Scaling Loan Growth and Portfolio Management with ESL

Transcription:

Bryan Peckinpaugh (00:09):

All right. Thank you very much. Appreciate everybody being here and for putting on a great conference. Fantastic panel to lead our way into this conversation. My name is Brian Peckinpaugh. I'm a Senior Vice President at Baker Hill. I run our account management team as well as our marketing and analytics business. And I'm joined today by Keith Cleary from ESL.

Keith Cleary (00:34):

Good morning. I'm Keith Cleary from ESL. I'm Senior Vice President, Director of Business Banking. Been around banking for about 35 years, but been at ESL almost 10.

Bryan Peckinpaugh (00:44):

Awesome. We're going to talk a little bit about segmentation specifically on the lending side. So again, some great conversations from the panel that I think you'll see reinforced here as we dive into the specifics about lending and portfolio management. I'm sure I don't have to convince many people in the room of some of these ideas, but thought it'd be good to just kind of hit on a couple thoughts from the industry where even though we're seeing some credit crunch and some risk avoidance in the small business space, it's still an opportunity for those community banks to drive business, to bring those loans in the door and focusing on those communities that you serve to again, grow that book of business where there is a market to attack. But when we're going to bring in that volume where we're going from a large, complex commercial deal and focusing on more throughput of a small business loan, we have to make sure we do it profitably. We make those yes and no decisions as effectively as we can. So when we talk today about the success that ESL has had, we're going to focus in on that use of technology to allow bankers to be bankers and not processors, to spend time with the borrowers to give that advice that the panel was talking about earlier this morning.

(02:10):

So what we want you to take away from today are some key thoughts around how you can view segmentation within your portfolio, what factors you may want to consider as you break that portfolio apart, how the technology can interplay across the entire life cycle of that borrowing relationship. And finally, all of those applied to ESL and their story of launching member business lending 13 years ago. So when we think about segmentation strategies at Baker Hill and with our partners like ESL, we really focus on that complexity of deal and how we can streamline the lower end as effectively as possible. Again, to free up time and capacity within your organization, everybody wants to focus on those large deals. Those are the fun ones. Those are the ones that have the highest interest margin on them for the bank, but you do have a lot of volume coming through that we need to make sure we focus on and bring tools to bear again to do that profitably.

(03:17):

So we focus in on again, that size of deal, that complexity. If you're treating all of these loan types the same, if your processes are the same, if your workflow is the same, you're going to spend as much time focused on that $25,000 operating line as you would. That $10 million hospital expansion that's going to really drag on that capacity of the organization, make it difficult to put that time and energy on the larger deals that require it and tie up your resources, getting those smaller deals through quickly, which is what the expectation of the borrower is today. So we segment based on a couple of key factors that allow you to, again, delineate between what might be a more complex commercial loan and a streamlined small business loan. So what are some of the characteristics that we look at as we think about a commercial loan, those large dollars as a given, but really about the structure of the loan itself, what goes into understanding the credit risk, looking at all of the borrowers relationships that are there, what the business structures are, how we bring in collateral to support it. Lots of things that take time to understand, put together and review. Those also lead to, from a portfolio perspective, a more complex review on an annual basis. So we have to go back through all of that as we review that loan annually, quarterly, whatever your cadence might be

(04:51):

In the small business space. And what ESL was able to do with our supporting technology is carve out portions of this business to focus in on the things that can be automated so that you can again, build that capacity into your process, let the technology help you wherever possible with the lower end of your portfolio. So again, you can spend that time on those bigger deals. Keith, I don't know if you want to talk a little bit here, I know you will in a few minutes about some of your specific strategies, but what of these were kind of most important as you guys were launching this 13 years ago?

