Stewart: Welcome to another edition of the InsuranceAUM.com Podcast. My name is Stewart Foley. I’m your host. I’m going to be joined today by Phil Dauer of Grandview Analytics, and we are going to be talking about five investment data challenges for insurers. Phil, welcome. Thanks for being on.
Phil: Glad to be here. How are you doing?
Stewart: Great. It’s nice to see you. We are both in the Chicagoland area, so we’re both looking out at some kind of halfway-decent skies today. Before we get going too far, and you’ve listened to a couple of our podcasts, I think you actually said you subjected your family to some, which is, that’s cruel and unusual, but nevertheless, what’s your hometown, your first job of any kind, and a fun fact?
Phil: Sure. Hometown, well, I was born in Skokie, Illinois, but I grew up in Harrisonburg, Virginia in the beautiful Shenandoah Valley.
Stewart: Oh wow. Beautiful.
Phil: And my first job was as a paper delivery boy. Six mornings a week for numerous years through the sun, the sleet, the hail, the snow.
Stewart: Character building?
Phil: Very much character-building, or at least that’s what my parents kept telling me. Fun fact, I’ve been through quite a few hobbies in my life. One of those about a decade ago was I moved to the Florida Keys for a while and I took up spearfishing. I got certified as a free diver and I started swimming around and shooting fish. Never got very good at it, but it’s quite a bit of fun.
Stewart: Very cool. So before we get going too far, Grandview Analytics, based in Chicago, Illinois, close to where we are. Tell us a little bit about Grandview, your focus on the insurance industry, but give our audience just a little bit of background before we get into these specific data challenges that you have solutions for.
Phil: Sure. So Grandview Analytics has been around for about eight years now. It was formed with a focus on providing the kind of combination of business and technology expertise that’s needed to solve modern data and technology problems. So we work with strictly financial services firms with a strong focus on insurers. We’re brought in to help with things like system implementations or integrations, data warehousing. A lot of that takes the form of moving data to the cloud these days.
I’ve been on a couple snowflake projects of late. We have consulting expertise across everything front to back office. And the key differentiator for us that we like to say is everyone on our consulting staff at least has at least a decade of experience in the industry. They’re all well-versed in the business of an insurer or an asset manager, while also possessing some very strong technical skills with data.
So that’s the consulting side of the business. We also have a managed data offering on the technology side of the business that is strongly suited to insurers as well. And this is an offering that helps to manage clean report your data, called Rivvit.
Stewart: That’s cool. And so, let’s get into it right away here, because if you’ve managed money for insurance companies or been a CIO for at least five minutes, you’ve run into all five of these and I want to get at it. So the first one we’re going to talk about is the data rules between the investment book of record and the accounting book of record, affectionately known as IBOR and ABOR. How are you addressing those challenges?
Phil: Well, there’s numerous challenges there. So we can focus on a couple, and there’s obviously esoteric. There’s an esoteric nature to this for each individual client, depending on which systems you’re running on, what your data flows look like, what your platform is, et cetera. But there are a couple of high-level generalizations that we can draw from all that.
First is that in insurance, more than other asset management, it certainly seems that accounting information is very important for everyone straight up to the front office. So getting CIOs and portfolio managers information—they’re typically interacting with their IBOR system. That’s their daily workflow, looking at their analytics, looking at their trades, all that stuff.
But for them to make informed investment decisions that IBOR data needs to be enriched with information from accounting. And if you think about insurance, a lot of that takes the form of book data, your book prices, your book yields, that kind of thing, broken out in the tax lots where possible.
The places where this becomes difficult is where the granularity of the data, for example, might not be the same where in your accounting system you’re almost always, or you’re going to always have tax slot granularity across all your positions. But in your IBOR system, depending on how that’s been implemented or maybe some legacy challenges, you may not have the same kind of granularity there.
So there’s some transformation of this data required to properly move it from accounting to IBOR. So that’s one thing we come across.
Stewart: And where that comes into play, Phil, I mean for me is: I’ve managed a zillion insurance clients and gained loss constraints. That’s something that is straight down the middle of the fairway for insurance companies is like they’ve got to know that information and that sometimes is not in the IBOR. I mean, I think that’s a great example, specific example of exactly what you’re talking about here.
Phil: Absolutely true. The whole ‘don’t sell anything at a loss’ type paradigm, which is specific to an insurer where a hedge fund or an asset manager or something like that might not have the same kind of constraints and maybe even just less worried about tax slots in general. But yeah, you’re absolutely right. And the other thing that happens in this kind of integration of these two systems is that oftentimes there’s overlapping data points where your accounting system may provide a duration, for example, for an asset that’s based off of a series of cash flows that the accounting system sees for that asset.
