Good morning. My name is Samantha, and I will be your conference operator today. At this time, I would like to welcome everyone to the M&T Bank Q3 2019 earnings call. All lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question and answer session. If you would like to ask a question during this time, simply press star, then the number one on your telephone keypad. If you would like to withdraw your question, press the pound key. Thank you. I would now like to turn the call over to Don MacLeod, Director of Investor Relations. Please go ahead.
Thank you, Samantha, and good morning. I'd like to thank everyone for participating in M&T's third quarter 2019 earnings conference call, both by telephone and through the webcast. If you have not read the earnings release we issued this morning, you may access it along with the financial tables and schedules from our website, www.mtb.com, and by clicking on the Investor Relations link, and then on the Events and Presentations link. Also, before we start, I'd like to mention that comments made during this call might contain forward-looking statements relating to the banking industry and to M&T Bank Corporation. M&T encourages participants to refer to our SEC filings, including those found on forms 8-K, 10-K, and 10-Q, for a complete discussion of forward-looking statements. Now I'd like to introduce our Chief Financial Officer, Darren King.
Thank you, Don, good morning, everyone. As noted in this morning's earnings press release, M&T's results for the third quarter include several items that we think are worth highlighting. Total revenues grew from the prior quarter and the year-ago quarter, notwithstanding the lower interest rate environment and associated pressures on net interest income. Although we recognized an additional valuation allowance on our mortgage servicing rights, which reflects higher expected prepayments arising from lower interest rates, we also recorded an impressive increase in mortgage banking revenues. This demonstrates how mortgage loan originations can act as somewhat of a partial hedge for the mortgage servicing business. Loan growth continues to be steady and in line with our expectations for low single-digit aggregate growth in 2019. Credit quality is consistent with our recent experience, with net charge-offs stable at rates well below our long-term average.
A further decline in criticized loans was accompanied by an increase in non-accrual loans, primarily the result of one large loan previously reported as criticized. Let's take a look at the specifics. Diluted GAAP earnings per common share were $3.47 for the third quarter of 2019, compared with $3.34 in the second quarter of 2019 and $3.53 in the third quarter of 2018. Net income for the quarter was $480 million, compared with $473 million in the linked quarter and $526 million in the year ago quarter. On a GAAP basis, M&T's third quarter results produced an annualized rate of return on average assets of 1.58% and an annualized return on average common equity of 12.73%. This compares with rates of 1.6% and 12.68%, respectively, in the previous quarter.
Included in the GAAP results in the recent quarter were after-tax expenses from the amortization of intangible assets amounting to $4 million, or $0.03 per common share. Little change from the prior quarter. Consistent with our long-term practice, M&T provides supplemental reporting of its results on a net operating or tangible basis from which we have only ever excluded the after-tax effect of amortization of intangible assets, as well as any gains or expenses associated with mergers and acquisitions when they occur. M&T's net operating income for the third quarter, which excludes intangible amortization, was $484 million, compared to $477 million in the linked quarter and $531 million in last year's third quarter. Diluted net operating earnings per common share were $3.50 for the recent quarter, compared with $3.37 in 2019's second quarter and $3.56 in the third quarter of 2018.
Net operating income yielded annualized rates of return on average tangible assets and average tangible common shareholders' equity of 1.66% and 18.85% for the recent quarter. The comparable returns were 1.68% and 18.83% in the second quarter of 2019. In accordance with the SEC's guidelines, this morning's press release contains a tabular reconciliation of GAAP and non-GAAP results, including tangible assets and equity. Recall that both GAAP and net operating earnings for the second quarter of 2019 were impacted by a $48 million write-down of M&T's investment in an asset manager, which had been accounted for using the equity method of accounting. The write-down amounted to $36 million after-tax effect, or $0.27 per common share. In July 2019, M&T agreed to sell its investment in the asset manager, which had been obtained in the 2011 acquisition of Wilmington Trust Corporation. The sale was consummated in late September.
There were no such noteworthy items in 2018's third quarter. Turning to the balance sheet and the income statement. Taxable equivalent net interest income was $1.04 billion in the third quarter of 2019, down by $12 million, or 1% from the linked quarter. This reflects a narrower net interest margin, partially offset by growth in both loans and total earning assets. The margin for the quarter was 3.78%, down 13 basis points from 3.91% in the linked quarter. Contributing to that decline were several offsetting items. On the positive side, a more favorable mix of interest-earning assets, specifically a higher proportion of loans, added about two basis points to the margin. A higher level of cash on deposit at the Fed accounted for an estimated two basis points of the decline in the margin.
We estimate that lower short-term market rates, primarily LIBOR, accounted for some eight basis points of the decline. This is consistent with our expectations of a four to nine basis point decline in the margin over the ensuing 12-month period, following a hypothetical 25 basis point cut in the Fed funds target, and by implication, LIBOR. A higher cost of interest-bearing deposits, primarily mortgage escrow deposits, accounted for approximately five basis points of the decline. We continue to see inflows of these escrow deposits, a result of higher prepayment of mortgage loans we service or subservice on behalf of mortgage-backed security investors. Absent the higher level of escrow deposits, the total cost of interest-bearing deposits would've been approximately flat as we manage deposit rates lower.