Keith Cleary (05:27):

Yeah, so we launched, we're about 103 years old as ESL. We have roots. George Eastman who started Kodak, also started ESL to help his employees purchase homes and stay in the area and attract good talent 103 years ago. And in Rochester, New York, unfortunately, Kodak used to employ about 30,000. Now they're down to about 2,400. So we were about a dozen companies employing hundreds if not thousands of people. Now, because of what's happened over the last 30, 40 years, we're thousands of companies employing dozens of people. So the needs do line up. We can talk about dollar amounts. They're usually just a proxy for complexity. And when you build for the rule, not the exception, they generally fit in terms of whether you're cutting it by sales size, by credit size, or by employee size. You're just trying to find what works for your organization, your community in terms of what segments make the most sense.

(06:25):

What we find in Rochester, we're about a 10 to one under million dollar, an over million dollar population. I suspect a lot of us in our communities share about the same thing. We have about 30, depends on what data set you use about 30 to 40,000 businesses under a million dollars and three to 4,000 over. And that over a million goes from a million to billions. We have some large publicly traded still. So we just had to find a segmentation that works for us. What do we want to be? And that's what I can tell our story, but it needs to work for your organization and what you're trying to accomplish needs to align with your strategy, with your community and with your opportunity. And I guess I'll just kind of touch base. We're about a $9 billion credit union, nine and a half billion now.

(07:13):

We have 900 people we just crossed. And actually the good news is at the end of October, we launched business banking in 2010. We had 3000 small business customers already, which tells you the characteristics of small business. We had plumbers, bakers, daycare operators, all operating their business out of their personal accounts, home equities, personal checking or whatever. So when we were able to bring a product set to them, they just migrated over to our business banking platform set. We've grown, and actually the good news is we're not 15 nine. We're kind of embarrassing to put that on the slide. We crossed 16,000 at the end of October. So we've grown significantly in those 13 years. And generally those are small businesses, and I'll talk about our journey in terms of engaging our branches. But when we launched business banking as a 9 billion credit union, we have scale.

(08:08):

We can do large loans. We just crossed a billion dollars in exposure in our commercial banking group. A large part of that is commercial real estate. It's hard to make. Well, it's great news. When you make a 10, $20 million commercial real estate deal, that's a lot of $50,000 loans, a lot of a hundred thousand dollars loans to kind of offset that. So our focus has been on not only making sure that pie reflects the risk that we want, but also that pie just keeps getting bigger and bringing more C and I and more of the profile of our community into that. So we've focused on households a lot. We did focus on the SBA, I'll always remember, well, I wasn't there at the very beginning, but our first year as an SBA lender, we made three loans for $385,000. Since then, we've made hundreds of SBA loans for approaching well over 30, I can't remember how many millions of dollars, but it's about 30, $40 million.

(09:02):

So that just emphasized that's where some of our focus has been. We have a great team supporting us. Sometimes these numbers are cool. They tell our story. We're the number one credit union in all of New York for SBA lending. We're number five when we think about the number of loans. And it doesn't look like we're on there. I think we're number 17 for the dollar amount of loans that are made across the country. So yes, segmentation has been important for us. And I will tell you one of our journeys, I heard some numbers up here in Rochester, New York when we were all hands on deck when we launched in 2010. Then we, as the portfolio grew, we did segment our teams. We picked the $10 million and over in revenue probably because we were a bunch of commercial bankers, and that's what everybody else did. So we picked 10 million and what we found was there's just such a big difference in complexity between a million and under a million and over. So we've reduced that. So now we have commercial bankers calling on 5 million and over revenue companies and business banking focused on 5 million and under. But we also created a business banker, one team that focuses almost exclusively with branch partners and million dollar and under businesses with a specific product set and just their day and activities is managed a little bit differently than the other bankers.

(10:30):

That's how we approach segmentation. If we look at the three legs of a growth stool, we've got our outstandings, we're up over a billion dollars in exposure and about $800 million in outstandings. We've become one of the larger SBA lenders in the state. We're number two in our SBA market too, right up against some of the larger commercial banks. And we've also brought in a number of households. I guess maybe I'll just keep on going. In 2015, 16, we had a couple of big moments that helped us grow our households, especially in that million dollar and under, I think you mentioned that we brought in Q2 to bring a mobile banking solution. We were kind of piggybacking when the retail bank has a 90 year head start on you and about 300,000 customers that's grown to 400,000, you're asked to use a lot of their stuff when you come on board with 3000 customers.

