agilon health, inc. (AGL) Earnings Call Transcript & Summary

[Audio Gap] Company's CFO, Timothy Bensley and the company's head of IR Matthew Gillmor. So I know, Steve, I know you got some slides out today that are going to walk us through kind of what's been going on at the company and then we'll open it up for Q&A.

Great. Well, thanks, Justin. It's great to be here. Appreciate the opportunity to talk about Agilon, our '23 performance and really what the forward implication is for 2024 and beyond. If you go a couple of slides back, let me start at a macro level. One more. So our vision at Agilon is to empower primary care doctors and really transform the delivery of senior healthcare in this country. We've built a sustainable primary care model that's improving patient experiences and outcomes, and we see that as an important response to what the country needs. When you look at the numbers on this page, you can see how this is resonating with primary care doctors, patients and payers, both health plans and CMS. Since our IPO in April of 2021, we have tripled the number of primary care doctors on our platform and the number more than that in terms of the number of patients on our platform. We've been able to do this growth in a very capital-efficient way. We are well capitalized today and don't need any additional financing needs to support our growth. If you go to the next slide. So this is our financial model and long-term commitments. Today, we are reiterating our 2026 adjusted EBITDA target. This is our guidance with Geo entry costs, which is north of $530 million. That's per the revised SEC guidance that we received earlier this year. When you add back an expected geographic entry cost in '26 of $70 million, that nets to greater than $600 million, which is what we told you at our Investor Day back in March. On the right, here are the adjusted EBITDA drivers that we've consistently talked to you about. Our membership, as we will show you, is meaningfully ahead. That's important not just for '24, but for '26. ACO REACH is modestly ahead. And then our medical margins are slightly behind. And given the questions we've got in the last couple of weeks, we'll spend more time on that kind of walking you through it. So you can go to the next slide. So how have our medical margins differed from expectations in 2023? First of all, just a definition, Medical margin is revenue less cost. This year, we have seen higher revenues than we expected, both with true-ups from last year and midyear true-ups this year as well as higher costs. On the left, you have our original guidance, $547 million at the midpoint. On the right, you have our most current guidance, $465 million. Walking through that, you back out our Hawaii market, which was in for $25 million in the guide that we originally gave you, which takes us to $522 million. There have been 2 large in-year issues that have taken down our medical margin. First, our supplemental benefit costs. Those are benefits that are filed by payers. Those have come in at $24 million higher in cost than we had projected. Those are largely cash cards, specifically for dental and for OTC. The second area is really around core medical services. Again, the difference between revenue and cost in that bucket has been $13 million off what we initially told you. When we isolate what is the driver of that, we are seeing 3 of 25 markets that are off and very concentrated in the subset of services, specifically outpatient surgery and associated specialist costs and Part B drugs. That gets you to $485 million for an in-year guide for 2023, but we also have had prior period of $22 million this year, which takes us to our guide that we gave you recently at $463 million. Go to the next page. So now we're shifting from medical margin to specifically talking about cost trend. And what are our assumptions for the remainder of the year, specifically August through December. Our full year cost trend guidance is $90 million higher than our initial expectations. We are 90-plus percent complete on May, June and July and those costs were $30 million higher than expected. But as you can see, they are moderating, May at $19 million, June at $8 million, July at $3 million. We've taken that information and our guidance on a forward basis now assumes $10 million a month for August through December or $50 million of higher costs. Just to foot to the $90 million I talked about at the top, our prior year development has represented an additional $10 million in cost during the third quarter. So what levers, if you go to the next slide, do we have to improve our performance in 2024 as we build to that guide I gave you for 2026. We are expanding programs