As expected, the migration of deposits into higher-yielding categories, notably commercial deposits into interest checking and on-balance sheet sweep, has slowed, and rates offered on new certificates of deposit have declined. Average loans grew by 1% compared with the previous quarter. Originations remained solid, while payoffs and paydowns remain consistent with levels we've experienced in the first half of 2019. Looking at the loans by category on an average basis compared with the linked quarter, commercial and industrial loans were roughly flat compared with the linked quarter as the usual seasonal softness in loans to auto dealers to finance inventories was offsetting growth in other categories. Commercial real estate loans grew 1% compared with the second quarter. Residential real estate loans declined by less than 0.5% compared with the linked quarter.
As was the case last quarter, the continued comparatively steady pace of paydowns of mortgage loans acquired in the Hudson City transaction was partially offset by higher levels of loans originated for sale. Holding originations for sale aside, we expect the portfolio of acquired mortgage loans to continue its low double-digit rate of principal amortization in future quarters. Consumer loans were up 4%, as growth in recreation finance loans continues to outpace declines in home equity lines and loans. There were no particular standouts, positively or negatively, in our community banking regions from a loan growth perspective. From the line of business view, recreational vehicle financing, as well as residential and commercial mortgage banking were particularly strong. Average core customer deposits, which exclude deposits received at M&T's Cayman Island office and CDs over $250,000, grew an estimated 3% compared to the second quarter.
This primarily reflects the escrow deposits we referenced earlier. Turning to non-interest income. Non-interest income totaled $528 million in the third quarter, compared with $512 million in the prior quarter. Mortgage banking revenues were $137 million in the recent quarter, compared with $107 million in the linked quarter. Residential mortgage loans originated for sale were $835 million in the quarter, up from $723 million in the second quarter, reflecting a new wave of refinancing activity in the face of the lower longer-term interest rate environment, as well as seasonal strength. Total mortgage banking revenues, including origination and servicing activities, were $88 million in the third quarter, improved from $72 million in the prior quarter.
Clarify that as residential mortgage banking revenues.
Thank you, Don. In addition to higher gain on sale revenues, the increase reflects reaching the run rate of the residential loan servicing and sub-servicing that we acquired in the first and second quarters. Commercial banking revenues were $49 million in the third quarter, compared with $35 million in the linked quarter, reflecting notably stronger origination activity. The $30 million or 28% increase in total mortgage banking revenues brings with it higher expenses, notably $5 million of compensation costs, as well as the $14 million valuation allowance recorded during the quarter on our mortgage servicing rights. Trust income was $144 million in the recent quarter, unchanged from the previous quarter. Recall that the second quarter results include $4 million of seasonal fees earned assisting clients with their tax filings, which did not recur in the third quarter.
Trust income continues to grow in the upper single-digit range over the prior year. Service charges on deposit accounts were $111 million, up from $108 million in the second quarter. The recent quarter included $4 million of securities gains, representing valuation gains on equity securities, while the second quarter of 2019 included $9 million of similar valuation gains. Turning to expenses. Operating expenses for the third quarter, which exclude the amortization of intangible assets, were $873 million, compared with $868 million in the prior quarter. As previously noted, the prior quarter's results included a $48 million write-down of our investment in an asset manager acquired in the Wilmington Trust merger. The two most recent quarters results reflect an addition to a valuation allowance on our mortgage servicing rights as a result of lower long-term interest rates.
Those additions amounted to $14 million and $9 million in the third and second quarters respectively. As noted earlier, those same lower rates have prompted a notable uptick in residential mortgage loan originations and associated gain on sale revenues. Salaries and benefits were $477 million, up $21 million from $456 million in the prior quarter. Contributing to the increase was one extra compensation day in the third quarter, amounting to $5 million, as well as $5 million of compensation costs arising from the uptick in residential and commercial mortgage loan originations that I just referenced. In addition, the third quarter results include another $10 million of costs that we would not expect to recur in the fourth quarter. The efficiency ratio, which excludes intangible amortization from the numerator and securities gains or losses from the denominator, was 55.9% in the recent quarter, relatively unchanged from 2019's second quarter.
The ratios for both quarters include the additions to the MSR valuation allowance, while the second quarter figure includes the write-down of the investment in the asset manager. Next, let's turn to credit. Overall, credit quality remains consistent with our recent experience, given the continued strength of the economy. Annualized net charge-offs as a percentage of total loans were 16 basis points for the third quarter, little changed from the 15 basis points in the first half of 2019. Non-accrual loans increased by $140 million at September 30th, compared with the end of June, reflecting one large commercial loan to a wholesale distributor that was previously included in criticized loans. The ratio of non-accrual loans to total loans rose to 1.12% at the end of the quarter. Notwithstanding the increase in non-accrual loans, total criticized loans decreased further from the level seen at the end of June.
The provision for credit losses was $45 million in the recent quarter, exceeding net charge-offs by $9 million. The excess provision primarily relates to the non-accrual loan to the wholesale distributor, net of the decline in other criticized loans. The allowance for credit losses increased to $1.04 billion at the end of September, compared with $1.03 billion at the end of the previous quarter. The ratio of the allowance to total loans increased by one basis point to 1.16%. Loans 90 days past due, on which we continue to accrue interest, excluding acquired loans that had been marked to a fair value discounted acquisition, were $461 million at the end of the recent quarter. Of those loans, $434 million, or 94%, were guaranteed by government-related entities. Turning to capital.
