Download the Article
Stewart: Welcome to another edition of the InsuranceAUM.com podcast. I'm Stewart Foley. I'll be your host. Today's topic is what insurers need to know about CECL. And we're joined today by Theresa Meawad, senior director and head of Solutions Consulting at SS&C EVOLV. Theresa, thanks for joining us. Thanks for taking the time.
Theresa: Thank you for having me.
Stewart: We're thrilled that you're here, and I think a lot of people don't really know necessarily all the facets of SS&C Technologies. It's a very large firm, like over 27,000 employees. And I want to know about how SS&C EVOLV fits into the equation here. But first, where did you grow up? What was your first job? Not the fancy one. And what makes insurance asset management so cool?
Theresa: So I grew up in Alexandria, Egypt, so that's a little ways away. My first job was probably babysitting. I did a lot of that when I was young. And what makes insurance and asset management so cool is that there's a lot of opportunity to learn a lot of different facets of a lot of different business models, and a lot of different businesses and to really think about asset allocation.
Stewart: That's so cool. So as somebody who's not familiar with CECL, can you start off maybe just give us a little background on SS&C EVOLV, and then let's talk about what is CECL and why does anybody care? So first, SS&C EVOLV, give us a little background.
Theresa: So SS&C EVOLV is a business unit within SS&C. And we focus on all types of cashflow-based financial instruments. So loans, fixed income securities, reinsurance recoverables, anything that's really a cashflow-based financial instrument. And we do accounting across the board for those kinds of instruments. So we do GAAP accounting, we really bridge the system between the core system and financial reporting. And we go all the way through disclosures.
We say we're data to disclosures, because every step of the way is important to us, the data inputs to the calculations, to the subledger, and then finally feeding the GL and also producing disclosure. So we really are holistic approach, and we fit within SS&C. We work with a lot of the core providers, like Precision LM and Singularity that are really the core accounting systems, and really layer on top the GAAP adjustments that are required to go from the pure servicing view, to the GAAP accounting view that would be required for disclosures.
Stewart: That's fantastic. And one of the things that I learned is that SS&C provides the backbone for so much of the financial services community that folks often don't think about. And this is one of those things. And what I think is interesting is that we've gotten a number of folks who are managing residential whole loans as a growing asset class on insurers balance sheets. And what you do provides a loan level solution for that asset class. Is that true?
Theresa: Yeah, that is true. We actually started working with mortgages. We're very, very heavy when PCI accounting was a big deal for insurers and banks. And we got really, really good at projecting cash flows and doing a lot of the basic adjustments that are required for those asset classes. So absolutely, those are our backbone.
Stewart: That's fantastic. So without further ado, what is CECL, and what type of instruments does it apply to?
Theresa: So CECL is the current expected credit loss accounting standard, or alternatively known as ASC-326, which came out in 2016. It really required holders of long-term financial instruments to think about the allowance calculations differently. So instead of looking at a looking one year horizon and looking at what losses had already been incurred in the past and reporting on those, CECL required a different view of a future-looking forecast based methodology that looked at the life of the instrument.
So whether you are a day one, when you acquire or originate the instrument, or five years into it, or in the last few months of its payment, it's still requiring you to look at the expected cash flows, looking at what the expected amortized costs are, and then looking at what the expected loss is. And then the second thing it does is because you are a forward-looking now instead of backward-looking, it requires adjustment for the macroeconomic forecast, the impact to the macroeconomic environment as well. So as we're looking at the future with a potentially looming recession and higher inflation and higher interest rates, CECL is requiring us to think about how that will impact a portfolio that an insurer holds, and to really account for that quantitatively as well.
Stewart: That's fantastic. And this is solid gold for a lot of people who... Like this is your world. So what types of data do insurers need to track, to make sure that they meet CECL requirements?
Theresa: So one of the interesting things about CECL is, unlike the old allowance methodologies, it's really expanded the types of instruments that it applies to. So no longer is it just loans, it's really expanded into things like health maturity, securities, AFS securities, even reinsurance recoverables. So it's really even for people who are leasers, it's also expanded into leases. And sometimes even reinsurance premium depending on the lengths of those insurance periods. So depending on the instrument, you may need different types of data that's required to calculate the allowance. But I'll actually, the way that I think about, it's not just the instruments that are important, but what type of data do you need? In two different ways. One is the calculation of allowance itself. So for that, the data tends to be pretty simple. You need repayment terms because you need to know the life of the instrument, and you need to know what the current amortized cost is so you have something to compare that to.