Your IBOR may be doing the same thing and they may be getting different answers. So that establishes a situation where you have to build a hierarchy and to establish the source of truth. So you have to look across whether those overlapping data elements and figure out where is our true source, what are the limits at which we evaluate that, and how do we surface that as the one true answer in the end.
That’s a very common problem, very common, not just in the integration of ABOR and IBOR, but the confluence of any data sets really where you have overlapping stuff and common data warehousing type practice.
Stewart: Where I see that applied is in structured securities, and anything with optionality, where your IBOR is running some kind of an OAS model and coming up with an OAD, an option-adjusted duration where your accounting system may not be doing that, right? I mean to me as an XPM, I mean, that’s where my mind goes when you’re talking about those kinds of issues.
Phil: Absolutely. That’s absolutely right.
Stewart: What about a lot of insurance companies, outsourcing is a huge trend and has been for quite some time. That means that you’re going to have external managers. Sometimes there’s internal managers and external managers. Sometimes it’s all outsourced. But now you’ve got sources of data from different places, different data feeds, and so forth. What are some of the common issues around external managers and the data challenges that, that presents?
Phil: Yeah. I’m sure a lot of people in your community are familiar with this one as well. The whole use of external managers has definitely been on the rise. Well, there’s a lot of things that could happen there. So when you think about the general overall flow of data in your system, you’re going to have data arriving from external parties that is going to, number one, arrive in different formats. Number two, arrive at different times. And number three, potentially be pertinent to different parts of your portfolio. And you have to find a way to integrate all that stuff into one holistic view.
So there are a couple of best practices here that I think insurers should be mindful of. One, it’s imperative that there be a centralized what, I just will generally call a hub for this. There needs to be one place where all this data comes together that can handle all the different formats and all the different timing of this to move the data further on through your system.
What you’re going to find is that some of the same concepts we just discussed will also be applicable here, where, for example, there’s a lot of private debt and private assets in general and insurance portfolios these days. And since those private assets don’t always have some sort of common market identifier, they could be identified in different ways in different systems by different managers.
But it’s going to be imperative that if you’re receiving these trades or these positions from an external manager, that you know that a certain identifier there is actually the same thing as a different identifier that you have in your own portfolio and in your own in-house systems. And that’s where some sort of mastering hub that sits between all that external data and your internal data plays a pivotal role. You don’t want to see two positions, you want to see one position. So that’s a good example of where that happens.
And if you think a little further about this, I was thinking about this earlier today, that what insurers deal with when they’re going to these external managers and they’re doing their external manager selection, is what is the asset class we’re interested in? Where does the expertise exist and which managers are going to… Or who in selling this to me are going to best portray that they can provide either the tightest tracking to a benchmark or the highest alpha or something like that? And those are always going to be concerns that are paramount to most any sort of data concern.
But I would think that maybe there’s an opportunity there for some of these people who do the external management to differentiate themselves by also providing standardized sets of data that are deliverable according to SLAs and are holistic, and are robust that can facilitate the insurer’s use of that in a more streamlined manner. A thought I had this morning that some of your listeners might find interesting.
Stewart: Absolutely. It’s very relevant. So the third topic here, this came up at our symposium and I didn’t appreciate how challenging it is. And that is that insurance companies have broadened with a protracted yield and a low yield environment. The variety of asset classes that insurers own today is vastly broader than it used to be. And that includes alternatives and private asset classes. And that creates some challenges. So with regard to alternative investments, can you talk a little bit about some of the challenges that you’re seeing there and how you’re addressing them?
Phil: Sure. And you’re absolutely right. I mean, anybody that has paid any attention to insurance over the past couple of decades has noticed the same trend. And you look at the portfolios now, there’s lots of private assets in there that take the form of infrastructure or private credit or PE funds or something like that. These all have a lot of idiosyncratic type attributes about them. Your standard IBOR systems, and let’s just focus on the IBOR piece maybe, are not suited to handle all of those data points.
So there’s a lot of, let’s call it bolt-on systems or additional systems within an insurer’s investment division to handle that. And this is things like capital commitments and drawdowns and all that stuff that comes with some of these alternatives. All that is necessary, but unfortunately, it doesn’t relieve the need to also track these things in your primary IBOR.
When you’re doing that, some of the big concerns are the following, like how do you model these? You want to accurately represent the risk on these investments. So you think about something like a PE fund with multiple investments. What is the underlying risk there? How are you going to track to that? How are you going to represent that in your portfolio?