M&T's common equity Tier One ratio was an estimated 9.81% at September 30th, compared with 9.84% at the end of the second quarter. The modest three basis point decline reflects the net impact of higher loans, earnings retention, and capital distributions. During the quarter, M&T repurchased 1.9 million shares of common stock at an aggregate cost of $300 million. Turning to the outlook. As we enter the final quarter of 2019, our guidance for the year remains little changed from our prior comments. We continue to expect growth in total loans in 2019 to be at the low single-digit pace, with continued runoff in residential mortgages more than offset by aggregate growth in the other loan categories. The reductions in short-term rates implied by the forward curve will continue to pressure both net interest income and the net interest margin.
We still expect modest year-over-year growth in net interest income for 2019. All else being equal and holding aside volatility in escrow deposit balances and associated cash balances placed at the Fed, each hypothetical reduction of 25 basis points in the Fed Funds target should result in four to nine basis points of margin pressure over the ensuing 12 months. The servicing and sub-servicing acquisitions we completed, combined with the strong third quarter origination activity in both our residential and commercial mortgage banking operations, has resulted in mortgage banking revenues growing better than expected. While trust income has been in line with our expectations, growing at a little better than a mid-single-digit pace. We would not expect mortgage banking results for the fourth quarter, either residential or commercial, to match those seen in the third quarter.
The remaining fee businesses continue to perform in line with our expectations, growing in the low single-digit range. Expenses for the year have grown a little more rapidly than we'd previously indicated, driven by two primary factors, growth in the mortgage business and investments in the bank, notably IT staff. Higher expenses associated with the servicing and sub-servicing acquisitions, as well as from the compensation expense associated with strong mortgage originations activity, drove expenses above our initial expectations. We also continue to expect to see some offsets to the year-to-date additions to our IT staff through lower contractor and consulting expenses starting in the fourth quarter. We expect fourth quarter expenses to be lower than the recent quarter. Our outlook for credit remains little changed. While sentiment about a potential recession is building, we're not seeing or hearing signs of a slowdown.
Our customers' largest concern is their ability to find enough workers with the right skills to add to capacity. As noted, criticized loans will be down this quarter from the end of June. That said, the specific reserve taken on the wholesale distributor we mentioned could result in a notable charge-off in the coming quarters. Regarding the new loan loss accounting standard, known as CECL, we've completed our second parallel run and expect to disclose preliminary results in the third quarter 10-Q. To preview those results, and based on the current economic forecast, we'd expect the allowance for losses on loans and leases to increase by approximately 5%-15% upon adoption of the accounting standard. That, in turn, should result in an impact to our capital ratios of less than 10 basis points. Regarding capital, we expect to continue to execute the capital plan that we've previously outlined.
Of course, as you are aware, our projections are subject to a number of uncertainties and various assumptions regarding national and regional economic growth, changes in interest rates, political events, and other macroeconomic factors, which may differ materially from what actually unfolds in the future. Let's open up the call to questions, before which Samantha will briefly review the instructions.
Ladies and gentlemen, as a reminder, if you would like to ask an audio question, please press star then 1 on your telephone keypad. Again, that is star 1 to ask a question. Your first question comes from the line of John Pancari from Evercore.
Morning.
Morning, John.
On the expense outlook, I just want to get an idea. I know you had indicated that the fourth quarter expenses should be down from the third quarter level. Just trying to understand how we should think about the magnitude of that decline. I know your previous expectation had been that the second half expenses would be similar to first half. If that still holds, that would imply a pretty sharp drop-off in fourth quarter. Just want to get an idea what type of decline we can expect. Thanks.
Sure. As we look at the fourth quarter and think about some of the things that happened in the third quarter, I'll start with that. The mortgage business obviously had a great quarter, and there was about $20 million of expenses associated with that which we didn't account for, to be honest, because we weren't sure where rates were going to be when we gave the guide. That $20 million is, I would just kind of add it to the guide, but let's not forget that there was $30 million of revenue that came along with that. Then there was about $10 million of expenses that we recognized in the third quarter that I don't foresee reoccurring in the fourth.
At that point, we're a little bit above what the guide was, but we wouldn't expect things to be necessarily equal in terms of each quarter in the second half being exactly the same. We think that there's some room for the expenses to come down and get down to the level that would've been implied on an average basis in the guide that we gave before.
Okay. On that same topic, you mentioned that obviously the higher mortgage-related costs were a factor, but also higher cost than you had expected in your technology investments. Can you give us a little more color there? What surprised you there, and how much should we think about that when we look at 2020 and if it could be a factor again as you forecast expenses going out? Thanks.
Sure. I guess in the IT space, there's a few things, some which I would describe as timing related in the quarter. One you see in the other software or processing and software line where there's some annual licensing expenses that came in in the quarter. We wouldn't expect those to repeat. When you look in the other costs, that's where a lot of the professional services expense is.
In there, we had talked about before the pace at which the contractors roll off as the new staff comes on, and there were some projects that extended maybe 30 days or 45 days, a little bit longer than we thought to bring them to completion. We've seen those roll off as we get to the end of the quarter, and that's why we feel confident that we'll actually start to see the impact of the adds to staff and the reduction in contractors as we go into the fourth quarter, and start to recognize that into 2020.
Okay, thank you.
