Fair and Accessible Lending Practices

"Fair lending" has become an important topic—and concern—for financial institutions. The concern isn't just limited to consumer lending, but to small business lending as well. But what does the term "fair lending" mean and how do financial institutions achieve it? A generally accepted definition states that fair lending "guarantees the same lending opportunities to everyone" by "prohibiting lenders from considering race, color, national origin, religion, sex, familial status, or disability when applying for loans."

To help financial institutions understand the importance of the issue in order to meet the criteria of fair, and accessible, lending for small business owners, this session will address how banks and credit unions can use new technologies and new data sources to profitably expand the scope of their lending effort to traditionally underserved small business populations.

Transcription:

Okay, thank you. Thank you so much for joining. I hope you're having a lot of fun. So welcome to this session on a very interesting and challenging topic of fair and accessible lending practices. A topic that has been driving a lot of discussions given the recent regulatory announcements from 1071 to the changes that are happening on the CRA side. My name is Abid Hasan. I'm the head of product at Uplinq accredit decision support technology company, and this topic is very close to our hearts, given complete alignment to our vision and mission of enabling access to fair and ethical financing to small and medium businesses by helping lenders say yes more often, leveraging thousands of data sources in US and globally. To talk about this topic, I have this esteemed panel here today. So without further ado, I'll introduce the panelist. Derek Ellington, who is a Senior Executive for small and medium businesses at numerous large and regional banks most recently at Wells Fargo. Thank you Derek for joining. Bill Phelan, who is the General Manager for Equifax Commercial. Manny Tocco, Head of Business Banking and Credit Operations at Citizens. And last but not least, Patrick Reily, who's the Co-founder of Uplinq Financial Technologies and who's also going to moderate this session. So I'll pass it on to Pat, get this going. Thank you both.

Thank you very much, Abby, and hello to all of you. So how do we keep small business lending fair and how do we make it infinitely more accessible? Clearly this is not a 30 minute topic, but we're going to get right into it and move as quickly as we can. Fair is a concept, but fair is also a regulation. And today what we're going to try and do is keep it within that ECOA protected class kind of definition. Just to avoid scope creep, I think we will bring in Dodd-Frank 1071 and talk about it. We're going to be talking about the CRA changes, but I want to make it really clear. This is a practical discussion. It is not a dissertation on bank regulation. Accessibility is important every bit as important as fairness. There's not really the strength of definition there, but honestly, most lenders don't struggle with small business fairness because most lenders don't lend to small business, and that is fundamentally a giant problem.

I want you to consider this stat for a moment. By the SBAs definition, 44% of the US economy is driven by small business, while 3% of bank loan books by definition for the FFIC and FDIC are made up of small business loans, 44% and 3%. It's an enormous gap. It also represents an enormous opportunity to lenders to be able to take on that challenge and meet the need. The difficulty is though most banks struggle to find both efficiency and accuracy and underwriting capability, and therefore the bulk of that unmet need really falls to either a small portion going to non-depository lenders oftentimes charging 30 to maybe 160% interest or simply not getting met at all. But today we're really not going to talk about the issue. What we're going to talk about are solutions around the issue. I want to really start the conversation with Bill. I think, bill, you and I, this has almost become a life work for us and based on your experience, what can give the biggest impact to improving capital access to SMBs.

Sure, sure. Pat, thanks and thank you everybody. We can't really see anybody because the lights are very bright here.

My name is Patrick Riley. This is my social security number.

But I appreciate being on this panel with Manny and with Derek with you, Pat, and thanks Ron for having me out for this session. As you said, Pat, I mean I've been in this business for 25 years and I founded my company PayNet way back when to really help automate and get capital out to small, medium sized businesses. That was back in 2000. We've seen a lot of progress since then. And I think what I've learned over that time is, as you said, the biggest bang for your buck here comes from automation. And it's a little bit of a cliche, but we're going to get into some of the details there. But if you just look at the facts around underwriting a small business loan that I've learned and we've studied and done time series analysis and everything else, A small business loan doesn't really ROI for a bank under $700,000 roughly. That's roughly depends on allocations of dollars and whatnot, expenses, but the average small business loan is $250,000. So you've got this giant mismatch between what ROIs, what's profitable from a lending standpoint and what a business wants, and that that's kind of the crux of the problem. And so what we found is that actually the lowering the cost of transactions and getting capital out to these small businesses through automation kind of closes that gap.