to control medical costs, specifically around high-risk complex patients, programs in renal, palliative and high risk management, we see that as a $30 per member per month opportunity. We have very strong results in a palliative study that we have just announced. It shows for a 1,000 senior patients that were enrolled in that program, significant reductions in terms of hospitalizations, 39%, 48% less ICU days and 21% fewer ER visits. We are also focusing on high-cost settings. We have much better visibility on inpatient and we -- across all markets in Q1 of 2024, we will have a real-time inpatient capability, think admit and discharge. At the same time, as we think about some of these markets and where we've seen cost pressure, we have lower-cost ambulatory surgery center alternatives in some of those markets. And then finally, we always have the lever because of this primary care model that we've built to move our primary care physicians from the bottom quartile to median performance. And Justin, I think it's important that even in these markets in which we're seeing cost pressure, our top quartile primary care doctors are running well above $200 PMPM, which tells us there's a real opportunity. We're engaging with health plans, around some benefit exposure to decrease that. We've got data collaboration work and future contract structures to eliminate or mitigate that, and we've implemented a strategy to address prior year development. We do that through closing visibility and data gaps and reserving in the upper end of our actuarial range exiting 2023. So if you go to the next slide. We talked about being ahead on 2 key drivers for adjusted EBITDA. Membership is meaningfully ahead relative to what we told you in 2021, in 2022. Initially, we're doing about 25,000 a year in growth in these classes. At the time of the IPO, we were talking about 40,000 to 45,000. We are now looking consistently at 100,000 members per year that are coming on the platform. Each year, we are running 20% to 25% higher than what we initially told you. Not only are more lives coming on the platform in these very large classes, but they are coming on at higher levels. Our platform is getting better. Some of that's technology, some of that's a longer implementation process. And so larger classes bringing on more members at a higher starting point is impactful as we think forward. ACO REACH performance is also ahead of our guidance for this year. And at $39 million this year, we are ahead of our 2026 target. ACO REACH is delivering a similar level of medical savings as Medicare Advantage in 2023. And this is very important to our doctors. Medicare fee-for-service is incredibly challenging for them. And the ability to be able in a risk-based model that provides this type of surplus to them is meaningful and I think that's part of our accelerating growth. Next slide. As we look at a recent announcement, we did divest Hawaii due to a lack of strategic and financial fit. Strategically, Hawaii was our only non-partner market, really didn't fit. From a financial perspective, you can see higher OpEx, lower medical margin. On the right, we show you the adjusted EBITDA performance, where do we expect it to be and where is it running year-to-date. The sale enables us to focus on our core partner markets and as a stock sale, we do not retain any liabilities. So let me conclude with one more slide. How do we put this together and how do we think out to 2024 and beyond. So we've told you that our medical margin growth in 2023 is modestly lower than our initial guidance. There are 255,000 members that were in year 2 plus classes in '23. They saw a 23% increase in medical margin year-over-year and we would expect them to continue to grow next year. We will always have same geo growth that runs above the market level but what's really meaningful when I talk about the power of membership is these large year 1 and 2 classes. They are the largest in our company's history and they contribute meaningfully. The class of 2023, 130,000 MA members that was strong this year and will ramp medical margin up materially in 2024. In the class of 2024, 110,000 members or more starting at the high end of what we typically talk about from a medical margin range. So when you look at the size of these 2 classes, they are literally as large as the prior 5 classes. We've doubled the company in the last couple of years and these members that are coming on the platform are starting at higher levels than what we've seen historically. And finally, we expect to build on our '23 ACO REACH performance and we will be providing our guidance or preguide at JPMorgan and [indiscernible]. So with that, go to questions.