(11:27):

So we leveraged off of a cash management system. We leveraged off of mobile banking system, and as we grew, we just found that these needs aren't there. I don't know if anyone's tried this, but our largest bill pay a few years ago was $9,999 because that's what our consumer limit was, and we couldn't exceed it on the commercial side. So we've had to chip away at all those either internal rules or products that limited us or small businesses from using them. So we brought on Q2 as a partner for our business mobile, and I think you were mentioning, I want to make sure we have that age old question. Do we let our DBAs use both put their consumer, we have a different consumer system. So we've had to solve for all those with the focus on the DBAs and the small businesses. But then we also, at the same time, we engaged with our frontline, our branches we're up to 24 branches at the time.

(12:23):

We had 19 where the small business owner wants to do business in their community and their branch, that's where they go probably for their retail business. So they wanted to talk to our branch managers for that. So we had to train people. And then we worked with Baker Hill to put a branch specific under a hundred thousand dollars loan request term loans and lines of credit in the branch's hands and designed it with Baker Hill to make sure we had the right questions and the right process to make it easy. Because if a small business owner is in a branch, many of us know this, I spent about eight, nine years in the branch. You are kind of pairing up busy people with less experience in commercial and business banking. You're a small business owner, they're great at what they do, let's face it. Right?

(13:10):

And then our branch partners are asked to do so much in a day. They never know who's going to come and sit down. They can be dealing with an IRA student loan, a home equity, and then a business business opportunity. Hard to train and hard to get good at all of those things, right? So we wanted to put the tools in hand that made it very simple for them at the same time making it easy for the customer. So we put an auto decisioning application in the hands of the branches, and that's where we started to see a spike. We were walking along at 3000, 4,000, 5,000 in 2016 or so. That's when we went from 5,000 to 8,000 to 8,000 to 10,000 and now up to 16,000 households.

Bryan Peckinpaugh (13:55):

Yeah, it's been a great journey with our partners at ESL. We have a heavy focus on what we call that gray area strategy. So as Keith mentioned, going from the fits and starts a little bit at the beginning to starting to put that auto decision in and making sure we ratchet in those metrics so that we're making sure that the system is taking on as much as possible without causing doubt in the decisions that are being made. It's very easy to set a high end and a low end that are very exclusive and continue to judgmentally decision everything in the middle. But that limits your output of the technology. You really need to look at what good looks like for your institution, what are the models that you're using, telling you on the credit worthiness of the borrowers that are coming in and getting that as tuned in as you can.

(14:50):

So again, you take the risk out while still making good sound credit decisions. And as you'll see here at the last bullet, it's been very effective for ESL in continuing to increase that volume, increase the applications that are going through that auto decisioning process without any impact to the credit quality that they're seeing. So it's been a very effective process. And we also have started to deploy that across the reviews of the portfolio. So it's one thing to put these tools in place during your upfront origination and decision making, but if you can tie that into the performance over time of those loans within the portfolio, you start to really get a great picture of the credit decisions that are being made and creating a feedback loop that allows you to take those same review processes, delinquent loans or charged off loans, flowing that back through to that gray area strategy to again, further tune based on what you're seeing from a performance perspective. So kind of going beyond just a predictive score and creating a model that makes sense for your financial institution. So if you want to talk a little bit about what you guys are doing with those portfolio loans as you review them.