And then potentially a loss history if you have it or a model that can project your expected losses. Those tend to be the data that's required of any insurer. And then depending on the model that they use, there may be other data requirements for the models, but those for just calculating allowance, it tends to be pretty simple. However, because CECL is so discretionary and applies to so many different asset classes and the models can vary dramatically from institution to institution and from asset class to asset class, the disclosures that are required under CECL have expanded dramatically. So the disclosures require a lot of additional data. So we're seeing a lot of people disclose things like, what is the average life of the instruments? How are they segregating their assets so that they can apply a model to them? What kind of credit attributes are driving the portfolio? Is it an external credit risk rating? Is it delinquency? Is it geographic location? So those things are now needed for disclosure, as well as things like vintage, any charge-offs that are being taken during the period, the amount of time that an asset has been underwater. So those are the kinds of things that are required for quantitative disclosures. And then CECL also, to make things even more complicated, requires a qualitative narrative around what's driving allowance. So management really needs to get into details and understand what's driving the allowance and then think about how they want to communicate that to investors. So reporting is actually a much bigger part of CECL, and it's been a huge area of focus as the macroeconomic environment has been volatile, people are disclosing more and more data to allow investors to come up with their own decisions, because there's so much management and judgment in CECL, and so much variation between the types of models and different asset classes that are in scope.
Stewart: That's really helpful. So I'm an insurer and I am faced with needing to adopt CECL. What challenges am I going to face as I take on what sounds like a fairly daunting task?
Theresa: It definitely can be. I would say that there are ways to definitely reduce the kinds of challenges, like thinking about the process holistically. But there is definitely a couple of challenges that we see across the board. So the first is really recognizing that CECL is a fundamental change from the incurred loss model. Originally when CECL came into play, a lot of people thought that CECL is just taking the incurred loss methodology and making it longer, but that's actually not true. When you look at what insurers should be thinking about is, it's not just a pivot from a short-term horizon to a long-term horizon. It's actually a fundamental pivot from a backward-looking model to a forward-looking model. And that it actually requires now the incorporation of economic data. And what that means is that there's actually many, many drivers of the change in allowance and of the allowance itself.
So previously under the incurred loss model, it was pretty simplistically a rate volume analysis that could account for allowance for the period and the change in allowances. But now you have many different drivers such as the life of the instruments, the macroeconomic conditions, things like that. So really understanding that it's a fundamental change and you'll need a fundamentally different process is the first one.
The second is data input complexities. Because it's a forward looking forecast, insurers need to track more data elements to be able to calculate the allowance. So amortize costs is one, repayment terms, understanding what prepayments or call rates are going to be in the future is another... Thinking about what the drivers of inherent credit losses in the portfolio really are, because sometimes the history is not reliable because we're in such a different macroeconomic environment. So that kind of data input complexity is really important.
And then third, I would say really importantly, there's a lack of standardization. So even though one of the really great things about CECL and one of the really horrible things at the same time is that it's a concept-based standard. So it really allows a lot of management discretion, management judgment, and that's led a lack of standardization. And it seems like regulators and auditors are getting comfortable with that, but that is something that is a challenge for insurers because they're not able to compare themselves to others as easily. So that's definitely something that will come into play.
And then fourth, making sure that the models and requirements are understood and that they're being applied appropriately. CECL is a brand new standard. The first wave of adopters really applied it only four years ago. And so things like post-implementation reviews are still going on and requirements are evolving. And then so the models are also changing and adopting to the fact that the past does look different than the future. And has since the adoption, because it was adopted during the middle of the pandemic unfortunately. And going along with the models and requirements, it's really important to understand what kind of trade-offs making when it comes to selecting a model.
So one thing that my mentor always told me is, all models are wrong, but some models are useful. And so you want to make sure that you're picking a useful model, and that it's accounting for the things that are inherently difficult, but that you're making trade-offs that are conscious. So a lot of times insurers are picking a model because of the name of the model provider, but it's a black box and they don't know what's going on in the model. And so it's very hard to explain. It's very hard to understand if they update their model, for example. Or they're picking a model where they have a lot of flexibility in terms of what they're picking as economic drivers for each instrument or each asset class, but it's not supported by the vendor and the vendor's not helping with explaining what the changes are. And so, it's left to management and that's a very hard model to support.
So finding the right balance and what your specific team can handle is really, really important. And then finally, disclosures are also a challenge where you talked about because of how much flexibility there is, the disclosures and explaining what's driving the period is becoming more and more important and much less understood if you have a black box model, or if you have a model where there's so much flexibility and what you pick as a variables that it's very hard to explain why you landed where you landed. So those kinds of things are really important and they're really emerging as challenges for insurers and other financial institutions that even have potentially have adopted CECL for years now.