Your next question comes from the line of Kenneth Zerbe from Morgan Stanley.
Great, thanks. Good morning.
Good morning, Ken.
Actually, just had a quick question. In terms of the mortgage banking business, obviously with the servicing that you've taken on. Obviously, this quarter was a really good result. If we think about the ongoing run rate from this, I know you said fourth quarter is probably not going to be as high as three Q, but what is kind of the right level? If you can kind of sort of pick a number. I'm just trying to get a sense. Are we at some meaningfully, sustainably higher level given the servicing assets? Thanks.
Sure. We're definitely at a higher level than where we would have been if you compared to the third quarter of last year. The third quarter of this year would reflect pretty much a full quarter's run rate of both the servicing rights that we acquired as well as the sub-servicing. That should be fairly stable. Of course, servicing is a declining asset, and so the fees come down a little bit each quarter based on the unpaid balances. Then really the volatility that we wouldn't expect to repeat, but is also a bit of a function of rates, is just the origination activity. Right now, I'm talking specifically on the residential side.
Now, one of the things that does tend to happen when you buy servicing rights is as you take the impairment or set up the allowance for prepayments, when those things prepay, oftentimes it shows up down the road in gain on sale. That was part of the uptick in this last quarter. In the commercial space, we had just a fantastic third quarter. I think it was the highest origination quarter in our history in commercial mortgage, driven off the originations.
I think you saw a lot of activity in the marketplace where rates had come down and where there was some concern about the amount of business that Fannie and Freddie might do. Towards the end of the quarter, they reaffirmed their appetite for the coming five quarters. They'll probably take a little bit for the pipeline to rebuild. It should be a solid quarter, but what I would consider a more normal rate of gain on sale in the commercial mortgage space versus what we saw in the third quarter. Those are really what was part of our comments about some of the things that were a little bit outsized in the mortgage business in the third quarter of this year, was really about originations as opposed to servicing. Servicing should be pretty consistent as we go from here.
Okay, great. Can you just remind us, what is the dollar amount, roughly, that the acquired servicing and the sub-servicing rights contributed to this quarter's earnings?
When we talked about it in March, we talked about an additional $60 million of servicing revenue for the year, with it kind of building and peaking in this quarter. We're probably in the range of $17 million-$20 million, somewhere in there. Maybe closer to $17 than to $20, in terms of what the run rate is on that acquired servicing. I guess I'll reiterate that you earn a fee based on the unpaid principal balance, and each month, those decline as people make their payments. As those decline, that fee income will move down at a similar pace.
All right, perfect. Thank you.
Your next question comes from the line of Matthew O'Connor from Deutsche Bank.
Hi, guys. I just want to follow up on expenses a little bit. It's been a pretty busy day. Just as I kind of try to add in the comments that you made about expenses in the fourth quarter backing out the $30 million and kind of coming to the maybe average of what you were thinking before. What does that imply for 4Q dollar expenses if we do that math?
I guess if you do the math and you look at what we would've implied in the guide, it would've been right around $830 million-$835 million a quarter, and you're probably in that range for the fourth quarter.
Okay. A follow-up. It was asked earlier, but as you think about the underlying expense growth at the company, obviously mortgage was unusually high, and you talked to some lumpiness in the investment. As you think about kind of whether it's full year 2020 or a medium term underlying expense growth, where is that? Where I'm getting to kind of the next question is I think there is some concern that banks your size may not have some of the scale that you need, and that's why we're seeing some expense pressure versus the big guys, not so much. Obviously with the rate headwinds out there pressuring revenues on that side, it's just more magnified potentially. If you could just address those two topics. Thank you.
Yeah. I guess I'll try and remember them all. You said a lot there, Matt. I guess overall, when you look at the expenses at the bank, this year was a year of expense growth that was very atypical for M&T. Obviously, a bunch of it was related to the mortgage business and the servicing that we acquired. That was $40 million. We've so far added about $22 million to the valuation allowance, which we might not have anticipated when we first agreed to acquire those loans back almost a year ago. Those have driven the expenses up. The other thing that's been happening, obviously, is the investments we're making to change the way we deliver IT at the bank. I think IT is probably one of the big things that is talked about in terms of scale.
What we're doing is, I think we're positioning ourselves to actually better compete with our competitors, both regional and large national players, by the investments that we're making. By shifting more of the resources, more of the IT team on staff, we think we can increase capacity in terms of what we're able to deliver for the same expense base. That's part of why we're making that move, and also the move to excuse me, more agile approaches, which should bring new capabilities to market a little more quickly than what we may have done in the past. Part of the shift is to make sure that we're competitive and, in the time that you're building that capability, you've got other projects going on.
We've talked a little bit before about the kind of double expense that you incur while you're making that transition. It'll happen a little bit, each quarter as we go through that, we'll start to see some of the payoff of that in the coming quarters, we think we'll start to see it in the fourth quarter with the professional services cost going down. When you think about where expenses are, and have been for M&T over time, we'll give you our thoughts on 2020 when we get to January. When you look at M&T through time, we've been a low nominal expense growth player. Generally kind of 2% a year or less. We certainly think that we're in an industry where that's warranted and we think about that all the time.