In many ways you’re going to have to proxy it. What are the cash flows on some—you mentioned structured products or a private structured product. How do you keep those up to date to ensure that your analytics are representative of what you of truly think these future cash flows, or lack thereof, may look like? Default assumptions and stuff.
So you’ll notice that there’s also been a trend, or maybe the start of a trend, in the industry where some of these big legacy enterprise systems have started to realize this as being a real pain point or asset managers and they’ve started to consolidate. There’s a lot of talk of front-to-back systems, multi-asset systems. There is progress being made. I don’t think that we’re anywhere near the end of that, but yet certainly represents a challenge. And not just an upfront challenge, a lot of that has to do with ongoing vigilance and diligence to maintain a proper data environment to facilitate the proper outputs.
Stewart: Yeah. I mean, I absolutely believe that the need… There’s a lot. I mean, and it’s not all these small firms. I mean, large firms have these legacy issues too, right? There’s been firms being combined and there’s M&A going on and consolidating all that together is easier said than done.
Stewart: I remember this in an economics class, money and banking, I think 500 years ago when I was in college, that the professor said, “Insurance companies and banks manage their assets versus their liabilities.” I dutifully took that note down in my notebook, but the reality of doing that is a real challenge. Insurance companies’ investment portfolios are unique in that they reside inside the belly of an operating insurance enterprise. And while the concepts of risk in investments and on the liability book have a lot of commonality in concept, the vocabulary is completely different.
So the idea that the duration of the liability, for example, and we had a podcast with Rip Reeves at AEGIS who talked about that being the starting point for the riskless position of a bond portfolio. It’s not easy to do. Right? Again, easier said than done. Can you talk about a little bit around asset liability integration and ways that you’re solving that challenge?
Phil: Yeah, I can speak a little bit to it. I mean, I’m no actuarial scientist here, but what you’re getting at is absolutely true. A key data point would be your key rate durations. And this year more than any other in recent history, people have probably been keeping an eye on this stuff with our elevating rate environment. When you’re thinking about the data necessary for this, a lot of what you’ll find if you want to get the best guess or the best accuracy you can out of these matching type analyses, is that you have to ensure that you’re using the same central library of data across all of your systems to do this analysis.
We don’t work a ton with the actuarial side, the liability side, but what we’ve seen is that there’s people responsible in organization for that liability side that are also taking in information about the investments and they’re doing the actual matching. But sometimes the information they’re taking in for those investments to do it may not be the same as what a portfolio manager is seeing.
So what you have in your IBOR and the analytics that is generating may be different than what this matching analysis is showing. And it’s because you may not have a standard library of cash flows. Going back to our conversation about the privates. There’s a key pain point there. It’s now like these privates, the cash flows for those are not managed the central location and they don’t flow through the organization in a seamless manner.
And then your IBOR is a separate system from whatever other system you’re using to do this kind of matching. So you could be dealing with basically different foundations of data and you’re getting different answers. And your portfolio managers operating in a way that your actuarial department may think is unwise, simply because looking for numbers.
What ends up happening is if your actuarial department is using investment cash flows that don’t match to what the IBOR is housing, which is common. It’s totally possible that that’s happening, that you’re working off of different assumptions and your ability to match to those liabilities, or at least your reporting on it is going to diverge.
Stewart: Yeah. I mean, and the thing that comes to my mind is again, the option-adjusted duration around residential mortgages, for example, that when rates have gone up and those are very negatively convex instruments and they’ve extended dramatically. And then is that captured by the actuarial team in their cash flow testing, for example? It hasn’t been as big a deal, but when you see this extreme rate move that happened in 2022, I think a lot of people, as you mentioned earlier, looking at key rate durations and really trying to figure out where those differences lie. Right? It’s important.
And the last thing I guess that I’d touch on here is chief investment officers and I have been fortunate to have a lot of CIOs as friends and it’s a very challenging job and a very challenging environment. The people who write or approve investment policies often don’t appreciate how challenging it is to monitor compliance. There are ratings, upgrades, downgrades, duration extensions, contractions, all sorts of things that happen on any given day in the investment world that create compliance violations. Some small, some large, but managing and oversight governance and overseeing that investment policy statement is a monumental task in the complex and diverse portfolios being run by insurance companies today. Can you talk a little bit about compliance validation and how you’re addressing that?
Phil: Yes, I can. I completely agree with you. Compliance definitely represents a pain point for insurers for many reasons. Well, let’s first talk about compliance in general. There’s a couple different ways compliance can be termed. A lot of time we think about pre-trade compliance and then we think of overnight compliance. From a data perspective that’s probably suitable. There is a third variation potentially, which is more trading controls and who has the reins for certain assets or the ability to trade things, which just puts forth yet another level of complexity to this.
