Interested in learning more about PrecisionLender? Visit PrecisionLender.com
Jim Young sits down with Rollie Tillman, Managing Director at PrecisionLender, and David Andrukonis, Director at RiskSpan. They discuss what CECL means for banks and how you can best prepare for it.
Helpful Links
Podcast Transcription
Ashley: Hey everyone, Ashley here. We're so excited to announce that it's that time of year again. It's time to think on purpose. Think on Purpose is a conference based in the belief that the path to customer success is built on a customer centric foundation. It brings together the best and the brightest minds in thinking in Austin, Texas on April 25th to the 27th. You could use the code podcast18 to get 10% off your registration and to show your pride for the Purposeful Banker podcast. It's time to stop reacting and get back to what makes banking great. It's time to think on purpose.
Jim Young: Hi and welcome to The Purposeful Banker, the podcast brought to you by Precision Lender, where we discuss the big topics on the minds of today's best bankers. I'm your host, Jim Young, Director of Communications at Position Lender, and I'm joined today by two guests. The first is Rollie Tillman, managing director for Regional Community banks, and our client's success team at Precision Lender. And the second is David Andrukonis, the director at Risk Span. Now, we've brought on David to talk about the implications of CECL, the acronym for the Current Expected Credit Loss model that most public filing banks are expected to begin implementing for the 2020 fiscal year, and most other institutions for the 2021 fiscal year.
But before we get into that, David can you tell us a little bit about your company, Risk Span, and what you do for them?
David Andrukonis: Yes, thank you, Jim. Good to be with you, Rollie. It's an honor to be on the podcast. Yes, so for lending institutions and for fixed income security investors, Risk Span covers that whole supply chain of data, models, and reporting. For institutions that lack internal data or are thin in a particular area in internal data, we have a library of industry performance data. For institutions with messy and disparate internal data, we help organize that. For institutions that need forecasting models, we provide models and model documentation, and a model execution platform. And then for institutions that need to generate reports explaining portfolio composition, and explaining model results to various stakeholders, management, investors, auditors, and examiners, we provide those analytics and reporting capabilities as well.
Then, as Rollie knows, for models that we don't build, we also have a model validation group that provides model validation services, too. I know that covers a lot of ground. Maybe a good way to think of it in summary is for institutions that have pain around data, pain around models or pain in the area of reporting, we can probably help.
Rollie Tillman: Yeah. That's great, David. I think it was through a mutual client that we came to know each other through the model validation exercise. But through that exercise, we came to learn about Risk Span's expertise in the CECL space, both in an advisory and software solution, and you shared some of that with us. And of course, we hear from our clients that they're keenly paying attention to the upcoming changes from an accounting perspective, shifting from a triple L to a CECL methodology in their risk estimation in accounting. So, we definitely wanted to have you on. We hear about it a lot. Wanted to hear more about your focus in that current expected credit loss space. We're interested in it, and our clients certainly are, so thank you for being on.
David Andrukonis: Sure. It makes sense and happy to talk about that. I think it's certainly time for folks to be focusing on it. I think for many institutions it's kind of now, and Larry Sherrer of the Saint Louis Fed made this point at a conference I was recently at, that it's time to be moving beyond the education and awareness stage into the action and engagement stage in terms of preparing to implement CECL. It's a huge change. The change from the outgoing incurred loss allowance standard to the incoming CECL standard is a huge shift. It's not just a change in the rule, but it also just calls upon expertise and resources that many institutions don't have on staff, never had on staff, and haven't had any reason or certainly haven't had the same level of reason to have on staff as they will need to have under CECL.
So, our goal though is to provide resources that lending institutions need, and to distribute the costs of those resources across the industry so that rather than every institution individually needing to go out and buy expensive data, hire full time quants, write very thick lengthy documentation, and then build a program that automates every quarter or even every month, all the necessary reporting requirements, the supporting reports around the allowance estimates each quarter. And you have like 7000 banks and 6000 credit unions all spending the resources to build 13,000 CECL solutions. Our view is we can help bring together the relevant expertise and existing capabilities that we've already had to meet the market's need for CECL in a cost efficient way.
As your customers and the institutions that you're talking to are probably generally aware, the four big challenge areas that we see for most institutions on CECL are number one, the data. That's the foundation of everything else. But even just knowing what data elements you need, where to get it, how to organize it, that is a challenge for many institutions. To support that, we maintain data libraries of industry performance data across a variety of asset classes, and that's enormous for institutions that lack sufficient internal data. Then for institutions that have enough raw quantity of internal data, still data accuracy, data organization is almost always a challenge. We've got data experts, teams of data experts who can assist on a project basis to help banks get the right data, get it cleaned, get it scrubbed, get it organized, and have data experts that are efficient at doing that.