Great. I appreciate all the detail. I know the questions have been nonstop since the third quarter report there. Maybe we just start with some of the commentary on Medicare Advantage and medical cost trend. So $19 million in May to $8 million to $3 million in July. So first, what was different about May, right? We all know that trend was up, right? I think it's been pretty broadly disseminated at this point in Medicare Advantage. But what did you see in May that moderated in June and then moderated further in July?

Yes. Basically, we saw -- as Steve was talking about, we saw a pretty big spike up just in core medical utilization across those 3 categories that we were talking about. A little bit hard to understand exactly what was driving that to spike up in those categories across things like outpatient specialist part of your drugs specifically. Clearly, as we came through July and June, those same categories continue to moderate -- moderated back down. But one kind of interesting thing is when you look across utilization overall, one thing that we've had pretty good success in and this is really, I think, kind of a hallmark of our model is our inpatient utilization has continued to be down versus a year ago. And so that's nice. We didn't see like a huge spike up in utilization that's continue -- that's really -- when you think about some of the clinical programs that we have that we put in place, it really is usually there to drive -- one of the outcomes is to drive lower [indiscernible]. But those kind of 3 categories specifically spiked up quite a bit and all 3 of them moderated back down.

And Justin, I guess what I would add, what I tried to call out is this is concentrated in a subset of markets. So 3 of our 25 markets in particular and in those areas around outpatient surgery, more hips, more knees, specialist costs associated with that. Some Part B drug being up, particularly in the area of oncology are the areas I would talk about, also supplemental benefit costs. Those cards, the way they're set, there are limits, they can be annual limits, they can be quarterly limits, they can be monthly limits, and we did see a surge in some of that in that second quarter.

Yes. So what is your visibility on this. I know what I always tell people is you're a step away from the plans and the plans don't always know, look, look at Humana, Aetna, both came and said, look, third quarter was even worse than we expected, right? So they clearly don't have perfect visibility. You're one step away from them in terms of the claim. So how confident are you in these May, June, July numbers?

Well, I think we're at this point, we're 90-plus percent complete for those. I think we're clearly on a lag to your point. And I think what we've forecasted in terms of increased expense, the back half of the year at $10 million a year. When you saw June and July come down to 83% reflects that. I think we're trying to be cautious around that and be appropriate as we think about it. In terms of visibility, where we do have visibility in select markets is on the inpatient side. And to Tim's point, we have been performing well on that. And then, we are working with our payer partners on better visibility around the supplemental benefit side. So there's more real-time exchange around that as well as on the claims. We do have a gold standard payer that is on a 1 to 2 month lag that's sort of the best and we're pushing our others to get there.

Got it. So then flipping over to those supplemental benefits, it's a little strange to see those kind of spike up and then moderate to your point, like those dollar cards typically get loaded every quarter, right? Do you see the dental benefits kind of people just run through them and then stop? What do you think is driving the volatility is there and why do you think it's kind of moderating?

Well, I think there's use and how much do you need? I think when you do dental work, it can come in larger episodes around that? But I think as we look at it and as we talk with the plans, I think the expectation is probably a step down in the second half relative to the first. But our expense expectation doesn't assume that. So what we put in is rather than trying to say we think outpatient surgery will sit at this level or specialist supplemental, we're saying, here's what we're seeing in aggregate and we want to make sure that we're prudent around that.

I think the nature of some of the cards also [indiscernible] towards the first half because not all of them are just monthly limits. But having had a whole very good experience now of 6, 7, 8 months, year, I think we've done a better job now going back with our payer partners and say, hey, let's understand what your trends are? Why are [indiscernible]. So as we built our assessment for full year expectations, one of the things we did was we've got a better conversation with the peers about where the double mount benefits are going year-to-date and now a better job on that.

You're actually seeing them decline. When you say supplemental benefits, are you talking about vision, dental and then these dollar cards, so it's all kind of bundled in.

That's right. But the outlier has really been around the cash cards specifically on dental and overly.

That's why I find it were. Like so many of those cash cards I see are quarterly refills. So I would expect it to be high and, let's say, April, right, when they get refilled and then moderating. The fact that it didn't -- that's moderated in July and doesn't look like it's back up. It was a little bit strange.

Let's play it out. Let's say it is higher in the second half. I think what we're saying is the expense guidance that we've given for the second half would be sufficient to cover that. So that's the plan.

