Flagstar Financial Inc (NYCB) 2020 Q3 法說會逐字稿

Greetings. Welcome to NYCB 3Q 2020 Earnings Conference Call. (Operator Instructions) Please note, this conference is being recorded.

I will now turn the conference over to your host, Salvatore DiMartino. You may begin.

Thank you, and thank you for joining the management team of New York Community Bancorp for today's conference call.

Today's discussion of the company's third quarter 2020 performance will be led by our President and Chief Executive Officer, Joseph Ficalora; and Chief Financial Officer, Thomas Cangemi; together with Chief Operating Officer, Robert Wann; and Chief Accounting Officer, John Pinto.

Today's release includes a reconciliation of certain GAAP and non-GAAP financial measures that may be discussed during this conference call. These non-GAAP financial measures should be viewed in addition to and not as a substitute for our results prepared in accordance with GAAP.

Certain comments made on today's conference call will contain forward-looking statements that are intended to be covered by the safe harbor provision of the Private Securities Litigation Reform Act of 1995. Such forward-looking statements are subject to risks, uncertainties and assumptions that could cause actual results to differ materially from expectations. We undertake no obligation to and would not expect to update any such forward-looking statements after today's call.

You will find more information about the risk factors that may impact the company's forward-looking statements and financial performance in today's earnings release and in its SEC filings, including the 2019 annual report on Form 10-K and the second quarter 2020 quarterly report on Form 10-Q.

To start today's discussion, I will now turn the call over to Mr. Ficalora, who will provide an overview of the company's performance before opening the line for Q&A. Mr. Ficalora, please go ahead.

Good morning to everyone on the phone and on the webcast, and thank you for joining us today.

Earlier this morning, we reported diluted earnings per common share of $0.23 for the 3 months ended September 30, 2020. That's up 21% compared to the year ago quarter and up 10% compared to the previous quarter. We are very happy with our third-quarter performance, especially given the uneven economic recovery and the lingering effects from the COVID-19 pandemic. Not only did our fundamentals and operating results continue to improve, but we also experienced better asset quality metrics, as NPAs declined, along with a significant decrease in the number of loans on deferral.

Our third quarter results reflect strong top line revenue growth and earnings per share growth on both a year-over-year basis and on a linked-quarter basis. This growth was driven by solid loan growth and continued double-digit net interest margin expansion. This, along with well-contained operating expenses led to higher levels of pre-provision net revenues. At the same time, our provision for credit losses declined compared to the prior quarter.

More importantly, the amount of loans in deferral declined significantly. As previously disclosed, our deferral program is somewhat unique in that it is for an initial 6-month period as opposed to a 3-month period. Accordingly, the overwhelming majority of loans on deferral were eligible to come off their deferral period during October and November. From June 30, 2020 to October 22, 95% of the $3.1 billion of loan deferrals eligible to come off of their deferral period returned to payment status. Both multi-family and CRE deferrals dropped 48% compared to the levels at June 30, with most of the improvement in the CRE portfolio occurring in the retail and mixed-use segments.

As of October 22, our deferrals represented 7.3% of total loans compared to 14.4% as of June 30, 2020. We attribute the substantial improvement in loan deferrals to our conservative underwriting guidelines and our proactive borrowers outreach program earlier on during the pandemic. Looking ahead, we have an additional $3.1 billion of deferrals scheduled to come off their deferral period, $2.9 billion of which are scheduled to do so during November. Given October's strong payment performance, we remain confident about deferral payment trends going forward.

Moving now to our third quarter results. We experienced another quarter of strong margin expansion. Excluding the impact from prepayment income, the net interest margin expanded 11 basis points to 2.20% on a sequential basis and was up 32 basis points on a year-over-year basis. This marks the fourth consecutive quarter of margin expansion. Since bottoming in the third quarter of last year, our margin has improved 32 basis points.

The margin improvement continues to be driven by a substantial decline in our funding costs. Our overall cost of funds during the third quarter declined 22 basis points on a linked-quarter basis and 80 basis points on a year-over-year basis. The decrease in funding costs was primarily driven by the downward repricing of our CDs, which resulted in an average cost of deposits of 85 basis points, down 31 basis points sequentially and 104 basis points year-over-year. At the same time, loan yields declined only 5 basis points compared to the previous quarter, owing to continued stability in loan pricing.

Net interest income for the third quarter rose 6% to $282 million compared to the second quarter and was up 19% compared to the year ago third quarter. During the last easing cycle, our cost of deposits bottomed at 54 basis points. Given the current rate environment and expectations, we believe that this cycle, our deposit costs will bottom below the levels experienced during the last cycle. Given this, and further repricing opportunities on the funding side of our balance sheet, we are well positioned for additional improvement in both net interest income and our margin.

On the expense front, total noninterest expenses were $129 million, up 4% compared to the previous and year ago quarters. The increase was largely due to our bank-wide systems conversion and COVID-19 readiness as we reopened our branches and 25% of our employees returned to their offices. Despite the modest uptick in expenses, the efficiency ratio improved to 43% during the quarter and we continue to benefit from positive operating leverage.