It's a big deal. It's interesting, the World Bank has an interesting term for it. They refer to small business lending as the missing middle. You've got this very high level of efficiency and automation down at the consumer end and at the commercial end, you can afford to put people around and get bespoke solutions and do it. But I think you're exactly right. I mean, the vast majority of people on our platform today, decisions are made in about two seconds and 99.6% of them are all handed in automated fashion. When you do this, you drive out the cost and you drive up access in availability. I want to ask Derek, I mean Derek and I both worked for some really large banks, Derek, certainly more recently than I have, and supporting getting small businesses served up to scale and really doing this in volume. And so I'd like to ask, what are the challenges that you face in delivering fair and accessible lending and what kind of solutions do you see big banks looking at to tackle these problems?

Yeah, thank you. And just for perspective, at the largest institutions, there are a few challenges that because of the need to deliver at scale large institutions, if they are not focused on simple set of products, repeatability and the reliability of systems to deliver those capabilities, when you start to try to customize to small groups of customer needs, that's when big institutions get into trouble. So everything is focused on what can we do at scale, what can we repeat the experience over and over again, ultimately do it at a profit. And also think about it, large institutions when you talk about a small business customer is every customer from idea and opening a DDA to a customer that may be 10 million in annual revenues. So that five to 10 million annual revenue customers needs are very different than that micro businesses needs. And there are a lot of micro businesses, but delivering a fair and equitable experience across that spectrum at scale.

And also when you think about the cycle we're in right now, large institutions don't want to have high delinquencies and high charge-offs. So we've also got to execute while managing the risk appropriately. So in many cases, there's a lot of data we have on our customers, but how do we harness that and how do we enhance that data that we have internally with other capabilities externally to really round out managing delinquencies and charge offs? And today, a lot of the large institutions haven't completed that circle yet. They're using their data to the best of their ability, but they haven't maximized, maximize external solutions that can enhance what they have access to today.

Yeah, I think it's a really interesting observation that one of the unique things about small business space is it's not homogeneous. If I just get you to think for a moment, what did Covid mean for a plumber? What did Covid mean for a restaurant owner? It's an entirely different experience, right? Then we can talk about size and we can talk about a whole bunch of different issues around this. And so sometimes when people think about automation, they think about sort of a one size fits all template piece, but that's not where automation is today anymore. It's this capacity to deliver bespoke in a very efficient way into a space according to the need of the customer. And so I think that's probably one of the most exciting evolutions we see right now. But Manny, you're actually doing it right, and Manny is over this category for one of the largest super regionals around. And so talk to me a little bit about what technologies are banks utilizing to make this happen and how are they reaching specifically underserved communities?

Yeah, I think Derek nailed it, right? It's around permission-based software solutioning, and there's a lot of good companies out there that do it exceptionally well, but it's the transactional analytics being able to harvest that data, which is incredibly rich. And I think you can deliver something that's bespoke when you're able to get at that sort of data. You augment that with foot traffic analysis, card analysis, website analysis. How many times does a customer or a potential customer go into a business's website? When you start to double down on that holistically and you take a very traditional approach, all the data that you're getting from SBFE, from the credit bureaus, from your personal FICO, and you augment that with very rich alternative data resources that I just described, you're really able to end up reaching l and m communities in a much more profound way. And we'll talk a little bit about 1071 in a minute, I think, right? That's on everybody's minds.

Yeah, hold your breath 1071, I know all of you're just saying 1071, 1071.

But in all honesty, I mean whether it's the micro mom and pop dog walker or it's the 10 million operator that Derek had mentioned, oftentimes in business we're working with imperfect information. And so this kind of levels that playing field in a much more profound way. And so we're getting after that data and starting to look at how are we partnering with some fintech's to deliver that on a very simplistic chassis that people can interface with and deliver something that's distinctive.

