With risk mitigation on the minds of commercial bankers everywhere, risk ratings are more important than ever. But at some banks, those ratings may not be helping much - in fact they may be hindering the ability to price appropriately for risk. In this episode of The Purposeful Banker, Dallas Wells and Jim Young look at ways those ratings may come up short - and steps banks can take to improve them.
Commercial Risk Levels & Bank Behavior During COVID-19 (Webinar)
COVID-19 Market Updates & Resources (PrecisionLender)
Questions? Comments? Email Jim Young at email@example.com
Hi and welcome to The Purposeful Banker. The podcast brought to you by PrecisionLender, where we discuss the big topics on the minds of today's best bankers. I'm your host, Jim Young, director of content at PrecisionLender. And I'm joined again today by Dallas Wells, our EVP of strategy. And today's topic is risk ratings. It should be a pretty hotline given the current circumstances. But the question here is really whether they're actually helping your bank to well, manage risk.
So Dallas, before we dive into this, I want you to indulge me here and let me broach this topic via a sports analogy. I know that that's a different sort of thing for us to do here is to talk to via sports analogies, but let me give this a try.
Yeah, that's shocking, Jim.
Yeah, exactly. All right. So back when I was a sports' writer, I had the joy and privilege - sarcasm font - here of covering college football recruiting.
It is a huge industry now and it's built upon scouting and reading thousands of high school football players, not an easy task. And caveat here for those of you who actually still really follow this closely, and I'm sure we do have some listeners that do, this may have changed a little bit, but back when I did it, they had the ratings basically on a star system from two to five. And the fact that they don't start at one tells you a little bit about the way it works, but five stars was the most elite level, probably about 100 or so kids get a five star rating. Then the next level is four star, maybe about 300 or so in the country get that one. On the bottom end, you had two stars, which was sort of like, "We've seen this guy once or twice. We've heard of him. We really don't know much. So we're just going to put them as a two star."
In the middle there's this is vast ocean of ... I don't even know the exact number, hundreds to probably a few thousand kids that are on the three-star system. So it's an enormous pool of players. And because it's so big, the talent in it varies wildly. And then you factor in regionality here, basically you got say 54 star kids out of the state of Florida, the 51st kid ... they don't want to keep throwing four stars to one state. So the kid that's just barely missed four star in Florida and is a three star is vastly better than a kid Who's a three-star from the state of Massachusetts. The point is this whole system, when you say a kid is a three-star on the ratings, it doesn't mean much because it's not really a well defined group and it's way too broad.
So to bring this whole torturous analogy around, Dallas, when I think about risk ratings at banks, college football, recruiting rankings, and specifically the three star rating is what I think about. So go ahead and tell me, shoot that analogy down or tell me if I'm on the right track.
No. I think this one works, Jim. I think you actually got one.
I think the biggest difference is that for banks, there's really not a universal rating system. There is no five star system, but there are some general buckets that more banks than not tend to use. And there are some divides roughly by bank size as well. And so a lot of smaller community banks, credit unions essentially use the regulatory eight or nine point scale, but whatever number, letter, rating, wording, and terminology you use almost all banks have that same sort of distribution. And they'll have somewhere in the past credit section is where the vast majority of their loans are actually set.
We work with a couple hundred banks here at PrecisionLender. And so we see how they grade all those loans. And some banks are pretty good at it. If they're using a nine point scale, there's a definite bell curve to it, normal distribution, where there's a spike in the middle around those past credits, but they do actually use some of the other numbers.
There are others where it's a nine-point scale and they literally don't have any ones or twos. Everything is a three, four, and maybe a couple of fives. So to your point, those number systems don't really tell you a whole lot. Beyond that you get into, there are some banks that use a dual risk grading system where there's one rating for the actual borrower of how likely are they to default, and then there's a second kind of facility grade of based on the structure of the deal itself. If they do default, what are your actual losses? And a lot of banks still just lump those together into one risk grade where they try to factor both of those together. And that gets really tricky.
So there are issues both ways that we can get into. But I think for starters, your analogy there is a pretty good reference.
Yeah. And I probably could have ... the writer in me is going to slap my wrist here because I probably could have just said basically, there's this one point in it in which it's like, if you're not really sure you just throw it in there and that's the three-star. And so because of that, you can have all sorts of stuff rather than really take the time to get specific and break it down and you go, "Yeah, probably a three star."