An important metric to focus on is pre-provision net revenues or PPNR. PPNR for the third quarter was $167 million, up $10 million or 6% sequentially and up $30 million or 22% year-over-year.

On the balance sheet front, total loans were $42.8 billion, up $523 million or 5% annualized compared to the second quarter of the year, driven by continued growth in the multi-family portfolio and a rebound in specialty finance lending, offset by a decline in CRE. On a year-to-date basis, total loans are up $935 million or 3% annualized. The multi-family loan portfolio showed continued strong growth during the third quarter, increasing $504 million or 6% annualized compared to the second quarter. Multi-family loan growth has been higher in each quarter over the course of the year. This growth is the result of the following factors: increased refinancing activity as many loans are nearing their contractual maturity date or option repricing date; higher retention rates; and market share gains as some bank competitors have either retrenched from this market or have deemphasized multi-family lending.

We were also pleased to see a rebound in the specialty finance portfolio during the quarter. Specialty finance loans and leases increased $138 million compared to the second quarter of the year. On a year-to-date basis, this portfolio has grown $438 million.

Our originations for the quarter were robust. Total originations, excluding PPP loans, were $3 billion, that's down 9% compared to the previous quarter, but were up 31% compared to the year ago quarter. Third quarter originations exceeded last quarter's pipeline by $800 million or 36%.

As for our loan pipeline, we continue to see strong borrower demand for our loans. Our current pipeline heading into the fourth quarter is $1.8 billion. Included in the current pipeline is 61% of new money.

On the funding side, total deposits of $31.7 billion were relatively unchanged compared to both the previous quarter and to the levels at year-end 2019. In line with our strategy to reduce our cost of funds, CDs declined $1.1 billion during the third quarter and $3.2 billion during the first 9 months of the year. The majority of this decrease was offset by growth in each of our other lower cost deposit account categories.

Our wholesale borrowings assisting primarily of Federal Home Loan Bank advances increased $675 million compared to the previous quarter as we continue to take advantage of the low interest rate environment and favorable capital market conditions.

On the asset quality front, our credit metrics improved during the quarter, reflecting the underlying strength of our loan portfolio. During the third quarter of 2020, our segment of the New York City real estate market, the nonluxury, rent-regulated, multi-family segment continues to hold up very well. Rent collections in this segment continue to be strong and have returned to pre-pandemic levels. Also, we have not witnessed any material changes in residential vacancy rates in our multi-family portfolio, with vacancy rates below 3% as of September for the current quarter.

The provision for credit losses under CECL was $13 million compared to $18 million in the prior quarter and $21 million in the first quarter of the year. This quarter's provision exceeded net charge-offs by $14 million. During the current quarter, we had a net recovery of $901,000 compared to net charge-offs of $4 million or 0.01% of average loans last quarter.

Nonperforming assets as of September 30, 2020, were $55 million compared to $63 million in the previous quarter. This reflects 10 basis points of total assets compared to 12 basis points in the previous quarter. Excluding nonperforming taxi medallion-related assets, NPAs would have been $23 million or 4 basis points of total assets compared to $30 million in the prior quarter or 5 basis points of total assets.

We will -- we also announced that the Board of Directors declared a $0.17 cash dividend per common share. The dividend will be payable on November 17 to common shareholders of record as of November 7. Based on yesterday's closing price, this translates into an annualized dividend yield of 8.2%.

Lastly, I'd like to end my formal comments by once again thanking all those who continue to serve on the front lines of this crisis, and to our employees, management and Board who have shown extraordinary dedication and commitment during this difficult time.

On that note, I would now ask the operator to open the line for your questions. We will do our best to get to all of you within the time remaining. But if we don't, please feel free to call us later today or this week.

(Operator Instructions) Our first question is from Ebrahim Poonawala with Bank of America Securities.

So of the $3 billion that came up, $157 million asked for another additional assistance, is there any reason why the next $3 billion that's remaining, we should not see a similar outcome where 95% go back to paying so that we end year-end at deferrals of about 80 basis points or at least less than 1%? And then if you could add to that in terms of what's your expectation around the outcome of the ones that need the additional deferral into 2021? What's the loss content around those that you expect?

Ebrahim, it's Tom. So let me just update you and clarify. So we -- as of last night, we have 97% that has paid, that brings the number down to $90 million, which remains unpaid, and it's still probability that we'll collect on this money and there will be no second-round deferral. So we're very optimistic that this typically runs in line with our normal payment structure. And it's not unusual after a month after quarter end. So we're very pleased with the outcome. There may be some customers that decide to work with us and some may not pay, but the reality is that we do have flexibility under the CARES Act to deal with this $90 million that's currently unpaid.

And going forward, just to go back to your additional question, we are very optimistic given that we've spent a lot of time on knowing what's ahead of us in the next 3 weeks of customers' reaction to the various communication that we sent to them regarding payment status. So we feel very optimistic that this will continue. The first slug of money that came due, which was that approximately $3 billion in October, the results are much better than what we all expected. So we're very pleased.