Given the growth this year, and investments that we've made, we'll look to see if we can't be in or below that range as we go into 2020. Like I said, we'll give you more details on that in the January call. I don't think we need to run at this kind of expense rate on an ongoing basis. We absolutely have some ways to improve productivity, based on the investments that we've made, which will help us stay competitive. Obviously we will react to the changing rate environment, which we've done successfully in the past.
Okay. That's helpful. Thank you.
Your next question comes from the line of Frank Schiraldi with Sandler O'Neill.
Good morning. Just wondering, Darren, if you could give a little more color on the credit that moved into nonaccrual in the quarter, in terms of industry, geography, collateral. Any color you could give would be interesting. Thanks.
Yeah, sure. Obviously, we're not going to talk about specific customers per se. When we look at this one customer, they're a wholesaler. Some of the financing is tied to inventory and receivables, and they've had some challenges internally with their management and that's put them in a bit of a cash bind. That's why we've kind of moved it from criticized into nonaccrual. The question will be the ultimate value of the receivables and collectibility of those, as well as the inventory. We obviously don't think it's zero, but we do expect that there will be likely some loss on that as we work our way through.
Obviously, given the magnitude of this relationship, and the industry it was in, we went through all of our relationships over $100 million, which fortunately I can count them on both hands, and looked through them, and they're not in similar industries and are in fact quite healthy. We looked through for other wholesaler and distributor type relationships, and we didn't find any others that had this similar situation. When I look through the loans that went into criticized through the first and second quarter, this was one of them. When we talked about it at the time, we looked for any common themes in those criticized loans, either in terms of industry, geography, or the like, and we couldn't find anything, other than they tended to be situational specific and oftentimes related to management. I would describe this as in that same ilk.
Okay. As a follow-up, you clearly mentioned you're not seeing broader weakness in credit. Just wondering if there's any areas that you're seeing as becoming more frothy, areas you guys are shying away from in your various geographies?
Nothing in particular. Obviously, we've paid a lot of attention all the way along to some of the office space, a la things going on recently in the industry. We don't have much there. When we look around, there's nothing in particular that I would point to that is what I would describe as very frothy. There continues to be lots of competition for lending. Pricing seems to be fairly reasonable. When I look at our spreads, they've been fairly consistent for the last three quarters. They did drop after tax reform. You kind of saw a resetting of the industry. You saw resetting of margin after tax reform, but that's stabilized since then. Overall, things seem really good. Structures haven't really changed much from what we've been dealing with over the last few quarters.
As we look, obviously New York City is a place we pay a lot of attention to, and we don't see anything there that causes us concern. Nothing is really standing out right now, which always makes you wonder what's going to pop. We're continuously looking through the portfolio to see where there might be signs of weakness. At this point, there's no one geography, industry or loan type, be it C&I, permanent mortgages, construction, where we have any concerns.
Okay. All right. Thanks for the color.
Your next question comes from the line of Ken Usdin from Jefferies.
Thanks. Good morning. Hey, Darren, can we talk a little bit more about the net interest income? I heard your comments about the year-over-year growth expected still for 2019. To the points about the cuts that have already happened and your guidance about what a cut does, the four to nine on a full year basis, can you help us understand the pushes and pulls with just the remaining burden from the cuts and then the excess liquidity in terms of how you expect the NIM to traject from here in terms of magnitude of compression as you look to the fourth?
Right. Now it's a bit of a moving target these days, isn't it, Ken, with the pace of LIBOR and where the Fed might go. That's really why we try to help give you guys the guide by talking about the four to nine basis points. That way, you guys can figure out what you think the rate curve might look like. I guess a couple of things that I think are important when we talk about the margin and the net interest income. When you look over the last two quarters at the margin compression, the print doesn't seem great, obviously. When you look at net interest income, it was down $9 million last quarter and down $12 million this quarter.
When you look at some of the things going on underneath that might take you beyond the four to nine, you look at increases in cash balances, and the cash balances carry a positive spread, albeit small. They don't harm net interest income, but they certainly harm the margin. You would see the same thing with the escrow balances that have grown last quarter and this quarter as well, that they're basically a slightly positive, and in some cases slightly negative carry, they don't have much of an impact on net interest income, but they do on the margin. As we look forward, the four to nine per 25 seems fairly reasonable to us.
The actual margin will move around a little bit depending on what happens with escrow balances, which can be as much a function of the rate environment and prepayments, as well as other cash balances at the Fed. There's a bunch of pieces that are moving there. At its core, it's four to nine. The one thing that I think is encouraging is we've started to see deposit costs bend over. If not for the escrow balance growth this quarter, we would have seen the increase in interest-bearing deposit costs be basically zero. We would start to see some traction on repricing activities as we go into the coming quarters, which will help moderate the impact of any further decreases in LIBOR. It still keeps you within that four to nine range as we go forward.
Okay.
Yep.
Yeah. I guess my follow-up would just be then to your point about the modest downward trajectory of NII dollars. Is that the trajectory that you'd still expect given, and I hear you on all the moving parts of it, but do we still just have to expect that modest slide in NII as you go forward from here, or is there something that can change with regards to that trajectory?
The trajectory, it might move around a little bit in any given quarter, depending on the pace of deposit repricing. Those are over a 12 month average is kind of the four to nine. Obviously the biggest driver is just LIBOR and how fast LIBOR moves down and how many rate cuts we get. Most of the two we had this quarter are in there. There's probably a little bit of residual from the last cut just because of when it happened in the quarter. We'll see some of that in the fourth quarter, and then we'll see what happens with LIBOR and the Fed over the coming meetings.