Stewart: So I've been writing notes feverishly while you were talking, and I've got, so challenge one, fundamental pivot from the previous incurred loss model. Two, data input complexities. Three, a lack of standardization. Four, models and requirements that are still evolving. And five, disclosure requirements. So those five challenges. You mentioned the selection of a CECL model and a CECL provider. If I'm in that situation and I need to make that decision, what consideration should I be thinking about and what mistakes have you seen others make that you could help me avoid?
Theresa: So the first thing I would do is really think about CECL as a process and an accounting standard first, and then as a model second. So what I mean by that is really think about what controls will you need to have in place? How much time is it taking from your team? How many points of reconciliation will you have? So that's something that's really important. And then when you're picking a model, I would say there are two very things is, how much flexibility do you have in the model? So you should have some flexibility, of course, to be able to exercise management's judgment, but know the trade-offs between the flexibility and the supportability of that model, and how much effort it’s going to take for your team to address and audit our questions and regulate our questions.
So if you have a model where you're really picking each individual, so let's take loans for example, CRE loans, because that's pretty big in insurance. If you have a CRE asset class and you're picking what variables are the macroeconomic variables that are driving a losses for CRE, and you pick unemployment, GDP, and the CRE price index, but a model is not built for that, you're actually just picking those variables. An auditor will come and ask you, how do you know that these are the right variables to pick? How do you know that you should be looking... What period you should be looking back at for those variables? So are you looking at just the backward, what happened over the last year? What happened over the last two years? What's going to happen over the next two years? So there's a lot of discretion that will go into it and there'll be a lot of auditor questions. And the insurer's team is going to have to be the ones to answer those questions.
As well as things like, are you looking at a point in time estimate? Are you looking at a change? Are you looking at what, for example, for unemployment, are you looking at where the unemployment level is going to land? Or are you looking about how much it's changing over time? Because that's a fundamentally different change, a fundamentally different equation. So things like that that the insurer will have to address that sometimes vendors can help with. If you have a prebuilt model. That may limit flexibility a little bit, but you more than make up for it in the fact that they're supporting their model themselves.
The second is, I would say very strongly from the modeling standpoint is transparency. This is a really big one because as a macroeconomic environment is volatile, people are trying to hone in their models and make them more useful. And while they're doing that, they're making changes. So if you have no transparency into what the model was in the first place and what calculations it was doing, and then you change something and it fundamentally materially changes your allowance, are you able to explain what changed?
Because a lot of times vendors are giving you a black box model and they're expecting you to rely on their judgment. But with CECL and the ability to apply the macroeconomic model quantitatively or to apply a qualitative factor, you really need to understand what the model is saying and why it's saying what it's saying, and be able to support it from a management perspective. And I would say that goes into some of the... Those two things, I would say are both, we're seeing people make mistakes on them. So people are choosing a model based on its name, for example. So there are some vendors that have a really big name and modeling and there's no question that their models are good, we can support that their models are good, but if you don't know what the model is calculating, it's hard to know if it's good for your portfolio, and it's hard to explain that to an auditor.
And then it's also... People are also picking models that are just the model. So they are having to pull the input data from one source into the modeling stage, then pull the input data into a cashflow projection stage, then bring those two things together to apply a loss rate to the cash flows. So that's causing a lot of headache for the teams because there's no built in controls and they're constantly having to reconcile those sources. And then there's no single point of truth. If those things don't agree, how do you know which one is right? Which one should you rely on? So those kinds of process things are really important.
And then we're also seeing people go the opposite direction where they want to pick their own variables, which is great, but then you're left with having to explain why those variables are appropriate to an auditor or regulator, and you're left on your own to do that, which is also taking a lot of time from the team. So I've seen us picking somewhere in the middle where it's an all-in-one process, where data's provided once and then the system is calculating and reconciling things internally, running the models internally. And then the models are prebuilt, they're statistically significant, and the vendor can show that. And then so that the results are reliable but understood by management, because the model's documented and the changes in the model are documented.
Stewart: Yeah. I think it stands to reason that the regulator and the IC staff want to know that you know the risks that you're trying to assess and what's being fed into those models, because ultimately the regulator's concerned about your ability to keep track of your risks. So how have you seen CECL impact insurers to this point?