I think the reality is that the underwriting standards that are so strong and the low LTV and the fact that when reanalyzing the cash flow, when there's a request for additional relief is a very high bar now, where the original COVID-19 relief was pretty much offered to all customers if they choose to. So as I said in the previous call, as an abundance of caution, we were very liberal on allowing people to have the CARES Act relief. But going forward, the bar is very high, and we do a significant amount of cash flow analysis to assume that if they're going to get more relief, they have to really need the relief.

So the good news is that, as Mr. Ficalora indicated, the payment structures that are coming in and the collections are very high. The delinquencies are very low right now for our core business model, and we're very pleased with the current results of the first chunk of large deferrals coming due. By mid-November, we'll pretty much be through 96% of all deferrals. And we anticipate on updating the marketplace by the end of the -- sometime by the end of November on those results. But we believe it will be consistent. So we're very pleased with what we're seeing right now.

That is helpful. And I guess just taking a step back, Tom, your stock's down about 25% since July earnings. There's a lot of nervousness around the New York City economy and real estate market looking out next year. Obviously, you've been growing your multi-family book. I think you mentioned rental yield collections are back to pre-pandemic levels. Just talk to us as we kind of fast forward to the next 6 to 12 months, what are you seeing from your borrowers as maybe people move out of New York? Like, what's the level of concern around credit and -- as it ties to -- as you're extending new loans? And just how we should think about the losses for NYB?

So again, we feel that the losses will be de minimis. We obviously adopted CECL this year. So obviously, we'll not typically ever have any real significant losses historically for the company, but we put up sizable provisions for our historical charge-off history. So clearly, we feel highly confident that the provision will continue to be lower, assuming that the macroeconomic backdrop continues to improve as we're seeing. So as you can see from the first quarter of the year where our CECL adoption was, and going forward each quarter it's been slightly lower, I can see that trend continuing. And possibly, depending on how robust this collection effort and the amount of loans that are no longer on deferral, we may see some recoveries sooner than we expect given the magnitude of the collection efforts. I mean, we really have a tremendous amount of people back on full payment status. And that's not a guarantee for the future. But clearly, what we're seeing in October, we believe that trend will continue in November.

And as far as the overall risk we see, I think the risk has shifted a little bit. Initially, when we had signed up a lot of customers on deferral, we were worried little bit about mixed-use. And if you look at where the mixed-use is right now, at the $701 million commercial mixed-use portfolio, we're down to $57 million on deferral, that's a pretty sizable reduction, and we should hopefully work through that in the next month. And then when you look at what's left on retail, retail has dropped substantially as well. We had $1.7 billion in retail. And right now, we have about $189 million left on deferral, which pretty much will work itself out in November.

So those not -- those slides say the shift may be into office. And obviously, office -- there has been a lot of noise in office in New York City. So we do have approximately $800 million left on deferral with office. Interesting story with office: We had a large customer that we've had indication, I mean, he may have difficulty paying and we contacted that customer. And at the end of the day, he paid 41% LTV, family asset class. It's something that his family generated over time, and it's a very low LTV. And ultimately, we have buyers for the note. So if in the event, for some reason, we had to go that path, we have people lining up at par to take the asset. So we're a low leverage lender, in particular, in the New York City marketplace in Manhattan. And given that there's been a little bit of shift into office, but retail has performed better-than-expected as did mixed-use. So we're very pleased with the portfolio we expect outside of our core product being multi-family. As far as the multifamily is concerned, it's been very robust on the collection side.

Got it. And if I can sneak 1 more in, just on the margin outlook, you've given some disclosures around the debt and CDs maturities. If you could, one, talk to us in terms of margin outlook for the fourth quarter and maybe looking out further? And what are the new CDs coming on at? And what's the new debt coming on it?

So let me be clear, I'm not going to go into 2021 yet. Let's just get through 2020. And we said all along, we anticipate it for the year to have continued double-digit margin expansion throughout 2020. We believe, in the fourth quarter, we'll continue to see the similar trends. We'll have double-digit margin expansion going into the fourth quarter of 2020, which is very encouraging, driven by, as Mr. Ficalora indicated, the substantial drop in our cost of funds, both CDs and borrowings. We have approximately $5.8 billion coming due in 3 months at 1.46% cost of funds, marketplace between 46 and 70 basis points. That will probably roll somewhere in between that level. We have $11.6 billion coming due from 12 months on an outlook on CDs at 1.21%. You bring it down between 50 to 70 basis points depending where interest rates go, which we believe rates will be going lower depending on the overall economic climate and a lower-for-longer reality in interest rates.

So if that's the case, and the nontraditional banks continue to lower their overall cost of funds, we will enjoy that benefit. In addition to that, as you indicated, we have about $1 billion of borrowings coming due next year at 2.03%. That money is anywhere from 30 to 70 basis points if we refinance today. And we have another $725 million in the fourth quarter at 1.18%. These are all positive attributes to the continuation of margin expansion, which we will enjoy in the fourth quarter.