Number one is the data. Two, the models. Building models and especially if you would want to build multiple models, which is a really, really strong place to be in, so that you can explore results across different modeling methodologies. That's a major task. Most banks that we talk to, these staffs, they run lean. They don't have a lot of people sitting around playing Minesweeper just waiting for a model building project to come in. Models is another area where we can help, and we've got standard models that we've already built for all the basic methodologies that community banks and credit unions are likely to use under CECL. So, we can provide those efficiently and also have a variety of statistical models covering a variety of asset classes that, you know those are the kind of methodologies that larger institutions tend to use, and we can provide those, too.
I mentioned the data library earlier. We also host a model library. The pedigree that we're bringing into this here is that for a variety of use cases, so CECL is a use case that calls for life of loan econometric forecasting models, but there have been and are other use cases for that expertise. Fair value estimation, both for accounting and just investing and trading decisions, and for stress testing purposes, too. We've been provided these life of loan econometric models for a long time, and so we have them in place and our message is we can efficiently leverage and deploy these for institutions to help them with CECL.
The third one is documentation, just all the documentation surrounding the data. Why it's relevant. Surrounding the models. Why they're valid. Just writing that documentation is a major burden. Documenting, again, the appropriateness of the data and the models and preparing sufficient documentation to support testing and model validation. That's a huge exercise, and is not only time consuming, but also just no fun except for a very special breed of masochists. So, we can take that burden off.
And then lastly is report automation. So, in addition to the final allowance number, whatever you come up with, there's a whole set of disclosure and explanatory reports, and all the bankers and finance folks listening will be familiar with it. But there's a whole set of accompanying reports that has to come along with your allowance number every quarter. Management, investment auditors, examiners are gonna pour through that. If you intend to produce those reports efficiently, that either takes some very efficient programming ahead of time or a lot of manual work each quarter. Just automating that report set is another burden that we can take off management.
The hope of all that is let management get back to focusing on what, I think the same thing that you at Precision Lender try to help management be able to do, which is get back to focusing on making the business and economic decisions that drive profitability.
Jim Young: I'm trying to understand, is this just a really, another box to be checked, or is there something fundamentally better about CECL that prompted the move to the change?
David Andrukonis: Yeah. Very good question. I think that's a subject of a ton of debate actually. I mean, there are ... I was recently at a conference with Hal Schroeder speaking. He's one of the FASB board members that voted for CECL. So, CECL was approved by a five to two vote. There was, of the seven member FASB board, five voted for it, two voted against it. Even amongst this supposedly the smartest of the smartest from an accounting standpoint, there's still disagreement as to whether it's a good thing. There are many people who feel very strongly that it's not going to improve things and is not gonna fix what it was intended to. I think that's very, from a curiosity standpoint, very interesting to think about. And I would encourage anybody who's just curious about that side of the argument to go read the dissent in the standard itself.
I mean, the bottom line, I think for everybody on the ground, is this is not going away. This is not going to be delayed. Larry Sherrer, at that same conference, which was the Center for Financial Professionals second annual CECL ... I said second annual. It was their 2018 CECL conference. Larry Sherrer from the Saint Louis Fed was there and really emphasized, this standard's not going away. It's not gonna be delayed. Regardless of what you think, it's here. What it was designed to fix is this too little, too late problem where the allowance was backward looking. So now under CECL, institutions take into consideration the macro economic outlook that they have for the future and they make estimates not only of credit losses over the short term, but credit losses over the long term.
And certainly, I think the argument for, something that's good about the standard certainly is it promotes awareness that there is always credit risk over the life of an instrument and that macro economics influence it. Anyway, there's so much actually to get into, and Risk Span's blog is one place, at RiskSpan.com, where we point people to some of those discussions. There is certainly value in institutions ... Institutions should make that choice for themselves of, are we trying to build a solution that actually gives us loan performance forecasts that we're confident enough about to manage our institution around them? That's a very different question from, are we just trying to put answers down that our auditors will not criticize us for? And both types of banks are out there. I understand both points of view, and both points of view can make sense depending on your perspective.