Okay. Last quick one. You mentioned how much higher are the escrow deposits priced versus I guess the average interest-bearing deposits at 85 to your point that the 85 would have been closer to flat without the escrow. Can you just give us an understanding of the difference between where the escrow deposits are coming on versus where your average is underneath? Thanks.
Yeah, sure. Round numbers, when you look at the escrow balances, they're priced off an index, either off of Fed Funds or off of LIBOR, so they're around 180 these days. When you look through the rest of the portfolio, it's primarily commercial interest checking that is driving the interest expense on that line. Within there's a range of rates. Those are individually negotiated with each customer based on the relationship and the magnitude of it. My recollection is that the average there is around 70, 75 basis points. You probably double in terms of what the escrow balances are earning versus what the other ones are. The nice thing about the escrow balances is because they are linked to the index, as the index moves, so too does the cost of those.
Understood. Thank you, Darren.
Your next question comes from the line of Steven Alexopoulos from J.P. Morgan. Steven, your line is open.
Hey, Darren, can you hear me?
Steven, you might have yourself on mute. Okay, your next question comes from the line of Marty Mosby with Vining Sparks.
Well, thank you. I wanted to touch base with you just to summarize these moving pieces, because when you think about your expenses, you really have, one, the mortgage valuation, if I'm getting you right, in other expenses. Typically, that kind of gets netted out in the revenues, you wouldn't see that grossed up expense. That's driving your expenses higher this particular year. You've got this transition going from external technology support to internal technology support. Right in the middle of that doubling up of expenses, you're seeing the benefit of those two things rolling off as you kind of guide it towards a $40 million-$50 million reduction in expenses as you go from the third to fourth quarter.
As you go into next year, if you take that dynamic forward, it seems like your expenses could actually drop as you go into 2020 once you get the benefits of not having the external tech spend and you don't have the, let's say, the long-term rates are flat, so you don't have this valuation situation.
I guess if you look forward, Marty, I think the way you're thinking about it is exactly the way we're seeing it and thinking about it. Really the only thing that would be a timing difference on when we might see the expenses actually drop versus the growth rate drop very dramatically. The latter, we're definitely expecting. To see actual decrease, we still got the transition going on with that tech team. That's going to continue through 2021. We're looking to bring on a large number of folks, and they'll come on in chunks of probably 100 to 150. That kind of happens consistently over maybe a quarter to every four months. When you're in that time period, you're bearing the double cost. We've started down that path. We did our first wave this year.
Probably took us a little bit longer to get up to speed and a little bit longer to get some of the contractors out than we might have anticipated going on or going into it. We're getting smarter at how we do that and better able to match the timing. We shouldn't carry it quite as long as we did this year, but we will continue to see the cost of that transition happen through 2020 and really start to see the benefit in 2021. On the mortgage side, like you pointed out, that stuff should wash itself out. As we go through 2020 compared to 2019 in the mortgage portfolio, holding the valuation reserve to the side, that should start to normalize itself and the expenses will obviously move in relation to how the unpaid balances are performing.
When we watch some of the other costs and some of the other professional services, there were some other activities that we had going on this year that we wouldn't expect to repeat. Those should help expenses. The offset there is just compensation costs for folks that have been added to the team. When you just give people raises, that raises your run rate, and you got to offset that. There's a couple of things going on back and forth, but when you think high level over the next couple of years, the way you're thinking about it, Marty, is exactly the way we're thinking about it.
On the credit side, because really the variances in this particular quarter are, in my mind, are expenses and then up in the credit side. The $9 million you have put into the allowance, and you put some into it last quarter as well. Has that covered what you expect to see in the charge-off from this large relationship that might get charged off in the next couple of quarters? Have you kind of pre-funded that, and so when we see the event, you'll actually just be drawing down reserves to pay off the expected loss?
What happened there, Marty, was that loan was obviously criticized before, and we started to reserve for it at that point. There were some other criticized loans that either paid off or became performing. What we had put aside for them helped cover the increase in the one that went non-accrual, and then we added to it. Based on what we know today, it's in the provision, and we feel good about it. This one is a little bit more of a fluid relationship just because of the nature of the collateral.
Depending on how things move with that organization, the collateral values could move around a little bit on us, and that's why we moved it to non-accrual and why we set up or added to the provision or added to the reserve, and let you guys know that we'll probably have something in the future. The exact magnitude and timing is still a little bit unclear. Outside of that, as you point out, criticized coming back down, and the rest of the portfolios, and we look at the delinquencies and what has been happening with the charge-offs. Knock on wood, everything has been fairly stable and predictable. All else equal, one would expect that as rates come down, that would help debt service coverage ratios and customers' ability to pay, which wouldn't lead you to believe that the charge-offs might tick up.
The counterargument, obviously, is if GDP slows and growth slows, that while the interest costs might come down, you got to keep an eye on revenues and the ability to service the debt from that side of the equation.
Thanks.
Your next question comes from the line of Mike Mayo with Wells Fargo.
Thank you. I have a balance sheet question and a tech follow-up question. For the balance sheet, the long-term debt has been trending down for the past three to four years. Is that going to stabilize, or do you think that can continue to run off as your liquidity gets kind of smoothed out unless you have another deal?