Theresa: So there's a couple of different ways. One is it's adding of course, additional complexity to their calculations. So having their internal teams have to understand a new standard and having to understand a process of doing things definitely is impacting their back office. But from a more positive standpoint, getting their insurance finance risk data all in one place, so they're able to do the types of analysis that they haven't been able to do before. And to provide that analysis to a host of stakeholders. So investors are really interested in CECL's results. It's telling them a lot about the credit quality of the portfolio that insurers are holding across a multitude of different asset classes. And then because there's so much disclosure, they're able to compare themselves to other insurers much more easily.
And then third, because they have a life of instrument projection that's now taking into account macroeconomic variables, hopefully from a trusted source that's helping insurers really understand what the underlying credit quality of their portfolio is and what the drivers of the change in the credit quality of the portfolio are, so that they can make better asset allocation decisions. Obviously, there's a lot of goals that asset managers are looking at from an insurance perspective. And so when you're making decisions, it's helpful to know over the life of the instrument, where our credit losses lie, when are they going to be taken, when do we expect them so that they can manage the portfolio risks and the portfolio rewards appropriately.
Stewart: So Theresa, do you think that the NAIC adopts a CECL loss allowance model similar to GAAP? Or do you think they're going to stay more with the current impairment model?
Theresa: That's really the big question. We know so far that the NAIC has no immediate plans adopt CECL, but they have been doing outreach to the industry to get feedback from all types of stakeholders around what they want to see and how they would feel about CECL adoption. They've indicated that they don't have any immediate plans to adopt, but they are waiting for the next wave of CECL filers, who are adopting this year, to do the next wave of questions.
And while they may not adopt the entire standard, we have seen in the past where they may adopt portions of the standard. So, maybe not the whole concept of life of loan and an economic model, but maybe just the life of loan concept would be something, or an incorporation of the economic forecast maybe in others. So, they may go one way or they may adopt completely, they may stay with the current standard, but more likely they'll end up somewhere in the middle.
Stewart: So if I'm an insurance company and I'm having to deal with CECL, what parts of it can SS&C help me with?
Theresa: We can help with everything. We're an end-to-end CECL solutions. Our platform accounts for all kinds of... All the instruments that you would need to be accounted for under CECL. Whether that's loans, HTM securities, AFS securities, even reinsurance recoverables, or premiums even. We can take care of all of that, and we do for several clients. We do a cashflow analysis for the life of the instrument, we apply the macroeconomic model, which is third-party validated and documented so that you don't have to worry that the model is actually working. We also allow for management judgment to be layered on top of the model at any level possible. And we have a best in class analytics and reporting tool that has all the CECL required disclosures that are built in, click a button out of the box. Once you're done with your process and you're comfortable with the results, that automatically feeds your disclosure. So you can just export them from the system in either Excel or PDF format, and then report those directly into your K's and Q's.
Stewart: And that dovetails seamlessly with Singularity, right?
Theresa: Singularity and Precision LM. Correct, yes.
Stewart: Very cool. All right. So I've learned a tremendous amount. I honestly didn't know what CECL meant when we started this podcast. So you've done a lot of good for me. I want to offer you two questions. You can have the option of answering either or both. Here we go. What's the best piece of advice you've ever gotten, and or, who'd you most like to have lunch with alive or dead?
Theresa: Those are both tough questions. I'll answer the lunch one. I'll have lunch with Snoop Doggy Dogg, because-
Stewart: I love that. Would you want Martha Stewart to join or no? They're like friends.
Theresa: I mean, if she's available. If she's available.
Stewart: Heck yeah.
Theresa: The reason I want to have lunch with him is because he's reinvented himself so many times. It's almost unrecognizable from where he started to where he's now. So I'm fascinated.
Stewart: I mean, you're so right. He's such a phenomenon, right? It's like, and he has such universal appeal. I mean, he has appeal across the age spectrum. He has a tremendous brand. And I second that one. I'd love to have lunch with Snoop as well. So thanks so much for being on. I really appreciate it. I've learned a bunch today, and thanks for taking the time and sharing all this information with us.
Theresa: Thank you so much. And thank you for your time as well.
Stewart: My pleasure. We've been joined today by Theresa Meawad, who's the senior director and head of solutions consulting at SS&C EVOLV, part of SS&C Technologies. Thanks for listening. If you have ideas for a podcast, please shoot me a note at podcast@insuranceaum.com. Please like us, rate us and review us on Apple Podcast, Spotify, Google, Amazon, or wherever you listen to your favorite shows. My name's Stewart Foley, and this is the InsuranceAUM.com podcast.
The InsuranceAUM.com podcast is available on