Keith Cleary (16:09):

Sure. I just want to emphasize as an organization, you do have to get very clear alignment between what Brian mentioned. What do you want your sales growth to be? Where do you want your credit quality to be and where do you want your folks spending time? And I think when you put together continue to pair that small business owner with a business banker one or a branch manager, they don't need a lot of process. They need advice, and that's where you want, it pains me to talk just about credit after so many great presentations about total relationships and deposits are important. I will limit my comments just to, because as a coach, this never gets out that I'm only talking about credit. But on the credit side, when you think about where we've either grown up or a relationship manager thinks they spend their day getting a streamlined credit process for some of the smaller requests is an enabler to do all those other things.

(17:06):

When you're in front of the customer, you want to be talking about what they need asking some of the great questions that came up this morning. You don't want to be asking for two years of tax returns and an AR aging and the latest interim QuickBooks statements. Number one, a lot of people don't have them. They're busy, and I don't know about you, but nine o'clock at night, I'm done. I'm toast. You're asking a plumber, a baker or a daycare operator to not only do what they do 12 hours a day now go home and figure all the stuff out that the bank needs. So I think the more you can blend into the background the better. But also you do have to watch your credit risk. But I'd also say one of the lights that went off at ESL very specifically, and I would just ask everyone to ask this at their own organization is if you're thinking about 25, 50, 75,000 loans, where else in your organization are those happening?

(17:57):

Right? A Chevy Tahoe is going to cost you $90,000, right? So you're making $90,000 loans in your organization, you're making $25,000 loans all the time. Why is the risk in that loan any different than the risk on a small business? Those of us that grew up in small business can go for the next 15 minutes as to why it's different. But a Chevy Tahoe loan, if you're doing that on an indirect basis to an individual, you're usually getting a W2 or a pay stub when it comes to the business side because they put some stenciling on the door. All of a sudden you're asking for two years of tax returns, you want to know how they're going to use it, now you're going to set them up for an interim and all this stuff. Why? You just have to add, now, it might be right for your organization, but there's probably a lot of 25 to a hundred thousand dollars loans happening in your organization today.

(18:48):

You just have to make sure you're worrying about them the same and using, that's the same process as much as you can. And then you do have to say, how good do we feel? Do you set the right? And I always use this visual, you set the right dials, you set the dials on where you want your bad credit experience to be, whether you want targeting a certain amount of delinquencies or a certain amount of charge offs. And then you can start to play around with those dials once you put an auto decisioning program in place. Got to have data that came up this morning. You got to have your own data. What's happening inside of your book? Are there things that you should tighten? Or if you're not getting enough throughput, where else can we look for an opportunity to maybe look at this a little bit differently?

(19:35):

And then the last thing is, like Brian said, the utilization of auto decisioning. I will tell you firsthand, I just bring a lot of anecdotes with me. I think when I heard neem talking about the branches, nobody wants to say no. Right? And a lot of our branches, when we do deliver a no, when we do an auto decision, I don't think I've ever gotten an email or feedback that said, Hey, this approval, I can't believe it happened. But I do get a lot of emails and feedback that says, boy, can you take a second look at this decline? Even though it's a five 80 credit score, there's extenuating circumstances. So a lot of people want to challenge the declines. And then versus again, nobody's ever called me and said, Ooh, I can't believe that one got approved. So you just have to make sure you're ready for all of those conversations.

(20:26):

And then I would just say what we've done a lot at ESL since we've had this for five or six years now, we've used our own data to take a look at that gray area. That gray area is where you're not using your auto decision tool. You've built an area that says, Hey, there's enough attributes here that make us want to take a second look. You lose your efficiencies if you've got that gray area too wide. So you're defeating your purpose, you're opening the door to a lot of discretion that you may or and may not want in there. So you just have to take a look at, and this is one of the things Baker Hill has helped us do. You look at your data, you look at what's been in the gray, but approved once somebody touches it, and then you look for those attributes and say, are those attributes that I can build into my model now that says, Hey, based on this next code, based on this personal credit score, we might as well turn those into approvals.

