How Bank Lending Technology Is Changing the Way Small Business Loans Are Scored

Mitch Woods: Welcome to today’s episode of Lending Made Easy. Today we’ve got Brian Peckinpaugh and David Catalano, and we’re going to discuss how to approach scored small business lending. 

There are several benefits to scoring with bank lending technology, but it can also be pretty daunting.  Throughout the first question, I'll ask when the right time to start looking into scored small business lending is.

Bryan Peckinpaugh: As with many things, the right time to look into it was six months ago. This one in particular because a lot goes into a good, and I’m going to lump it into a bigger bucket, maybe, Mitch.

I’m into auto decision lending or factor-based lending because to be good at that, I need to understand what good credit looks like in my customer base. Because I can be two very similar-sized institutions a block away from each other, and what good looks like in the market that I happen to serve might be wildly different than the next bank’s market, all based on what focuses I have, the types of businesses I serve, et cetera.

And so a great first place to look is in the decisions you have been making. Looking back at the last six months, 12 months—maybe more—to say:

• What were the easy decisions?
• Where was the simplicity in what I did that I can replicate with a system?

That requires you to have a loan origination system that tracks that data so that I can go back and look at the decision factors and criteria that went into making that loan decision.

And then, I’ll use a phrase that many clients and prospects get frustrated with. But this is relevant to this conversation, and you must take a crawl, walk, run approach with bank lending technology for small business loan scoring. 

Now, that tends to drive people a little crazy because they associate crawling with slow. And that doesn’t have to be, right? It just means starting small. Don’t try to eat the elephant in one bite. Figure out the best strategic way to eat it in multiple bites. And it’s looking back on what are those simple factors.

The example I always use is, why would I want to go through a complete underwriting process to give a $25,000 loan to a business owned by somebody who has millions of dollars in the bank with me? That should be a no-brainer “yes” every day of the week, just from a customer service retention perspective and making sure I’m doing what’s proper and necessary to retain my best clients.

Then you can build on that over time, and you may build quickly, right? You may find many things in your particular decision processes that are easy and fast to replicate using bank lending technology. 

Then you can eventually work your way up to, as you were saying, Mitch, the score-based concepts and determine the true predictive nature of those to my client base. The scores aren’t the same for everybody, even though it’s the same result, right?

How I view a score of 400 could be wildly different than how David thinks about a score of 400 based on how we’re looking to apply it. Again, long story short, you should have started six months ago if you want to be doing it now. But the next best option is starting now. It is beginning to collect that data and build your analysis for how you will make those decisions soon.

David Catalano: Yeah. The other exciting approach you could take is to take your existing portfolio and score it.

Bryan Peckinpaugh: Yeah.

David Catalano: Then score it again in six months or quarterly if you want. But give it some time because it is a lagging indicator. Score it again and then compare those scores. And look for deterioration or improvement and then dig in and see what’s happening there.

The other thing it’s culturally important to understand is that you have to rely on data.

If you’re going to make a decision based on a score using bank lending technology, you have to trust the score.  you’re going to need to get everyone comfortable with that. That could take time because people like to use their critical thinking skills on the credit side and analyze coverage and collateral and industry and all the good stuff that we’ve come accustomed to looking at and putting and assigning good or bad too, and all that goes out the door.

Sometimes you get a score and wonder, why is the score so high relative to this score? This package has all this stuff going for it, but this one doesn’t yet. The score is higher, right? Yeah. Okay. But there’s a lot in scoring, right? The guys that create these scores are, you know, they got a lot of data behind them. And they’ve got a lot of experience behind them. Scoring it and getting used to those numbers and scoring your portfolio.

And then we have a—this isn’t a shameless plug—it’s just a plug. We have a product called portfolio monitoring, and we actually grab the scores, pull them in, and just compare them for you and tell you which ones to talk to based on what you’re trying to determine.

Is a score higher? Talk to them. Is a score lower? You should talk to them. But it’s just a way to analyze the data. You must get comfortable with the data. Also, if you’re in a niche, you want to think about how you can apply auto-decisioning to your niche.

I spent 11 years in niche lending. Dental was one of our niches. Absolutely a wonderful place to score.

I could immediately look at a dental deal and tell you if there’s any good. But with a score from bank lending technology, you can make that easy, and you can get your decision dollar amount pretty high with a score in a particular niche. Especially if you’ve had to disposition some bad credits in the past, you understand the industry, how it works, where the risks are, and how to dispose of the asset.

But, most of the time, you’re running into situations where the loans are performing well, especially in the last decade.

What a great opportunity to score! If you’ve got a niche. Start looking at scoring in those niches. All niches that seem to be similar are not.

That’s what I found in the past—because we did niches nationwide—is that they just behave differently. Understand your niche, and then I would recommend you start scoring them. After that, you can do your regular underwriting, right? You can just score and compare your scores to the underwriting and then look at the results.

Building up scored loans over time and comparing how they’re performing versus how your decisions went, compare and contrast what a high score would be versus your decision. Did you have bad loans on your manually decisioned judgmental type decisioning process?