Rollie Tillman: David, two things that you mentioned struck a chord with me. One, at Precision Lender, I think the first point is where there's a line in the sand between our two companies. So, I think at PrecisionLender, we often say that we're focused on the interaction between the relationship manager, the customer facing banker, and the customer. And the problem we try to solve is one more of an economic pricing and deal structure solution. We describe that as being front of the bank. Clearly, and I think you described it as a special kind of masochist, you guys are focused on very painful problems, but they're real problems. They're covered by FASB and GAAP accounting and regulatory constraints, and you have to cross every t and dot every i.
I think the important part is, is that at Precision Lender, we try to put a tool in the hands of a relationship manager that all the folks in the head office, in those special disciplines that have to deal with those real life problems can have the right input assumptions and that Precision Lender to make better pricing decisions. I think that's where there's a natural hand-off. We try to take all the expertise that you help your clients develop in terms of complying with things like CECL to make better pricing decisions.
The other thing that struck a chord with me is that we've often taken the approach because we're trying to make pricing decisions, we always describe our view as a windshield-looking view. Actually I think we're probably more aligned with the thoughts around CECL compared to, as you described, a triple L, which is more of a rear view mirror look. So, while we never have or probably never will put ourselves out as a CECL solution, we've taken more of a forward looking view at expected credit loss and things that can go wrong when evaluating the prospective value of a deal. So, I think there are a lot of synergies around at least the core thoughts of CECL and what it's trying to accomplish with how we've tried to guide our clients that think about valuing the potential profitability of a loan.
So, it's interesting to hear you say that. I think you mentioned that you were just in New York at the, I think it was the Center for Financial Professionals CECL's 2018 conference. I was kind of curious, since you just were with all the brain trusts and the folks who are interpreting and working to implement these things, if there was anything hot off the press that you could share with us for the benefit of your listeners?
David Andrukonis: Yes. I did. I picked up some off the record information from various banks about indications of methodology choices and key assumptions that they will likely be employing, which I can share on an anonymous generally aggregated basis. Then also some on the record comments from some of the regulatory authorities here. There was, one of the headline panels of the day was, featured three folks on there. There was Hal Schroeder, who I mentioned a minute ago. He is one of the FASB board members who voted in favor of the CECL standard. Larry Sherrer was there. He's a senior examiner at the Federal Reserve of Saint Louis, and then Sybil [Syrrahkyah 00:17:41] of the FDCI. I apologize to her if I'm butchering the pronunciation.
For me, the single most groundbreaking statement was from Larry of the Saint Louis Fed. What he said was, and I mentioned a minute ago that he was saying the standard's not going away. It's not going to be delayed. But what he did say, which was really interesting is there may be some relief coming in terms of regulatory capital treatment. CECL is generally a more conservative allowance approach than the incurred loss methodology. But if you keep the same regulatory capital ratios in place, and the same buffers in place, as we had under the incurred loss methodology, but then you make the left side of your balance sheet more conservative, then the effect will be restrictive on lending. So, if we felt like capital ratios were already safe enough prior to CECL, if they're safe enough today, if they'll be safe enough throughout 2019, and then CECL effectively adds more conservatism on top of that to the balance sheet, then we could reasonably loosen the capital ratios slightly and still get back to that same level of safety and soundness.
Larry's point was that the regulators have been doing a lot of listening to this argument, and that there is interest among the regulators in working on it. Larry thinks, anyway he's speaking for himself here, but he thinks there will probably be something in probably in tier two capital to address this concern. Such a thing would need to go through the public rule making process. He told the audience to look for a draft as early as Q2 of this year. In a very CECL nerdy wonky mindset, that was a major bombshell of news to me. I think many institutions will be very interested in tracking that. Certainly something that we'll track at RiskSpan.com on our blog.
And then Sybil from the FDIC, she made the point that conservatism can be a catch-all defense to model deficiencies and stress testing. Maybe that's putting it a little too strongly. You can lean on conservatism as a justification for model shortcomings in the area of stress testing because the whole point is to stress, but for CECL that's not gonna work. You can't address or excuse model deficiencies with using conservatism as a rationale. That just sharpens the need for models to be found.
Then, I also was mentioning conversations with individual banks about what practices are emerging. Banks are generally split on the ways in which institutions will transmit the influence of their macro economic outlook into their final reserve number. You have to do that. Your final allowance reserve must reflect your expectations of the future, your reasonable expectations of the future, but the mechanism through which you modulate your allowance number for that outlook varies, and there's two choices. There's having quantitative or regression type models that explicitly take macro economic inputs, but then the other option that folks will do is just through the Q factors primarily. So, through managements, judgemental Q factors, you'll take historical averages and adjust them up or down in the direction and magnitude consistent with management's macro economic outlook.