All right. Maybe I'm on a roll here now, so I'm going to press my luck with it, but I'll make this the last sports reference here. I understand why there is a lack of ... the regionality of college football recruiting is obvious and then it is by its very nature, a subjective system for rating, so I get that. I guess I'm just a little bit surprised why you would have sort of this lack of specificity, maybe not the subjectivity, but the lack of specificity with risk ratings, why aren't they more exact, I guess?
In some institutions they are. They are very formulaic and they're driven by some specific metrics and the number is what it is. It's whatever it comes out to be. And also, especially in larger institutions, there's a clear division between production and credit. And so credit is really responsible for rating those and they don't care a whole lot about what that means to the production side. That's their job is to kind of independently review and rate that. And those lines get really blurry in community banks, where a lot of times they report up to the same executive and it's just a little trickier to be truly independent for those banks that are not using a quantitative scale, where it's, is this a pass? Is this a strong pass? Is it a watch? There's a lot of opinion in there and it's pretty subjective.
Usually those scales will have some broad guidelines, but it's like debt service coverage ratio is kind of the only thing in there. And by the time you get done doing all the adjustments that you want to do for that particular borrower or industry, you can kind of cram that thing into whichever box you want to. They end up being pretty broad depending on how your system is designed.
I think that's what we're starting to see in the difference between some of these banks and in the report that Gita Thollesson is working on
that'll be coming out soon. She's actually identified a difference in those systems. And so the banks that are using more of a quantitative method that are actually measuring this with some real metrics to guide those probability of default ratings, they're actually identifying downgrades more often and sooner than those banks that aren't.
And basically you can't have an opinion about it. You can't say, "Well, I think they'll be okay. I know the numbers look a little ugly, but I think there's still a pass. If the numbers say they're not, then they're just not. So they are more frequently downgrading. And I think given the environment that we're in, probably just being more honest about where their portfolio really is and just letting the downgrade process do what it's supposed to do. So I think that's interesting that given how fast things have happened through this crisis, that there's one method that's allowing some banks to move a little quicker and to be a little more exact.
Tell me if I'm going too far here, but I'm just wondering if in some cases having risk ratings, the idea of which is to give you an idea of how risky this deal is and to help you ... I'm now explaining this to a crowd that pretty much understands what we're talking about here, but I wonder if you can have the opposite effect i.e you're increasing your risk on a deal by the rating itself. And I guess maybe even though we just talked about how it can be vague, it can be subjective, there's credibility attached to it. And so maybe it lulls you into a false sense of security. Well, the rating says this is a pass, so we're good here. And maybe it would almost be better in some of those situations, if you didn't have the rating at all.
Tell me if I'm going too far there.
No, I think there's some validity to that, especially when you think about how most banks operate, what happens is those problem credits, those ones that get downgraded, and especially those that get downgraded past a certain threshold, those get reported up the chain of command up through the executives of a bank and eventually to the board of directors, those are the ones that you have to parade out in front of the examiners when they come in to look things over. They're going to say what's your watch list and where the files on those they want dig into those every time. So if you have loans that should be on that list and aren't, I think there's a skewed perception of the real risk profile by the board, by the management team, even by independent auditors and perhaps from your regulators. There's some hidden potential ugliness there, and the fact that they don't see it, doesn't mean that it doesn't exist.
And I think that can cause some management teams maybe to make some decisions a little blind. And I think what you're seeing now as banks are doing some of their reporting on second quarter earnings is I think you're seeing some of that uncertainty where you see executive teams that aren't sure exactly what the risk really is. And the situation is different in that it's a pandemic. We don't have any way of kind of rolling forward what's coming at us, but every recession is like that to some extent.
But you're seeing these massive allocations to loan loss reserve. You're seeing banks willing to cut earnings more than in half from the same quarter last year and basically say, "We don't have a ton of downgrades. We don't have a ton of delinquency yet, but this feels bad. So here's a whole bunch of rainy day fund money to help us prepare for that."
So you can see the inexact nature of this. You can also see the fact that so much of this as a trailing indicator. By the time a loan gets downgraded, you may even be past the point of being able to do something. So I think that false sense of security is a real issue, both from strategy decisions at the bank, everything from do we allocate more to reserves? Do we pay dividends? Do we expand into a new market? All those things should be driven by sort of the health of your portfolio.
But it also affects deal level decisions. What do you do about this particular credit? What do you do when their line of credit renews? Should you shrink their line? Should you require more collateral? Should you change pricing? Those are decisions that you need a real evaluation of their risk to be able to make good solid decisions there. And I don't know that these backwards looking subjective nine point scales really help you do that at all. And I think that's the core issue that banks need to look at.