(21:17):

Or there are things that you'll always see that, no, even though I touched it, it's out of here. It's on the decline pile. So you want to just keep working on that auto decision utilization, get that as high as you want it to be, but then you just have to live with the output and then just kind of pick your lane in terms of do you want, what bad rate do you want? And then once you have that, you just have to kind of stick with it and deal with it. And I would also say just a couple of tips that I have is one of the things that happens, those of us in commercial and those of us that have managed the department, our book hasn't always been a billion dollars. And when I was thinking about, Hey, a few $25,000 loans went bad, it might add up to $150,000 in charge offs.

(22:06):

That doesn't feel good to anybody, right? You've kind of put a customer in a bad position, you've cost the bank $150,000 worth of decisions. But our CEO reminded me at one point, we're almost a 10 billion balance sheet that $150,000 loss, Keith, get over it. It's not just on your part of the balance sheet, it's on the entire bank. So it just goes back to you just kind of have to trade efficiency for if you've got really tight credit risk, you're trading efficiency risk. You just are on behalf of your employee and your customer. I just want to then, so one of the things that we, so the first five years, so we've been at this 13 years, the first four or five, it's a ton of fun, right? Everything's growth, everything's just, the charts all go up at 90 degrees. So everything's new. You're doing a lot of new business.

(22:56):

Once year five comes, now you're getting into annual reviews. You might get a customer that's stressed a little bit and everybody loves doing annual reviews. Your employees love them, your customers are down to you and maybe a cousin are the only people they ever send their financial statements to. So you look for the quickest way out of those and try to process them quickly. But as we grew, every customer, commercial customer is the gift that keeps on giving, right? You have to touch that every year if you choose. So using Baker Hills portfolio manager, portfolio risk manager, you can kind of take a look at some attributes of an amortizing mortgage, the old what could go wrong if you just manage that by delinquency and it works a lot of the time. You just got to factor that into your efficiency calculation and say, how many of your mortgages just chug along and pay you get a delinquency report.

(23:50):

So if there's a problem, you're going to catch that. Could you catch that earlier on a rent roll? Maybe. But how many? Who wants to go collect a bunch of rent rolls? So you just have to take a look at the tools that are available and say, listen, there are some indicators that say this credit is fine. It's as we approved it and we use portfolio risk manager, it puts some annual reviews out three years, so you're not touching it every year. You might touch it every three. Now you bill triggers in, right? If my score went down as an individual, we're pulling credit. We're just letting the back it happen. If additional liens pop on or there's a deterioration in credit in any way, we can jump in, but we're only jumping in when it's necessary. So one of the things we took a look at, because the other thing is it's fun to grow a business, but at some point you become BAU in the organization, you're subject to all the same staffing models, you're subject to all the same efficiency ratios, and you have to take a look at where you're spending your time.

(24:55):

Our credit department, if you treat everything the same, it takes a good eight hours start to finish to do a full annual review. We need to do that when we have 15 million of exposure. We don't need to do that when we have $75,000 of exposure. So we've been able to use portfolio monitoring to really take a look at where do we want to spend our time. And you could see we're spending anywhere from eight hours, three hours, 30 minutes to 15 minutes. So everything's not created equal. We're spending the appropriate amount of time on an annual review.

(25:32):

And the good news is, knock on whatever, I'm not a believer in jinxes or anything like that, but as of the end of October, we were at about a two and a half percent delinquency rate and almost no charge offs through some of this, the start of the turbulence that's happening now. And some of that has happened in the small business group, but it's only about a two and a half percent 60 day delinquency. It's not converting to charge off yet. So the tools are working as far as we're concerned, and we're able to do a lot more on the upside because our team is out there on the sales side because our team is out there talking with the branch partners, talking with the business owners and originating new deals. So they don't have to think, they have to split their day and put a little bit of who needs to coach a salesperson as to do their office time, get their reviews done, call the accountant, why don't we have the statements, send the letters and all this stuff. If you can eliminate that where it makes sense, it just helps.