Granted, we haven’t had a ton of bad loans in a while, right? We’ve had a robust economy for the most part.

We’re about to come upon a portion of the economy that will not be very good.  we’re probably going to have more there’ll be more stuff to look at, more testing of your judgmental process versus the scored process with bank lending technology.

If you don’t have a scored process, you won’t be comparing that. But a lot of opportunity for scored auto decisioning in lending, and it doesn’t have to be just the extremes, right? It could be your bread and butter lending under a certain dollar amount, but there are firms like FICO that claim their product has efficacy up to a million dollars.

There’s plenty of opportunity within commercial lending in the community space to provide auto decisioning.

Bryan Peckinpaugh: You hit on something in there, David, that’s, I think, critically important. And that’s before you start looking at doing this. You must make sure you’re culturally prepared to implement bank lending technology.

Don't do this if you’re not willing to deal with that stratification. It’s on both sides of the spectrum.

I’ve seen it bite plenty of organizations where you’ve got credit risk people looking at the approves and trying to back those out because of factors they don’t like.

Then you’ve got RM’s and others trying to do the deal that sees a decline and want to approve it for extenuating circumstances and factors.

If you set those boundaries and those thresholds, you’ve got to learn to live with them. And you must be culturally prepared to accept what happens inside of your approval and denial bans.

David Catalano: We need a great area strategy.

Bryan Peckinpaugh: Exactly.

David Catalano: What is that? Tell us. 

Bryan Peckinpaugh: Yeah, so your gray area strategy, if you think about those auto decisions, is, how do you think about the middle? What do you think about the things that aren’t an easy “yes” or an easy “no”? 

That’s where you need to play with those bands, right? Some organizations will start with only auto declines, right? My gray area is any yeses. Some organizations will start with only auto approvals. That doesn’t happen very often. That’s a lot harder one. It cuts way more than most people are comfortable with when using bank lending technology.

But most organizations will say I have a top end to which I’m always going to say “yes” and a bottom end to which I will always say “no.” Those that do it best are the ones that can tighten the band as much as possible. Because that’s where you’re maximizing your high-value resources. The more I can put into the hands of sensible auto decisions.

Going back to David to all the things you were talking about, I’m looking at all the things that will make me comfortable with those decisions. Not chasing too much business away because I’m saying no, but also getting the ones I should.

The more narrow that band is, the more time my critical thinkers will have to think about those deals and come up with ways to say yes to the transactions I want effectively.

It is a critical piece of this, really understanding your comfortability with the gray area strategy because some people just aren’t. They want everything to be gray. It doesn’t work in this in this space.

If you’re going to do this, you’ve got to have some black, and you’ve got to have some white because if everything’s gray, you might as well not do the process using bank lending technology.

David Catalano: Yeah. And—if you’re serving a niche—speed will be important, so you want to narrow that gray band. My niches are more forgiving than others, right?  it’s just less volatile industries tend to do pretty well even when the economy gets tough. They just continue to produce on the business side of things.

That gets back to that dental experience I’m talking about. You need to really understand how you will apply the tool. And again, if you’re in a niche and speed’s important, narrowing your gray area will be really helpful. And then potentially explaining to, before let, let, let the reputation of your credit get out there into that niche there.

People know when to come to you, right? Because you can take a lot of costs out of your system if you’re not underwriting that loan in around a 74- or 75-day turnaround time for a commercial loan and knock that down to 35 days. You don’t want to spend 75 days doing something you can auto-decision.

Because it’s a strong niche, it’s strong. It’s a strong borrower. You can do a score and be done with it. You’re going to do a lot more of those. You’re going to track those. You’re going to magnetize them.  I would encourage you to consider this evaluation quickly.

Just get it started. Just start collecting the data. It doesn’t hurt you to do that. You got to buy some bank lending technology to run that and collect it, but so what? It’s nominal cost relative to what you can gain on the business side, both internally on the efficiency side and then externally to your customer base.

Bryan Peckinpaugh: Yeah. David, you and I were having an off-the-mic conversation about questions like:

• Does volume matter?
• How do you need to think about this?

 That’s where it gets really interesting. Speed matters if I can use tools and technology to say “yes” to something. If I don’t have those tools, I will miss deals. I will miss out on some of those because I can’t make the judgemental decision fast enough.

Putting the tools in place, may only mean one or two deals. That’s enough to pay for a lot of bank lending technology. If I’m getting one or two deals a month or one or two deals a quarter, without a tool, I wouldn’t get because I’m not fast enough, I’m unable to compete with others that have the technology in the marketplace.

Those margins pay for a lot. If I can compound that, month after month, year after year, I don’t need a ton of volume to make this make sense.

If I’m capturing new markets, expanding into new and different businesses, and capturing larger deals going after them with speed and efficiency. I can pay for this in many ways that don’t require me to do hundreds of loans that fit this category every month.

David Catalano: Yeah. I would argue you can also pick up yield.

Bryan Peckinpaugh: Absolutely.