But let’s focus on the perks too. In insurers, a lot of insurers are operating multiple jurisdictions and most jurisdictions when we think of the US take the form of states and each state has their own set of regulations. Those regulations may use similar terms, but those terms may be defined differently.
They could use different measures of capital or different measures to define certain aggregations. All that stuff has to be encoded in some manner to be captured within the insurer’s investments. On top of that, most of that encoding takes place within your IBOR system in many cases. And also in many cases, that encoding is done in a bit of a proprietary language of the platform itself.
So it’s not easily shareable with the rest of your data ecosystem. So that’s just a couple of them. But there are more. I mean compliance is something that you want to track as granularly as possible. Not only do you want to know what you knew, you also want to know when you knew it, right? You were getting at. You can be in perfect compliance. So through no fault of your own, the market changes, downgrades happen, something like that, and all of a sudden you have a violation.
So you want to be able to separate those things that are based on decision-making, like your pre-trade stuff like we were saying. So those things that just happen in the market. So you need a robust data organization to do that and you need your data elements clearly defined centrally for all those calculations. You can’t have someone using a capital calculation somewhere that’s different than what somebody else is using for the same regulation.
All that stuff needs to be, once again, mastered and centralized in a governance framework. And the other thing that we come across at Grandview a lot with compliance is that oftentimes, and in almost all cases compliance is really exception-based. So it really only raises its head when something out of whack shows up, be it in pre-trade or overnight or something like that.
But it’s also imperative that for every one of your transactions, every one of your trades that you have some audit trail to say that at the time that we did this, there was no violation. These were the outputs of all the compliance checks. They were all within allowable ranges. And that does end up being a pain point for people when they need to go back and provide some sort of audit to say, show us your trades and that you followed your proper compliance procedures. That’s often a blind spot.
Stewart: Yeah. I mean, boards of directors and investment committees are relying on the investment team to tell them, “Yes, we’re in compliance.” But that’s challenging. We’ve talked about that when you’ve got the kind of market declines that you’ve seen this year just outright declines in asset classes and equities and core fixed income that changes the denominator fairly significantly that can create some compliance issues with regard to issuer concentrations. What’s commonly known as room and a name, to your point about pre-trade compliance.
So a huge deal and terrific that you… I mean, you’ve outlined… Just to kind of recap data rules around IBOR and ABOR, challenges around external managers, the data, the timing and the integration. The challenges around alternative investments and private investments. We’ve talked about that. Asset liability integration. And then last but certainly not least, compliance validations. So five great topics, and thanks very much for going over those. So Phil, on top of those five, what didn’t we cover?
Phil: There’s so much. I mean in the data universe, there’s just so much. Off the top of my head, we didn’t really talk about capital calculations in terms of your RBCs or anything like that. We didn’t talk specifically about things like NAIC ratings. CSOL, we didn’t touch on that. Top of mind for a lot of people: ESG. And ESG is its own unique kind of data set. Not every organization has really figured out how to integrate that into their deal pipeline, their trading, their compliance like we just talked about. It’s really a pain point. We didn’t touch on it today. Probably be a whole other podcast.
Stewart: Absolutely. We will do it. I will say this, you’ve helped me kind of walk through a lot of these issues and bring up a lot of things that I’ve dealt with earlier in my career. When I talk with you, and I think you know this, I taught as a professor for a number of years and I’ve always got people who are early in their career in my head. I would be surprised if when you graduated from school that you thought, “Well, I’m going to be a senior director at Grandview Analytics working on insurance data problems.”
So I’m thinking that this maybe wasn’t on your radar screen at that point. As you look out today with the vast data challenges that are sure to be more complex going forward, what would you tell your 21-year-old self? What advice would you give that person?
Phil: That’s a good question. I think probably two things. One, maybe for someone like yourself, you’ve spent a career in this industry too. Relationships are critical. Just absolutely critical. You’ll see the same faces in different places throughout your career. As big as this industry seems, it’s actually quite small in the end.
Stewart: It’s very small.
Phil: So relationships are critical. And the other thing that I think has benefited myself that I would give advice to anyone is that intellectual curiosity. Just learning for the sake of learning. Knowledge is so important. Just take it upon yourself to learn and to teach yourself things and to read and to challenge yourself. That’s how you grow. Right?
Stewart: That’s great advice, and a great education. Thanks for being on. Phil Dauer, senior director, Grandview Analytics. Phil, thanks for taking the time.
Phil: Thanks for having me.
Stewart: Thanks for listening. If you have ideas for podcast, please shoot me a note at email@example.com. My name is Stewart Foley, and this is the InsuranceAUM.com Podcast.