So, smaller institutions are more likely than large institutions to continue to make those adjustments by the Q factors, so your forecasting technique is essentially adjusted averages, as opposed to via statistical models. But points to this question of how long ... There's this reasonable and supportable period concept where the length of your explicit macro economic outlook should only go out as far as you can reasonably and supportably support, and then beyond that period, you should be reverting to long term history at a segment specific level.
The length of the reasonable and supportable period is not prescribed by the standard, and the length of time or the shape of that reversion after the reasonable and supportable period back to your long term history is also a question mark under the standard. I've been very interested in what different banks are gonna do there, and we picked up some market intelligence on that while we were there. Everybody I heard from will be in the two to five year range on reasonable and supportable period, and I'd say most are closer to two years. I think the single most common answer that I heard for the length of the reasonable and supportable period that people are using is two years. If you're doing anything else, it's probably three or five. And then some folks are gonna revert immediately to long term history without any gradual reversion period beyond that, and others, the other answer I heard was a one year period for that gradual reversion period.
Jim Young: Well also, David, as you mentioned, whether or not you eat or sleep CECL is secondary now. You're gonna have to. CECL's gonna be part of your diet going forward, so you've got to digest this stuff. Thanks for taking that conference and boiling it down for us. Finally, and you touched on this a little bit, David, but people listening, they need a lifeline on CECL. What should they do to find out more about the change? I don't know that it's possible they would need to know more than what you've just delivered to us, but on the chance that they do, where should they go?
David Andrukonis: Good question. In terms of any upcoming changes with respect to CECL, the standard was released in 2016. What we will continue to see in terms of any changes to the requirements or the guidance around the requirements, there is a transition resource group for CECL. That's a group that FASB puts together to hear industry questions on CECL, comment on those questions. We maintain a regular blog series on our website, RiskSpan.com. It really consolidates a lot of this information for folks, tracks emerging practices, and tracks regulatory guidance on CECL and it's freely available on our website, and I think a good place for folks to go. Updates from the transition resource group are there. We actually did a webinar a couple months back with Graham Dyer of Grant Thornton, who's a member of that transition resource group, which is also available on our site, and reviewed the latest transition group updates there.
And I think, well we'll see. Again, we'll see accounting practices emerge. Those are also things that Risk Span will track. And then lastly, it'll take a while, but it will be very, very interesting and informative in 2020 and 2021, once we begin to get feedback from the audit community, and ultimately the PCAOB after CECL adoptions have happened and auditors begin to sink their teeth into them and make their comments about them and give feedback, that could be a big wake up call for some. There will be practices that change and emerge following the early rounds of feedback from the audit community. And those lessons and the instructions that come from them will also be things that will track. I would encourage folks to visit Risk Span's blog. We've got our team of CECL experts pulling together advice and learnings and commentary from around the industry on CECL.
Rollie Tillman: David, it strikes me too. I think you have a deep thought obviously and you're plugged in and have your finger on the pulse at Risk Span. But from a practical perspective, and we've seen a demonstration of this, I think you referenced it earlier in the call, from a practical perspective, you actually have a software solution, a soft solution to this to help folks organize their data, to help them actually accomplish the needs of reporting. I just wanted to point that out, that I thought that was an interesting aspect as well. I don't want it to get lost.
David Andrukonis: I appreciate that, Rollie. Very kind of you to bring up our end-to-end CECL solution that we offer. Yeah, we're proud of it and excited about it. We've got a great team of data scientists and quants here that have built an end-to-end data consumption model execution report generation system within Risk Span's technology platform that is also certainly something that can help folks as well. So, thank you for ... I appreciate you bringing that up.
Jim Young: Thanks, David. You never have to apologize for guiding people to your blog, because that's what we do at this point in the podcast every week, in which I tell people if they want to listen to more of our podcast, or check out more of our content, they can go visit us, visit our resource page at PrecisionLender.com or you can just head over to our homepage to learn more about the company behind the content. That'll do it for this week's show. David, thanks so much for coming on.
David Andrukonis: Thank you.
Jim Young: And I want to remind our listeners, if you like what you've been hearing, make sure to subscribe to the feed in iTunes, Sound Cloud, Google Play, or Stitcher. We love to get ratings and feedback on any of those platforms. Until next time, this has been Jim Young for David Andrukonis and Rollie Tillman. And you've been listening to The Purposeful Banker.
About the Author
Follow on Linkedin More Content by Maria Abbe