I think one of the things you'll hear here is this idea that if you think back to, I used to work for Equifax, and so I'm not sharing any trade secrets, but A FICO model has a few dozen variables in it. And the challenge in small business space is sometimes I have a piece of data and sometimes, sometimes a piece of data is applicable to a particular type of business, sometimes it isn't. And so the type of modeling that we do to actually address this problem is somewhat more complex than that. So we shouldn't be surprised that traditional scores sometimes don't work very well. But this is absolutely an addressable issue, and what I can tell you is very interesting is product fitment, we get product right, and we can oftentimes reduce risk exposure by a factor of three. That means going from a 3% charge off rate to a one or a one and a half down to 50 bips, doing that means no turns. Yes. That's really powerful. It's really important. It sort of underscores getting this. And if I go to Bill, I think you've actually written on these subjects of automation of the underwriting process. Can you kind of give some tips or explanations for how banks can start to do this?

Yeah, I think it plays off what Manny and Derek were saying, Pat, it's really the recipe for automation. It's data, as you mentioned, Manny, it's software as you mentioned, and it's analytics. Obviously analytics have to be a big part of that because the data by itself just doesn't give you the answer, as you said, Manny. And so that's the three pieces of the recipe for automation. And so let's just walk through those real quick. I mean, obviously there's great software out there, Pat, you're very familiar with. And your ability, you might know one to help the decision process through great systems like Uplinq, but also the ability to safely analyze and score and come to a result you want. I think as Manny said, you don't want to put these things on autopilot. You want to have these things actually, these scores, these analytics, you want them to be obviously tested upfront, rigorously tested through a model risk management process, and then actually validate it on a continual basis.

You don't put these things on autopilot, but the technology for scoring in small business has come leaps and bounds since I started 20 years ago. In this business, we're seeing IRC's, which is a measure of statistical power in the eighties, which is very good. We're seeing stability measures consistent over time. And then obviously as you mentioned, the data, I mean, one good metric in data is the database that I'm managing at Equifax, which is about 7 trillion of small business loans. It's a lot, and you can harvest a lot of information. Again to what Manny said about, and Derek said about targeting how you lend to a farmer is different than how you lend to a law firm or a manufacturing company or a trucking company. And so you can start to, as you said, get bespoke, but with scale by looking at that data, that 7 trillion and being able to understand what's going on. So that's the recipe that I've seen, Pat, that really works.

Yeah, I think it makes it, when I'm not looking into the lights, I can see will whiteside's over here, will whiteside heads up our modeling practice? And what will tell you is before he ever touches the data, the work that goes into getting data ready is when we step into an institution, the most likely thing we're going to see is how do we create fitness in the data infrastructure? Because most enterprises are collecting information kind of in an old school small data format, and lending into small businesses really means embracing a lot more information. The great news is that information is available. It is thinking about how I put it to use and how we take advantage of it and put it to work is a process and takes thought. So I would always say step one, think about your data environment. How do we get higher transparency in data environment fluidity, and how do we enable analytics in a safe capacity as well, right? You've got a lot of information you need to make sure it's well protected. As we're starting to talk about things that sort of buffer against the edge of regulation, I want to pivot a little bit and maybe talk about how this all dovetails into regulation. And maybe a question for you, Manny, is how do banks prepare for 1071 and really what are the long-term implications embedded in all that?

Yeah, I think we're still a little bit in the journey just given that it's still a little bit of a longer tail. Yep.

Clear gun hasn't fired yet.

Yeah, right. We're getting closer, but I mean we're going through our governance compliance. What does that need to look like? Design work internally with our own lending platform in concert with the FinTech that we use for scoring models up to 250,000 in exposure. And then I would say in concert with that, trying to ascertain the firewall provisioning that needs to end up taking place, the cover transactions, and then how do you end up reporting on all that as the examiners are going to start to come in and want to have visibility into this in terms of how it changes things. I think it really is going to galvanize the organization. Patrick and I were catching up a little bit before this event here, and in some respects, and I think as an industry, especially if you're in credit, it almost feels a little bit overwhelming that you got to implement all this.

It's just additional regulation. It's another thing put on top of the heap, but I do think it's going to end up making us, we're at a very important inflection point where it's going to make banks start to evolve and leverage alternative data and scoring models to really reach those low and moderate income census tracks. So I think between that, what it's going to do for CRA and really just put more deals top of funnel and then have just better hygiene around the portfolio. I mean, we were also catching up earlier and I said, oh, this is what a potential recession looks like. Delinquencies are up and NPLS are up, and we haven't seen that in some time. So I think all of these, when you start to couple it together, is really going to be meaningful.