And obviously, going forward, we think that the spreads are holding up on the asset side. We're still getting north of 3% as far as our overall weighted average yield and that's subject to other markets. So we're still getting around that 300 basis point spread over 5-year treasuries for our core product mix on the asset side.

Recognize that in every prior difficult market, we have substantially less losses than the marketplace actually realizes. And when we look at our existing portfolio, there is a high probability that we continue to be paid on our portfolios and that the realization of the losses that we've already booked as reserves is highly improbable. There is no reason to look to our history and suggest that those losses, which are CECL accounting losses, that those losses will ever be realized. That's an important thing for us to recognize from the standpoint of future period performance. We have high probability of performing in the future period in the New York market rather consistently with how we've performed over every previous cycle. Yes, there are nuances. And yes, there is a difference with regard to some of the properties that we have. But still, the likelihood that we would realize these massive amounts of charges is highly improbable.

Our next question is from Brock Vandervliet with UBS.

I think there was a lot of concern initially going back to the Rent Act and now certainly more focused with COVID that lower building valuations, especially in your core multi-family space would make it much more difficult for you to pencil out when loans came up for refinancing. How have you found valuations holding in? And can you talk about that?

So Brock, it's Tom. I would say that for the most part, cap rates are still remaining significantly lower than I guess what was expected at the beginning of the year. So rates are still holding in on the capitalization side. Values are obviously being subject to some regional decline. So obviously, when we look at our peak-to-trough and when you look at CECL model with downside scenarios, we're running apartment houses down close to 30%, office down 30%, retail down 30%. So we're really looking at the impact of the cash flows.

And what's most important, when we're actually receiving the cash flows from the customers as we work out these ones that are on deferral, they're coming off deferral, they're able to pay their bills, which is telling you that they're cash flow positive. And in the event, we're going to refinance them at cash flow positive, we'll be able to do that with no real issue. It doesn't mean that every loan has been cash flow positive. But the results that we see here to get to that high bar standard of additional relief request for the most part, most customers were able to do their debt service coverage ratio adequately. It may not be at 1.67%, maybe they're at 1.1% or 1.2%, but the reality is the money is coming through, we gave them 6 months of balance sheet relief. And now they have excess cash flow, and they were very thankful for that because, obviously, it helped them out to work with their own customers on the ground floor, especially on the commercial side.

So I'd say that there has been no question, obvious reason, an adjustment in the marketplace. But given the low LTV that the company has, we have a 47% LTV and clean, the 48% original LTV in Manhattan. These are low LTVs. So in the event that they come back to the bank -- and which they will because obviously, they're relatively short-term assets -- we're very comfortable on refinancing them. And you can see by the origination flow, we've been doing that.

Our overall retention rate has been spot on all year, north of 50%. The goal was to keep it at 50% to shoot for that 5% net loan growth number for the year. This is coming from our own portfolio. We believe that that may even improve as we get into the fourth quarter because, obviously, we see a lot of these customers coming to us, in particular, we have about $1 billion that are up for renewal this quarter, which is our own portfolio. And we believe that all of these loans will have the ability to be able to refinance within the portfolio. So we're very confident there.

Okay. And I heard your commentary on loan yields holding in around 3%. It seems to be somewhat at odds with the drumbeat of lower yields really across the sector and many different loan types. How is that a factor of the competitive environment? Why are you so successful in holding yields north of 3%?

So Brock, I'd say on average, if you look at where the 5-year and the 7-year treasuries, tack on approximately 300 to that, around 3%, slightly north on average. We are going to do very well in competing. Obviously, the agency is a major force here, but the structure of an agency loan versus putting a portfolio loan has different nuances, and we're very competitive there.

So if we have to shave 0.25 point and 0.125 of a point for solid business, we'll do so. We're not going to do high LTV. We're not looking -- if customers are going to get financing, it doesn't make sense to the bank, especially in the midst of a pandemic, well, let the assets leave the bank. But the retention rate has been hitting our bogey, which is fabulous. We believe that will improve in the fourth quarter in particular, and we feel highly confident that all in, we can average around 3% loan coupons, which when you run that through our model, it bodes well for margin expansion in a very low interest rate environment.

Where I think there's a disconnect on the funding side is that bank -- short-term bank deposits are probably priced too high in the marketplace. The good news is that we're seeing some of the nontraditional institutions lower their rates to more incline to the market rates, where overall treasuries are trading plus a spread on the funding side.

And our next question is from Steve Moss with B. Riley Securities.

Just following up on the borrowings here, just wondering what is the ability to restructure any borrowings either in the fourth quarter in the upcoming year?

So Steve, I'd say that obviously there's always that opportunity that we look at. Obviously, there's a cost of capital for that. So we would evaluate that depending on profitability, depending on forward curve analysis. We do have approximately $1 billion at 2.03% coming due next year. So that -- so the pain in that restructure is probably not significant because it's due in 1 year. But we evaluate that from time to time. In the event that we have, let's say, sizable economic benefits in a given quarter, we will always look to monetize that opportunity on a core basis going forward to try to enhance future earnings going forward.

