Stewart: Benchmarking for insurance companies has been a problem for many, many years. And today we’re going to talk about something called benchmark snapping with Erik Troutman, Senior Insurance Strategist at Loomis Sayles. Erik, thanks for being on.
Erik: Thank you very much for having us, Stewart.
Stewart: We’re very happy to have you. I want to start this off like we start them all off, except I have a little bit of inside information: Home town, first job of any kind, not the fancy first job, and then fun fact, but the fun fact I actually already have on you. So can you just give me your home town and first job?
Erik: That is a tough one because I moved around a lot as a child. So, born in Poughkeepsie, New York. So I guess that’s the home-home town, but we moved many times while I was growing up and wound up, basically, finally in Wyoming and then coming to Boston in 1988-
Stewart: Wow. Good for you.
Erik: … and, basically, been here ever since. So I call Boston a hometown now.
Stewart: Good deal. How about first job?
Erik: That would be cashier at a Long John Silver’s in Colorado.
Stewart: There you go. A fishy start. I couldn’t let it go. I’m sorry. So I got a text this morning from one of your colleagues who said that you have five kids. So the fun fact for you, I’m going to go with: five kids. Tell me a little bit about the family.
Erik: It can be described as chaos at times. So, we have five kids, three girls, two boys, and they’re tightly packed, 17 is the oldest, and 9 is the youngest.
Stewart: Good for you.
Erik: And they go to a multitude of schools and it’s just one of those chaotic scenes. And we also have … I mean, the other, I guess, fun fact would be that we, basically, have a zoo at home as far as animals go.
Stewart: Oh, that’s perfect. What’s the animal situation? Give me a little bit of an inventory.
Erik: Three miniature schnauzers, dogs, three bunnies,-
Erik: … four guinea pigs and then various, I guess, reptiles. So we have leopard geckos, we have chameleons. We also have some more exotic ones like tarantulas and centipedes. And so-
Stewart: Wow. That sounds like a fun house to grow up in as a kid. That sounds like a cool environment.
Erik: They love it.
Stewart: That’s awesome.
Erik: And so-
Stewart: That’s the greatest thing you could do.
Erik: I just wanted one black Lab to sit on the couch and that’s what I got. So…
Stewart: That’s perfect. All right. So you’ve been in the insurance asset management business and working with insurers in an advisory capacity for over 25 years. Benchmarking, as you know, is a challenge for a myriad of reasons, some of which we’re going to talk about. Can you set the stage for benchmarking for our audience?
Erik: So benchmarking, I mean, in total return space, it’s very clean in many ways because you just put a certain amount of dollars into a portfolio. You have a benchmark, which a manager runs against, and out-performance is easily seen by having more money than what the benchmark would’ve given you after time. Performance attribution, all of that, is very simple, clean because you don’t have constraints on it.
But when you come into the world of insurance, that all changes because you’re getting money at specific times in terms of premiums coming in, which you are investing in certain ways to fund the liability payments and the growth of the surplus of the company. And those timestamps matter in terms of, one, when you price the product and determine the rates that you’re going to charge and then putting the money to work. And then holding that money in those investments over time and not necessarily worrying so much about the ebbs and flows of the market, and that can come in two different flavors. And I think of it as the constraints of the company in recognizing, realizing gains and losses, and then just being and holding the money in things like duration-matched ways so that when you price a product today, it is good for the life of the product.
So you don’t care about the ebbs and flows as much as the total in the total return space. And that means that when you’re investing, that timestamp matters very much to your assets and how they’re going to actually behave across time.
Stewart: And it also gets into … I don’t want to front run a bunch of our questions here, but there’s also all sorts of regulatory and ratings considerations. I mean, it is not as simple as running total return. Not only is it the timing of pricing the products and whatnot, but it’s also how you’re rated and regulated. It is not a purely economic equation of making asset allocation decisions and changes and reallocations?