I think it's interesting. Hopefully what you're hearing there is part of, so when I still wore my banker hat with great big giant bank, I would say I had two good days. Day one was when I nailed my financials, and day two was when the regulator left my office happy and the sooner they left my office happy was good. And so if you think about it, one can think about how do I automate regulatory compliance? Well, I'll give you one quick example that relates directly to this. So years ago, all the way back in 2010, I wanted to make sure that if a regulator ever walked in on a consumer exam to understand fairness, I was going to have no surprises. We just baked in BISG testing into every single loan origination, not post hoc after we had done the loan, but real time. Well, what's BISG, BISG is the statistical test that'll run if a regulator walks into your office and says, are you treating people fairly because of that, this ability to then take that capability and immediately apply it into small business and create that sort of hum to like 1071 view becomes intrinsic to what you do.

That's smart design. Because now the ability to comply from regulation, the cost of that goes effectively to zero. The better thing is you're treating people well. You have ability to understand what do we need to work on? How do we do it? How do we make it happen? So I agree with you a hundred percent. This is an opportunity really to step up and figure out how do we serve better?

I'm going to point the next to Derek, and what are some of the concepts as we look back to that big bank land that regulators are kind of positioning and thinking about? What are we going to do? How are we going to do, I'm thinking of things like CRAI, frankly, I don't know about any of you, but when has anybody read the 14 or 1500 page CRA reg? Maybe there's some writers. There you go. Okay, give that guy a prize. Okay, I'm only on page 382 or something. But yeah, there you go. And thank you for it. And everybody else pity that, man. No, practically speaking, this is really interesting. It's sort of saying there was this thing that happened called the internet about 30 years ago, and now we're going to redefine what footprints are based on not simply where branches are located, but that now it's a lot more than that, right? And

Well, it's a lot more than that because now it's non branch areas?

That's right. Non branch areas, right? This whole concept of I can market even where I don't have people. And so I think this is really important, right? It's fair. That's a really good part of the fair definition. But Derek, why don't you talk a little bit about what you see happening in the regulatory space and really what all this means.

Yeah, one of the things, I mean, foundationally, when we take our practical pill and we take a step back and we think back to the pandemic, it shined a spotlight on during the PPP process. The small businesses that were disproportionately impacted the most during PPP were our diverse small businesses. Those with any sort of diverse dimension had challenges accessing PPP, and that is common in all of our lending environments. Those customers tend to have more challenges accessing capital. So again, practically, when we take that practical pill, if you look at the CRA exams of most banks, the CRA exams are great, outstanding or some other very acceptable score at most banks. But then when you talk to actual customers, you look at independent surveys, they all say no. The customers that really need the support the most are having all kinds of challenges accessing and obtaining credit, and particularly credit fairly.

So the regulators have just finally said, we're putting a stake in the ground and we're going to reevaluate and we're going to look at CRA reform and really challenge the environment to get to a better outcome in terms of what's actually happening with those segments of customers that really need our support the most. And also it's created opportunities where they've gotten behind and pushed all financial institutions to become more creative in how we approach things. Look at special purpose credit programs as an example. Many institutions have launched those or in the of developing them to really target ways to serve more customers in these groups that really need the support. And if you look at the ideas that have been launched, they're all different. None are the same. Some deal in more commercial real estate, some are more based on lines of credit as the primary operating tool, but they all get at the heart of trying to create more accessibility to those that need the resources the most. And again, we all know practically, if we don't measure, nothing changes. So now they're going to hold us all accountable in the 1071 environment, and we're going to all share our results. What are our approval rates? How successful have we been at getting customers to apply as well? And then again, what are the outcomes? And if we measure that and publish it, things will improve because none of us want to look bad.