But clearly, the stuff in the back end is expensive. We have some expensive borrowings going out to 2028, 2029 with around 2% handle on that. And I think that would be much more painful as far as cost of capital and the earn back. So we do evaluate earn back. It has to make logical sense to take that charge. And we evaluate that on a quarterly basis.

Okay. I appreciate that. And then just as we think about loan growth here, what -- on the specialty finance portfolio, good growth quarter-over-quarter. Kind of wondering what your expectations are for the upcoming fourth quarter and into 2021.

Well, I'd say for specialty finance, we've really expected that rebound in Q3 given the decline in Q2 from literally being in the epicenter of the pandemic. That was a nationwide drop in general. So we have the recovery there. On the auto side, which is a significant portion of our business, it's booming right now. So there's a shortage of inventory. We see a lot of good drawdowns right now. So we have a nice uncommitted -- committed line of approximately $4.9 billion with outstanding of $3.3 billion. It wouldn't be surprising if you see more lines being drawn down at the end of the year, assuming the economic recovery continues.

So I would say that they will continue to grow. We're very optimistic of that group. Their credit is pristine. We have 0 loans on deferral. We don't -- we anticipate 0 losses in that business and it's performing beautifully. So we're very supportive of the business. Remember, this has not come with funding. So we're a credit buy up shop, and we choose what we want to finance. As far as the customer book that we have right now, we have 100% pay, and we have a great portfolio that should just grow from the fact that we have a substantial commitment line. But we also anticipate to be active in the market to grow it at levels that are consistent, which are very high CAGR.

So I would say it's going to be hopefully continued growth, that with the multi-family CRE right now has been relatively flat to slightly down as the marketplace, given the lack of property transactions and deals happening in the New York City marketplace, and that could change eventually as the economy recovers on the real estate side, but multi-family has clearly going to be the driver along with specialty finance in the future for the company.

Okay. Appreciate that. And if I could get sneak one last 1 in here. I think the press release said $2.8 million in branch reopenings and COVID expenses. Just kind of curious how to think about expenses on the -- for the fourth quarter?

So look, we were probably a little bit off in Q3. We had the conversion. We had some additional expenses there. The PP&E costs as well as testing for employees and the reopening for both the headquarters, the multiple operations that we operate out of as well as the branches, there's a lot of testing involved. So there's expenses to get reopened and that number, obviously, went into the third quarter.

I would say we're probably going to be around that $130 million level by year-end, so maybe up a couple of million by year-end. So for the year, we'll probably stay about $507 million for 2020, which was significantly lower than our original projection, going back to last year at $520 million. We'd like to see it lower, but obviously, we never anticipated the pandemic and there's cost associated with that pandemic. Not material costs, but clearly, we saw a little bit of an uptick in Q3. And we're probably going to see around $130 million level in Q4.

Our next question is from Ken Zerbe with Morgan Stanley.

Tom, you guys mentioned that your loan yields, I think, were either around 3% or maybe 3% over the 5-year treasury. But your average -- more average loan yields on your portfolio currently are around 3.60%. Is that 60 basis points the right amount of asset yield compression that we should expect over time?

Well, Ken, I would just -- you've got to back out prepay, right? So we reported 3.60% but when you back out prepay, it's like 3.49%. So that's where we are as of the end of the quarter, approximately. And then when you think about what's coming due, it's about a 3.40%, 3.41% that's coming due in the fourth quarter and $1 billion, next year is 3.44%. So yes, it's lower than the market yields, but the pain is not as bad as it was a few years ago. So I think it's manageable given the deposit costs and the borrowing costs coming down materially from there and given the repricing offsets because we're still liability-sensitive.

So yes -- so I think that we're getting around the low 3s on average. And I'd say 300 basis point spread that's certainly depending on the deal, depending on the duration, 10-year versus 5 or 7, IO versus non-IO, it's all reflected to the market. Competition is fierce with the agency, but we compete, and we stand by our solid customers on the asset quality side.

At the same time, I'd also want to make this very important statement. During this whole journey of pandemic, dealing with COVID-19, when we worked with these deferrals we were very focused on making sure to ensure we have the customers paying their escrow accounts, making sure that we have the operating accounts now in control. So in the event they come back for potentially a second round deferral, we really understand the behavior pattern of that building because we have the cash flow. It's no longer housed at JPMorgan, it's housed here. And that's very important when we entered into these deferral agreements. So the fact that we have the ability to be a little bit more flexible on rate for deposit relationship, we're enjoying a nice shift into lower-cost funds as well, which has been a cultural change for the bank. Historically, we've always been about the asset side and the deposit flows came from acquisitions, right?

Got it. No, no, understood, certainly. And then with the reserves, obviously, I know your credit is really, really good. So there's no question about that. But can you just talk about what the driver of the reserve build was this quarter? I mean, just given the 0 NCOs, I would have expected 0 provision.