Okay. If you look at the long-term debt, it's been coming down, but there's been a bit of an uptick in short-term, and then obviously an uptick in some of the deposit balances in the last little while, particularly driven by escrow. The mix of long term and short term, we will now start to reevaluate given that LCR, the ruling has been finalized. We were kind of using more short-term borrowing to manage the liquidity coverage ratio and our investments in HQLA while we were waiting for the rule to be finalized. Now that there's a little bit more certainty there, we'll look at how much securities balances we want on the balance sheet and look at the funding mix of those.
We've also got some long-term debt that is going to roll over next year, and obviously, we'll look to replace that, but subject to where we see the balance sheet going. We think we've got some opportunity there to manage the securities portfolio now that the rules have been finalized and given how we feel about the strength of the organization.
Regarding tech, I believe when I've talked to you in the past, it's around 10% or so of your revenues are spent on technology. It's around maybe a little bit less than $700 million, with about 60/40 run the bank, change the bank. As you bring more staff in-house, do you see either one of those changing?
I guess we haven't, I don't think, disclosed or talked about what the tech budget is in specifics. I think that the 10%-12% that you might be referencing is we talked about the compound annual growth rate of that part of the budget. The mix between kind of run the bank versus improve the bank, build the bank, whatever you want to describe. 60/40 is probably a reasonable estimate. It can move around from year to year, depending on what's going on. When we're making investments in new regulations like FDIC 370 or CECL, I don't know whether we would call that build the bank or run the bank. We're probably splitting hairs there. At the end of the day, the total tech budget in the short term is running a little bit higher because of the transition that we're making.
Our belief is that we will, all else equal, be flat with more on staff and less contractors as we go forward, and then the growth rate from there wouldn't need to be at the 10%-12% rate, but would probably continue to run at a rate slightly above the bank average. That the investments we're making in technology would be offset by cost reduction somewhere else as we gain productivity improvements from the tech spend.
The new staff that you're bringing on, are they focused on besides the FDIC and CECL, digital delivery, AI ML, back office, transitioning apps to the cloud? Where are you putting the new hires to work?
Well, we've got a whole host of things that we're working on. A bunch of them you named. Obviously, some of the regulatory changes have been a big consumer of our tech team this year as we get ready for FDIC Part 370 and CECL. At the same time, we've been making consistent investments in a lot of the systems that are used either by our customers or by our employees who interact with our customers. We've been doing that in the commercial loan origination space. We've been doing it in the treasury management and merchant space in the commercial part of the bank. We've been doing it in some of the M&A support that we do in our institutional client services business. We've been doing it in customer-facing things, both in our wealth management business as well as in the consumer business.
We continue to make investments in cybersecurity, in data. It runs a whole host. We are starting to migrate some applications to the cloud. Some of the newer ones often run on the cloud, and we're doing some of that. We've got a whole range of places where we're investing. It always kind of starts with the customer and works backward and looks at where are we in the competition and where we need to make sure we're positioning the bank, and then also looking for ways that we can obviously improve productivity or efficiency.
Thank you.
Your next question comes from the line of Brian Foran with Autonomous.
It's easy to get lost because it hits so many line items, and everything has a lag and timing issues. I guess if you just step back, was this a good acquisition, accretion or return on invested capital, however you want to measure it? If you just set aside the quarter-to-quarter stuff and where rates are and these escrow deposits, does this all work?
Ask that again, Brian. The first part of it we missed.
Just the mortgage servicing. It is hard to really parse apart everything. If you just think about it on a net basis, has the acquisition and the resulting business it built, did it work? Is it accretive? Is it going to be good for 2020, or is there some underlying slippage? There is so many moving parts and it hits so many line items.
Sure. I think the short answer is we are still happy with the acquisition of the mortgage servicing rights. It's probably not quite the return that we thought when we first did it. That's affected by the timing of some of the charges that we're taking. Overall, we feel it's accretive and above our long-term cost of capital, which is how we kind of evaluate everything. We had some wiggle room built in, which you always do, given the volatility of this business. When we look at it, to your point, stepping back, we still feel good about it and are happy we did it.
Great. Thank you.
Your next question comes from the line of Gerard Cassidy with RBC.
Hi, Darren.
Good morning, Gerard.
Quick question. There seems to obviously be concerns about your IT spending and maybe not keeping up with some of the bigger banks. What measure would you recommend investors look at to show that you are as competitive as the big banks? Obviously, the big banks can be splashy with the dollar signs that they're spending, and yours will be a lot less because you're a smaller bank. Is deposits the preferred area to look? Your deposit growth year-over-year was better than many of the big banks. Can you share with us what we should be looking at?
Sure. It's a great question, Gerard. I guess I think the best place to look at is customer-based measures. I would look at customer growth rates, customer attrition rates, satisfaction rates. It could be J.D. Power, it could be Greenwich in the commercial and small business world. It could be some of the Cerulli research in the wealth space. At the end of the day, the number one thing you're investing in technology for is to help provide a great experience for your customers and make sure that you're on par with everyone else on things that are just kind of what we would call hygienics or expected, and look for places where you can differentiate what you're doing.