All right. So it feels like we're piling on banks a little bit for not having better risk ratings. So maybe we should pivot a little bit in tone here and talk about solutions or we could continue to be judgy a little bit and ask why isn't this better? Why haven't they fixed this? Right now as you mentioned, there's no ... it's pretty obvious that you need to be able to have a really good idea of how to measure the risk at your bank. I don't think that's ... unless you happen to start banking sometime after 2008 and just had no memory of what happened before that, it seems like it's a pretty clear standard important thing for the bank to have. So why haven't they fixed this issue?
Well, it's hard. It's a big undertaking. Number one, there's just, if you change it, you've got thousands of credits than to sort of re-underwrite under the new criteria. It's the middle of a pandemic, employees at home, dealing with all the stuff they're already dealing with. Is that the time? Maybe not, but I don't know that it ever is, because it is hard. It's just a labor intensive effort and it has so much impact on how much capital should the bank hold.
You think that's hard at a $300 million community bank, imagine the scale of that at a JP Morgan or a Bank of America. We want to make an adjustment to our risk rating system. And that's going to drive how much risk based capital we're holding on all of these deals. And we have to pull that off in time and then explain it to the street if it impacts what capital levels we're holding.
So it's kind of an existential question of how do we measure this stuff? Now, that being said, I think it's hard to argue against banks. The fact that the banks should have a dual risk rating system, this trying to boil everything down to one number, what you end up with is you can have two strong rated one or two grade credits that are very, very different. One could be an unsecured loan to Berkshire Hathaway because it's Warren Buffet, and the other one could be the local muffler shop that is on the verge of bankruptcy, but it somehow is secured by cash. Your actual expected loss on both of those is near zero. Those are very different deals and the nature of the risk, and the way that you need to evaluate in the way that you need to manage it are very different, but you can't tell the difference with one system.
I think almost every bank we talk to during our sales process, as we demo our platform, we have the ability to handle both ways of doing that from a risk grade perspective. And a lot of banks say, "Well, we use one grade now, but we're going to move to a dual system." And a lot of them we bring them on as clients. And now it's four years later and they still haven't done it. It's on everyone's to do list, it's just when they actually going to get to it. And I think that this may be another one of those things that it's the impetus to actually get that done. You need to be able to differentiate between those things and to be able to take action and to be able to take quickly. So those sorts of things I think are just going to become instead of nice to haves must haves for the industry.
Yeah. Carl Ryden has a collection of a vast never-ending reservoir of annoying but accurate proverbs that get spouted out. And one of those is, particularly when it comes to big project like that, and with the sort of understandable hangouts of like, "Look, we're in the middle of doing this, it's going to be really hard while we're doing this to also change this." And you get the whole, "Well, you know what they say, the best time to plant a tree was 10 years ago. The next best time is now."
So understandable reasons why this is difficult, if you're going to, and I'll mix it in another proverb here, if you're going to start that journey of a thousand miles, then what is your first step in this process? Is there a way to say like, "All right, we're going to dip our toe in the pool with this. We're going to get this ball rolling without having people going, I can't do it right now."
So our take on this is that there are things you can do where you don't have to just rip it all out all at once. And the most important thing with risk ratings is, I think a lot of banks get tangled up in the accounting aspect of it, sort of what I was talking about before, how are we going to get all this done and get every deal that's on the books underwritten again and put it in under the new grading system? And that is a daunting undertaking. I would argue it's not the most important. Instead, this is something that you can take baby steps with and you can do it as you touch a deal. So as a brand new deal comes on the books or as you're doing a renewal, we've helped a lot of clients do simple things like just add a plus and minus to their numbers.
So if most of your deals are a three, just by the way you've set up the metrics, maybe let the banker, whether that's credit or however you have it set up at your bank, put their thumb on the scale, either direction with a plus or a minus. So if it's a three plus what we're aiming for is to change some behavior and change some structure. So a three plus allows you to be maybe more aggressive, a three minus makes you handle that a little differently from a structure perspective, require more collateral, shorten the term. That's what we really want to do is use those risk rating systems to guide behavior from your bankers to cause them to make a different decision. So through the pricing and structuring process, you can actually improve the health of the portfolio using that.
So a new deal comes on the books, if it's a minus, maybe it has higher ROE targets or maybe you put some extra restrictions in place. And if it's a plus, you let them be a little more aggressive. And you can do that with both new deals and renewals, and it doesn't have to change the grade that goes in your core system or in your credit system. It can still be a grade three. All we're talking about is the how you make the pricing and structuring decision. So it can be a part of your pricing approval process. It doesn't have to go through all of your credit policy and all the board level changes, sometimes regulator changes that have to go along with that.