Look, we're in the midst of a pandemic. We have a history of pretty much 0 loss, especially on the CRE side. So when you think about CRE, I'd say qualitatively and quantitatively, we have a 41 basis points quantitative reserve versus a 59 basis points qualitative reserve. And when you add that up, it's based on our history, it's considered as what we would say we're over-reserved, because we have a history of no losses. But we acknowledge the property price declines, in particular, our region, which is the New York City region, so we probably reallocate a little bit more towards the commercial real estate space, probably into the office book. And also into the retail book.

Well, I'll tell you this: given the history and the actual results of the loans that have been back on full payment status, you're probably more likely than not, going to see a reversal if everyone's back on full payment status because we took into account rating shifts, revaluation of cash flows and the anticipation that a certain degree of these loans would be in a second round deferral. But the fact that we're seeing 97% as of today for the first bucket in full payment status, and that continues, you're probably looking at an adjustment in the future that will be favorable for the bank because of the adoption.

Yes. And just 1 quick one on expenses. The $130 million you gave for fourth quarter, is there anything unusual in that? Or is that sort of the new base level that we should grow on as we think about 2021?

No, I'll give the Street some guidance when we come out in January, but I'm not getting give 2021 guidance yet. I mean, we feel highly confident that the conversion is complete. We have opportunity within the bank. We look at that very carefully in the fourth quarter. And obviously, as you know, historically, the company has always been very efficient. We beat our overall expectations in 2019 and '20. So we budgeted $520 million. We came in at $507 million, as anticipated $507 million. Assuming we hit $130 million at the end of the quarter, hopefully that number is conservative. But our goal is to be very efficient given the magnitude of the carnage in the marketplace. We're going to be always belt tightening.

So we'll have some guidance for us -- for you guys in next quarter. But clearly, we're very pleased to be at below $510 million. We'd like to be at $500 million, but clearly, the pandemic had some expenses. And the post -- and pushing up the conversion for a year had some consequences of expense. And I think that we should get the benefit of that going forward since we're all on one platform and lots of our systems now could be reevaluated as far as to see where our potential expense savings will occur. And after a process that we'll go through the year-end process and budget and forecast.

And our next question is from Steven Duong with RBC Capital Markets.

Tom, just your CET1 ratio, it dipped a little bit this quarter. Can you just explain what was going on with that?

Yes, it's Tom. Well, mostly growth, right? We were growing, which we continue to grow on a quarterly basis. We're very comfortable with our internal capital model, our capital plans. And again, this is solely driven from growth. I mean, our earnings are going higher, which is great. We forecasted for the year at 20% EPS growth from '19 versus '20. We're on target for that. So we're very pleased with the growth trajectory of the capital build now. So that's positive towards overall capital levels. We are growing.

Got it. Okay. And then just on your office exposure, do you have the exposure by the different boroughs, say, like Manhattan?

Yes. So look, I think Manhattan is the area that we're spending a lot of time on reevaluating these credits and cash flow and ratings reevaluation. And I'll tell you that if you break it down, I'll go right to the categories mixed-use. We had over $100 million going back to the previous quarter. We're now down to $40 million in mixed-use commercial in Manhattan. The retail is phenomenal at 99% pay, we're down to $4 million in retail in Manhattan, which is much better than we anticipated given the pandemic. Office right now is probably the high bar, where we have a total deferral of $800 million. Right now, we're at $762 million, that's in Manhattan. We're going to work hard in the next few weeks to see how this all plays out, but that is the area that has the highest percentage as well as -- and the dollar amount that's on deferral. Multi-family as of the end of the quarter -- actually at the end of October 22, which is the update we gave in our investor deck, it's $530 million that's in Manhattan.

So it's blended across the 5 boroughs in multi-family. The next borough will be Brooklyn at about $469 million for multi-family. But clearly, on the office side, that's the area that I would reshift the energy level towards the deferral issue that we have to deal with.

Got it. Appreciate it. If I could just ask 1 more. Just on your office exposure, what was your pre-pandemic coverage ratio?

So I'd say we anticipate -- I have -- just bear with me 1 second here. Office, we had about 190, just under 190 on debt service coverage ratio, LTV at 52%. I guess I used the example that we had in office of a very wealthy owner as an office -- has an office building, and it was due in October. LTV was 41%. And we had to make sure we get paid on this one. We felt highly confident that this is -- it's a flagship area, flagship location. And in the event the building was going to go through difficulties, we have people willing to work it out with the customer, and we sell the note at par.

So these are real, coveted asset classes. This is family wealth over many years, some of them at 1031 exchanges, that has created long-term wealth for property owners, and they're not just going to walk away. So the fact that we gave them 6 months relief, that's a lot of balance sheet relief, 6 months cash flow on a large credit, and this was a fairly large credit, and we ended up getting paid, which is great. And that's our approach. Our approach is at the bar going into the second round deferral, if any, is very high. We go through a very rigorous program that the cash flow analysis has to be presented, and if we feel they have the ability to pay, we're not going to give out relief when it's not justified.