Erik: No, you have to take the capital position of the company into account, what the rating agencies will think of that capital allocation and how much volatility you can actually, as a company, take – because while the liabilities may be fine, the surplus of the company could become strained, it could impact your business processes, your dividend policies and other efforts that you have around the firm as to how the assets perform. So you do have to take into account the actual overall volatility of the assets and all the regulatory needs that comes down on that.
Stewart: Yeah, I mean, versus, you’ve got an institutional asset management operation running inside of an operating insurance entity and you can never lose sight of that?
Erik: No, your client is the actuaries, the products, and the overall entity itself.
Stewart: Speaking of which, that’s where you got your start. You started on the actuarial side?
Erik: I did. I started in the actuarial development program at John Hancock Life and rotated around the company into several different locations, one of which was actually their mutual fund complex, which was quite interesting. And then eventually wound up in what they called their investment policy and research area. And the way I look at it is that was the third leg of the stool. So you had the product development actuaries, the pricing actuaries that developed the products, priced the products, and they knew what they needed, and then you had the bond department that actually put money to work.
And our role was to, basically, work with both parties to understand how we could invest, what we could buy, and then, basically, tell that to the actuaries and then also hedging overall, making sure that the ALM was aligned. That was my start, was growing up in that department.
Stewart: That’s a terrific place to start in this business. The same colleague, actually, that sent me the text message about you said the following, “The life insurance business has changed more in the last 2 years than it has in the last 50.” How has the insurance industry changed while you’ve been working with it? Maybe even more importantly, where do you see it going?
Erik: The biggest change I think has been the asset classes, sectors and investment policies, I guess, of the companies. They’ve gotten much more diverse, much deeper into all the different places in the asset sectors, not just your plain vanilla treasurys and public bonds. They’ve evolved dramatically from that. The ALM and hedging strategies have evolved a lot since then as well, across time.
I think the competitiveness has gotten much more increased over time, as well, so that drives in life insurance, in particular, it drives crediting rates and the aggressiveness with which you invest. One of the biggest fundamental changes has been the private equity parties coming into the life insurance space and how they have changed the game.
Now, it’s not that they can invest differently per se, it’s that because they’re pooling a lot of assets, they can tap into less liquid markets that take much more, I guess, size to get into and much more, say, relationships to get into because you need to be in the deal flow to actually place the money and they’ve done it for return reasons. They saw a huge capital base in life insurance and you had life insurers that wanted to get rid of lines of business so there was a natural synergy there that they could achieve their returns by investing in these more aggressive and, I guess, less liquid areas to, basically, price the business such that a life insurer would get rid of it.
Stewart: Yeah, it makes sense. I mean, so when you look out and you see public bonds where their trading rates have come up significantly, you mentioned that insurers are investing in a much broader set of asset classes than they have been prior. Do you think that’s going to continue? Do you see that trend slowing down any, with rates rising at all? What’s your take on the future direction?
Erik: I think there’s always a search for yield and so insurers will naturally … I mean, you think about it almost like water, it will flow out in search of every nook and cranny that it can find, too. I think there are limits on that because I think as more and more money flows into different sectors it causes the pricing to get maybe too rich. You’ve seen that time and time again where a new asset class comes along, the early movers take advantage of that and then maybe too much capital chases it and, basically, ruins some of the alpha, if you want to put it that way, in what they can achieve.
I think the banks are stepping back from a lot of the different lending areas and the more private equity and private capital direct lending stuff coming up. Insurers naturally can step in there because of the illiquidity. Life insurers, in particular, can step in there because of illiquidity. They have the deep capital and it aligns from an ALM standpoint. So they can naturally step in there. I’m not going to say we’ve reached our limit, but I think there does come a time when there will be enough money chasing certain areas that eventually there’s only so many assets and too much capital chasing.
Stewart: So when we talk about benchmarking in particular, all the capital aspects, the lines of business, the profitability, the tax position, the list goes on and on, it makes it virtually impossible to benchmark insurance companies one to the other. Can you talk a little bit about what you’ve learned in tackling this challenge and what you’re doing at Loomis Sayles to talk about ways to unravel some of the benchmarking challenges for insurers?