I'll give you two challenges. So in the Central Valley of California, we have a lending program in which the average FICO is four 80. The average default rate on that four 80 is about two and a half percent. So two and a half percent is a lot higher than a lot of lenders, but most lenders stop at 660, 620. Where do they end? And you can price, you can price to 250 basis points. There's not a lender in here who can't do that. In Nigeria, we run a program that processes probably 10,000 loans a week or something in that order of magnitude, they run at 80 basis points default, 80 basis points. The idea that this is unattainable is inexcusable. It's just a bad fundamental idea. And so almost, we've got a couple of minutes here. What I'd like to do is ask each of you, if you could give one bit of advice for anybody who's trying to approach this and say, how do we get better? Where do we start? And so maybe I'll start with you, Manny.

Yeah. I think it comes down to prioritization, right? Realizing that the need is acute, right? I think it's the right thing to do just to be innovative and stay out in front of the competition. But the last thing you want to do is be last to the finish line in the eyes of the regulators or not be able to drive meaningful outcomes as it relates to where you need to be from a CRA perspective. There's a million things that are happening with financial institutions these days, and I think it's just where does this fit in in terms of the pecking order around strategy? And I know I could say from where I sit, I know we're in a very tight environment and capital scarce. Where do you really want to end up leaning in to influence that sort of change? And again, it's around the prioritization.

Go ahead.

Well, I think we are getting better. Number one, I've been tracking this for a long time. I've seen, as you were saying, Derek, the credit approval rates have been consistently increasing year after year. Now it's not increasing at a rapid rate, but it is going up.

Remember Equifax, so he has a pretty good view.

No, but we're seeing a consistent over a long period of time, rates have increased on average. So that shows progress right there, number one. But I think to your point, it's a good one. How do you even take it one step further? We've got more to do. We talked a little bit about the data, the workflow, the analytics, just tactically looking at that, just make it easier, make the points of entry easier. I think you were talking about that a little bit, Derek, where make it easier for the small business. A small business is the chief cook and bottle washer, and they don't do their finances until 10 o'clock at night sometimes. And so making it simple for them to apply is a key simple thing that you can do. Just think about user experience, think about pre-filling data so they don't have to add, put a lot of data in. And then finally, I think we see a lot of the scoring, the model. Risk management is a key step because you got to get that right if you're going to do this well and do this safely. And by safely, I mean safe and sound. Absolutely. That's what I'm talking about. So just I'd say spend some time on that model risk management.

Awesome, Derek.

Yeah, and I'll say quickly, a couple of things. One, just the realization that this is not going away in the CRA reform, the focus is going to be more and more on small businesses because mortgage volumes are way down, right? Rates are way up, mortgage volumes are way down. So you're going to see over time the small business lending component be more of the calculation statistically as things evolve. So there'll need to be that focus there. And then second, as we talked about earlier, a lot of small business owners still today, you talk to anyone randomly don't understand the landscape of how to get credit ready and what are the right solutions that they should be applying for. So as an industry, we still have a lot on our shoulders to create opportunities for them to be educated more around the right solutions that they should be applying for. And then what are some of the criteria? What does a credit score look like that gives you a better chance of being approved, and what's the Path to getting to that credit score as an example? And what are some of the other criteria in institution that you're looking for in terms of maybe minimum time in business, et cetera, that if they know that before they apply, they can get ready to apply and have a greater success level.

Excellent. Excellent. Well, I'll leave you with this. If you're thinking about technology partners to make this kind of thing happen, every once in a while, I get to stand on a World Bank stage, and I always start off by saying, don't believe any FinTech. They're all liars. And I think the level of hyperbole is, I'm looking at Penny right now, who has to deal with the liars all the time is pretty remarkable. Make them prove it right? When we're talking about coming up with better solutions in this space, there is no problem stepping up to prove it and demonstrate that for my loan book, for the types of customers I have, I can do this. We almost compel each of our customers to do A POC before we do business with them, because we want to demonstrate the quality of the work that we do and make sure they understand that it is going to satisfy compliance, it's going to satisfy credit quality standards and everything else that needs to happen.

The last thing I'll say is that just a few weeks ahead now, the middle of January, cornerstone is going to release a white paper. Some of the people on the stage here have actually participated in that. World Bank, participated in it, some of the best international small business lenders have participated in. Its really exciting thing covering fairness and accessibility for small businesses in lending. It'll be a really interesting piece. I encourage you all to just ping away at it. So Cornerstone will be releasing that and I encourage you to take a look. And so any rate, thank you all today and it's been great.

Great job.