Bryan Peckinpaugh (26:31):

Yeah. And thanks for sharing your story with us, Keith. I mean, it's been a great success over the last 13 years of partnership that we've had with ESL, some fantastic results and positioned well in the economy that we're in to continue chasing those smaller deals through automation and reducing that effort on the backend. Again, very successful. I think the results speak for themselves, and we will leave it for a couple of questions if anybody has any for Keith, and we have a booth over in the solution center if anybody's interested in learning more about some of these solutions and how they could help your financial institution.

Audience Member 1 (27:20):

Thank you. Could you guys just say a little bit more about the gray area? So where somebody's been declined and then you look at fresh data and then you get to a Yes. Can you give an example of where that might happen or kind of data that might help make that switch around?

Keith Cleary (27:36):

Sure. So we have a couple of rules inside of our decision that we do take a look at all the time. It's a great question, penny. And we do, a lot of it has to do with length of time in business or a certain industry. And when we pair that with maybe an operator that has a seven 80 credit score and they've been in a related business, we find, and especially if it's in an industry that we, we find that as soon as we touch that, we're approving it almost right away because of the individual's credit score. So we then have to tweak maybe the order of the rules that we do to get that out of gray. The other thing we do is in the absence of a counter offer, length of time in business, as soon as an analyst picks that up, if we like everything about it, but it's only been in business for two years, then we'll take a look at it on an SBA track to say it's not quite mature yet, so let's pull that off of auto decisioning and approve it there. We'd love to find a way to have that built all right in, but you start to put way too many dimensions on your list of rules and try to, if I have an 800 credit score, I want that to go first and a business score to go, that's not easy to do, but you got to look for those opportunities. So hope that helps.

Bryan Peckinpaugh (28:53):

And from the vendor perspective, this is a big area of caution that our consultants always work with organizations on because you can almost get too greedy in the gray area of looking at too many of the nos and trying to turn them into a yes. The more you do that, the more you diminish the efficiencies that you're driving with the solution. So as Keith mentioned, it is really critical for each institution to look at what are those handful of factors that we're willing to entertain and making sure that some of them are hard and fast so that you don't erode that and move down towards not saying no to what you should. We don't nearly get as many yeses into into nose another one in the back there.

Audience member 2 (29:50):

Hey, can you speak to how much of your ADM is based on what we might consider traditional parameters, five Cs of credit versus some of the new Shaka zumo. We're going to go out there and look at how many web hits they've gotten and factor that into our decision making model.

Keith Cleary (30:12):

So I think you're asking about, we still use a fairly traditional credit scoring model. We use an IP score. So to the extent that we've looked into that, and I know they're using it in other parts of the bank, renters versus owners, those kinds of things, I think we're all looking for those alternatives that maybe fintech's leveraged a little bit more. I know at ESL we maybe, I don't think, it's not that we've rejected them or haven't bought it in them yet. I don't think we've spent enough time kind of delving into that. I'd love to hear if anybody else has.

Bryan Peckinpaugh (30:48):

And we're certainly seeing more and more of that and requests for alternative data and other sources to help with the decisions. But even more so I think than ever before, return to the five Cs of credit and really focusing on the basics is what most of our customers are asking for. I really do. I think that's all we have time for.

Keith Cleary (31:09):

I'm a firm believer in what NIM said is you have to set your customer up for success. And a yes is an enabler. Sometimes that helps them get the capital they need to grow the business and leverage their opportunity. But somebody that's not quite ready for that yet, and if you've stretched almost done, you've taken a half a step back with that customer and the best answer to give is not yet. And then have something built into your community to say, let's go see these people. And then when you're kind of done with that, and that's what frees up having, taking process away from both the customer and the relationship manager lets them think a little bit outside the box. So a no doesn't mean you're gone forever. A no could sometimes mean because again, we haven't instituted all the different attributes that fintech's might have. Sometimes it's just a conversation is the new information you need.

Bryan Peckinpaugh (31:58):

Appreciate everybody's time this morning. Thank you for your attention.