Erik: Yeah. The way I think about it is starting with the liabilities. The actuary’s always developing the product, they develop to be profitable. So they have a lot of things priced into the product already, such that if I can deliver on a yield that they make, if I make an assumption on a yield I can get on my assets, default should be taken out, profitability, all these other expenses should be, and I should get what I want at the end of the day. And then the question is, I take in my money, I’ve priced it where I thought I could, and make money. Then there’s the question of does the asset manager deliver on that, internal or an external asset manager? How do they deliver on that? And because of all of the, as you mentioned, the taxes and other constraints and regulatory capital, it’s not a total return question. It’s more of, “Can I put the money to work?”
And, there’s some simple metrics you can do to determine whether you’re actually at least achieving some profit, your targeted profitability and targeted yields. And that’s simply by saying you took in a certain amount of premium over the month and your priced products would’ve had your yield at, say, X, and your manager, internal or external, again, basically, put it to work at Y. And on the surface Y is bigger than X and, therefore, I should be fine. Now, you also want to layer in the various things that you mentioned like I need to make sure that my overall rating of the assets is within my tolerance such that my regulatory capital is good, my duration is aligned so that when this thing moves, I don’t have reinvestment rate risk. So you have to make sure that those metrics and other things are in alignment.
But, at the end of the day, my asset manager has put money to work at a higher rate than what I wanted or has at least achieved the rate I was hoping to get. And you can do that. I mean, you can simply … I’ve done this process many times where you, basically, look at either a model portfolio or an index as a guide for what you should get. And you, basically, say you can do it on any timestamp, the more granular, the better, you could benchmark your asset manager that way. The issue is that a lot of managers can … Yield can be a mis-rated bond and you get more yields and there’s a lot of downgrades or defaults or things that you suffer down the line, but it doesn’t come out in the metrics that you’re seeing at the beginning. And that’s where I think of this benchmark snapping process actually becomes a very interesting option or alternative.
Stewart: So talk a little bit about that, because I’m not familiar with the term of benchmark snapping. I think this is really the heart of what we want to talk about, is to really explain to our listeners exactly what it is and how an insurer can implement that.
Erik: Yeah. So snapping is, the concept is, I think of it as either snapping that chalk line, snapping a picture of what the investible universe looks like as of a particular timestamp and then, basically, using that picture, that timestamped universe, that snap, as a guide as I move forward in time. You can put a lot of nuances into that snap because if you think about it, if I use just a corporate index as a guide, well, there’s a lot of bonds in the corporate index that do not trade, that aren’t liquid. So maybe I want to think about a snapped universe that’s only of bonds that were issued within the last 60 to 90 days because my portfolio manager can only buy those bonds because those are the ones that are trading. So there could also be … Indexes can contain small bonds that you can’t fund.
You can put a lot of different things into this construction of your snapped benchmark. And then, you basically, do this large algorithm and you say, “On 6/30, we put in a hundred million dollars into this portfolio. The selection universe would have looked like this.” And you can build a portfolio out of it. Now, the investment manager will go and buy their portfolio. And then you can just simply move those universes – your actual portfolio will move forward in time and your snapped benchmark/portfolio will move forward in time. And what we do is we just, basically, build that portfolio and track that portfolio going forward.
Stewart: I see. So it’s saying, “Here is an investible universe at a point in time,” and then I’m going to compare what the portfolio I actually bought at that point in time and I’m going to compare those two?
Stewart: Because if I understood you correctly … And, I mean, I know this, but it’s a very helpful reminder that when you look at benchmarks, a lot of those bonds I can’t buy. I just cannot go buy them because they’re not trading. So why benchmark myself against something that I can’t own? So this is really a practical way to do this.
Erik: Indexes are a good start, but you’ll often get pushed back from managers which say, “I can’t buy these bonds. I can’t buy this area.” As issuance ebbs and flows in the universe, different sectors could be too large for what you want for your risk exposures. Ratings could be too large for your risk exposures. You may want to remove some of those mispriced bonds. Some that are triple B that are trading high yield or things like that. You just, basically, want to hone in on what the manager could buy, given your guidelines and so on.