Right. Right. And so you -- pre pandemic, you were at 190%, and your minimum is 130%, is that correct?

Yes. Yes, on the commercial side, that's correct.

Our next question is from Matthew Breese with Stephens Inc.

Just touching on the expenses, the uptick in G&A, how much of that was due to the systems conversion and is therefore permanent versus COVID and subject to fluctuation?

I'd say it's probably evenly split from the uptick. I think I guided it was 120 each, John, guided once we were off by a couple of million bucks. I'd say probably $1.5 million for the conversion, the rest was PP&E and testing. Remember, we have a lot more people going back to the various locations, and if there is an expense that's associated with that, dealing with getting the branches up and running -- so all the branches for the most part are up and running for the time being depending on the pandemic, there clearly was some expense there -- so I'd say it's split between conversion and PP&E.

And then -- I interrupted you, I'm sorry.

In testing as well. There's an expense to that. We don't have doctors on staff, so we have to do hire people to ensure the safety for our employees.

Got it. And then could you just update us on the strategy around securities? Where do you want it to be? What are -- what are new yields and what kind of investments are you looking at there?

So in my guide, up double-digit for the quarter. I anticipate the security yields to come down. So our guide for the quarter is up again, once again double digit, so assuming 10 basis points up in the fourth quarter, which is reasonable based on our modeling. We're assuming the current yields are coming down on securities. I wouldn't be surprised if we'll see another significant drop in security. So we're probably looking at somewhere in the 235 to 238 type level in the fourth quarter. And we're being very cautious given the marketplace.

So our securities portfolio as a percentage of assets, it's probably -- it's been pretty much at the lowest level. It's been like -- I think it's like 9.58%. Is that right, John? Yes, 9.58%, which is about a $5.2 billion book on a $54 billion balance sheet, low. So obviously, we don't want to jump into the securities market with significant investment now given that we have reasonable loan demand, and we want to be very cautious given where yields are. Just a wait and see moment, and we can't control interest rates. So we think that yields in general are going to be relatively low. You're looking at reinvestment yields probably in the low 1s. And it's not the opportune time to grow that portfolio. We may, from time to time, put on additional liquidity. That will be a business decision that we decide as how we manage our ALCO process. But clearly, it's not an opportune time to grow the book.

Understood. Okay. And then I know you guys feel very comfortable on the deferral front. Curious, what happened quarter-over-quarter to criticized and classified assets? What were the balances this quarter versus last? And have you seen any risk-grade migration?

Yes. I'd say special mention went up a little bit because, obviously, when you go through this analysis and before you even get you as much information you can from the bars that are going to the pandemic, we focused solely on the portfolio that was in deferrals. So we've done a lot of work in that for regulatory reasons. That's what we should be doing, and we've shifted a lot of these loans to special mention.

Okay.

All right. But again, the good news is that as they continue to pay, they go back to full payment status and they're cash flowing north of levels that we didn't expect, we're going to roll them out a special mention, and that will be the next phase of the benefit of that. So we do it as, out of an abundance of caution and conservatism as we should.

And then the last 1 is just you mentioned cap rates are holding up better than one might have expected. Could you just frame that a little bit better for us? Where are multi-family office retail cap rates today versus a year ago?

I'd say consistent. I don't see any movement at all. I mean we were initially -- our view initially was when the rent regulatory changes came in last year that we had a 30 to 75 basis point adjustment depending on the borrower. We write a much higher cap rate in the Bronx than we do in the Bronx versus Manhattan. So Manhattan cap rates were always lower in general. So we have a much lower LTV as an abundance of caution. But they haven't moved a whole lot. They're still in the low to mid-4s, depending on rent regulatory versus non-rent regulatory. I'd say for -- when you look at some of the deals that we've done in the past 12 to 18 months, since the change in rent law, I'd say a 5% handle on a 100% rent-regulated building is not unusual. But when you throw in commercial use space and other revenue streams, you start getting into that mid-4s. So it's been consistent.

So we do a deep dive every twice a year on rent regulatory stress testing. And what we find when we statistically look at what's going on in the marketplace, the cap rates were going lower, but we conservatively analyze it with an uptick and to stress the portfolio. But surprisingly, it's been resilient.

And our next question is from Christopher Marinac with FIG Partners.

Tom and Joe, you had mentioned earlier about the very strong credit quality on the specialty finance. It looks like maybe 98% or more is pass rated. So do you have to allocate the reserve at a higher level just given the pandemic? Or would you still kind of look at the reserve allocation similar to prior quarters?

De minimis allocation. We're very comfortable. This is all super senior secured credits. Even when a credit goes through difficulties, we're secured by inventory, equipment. So again, we're very comfortable that this portfolio has 0 loss content.

Your next question comes from Ebrahim Poonawala with Bank of America Securities.

Just 2 quick follow-up questions. One, on the loan-to-deposit ratio, how high do you see that running given just the attractiveness of the debt markets, if you can touch upon that, Tom?