One of the things to keep in mind when you compare the regional banks to the large ones is the large banks have broader-based businesses than we do. We don't have trading operations, for instance, which would consume a lot of dollars. I can't speak for everyone else, but what I can speak for M&T is through all the acquisitions that we've done, we've never maintained duplicate systems. We're not carrying that expense, and we don't need to incur the cost of running them or consolidating them. A lot is made of how much you spend. I think most important is, are you giving your customers the tools and capabilities and services that they're looking for? They let you know all the time. You can kind of look at deposits, as you pointed out, Gerard.
The only thing you just got to watch for, in my opinion, on deposits is you can move the deposit balance number with rates a lot faster than you can with technology. The one that's really hard to fake is transaction accounts or operating accounts, whether they're consumers, small business, or commercial entities. When we think about our tech investments, we're always thinking about those parts of the customer experience and how we can make it better.
Great. Thank you.
Your next question comes from the line of Saul Martinez from UBS.
Hi, good afternoon. Thanks for taking my question. I guess I want to parse through a lot of the moving parts around the NIM guidance. Try to put a dollar sign around it. I know there's a lot of uncertainty about rates and whatnot, but four to nine basis points of NIMs pressure for every 25 basis point reduction in the short end. Looks like the fourth quarter one-month LIBOR, the average should be, I don't know, I'll probably have to sharpen my pencil here, but maybe 40, 50 basis points lower than what it was on average in the third quarter, especially if we see an October cut. I mean, four to nine basis points.
Is it fair to say that it's going to be hard not to see some degradation and reduction in net interest income in the fourth quarter on a nominal dollar basis versus what you posted in the third quarter?
Yeah. I think that's right. The actual margin, you've got the math right in looking at where Fed Funds and LIBOR might be over the quarter. The NIM, the prints of the NIM might move around a little bit from that, depending on where cash balances end up and how much more escrow comes on. You're in the right ballpark. When you look at the dollar impact, it's going to come down. When you look at the guide that we've given, I think you should be able to figure out where we think it's roughly going to be in the fourth quarter. The other wild card is deposit repricing.
Sure.
Like we mentioned, we feel good about the progress that we made this quarter in fending over deposit pricing. We should start to see that come down this quarter. The other thing that's a little bit tricky to gauge is just the pace of loan growth in the quarter, in particular because the fourth quarter always seems to have some uptick. The question there is when does it occur? Right.
Right.
You should start to see a little earlier in the quarter in the floor plan, auto floor plan balances. Oftentimes we see a big December. You kind of wonder what drives seasonality. It's year-end, and people looking to close deals for the tax year. We always seem to see a spike in some of the loans that book in December, and obviously that will, depending on the timing impact the dollars of net interest income.
Got it. I guess if I could just get one more in, a little bit more of a broader question. The downside of having sort of best-in-class profitability, efficiency, deposit franchise is that it does leave you more vulnerable when the environment starts to turn a little bit worse. I guess the question is just how do you think about sustaining profitability and creating value in that environment? How do you think about opportunities where you can grow? I guess the gist of the question though is can you do that on a standalone basis? Can you really or do you need to do something more strategic to really be able to sustain sort of the best-in-class type of profitability and operating metrics you have?
I appreciate the question because it's a really good one. The strength of the franchise, as you point out, is both a blessing and a curse. When you look at how we run the bank and how we think about the bank, we've always started with returns. To talk about the returns, I was pleased to hear you say that because that's our focus. That thought process is what has kept us out of trouble and what has helped us make a lot of the investment decisions that we've done through time. Sometimes it was acquiring servicing or sub-servicing business because the returns made sense based on what was going on in the industry. It's helped us decide when to make loans, and how to structure them because if the returns don't make sense to us, we've tended to step away.
It's helped us price acquisitions and make sure that we're being thoughtful with capital and how we deploy it to grow with mergers and acquisitions. When you look overall, one of the key things for us is making sure that we maintain those returns. If that means we need to be patient in time and have a little bit less growth, we're okay with that because we would rather make sure we protect the profitability and the ability to generate capital, so that we can make some of the investments that we're making. If you look at really the last 12 months to 18 months, we had some outsized increases in the net interest margin because of the positioning of the balance sheet and the runoff of the Hudson City portfolio, which allowed us to grow the loans, deploy capital and increase the margin.
We took advantage of that, and we upped the investment that we were able to make in the franchise. The reverse is true in that as we see rates come down and our net interest margin relative to the peers kind of resumes its more normal position of slightly above the median. We should be able to anticipate bringing the expenses down commensurate with that. You look across the different alternatives that you have of making loans, investing in other businesses, buying other banks, or deploying excess capital to shareholders, and think about the returns. Obviously, the first place we want to invest is in the business and in our customers. This year we've been able to do that a little bit more with the loan growth that you've seen. We think we can continue to do that.
We don't feel pressure to have to get to a certain size or to buy some growth or either in the form of poor pricing or in poorly or ill-conceived acquisitions. We'll continue to focus on returns and being patient and making sure that we're running a good bank, and that will leave us, we believe, with opportunities when they present themselves.
Got it. Thanks so much. Really appreciate it.
There are no further questions. I would now like to hand the conference back over to Don McLeod for any additional or closing remarks.
Again, thank you all for participating today. As always, if any clarification of any of the items in the call or news release is necessary, please contact our investor relations department at 716-842-5138.
This does conclude today's conference call. You may now disconnect your lines.