Stewart: And how manual of a process is that? What bonds were issued in the last 60 to 90 days that I can actually buy? How do I get that universe put together?
Erik: Honestly, that’s where the computers come in. I mean, honestly, you build the selection criteria and then you turn it over to what universe do you want to layer these selection processes on? It can be … You often start with an index, say Bloomberg Corporate Index or a securitized index or something. And then you layer on top of that, these selection criteria. But the computers can do it all and they can go through and you can use trace data to access what did and didn’t trade. And then you can put that in and you can, basically, build the index and size the index. You can size it up to the dollar amount actually that you purchased and you can have things in there where you could conceivably also have minimal lot sizes and things like that.
So you can be … And it’s not necessarily a one-and-done. What we’ve found is that sometimes you get too granular and what you get is you cut the selection universe, still find that you get idiosyncratic noise and you want to back off. So in securitized space, in particular, we’ve found that somewhat of a less granular snapped benchmark might be a better option than targeting just a double A or single A or triple A, ABS, or things of that nature. So it’s a back and forth a little bit to work out some of the kinks.
Stewart: Okay, Erik, that helps. So let’s talk about with regard to benchmark snapping, can you talk a little bit about if I’m using a total rate of return benchmark and doing it, what I would refer to as ‘the traditional way’ of just looking at index returns, that is very likely, or has the potential to give me, a false reading or a false sense of how I’m actually doing. So what differences have you seen in using the benchmark snapping approach versus the traditional approach where insurers are concerned?
Erik: So the total return benchmarks are evergreen. So bonds come in, bonds go out and you do feel in the month that a bad event may happen, you will feel that in the total return, but then it just leaves the benchmark and you never see it again. And the new bonds come in and you can’t buy those and so on. And so the total return universe is never stacked. It’s always changing, but that’s not the way the insurance portfolios are managed at all. It’s not that it’s a buy and hold, but it’s very much a buy and hold for a long time, again, because of the ALM aspects and other aspects of just managing through an insurance liability.
So what the benchmark snapping does is that it will lock down the universe that you could have bought as of that timestamp. And it’s also great to get a reflection of just what was going on in that … You think about tight spreads versus wide spreads. It’s just poor underwriting standards at a time and good underwriting standards at a time.
So by locking down that universe, you can really start to benchmark whether the asset manager is any good at credit selection, because what you can do is, as you go forward in time, you have the universe that they could have bought and you can say, “We had 10 downgrades and we had 5 upgrades.” And then you can go over to the actual management portfolio and say, “Oh, I only experienced 3 downgrades and 2 upgrades.” Or you can actually get that comparison. Now, it takes time for that to develop. And that’s the one instant you know, where you can go back and you can use index yields at a particular date to, basically, do a book yield comparison.
This benchmark snapping process, basically, adds this time dimension where you can actually start to determine whether the manager can deliver on credit selection and get you increased yield with reasonable metrics.
Stewart: Well, to your point, and I’ve run a fair amount of insurance money back when the earth was cooling, and the way that you’re describing this is very consistent with how insurance portfolios are actually put together. At the time when the money is there, what is my investible universe today? And then if I can freeze that and compare what was the performance characteristics of what I actually bought versus what I could have bought? That seems like a very pragmatic way to look at benchmarking, particularly when you’re talking about upgrade/downgrade statistics, for example, which have, as you very well know, capital implications for insurance companies and OTTI implications for insurance companies and other things that total return investors don’t deal with.
Erik: Yeah. And I think another aspect that can come out of it is reinvestment rate risks. So if I lock the universe down, I call it aging the benchmark, if I age the benchmark and I keep the ecosystem flows, all the cash stays in there and as it comes off in principle and coupon, I have to reinvest it and I reinvest it.