So I would say that, historically, the company had a loan-to-deposit ratio as high as 165. So we're not there, but we've had some improvement. It's been a long time since the bank has executed on an M&A transaction. And obviously, our historical growth in deposit has been driven through M&A.

So the goal here in the long run is that eventually, as we continue to focus on our business model, which includes putting businesses together and getting that funding side shored up on -- through consolidation, that's always a possibility. But absent that, I think you're going to see probably, we'll call it nominal deposit growth, and we'll still choose based on what's best for the shareholders as far as cost of deposits and duration. And right now, the wholesale markets are very attractive. They're more attractive than what's offered out there in, we'll call it the hot money CD market. Clearly, you can put on 5-year money lower than you could on the CD market. If you go short, you're looking at between 25 to 38 basis points type cost of funds depending on your duration risk that you're willing to take.

So there is this business decision you have to make when it comes to repricing your deposits aggressively for growth. And again, our growth is manageable, it's reasonable. We're not targeting 10% type loan growth and asset growth, we're still shooting for that 5% net loan growth number driven by multi-family and specialty finance, offset by the CRE reduction. So I think with a reasonable growth on our internal deposit analysis, we can grow between a small portion of that on deposit growth as well as repricing the funds lower aggressively and benefit from a unique opportunity on the wholesale side, which is cheaper than retail in this current environment. With the proviso that in the event that we're fortunate to be able to put something together on the acquisition side, well, this business combination side, that's where this model can change dramatically to have a different funding mechanism.

Got it. And while I have you, just in the remaining $3 billion deferred book, there is no reason to believe by using the experience because there could be some pushback around, there's higher risk loans in that portfolio. When we think about the remaining deferrals, is there any reason to believe why they would behave differently than what you've seen so far?

Well, look, I would say that these are the ones that have filed for the deferral later, right? It's 1 month later. The other ones that waited 30 days to come back to what the bank offered. So look, we -- we're all over these customers, we understand the dynamic of it. We're very optimistic that this will continue. And I broke out the area of risk. I think we've shifted it more towards office versus retail and mixed-use, which is a good thing and we'll deal with that as it goes along.

But these office credits are -- the average LTV that we originally generated was 52%. So it's better than the retail. And I'd say not as -- obviously, mixed-use was the lowest LTV, but it's better than the retail. And the retail surprisingly did outperform. At $4 million that's sitting in Manhattan right now, we feel pretty good about that. The retail customers have stepped up in payment status. And the fact that we can reshift our effort now that office is the potential risk here, the LTVs are low. The customers are very strong. These are large property owners, and we focus on conservative underwriting, and the debt service coverage ratios when generated were extraordinarily high.

And our next question is from Collyn Gilbert for KBW.

Tommy, I just want to make sure I understand. So on the expense side and you've given obviously good color, but -- so the elevation that you're going to expect in the fourth quarter, and obviously, what we saw in this quarter, attributable to the conversion costs as well as COVID costs without specifics. So generally, I mean, are there -- should those expenses be coming down next year? Or are there other investments that you're going to need to make?

No. I think -- yes, I'd say the investments were made on the conversion. We spent a lot of time, efforts and energy over years to get this conversion done. It was the largest conversion the bank ever went through. It's all systems through Fiserv. It was a partnership. We collaborated together, we work together what the bank's needs were, and we adjusted their top offering, and we worked with them over the past multiple years. So we feel most of that expense has been absorbed.

Going forward, we're going to have -- I'm not going to give 2021 timing shift. But again, we're an efficient company. We want to get that efficiency ratio down to 42% and that was the goal for 2020. We think we'll hit it by the end of the year. And we're also driving profitability up. So if you look at where our EPS, CAGR was '19 versus '20, we should be up around 20% EPS growth with significant margin expansion. I think when I first gave my margin guide, people felt I was way too aggressive at being at 225 margin. I said we came off of a low, was it 188, John, 188 in margin. So we felt between 30-some-odd basis points is reasonable.

So here we are coming into the fourth quarter. I gave you guys 10 bps going into Q4 to get you to that 230 number. And hopefully, I'm conservative. And I think it all depends on what happens next year with cost of funds going forward and where the shape of the curve is. So the good news is that the spreads are healthy, that we have healthy spreads in our business model. If it's 275 or 300, it's not 110 like it was going into the financial crisis. And I think the competition has waned a little bit. I see people a little bit nervous to do lending in New York and we're open for business in New York. We're going to be there for our customers in these uncertain times, and these are cash flowing assets and the collections are high, and we're very comfortable on working with these historical property owners that have been doing business with us in good times and bad times.

And we have reached the end of the question-and-answer session. And I'll now turn the call over to management for closing remarks.

Thank you again for taking the time to join us this morning and for your interest in NYCB. We look forward to chatting with you again at the end of January when we will discuss our performance for the 3 months ended December 31, 2020. Thank you.

This concludes today's conference, and you may disconnect your lines at this time. Thank you for your participation.