So pick the pieces that are easier to change, but they're the ones that are important because they actually impact behavior and tackle those first. You can also at the same time, before moving to a full dual grade system, same thing, as you're structuring deals, you can kind of separate out that loss given default aspect. And really what we're talking about there for most community and regional bank deals as collateral and guarantees.
What's the collateral? What's the recovery rate? What's the math on, if it's an 80% LTV versus a 60% LTV? And this is where we get into what Gita refers to as a quantitative system. If it's more subjective and basically it's, "Hey, our policy is 70% LTV or better, and it's at 69%." It's all good. If it's at 40% still all good. Well, the math says those are different, and you can price and structure those differently. So that's kind of what we're getting to is let them ... another one of Carl's sayings, let's weaponize the math a little bit. It's not hard to figure out how much impact a different type and a different value of collateral will have on that deal. And same thing with a guarantor. Is it no guarantor at all or is it Warren Buffet? That should have a material impact on how you handle the deal versus just, "Oh, we required in policy. So all deals have it and just doesn't make any difference."
Measure what that difference is. And again, tie it to the actual behavior that you care about when you're either doing a new dealer or when you're renewing one that's rolling over. So that's where to start. And then pretty soon you'll have, when you go to actually make the big change over to all existing deals, a pretty healthy portion of your deals will already be labeled somewhere in your systems, in PrecisionLender if that's where you do it, as three minus three, three plus, you've already got some differentiation. And you can just map some of those things over to a different scale if you need to. You've also tracked the math on your loss given default stuff. So you can start to translate that to a dual risk grading system. And yeah, then you can do that when the world's not quite as crazy as it is right now.
Is there a way ... we talk about weaponizing the math to sort of get better at doing these things, but is there a way also to ... because this is a long winding road, and we know that sometimes big projects at banks, the longer they stretch out, the more likely they are to get derailed and mothballed, is there a way to sort of maintain momentum to sort of show like, "Hey, we took these initial steps and it worked. And that's why we need to take these next even bigger steps."
Always break it into bite sized chunks. And so again, if grade four is your past grade and that's where 85% of your credits are sitting, start by having a plus and minus just on that grade and give some rough criteria and heck make it pandemic related. If they are at risk of closure based on economic shutdowns from the pandemic, they get a minus. If they're not, they're a plus. Get some way of differentiating based on new risks that you should be evaluating that maybe isn't in your current grid. And do that for two months and then just go back and measure what happened. And what you should be able to show is, "Hey, we put pluses on these and minuses on these, and look, we got tighter structures on the minuses, we got better spreads on the minuses to justify the risk. This works. We're impacting behavior. Now let's do the other grades."
And you take that baby step. Then you start to look at, "Well, what if for again, brand new deals we do a dual rating system as we price it. We do the single one as we underwrite it and put it into the system. We're bankers. We can manage to keep three numbers in our head instead of just one. Let's do both." And again, then go back and look in a couple months. Did that change how those deals went on the books? And I think if you can show evidence of that, that's how you can get buy and up to the very highest levels of the bank, where you're going to need some support to get something like this done. Start small with some pluses and minuses, and then you can eventually do the big system wide overhaul. But if you break apart the credit and pricing approval processes, which you should do anyway, that should make it a little easier to pull that off.
All right. That'll do it for this week's show. And I can't promise this, but I will make every effort not to bring in college football recruiting talk into the next couple ... at least the next couple of podcasts, Dallas.
No need to go cold turkey, Jim. We can work it in if we need to.
Exactly. Baby steps, right? Baby steps here.
Jim Young: Right. All right. Well, thanks again for listening. And as with the last couple of episodes, this discussion is part of a larger conversation we're having this quarter with commercial banks about, what's your next move? If you've got thoughts on this that you'd like to share, feel free to reach out to us via our social media platforms or you can even send me an email. It's initial J-Y-O-U-N-G@precisionlender.com. And it'll be in the show notes.
And again, now for a few friendly reminders, if you want to listen to more podcast or check out more of our content, visit our resource page at precisionlender.com or head over to our homepage to learn more about the company behind the content. If you like what you've been hearing, make sure to subscribe to the feed and Apple podcasts, Google play, or Stitcher. I'd love to get ratings and feedback on any of those platforms. Until next time, this is Jim Young and Dallas Wells and you've been listening to The Purposeful Banker.
About the Author
Follow on Linkedin
More Content by Jim Young