will be corrected through a partial claim process only. Those that need a modification — either on its own or with a partial claim — could be bought out and re-securitized. There's a clear waterfall we follow, but we don't yet know how much will be in each bucket.

What we don't know is how much is in each one of those buckets.

Correct.

No, this is helpful. I'm thinking of it more as an opportunity than as a problem, because you have lots of liquidity and you have lots of capital. On gain on sale for the fourth quarter, the revenue number you quoted — could you give that again, Lee?

Yes. $200 million.

So that's obviously down from what we saw in the last two quarters. Is there going to be sort of a disconnect with locks being down because of seasonality, but closings still being fairly high, and then some cost compression? How do you think that plays out?

Closings will remain strong because fallout-adjusted locks were high in Q2 and Q3, so we'll continue to push those closings into Q4. Full out adjusted locks will decline in Q4 due to usual seasonality — fewer business days and winter months affecting the purchase market. Agency and MBA forecasts have volume coming down in Q4 as well, consistent with that outlook.

How are all these IPOs affecting the competitive landscape for you all right now? We've had several since August.

It's not impacting us from a business point of view. I'm thrilled to see investors looking at mortgage favorably, and I would hope that could bring more new investors into the Flagstar stock.

I can only agree with all of that. Thank you. Congrats on a great quarter.

Our next question will be from Steve Moss with B. Riley Securities.

Good morning. Most of my questions have been asked here, but on credit I was curious: where are criticized or classified loans these days?

We really don't see many criticized or classified loans in our portfolio. On a precautionary basis, when loans were in deferral, we took a stance that those loans should be considered past watch unless proven otherwise. But at this point in time, we're not seeing a significant amount of criticized loans in our own portfolio.

You don't want to get overconfident about that, which is why we're cautious about the future. That's also why we've increased our ACL based on CECL model output and qualitative adjustments. We're surprised and pleased by how strong things are holding up, but we'll remain careful.

And go back to my comments about our view on the economy and CECL. We're not yet convinced we've achieved a V-shaped recovery, which informed our ACL position.

Look at the balance sheet and identify what really are commercial loans in this organization. If you take away the residential loans available-for-sale and the warehouse loans — where we've had only $5 million of losses over the last 12 years, including through the great recession — the part of the portfolio that's really exposed to losses is small compared to the size of the company and compared to a typical mid-sized bank with $29 billion in assets.

That's helpful. On return on MSR, how sustainable is the high income? What would the downside be if mortgage rates come down another 25 or 50 basis points?

Historically, we've been satisfied with a 4% to 6% return on our MSR asset and have consistently achieved that or better. This quarter was particularly strong, but that level of performance quarter-over-quarter is unlikely to be sustained. For modeling, assume a 4% to 6% return on MSR.

I agree. This quarter benefited from favorable conditions, and we took conservative valuation actions in Q2 that set us up well here.

And I would add, we have really good people managing that asset.

One last question: expenses dedicated to the bank were up $10 million quarter-over-quarter. Is that a new run rate?

There were a few non-recurring items. In Q2, we capitalized origination costs related to PPP loans which we sold in July, and we had accelerated vesting of certain executive compensation tied to the recent secondary offering. Those items are not expected to recur. So Q3 included some non-recurring things that inflate Q3 expenses and deflate Q2 expenses.

So, to answer your question, it's not a new permanent increase.

That's helpful. Thank you very much.

Our next question will be from Bose George from KBW.

Hi guys. Good afternoon. You only spoke about the net interest margin a little bit, but can you talk about how it looks next year if rates remained the same, and as mortgage banking slows at some point, how that impacts the warehouse business and loans held-for-sale — basically how that impacts margin next year?

If rates remain the same, we feel we're in a pretty good position to sustain our net interest margin. It's hard to look beyond a quarter, but for the next quarter we feel good. Even if mortgage volumes decline, warehouse balances can stay strong because we're gaining share and being selective and firm on pricing. So prospects for warehouse being strong even in a weaker mortgage market are good.

I would add that a couple of quarters ago warehouse balances were lower and net interest margin excluding loans with government guarantees was still consistent. There's room for warehouse balances to rotate down a bit and still sustain net interest margin at recent levels.

Okay, great. One more on the Ginnie Mae buyouts. Some peers have purchased quite a bit of this ahead of curing. What are your thoughts on why institutions are doing buyouts already? Are they parking liquidity or seeing an opportunity?

We don't comment on other institutions' actions. Our approach is to buy them out where we can modify either outright or as part of a partial claim and re-securitize. We will buy out loans where it makes sense and where we can realize the gain on sale from resecuritization.

Okay, makes sense. Thanks guys.

I'm showing no further questions at this time. I'd like to turn the call back over to our speakers for closing remarks.

Thanks, Carrie. In my closing remarks for the second quarter I talked about how the calls for social justice in the world around us had inspired a cultural change at Flagstar, how in the past we had not communicated much with our employees about what was going on in the world, how we set out to change that, and how we're not turning back. In the third quarter, we continue to engage employees on issues that are important to them, and continue with our diversity, equity, and inclusion initiatives. This is not a passing fad for us. It's the real deal, a fundamental cultural change. We are changing the composition of our leadership team, and we will change the composition of our board of directors to achieve more diversity. We're putting money behind support for minority small businesses and nonprofits that support diversity, equity, and inclusion. We're taking a harder look at our hiring practices, and we're taking cues from our employees about how we can do better. I'm doing my best to put the arm on fellow bankers and business leaders through organizations I belong to, to support diversity, equity, and inclusion in their own companies. Why? Well, first, because it's the right thing to do. And second, because I firmly believe it makes for a better company. These past two quarters also happened to be the best in Flagstar's history. Is that a coincidence, or is there a connection with our cultural change? In my mind, there's no question about that. The seeds of the success that is playing out today were planted when we recruited top-notch executives to run our businesses, when we built an affordable risk management structure, when we carefully diversified into commercial lending, when we developed our fee and deposit-generating subservicing businesses, and when we leveraged relationships we had nurtured for years to make warehouse lending a profitable, low-risk operation, now one of the top five warehouse businesses in the country. But I have to think that what we're doing on the DE&I front is not just underpinning, but is accelerating our success. We're a much better company today for it, and expect to be even better in the future because of it. Last but not least, thank you to our employees who own the success. I'm in awe of what you have accomplished in circumstances beyond the imagination. And thanks to everyone who has taken the time this morning to hear the story of a most successful quarter. Talk to you in January. Please stay safe and healthy.

Thank you. Ladies and gentlemen, this concludes today's teleconference. You may now disconnect.