David Catalano: Speed. People will trade a bit of yield for speed—and convenience.

Mitch Woods: You all are bringing up some excellent points. Speed. Speed matters. 

Something else to consider with scored lending and auto decisioning using bank lending technology—since it’s so data-driven—is also the consistency factor, right? Anything that’s data-driven will provide more consistency, letting you focus.

Bryan, to your point, on that gray area strategy where it’s not the easy “yes” or easy “no.” With that in mind, how would you determine success with a scored lending and auto decision approach if you're a bank or a credit union?

Looking at that holistically from the idea of speed and how much it matters, but also with the idea of possibly increasing the consistency in your loan decisions as well?

Bryan Peckinpaugh: the most critical thing is you got to come at it from a multi-factor point of view. It’s easy to get caught up in looking at top-line revenue or simple factors. But you need to come at it from all of the elements.

You need to look at what it is doing to my credit quality, is my credit quality improving by leveraging the data?

Mitch, as you were talking about, you need to check if you are bringing more factors in that are non-judgemental. Am I taking the human emotional factor out of it? Do I see a change for positive in my credit quality? Or did I expect a decline in credit quality?

Some folks implement these processes and adopt bank lending technology to gain on the top line but are willing to sacrifice a bit on the credit quality because the additional business I bring in makes up for the loss on the back end.

You got to sit down and figure out why you’re getting into this. Ask yourself:

• What am I looking to accomplish?
• What are my key goals and objectives in implementing a strategy like this? 
• What is it that I want to measure?

You should be looking at the number of deals in aggregate. How is that trending? I should look at yield or margin on the deals I do. Is that improving based on what I’m doing?

It’s going to kind of sound funny—maybe at first—but I should also look at the results of the manual decisions that I’m making because, again, part of this is freeing up the people, so taking them away from the smaller—”easier”—deals and putting more time on the more complex deals. I should see improvement there as well. I shouldn’t see stagnation in that.

Make sure you factor into your analysis that part of putting a bank lending technology like this in place is getting better at the judgmental because I’ve got more time and capacity to do so.

You need to sit down and look at it in totality and break down all of the factors that go into how this moves the needle for your organization, and then you got a lot of awesome things to measure.

David Catalano: Yeah, the only thing I would add to that would be to look at what business problem you’re trying to solve by adopting bank lending technology. And will it solve it?

Here’s an example.

We were working with a bank. They were implementing a small business loan origination department. They hired a person from another bank who knows small business lending. Not SBA lending just scored lending. This particular bank was referring out every loan under a million dollars. Can you believe that?

That’s pretty easy.

How many loans under a million dollars are they now doing through this department? Right? You’ve got the business case. You’re trying to plug this hole in your process, and you hire people and put a strategy together to do that, and you just measure how you’re doing against that.

But that’s a very defined business problem they were trying to solve. And that’s how they should measure their success. There are many ways to do it, but again, it returns to what I am trying to solve and did I solve it. And everything that Bryan’s talking about’s dead on.

You should track those items because they are ancillary to the ideal solution that you put this in place for. It’s typically some type of bank lending technology, and then there are the additional benefits of the software. Maybe those other benefits allow you to pay for it over and over and grow in new and different areas.

The other thing I would do is look through your business and look for common industries. What industry types do you have, and do you have many of the same ones? Does that industry type lend itself to a scored approach using bank lending technology?

Bryan Peckinpaugh: Yes.

David Catalano: And see if you can apply it to that particular, you may have a, who knows? You may have the equipment from a commercial equipment supplier. You may be doing deals with them throughout a branch network. You didn’t realize you had a hundred on the books, and they all perform really well.

Could you apply small business scoring auto decisioning to that type of business and just transform that and triple that?

Bryan Peckinpaugh: Right. And I would also just continue to, maybe caution’s not the right word, but I’ll use it then make sure you’re always looking at those ancillary benefits.

Because you’re, again, spot on, David, and not forgetting that if I turn technology towards those specific industries, that does inherently free up the rest of my staff for other things.

And so, just judging, you’re right. You need to build that business plan. You need to identify what you’re trying to solve and the answer did you solve it? But then you got to look at what else solving that problem opens me up to do. Right?

Because if I solve that problem, and I can then redirect my resources to, say, now that you’re not bogged down on these smaller deals, now that we’ve given the organization the technology to do things in an automated manner with bank lending technology:

• Can you use that increased capacity in your day to get another big commercial loan? 
• Can you get me a CRE deal?
• Can you get me things that move the needle further because I’ve taken the low-hanging fruit off your plate?

David Catalano: Exactly. You’ve capacity in the system.

Mitch Woods: Yeah, some great ideas for any financial institution wading into this automated scored business lending approach using bank lending technology. 

That’s something we talk about with clients day in and day out. If it’s something your financial institution is looking at, please reach out to us. 

We’d love to have that conversation. Hopefully, you can take away some great concepts from this episode.

Brian and David—great topics today. Great conversation. And thanks to everyone out there for listening to today’s episode of Lending Made Easy.