And the way this works is, essentially – so if you go back to the benchmark snapping, I make a $100 million contribution on June 30th. And then on July 31st, I make another $100 million. Well, I just do another snap and then I meld the two together. And now I have a combined benchmark snapped portfolio. And, as you can see, it’s starting to grow and I can keep that going and then I can also recycle all of my cash flows back through it. And so I keep that close and I can use that as a gauge against my asset manager to also get the $100 million, two times, three times, four times, and keep that going.
And then if they are better at buying things that don’t call, aren’t prepaid, that if rates drop, for example. I’m thinking of securitized, in particular, there’s that reinvestment rate risk that I can get. And, essentially, these benchmark snappings will pick up on that. And if my manager’s better at managing that, then that will also start to come.
Stewart: I have to say, it seems like a very logical and very solid way to look at things. What sense does it make to talk about a total return benchmark that I can’t invest in, that I’ve got all kinds of constraints I’ve got to deal with? If I’m careful and I do a good job of my benchmark snap and I’m looking at a truly investible universe that I could have purchased on that given day, now I’ve got something that I can meaningfully compare to and say … Yeah, I mean, whether it’s an internal manager, external manager or combination of both, it seems I’m going to get a much better sense of manager performance, if you will. Not totally return performance is not what I’m talking about. Just what value did the manager add versus what the investible set was? Have I got that about right?
Erik: No, you’re spot on. And, honestly, the genesis of this process at Loomis Sayles actually was for buy and hold. And the Chief Investment Officer said, “Well, how am I supposed to benchmark you?” Because the total return isn’t going to tell the story. And they said, “Well, how about if I do the universe that I could have bought?” And that was the simple premise on which this was instituted years ago.
And when I came to Loomis and I saw the process and I listened to them how they described it, I was like, “Well, this is book yield benchmarking for a life insurer.” And we actually went out, we talked about it internally and we actually worked with a large life insurance company that had this dilemma. They had some total return manager, internal managers that were more total return focused, and they couldn’t come up with good benchmarking.
And so we worked on a process by which we snapped. We sent to them information on different … They have a pricing algorithm that has different cells that build model portfolios and we snapped benchmarks for all of them. And we started this a couple of years ago and we’ve delivered this on a weekly basis. So that’s a pretty granular time step for them because they’re putting money to work every week. And we’ve, basically, provided them all of the information and they have built a whole process on their side to do the performance measurement exactly as we’re doing it on our book yield benchmarks.
And it took us about a year to work out all the kinks in the process as far as that granularity stuff I alluded to and the amount of outstanding limitations, country limitation, industry limitations that go on top of this selection process, and they have gone live with it on their side. And their portfolio managers like the process because it’s very reflective of what they can do and buy. Again, it’s been great because it’s nice to have this willing partner in learning about it.
And we’ve been working on the aging process, as well. This part where, basically, as I go through time, I evolve the book values and I evolve the book deals. And we’re now working on … If you think about insurance portfolios, there’s new money and there’s existing money. The new money is easy, but how do I do this on an existing money portfolio and without building the history up? And the way we’ve actually … It’s not that we kicked the can down the road, but we simply said, “Well, we’ll build it up over three to five years and then we’ll call that the existing benchmark and then we’ll have a new money benchmark, which we have already, and we’ll have this existing one where we let some of the snaps drop off at the end and we keep adding to the front and we have a snapped process that spans three to five years.”
Stewart: It makes a lot of sense. I mean, it just makes … I mean, it’s very common sense. It’s pragmatic. It accurately reflects the value being added, or not, by the manager, whether it’s internal or external.
Erik: It’s a mountain of data. I mean, there’s no getting around that, but that’s where your systems come in and your processes come in and you need to have a lot of scalability in that process to do it and you have to be able to track things that might fall outside it. So if you think about one nuance that a lot of those indexes out there, when things move inside of a year, the bonds drop out. Well, now in my snapped benchmark or in my insurance portfolio, that’s not true.
Erik: So we have a process. We catch all of those so that we can maintain pricing and analytics and stuff. We catch all of that in our process. And we form a pricing portfolio out of it to keep them. A few slips come and go, as you get exchanges or other things. And we have some artificial intelligence processes to catch that, to help us as we age the universes that we can catch the CUSIP changes or other corporate actions that might have taken place. So, again, you have to have the systems there because if you try to do this on a manual basis, it will just implode from just pure weight.
Stewart: Yeah. That’s amazing. Well done. I love this stuff. I mean, I love doing podcasts with guys like you because I learn new stuff. You guys have done a phenomenal job, and I really appreciate being able to hear what you’ve been up to.
Erik: Honestly, because book yield benchmarking has always just not been there. There’s one or two products that I’ve seen out there that tried to do it. But this process, honestly, is, it was something that they didn’t know they had because they, honestly… When we first met with this insurer that I spoke of, I described processes that I had done at my prior firm. Here’s the way I used to think about it. And then when I was talking with them, when I came back to Loomis, I was talking with our IT guy. It was our IT guy. And he’s, “Oh, here’s this process we do for buy and hold.” And I was, “Oh, my God.”
Stewart: That’s it. That’s perfect.
Erik: The IT side is huge. That is huge. You’ve got to have just people that love to roll up their sleeves and can get in there.
Stewart: That is, at the end of the day, that’s insurance asset management. It is a roll-up-your-sleeves game. I mean, absolutely. That’s cool. Thanks for being on. I really appreciate it. I just have one last question for you. We kicked off the beginning of the podcast that you have five kids and I’ve got a soft spot in my heart, being a former professor for college students. And so you’ve been in this industry a while and I would ask you this, if you could speak to the Erik Troutman that was graduating from college out of your undergraduate institution, what would you tell your 21-year-old self? What advice would you give your 21-year-old self as you look out in the world today, the financial services opportunities today, the world, the insurance world, asset management, whatever it may be? Is there any advice you’d give yourself today?
Erik: Yeah. And it’s going to sound silly but I used to think that education was the end all and be all of everything. I loved to just try to consume as much education as I could, always tried to get the best grades, that sort of thing, but what I missed was the personal side, the networking side. And what I’ve tried to tell anybody that I mentor or talk to, or when I’m thinking of my kids and telling them how to do in school, I’m like, “Try to make connections,” and that’s as valuable as education in some cases-
Erik: … in my opinion. So when I think of that, I said, “Just network the heck out of things, in addition to doing good on your studies, and don’t take a singular focus on just knowledge.” And that’s one thing that I think. Honestly, that’s one advice I give because I think that’s the mistake I made.
Stewart: That’s great advice, and it’s been great to have you on and I’m very happy to have added you to my network. And it’s been really fun to get to know what you’re up to and learn about benchmark snapping and it makes all the sense in the world to me. And thank you for taking the time.
Erik: Thank you very much. No, this is fun. I liked it. I love talking shop.
Stewart: That’s great. That’s it. It’s about all we do here. So it’s good to have you. Erik Troutman, Senior Insurance Strategist at Loomis Sayles. Thanks for listening. If you have ideas for a podcast, please email us at podcast@insuranceAUM.com. My name’s Stewart Foley and this is the insuranceAUM.com podcast.
This material is provided for informational purposes only and should not be construed as investment advice. Any opinions or forecasts contained herein, reflect the subjective judgments and assumptions of the authors only, and do not necessarily reflect the views of Loomis, Sayles & Company, L.P. Investment recommendations may be inconsistent with these opinions. There is no assurance that developments will transpire as forecasted and actual results will be different. Data and analysis does not represent the actual, or expected future performance of any investment product. Information, including that obtained from outside sources, is believed to be correct, but Loomis cannot guarantee its accuracy. This information is subject to change at any time without notice.
Any investment that has the possibility for profits also has the possibility of losses, including the loss of principal.
There is no guarantee that the investment objective will be realized or that the strategy will generate positive or excess return.
Past performance is no guarantee of future results.
This material is not intended to provide tax, legal, insurance, or investment advice. Please seek appropriate professional expertise for your needs.