Okay. And then talk a little bit about what you're doing on contracting there to kind of move away from that kind of volatility in that benefit that I assume almost obviously, I don't think you can have -- your doctors can have much impact whether someone's going and buying toothpaste with their over-the-counter card.

Sure. So let me start with we've got a $24 million higher expense associated with supplemental benefits in our run rate. We carry that forward to next year, and we assume that's flat to down. So we're not expecting massive relief on that as we think forward. In terms of data work, we are working with the plans. That's typically a different part of the health plan in terms of how we get that information on a faster basis. So there's data work. And then these contract structures in which there's the opportunity to either not take risk on supplemental benefit or to cap that risk. That design is occurring across a very broad base out there. And Agilon and our partners and the associated senior patients are one piece of that. And so we've got about 1/3 of our existing members and payers that were renewing every year and then obviously, the new ones. So I think it's a combination of contract structure, data work and then also just sort of what is our underlying forward assumption.

If you had the ability to, would you just kind of wipe this out of your contracts, right, take a lower PMPM, but remove this risk given your inability anybody's ability to kind of manage it.

I would. I think that it may be difficult for some health plans to administer that. And so the alternative is you put a cap on.

So when you say put a cap on it, you're just saying this way, right now, what I've heard is people are utilizing them at greater levels than they have historically. So if they had $1,000 a year, they were using 60% of it before. Now they're using 90% of it, right, right? That's the problem.

Well, I think they're using more of it. I think the limits were also set at much higher levels. And I think if you talk to the payers because this is a pain for them as well. The adjustment to '24, for those who've seen the largest issues is in bringing those limits down. So I think it's a combination of those.

How does -- how would a cap work explain that?

Well, I think you could look at -- so historically, supplemental benefits you're getting an actuarial estimate from them. You're getting revenue, you're getting cost. And so you are looking at that, and we're saying above that level, what would that look like? And so that's the estimate that we're getting from them. And so let's cap it at that level.

Got it. So you're effectively saying, look, if you guys think this is going to be 70% or 90% let's cap it there. And if you're wrong, that's on you, not on us.

Correct. If you want us to take risk around this for something that we can't control, which is a senior patient going to a local drug store and purchasing $70 worth of tylenol. You should be the one that sort of sets the expectation and takes to.

And what's planned receptivity to that? So let's say you have a 30-year contracts up. How many of those are kind of capping this risk?

So I mean, we're in the process right now kind of working through with them on that, some more receptive than others. But again, our forward assumption is not that in '24, we have massive relief.

Got it. And then as you think about one of the things we're all talking to managed care about, and you're kind of, again, a second derivative of this is how did you set benefits for 2024 right? For instance, I know Humana as I believe is your biggest payer. Our analysis indicates they cut their supplemental benefits more than most. So how do you kind of feel about the potential, let's call it pressure release valve of benefit changes for next year given what plans are seeing?

Yes. I mean, I think it varies by payer. I think some payers took a meaningful step down. I think some payers did take a step up in terms of those. We don't have a ton of D-SNP business. So in the non-D-SNP business, when we look at it on a composite basis, our forward assumption is kind of flat to down around that. But it varies pretty dramatically across some of these markets, depending upon your payer mix and just your concentration with those different payers, Justin.

Got it. Got it. And then some of the other offsets you talked about, how meaningful can it be in [indiscernible].

Well, I think having ambulatory surgery centers as an alternative to hospital inpatient -- or hospital outpatient surgery can be meaningful. I think -- so that would be one for us that I would really call out. The supplemental benefit piece, you've got sort of our assumption around that. I think the biggest one is really these clinical programs. Right? Think about the palliative care data that I provided to you on that. We see those programs as an opportunity of $30 PMPM, we're not saying we're getting all of that next year. But there's a real opportunity as we get those into all of our markets. The markets that we're in, we are having very strong outcomes. To Tim's point, our composite inpatient is down. That's a function of the work that we're doing around that.

Okay. So as you think about kind of one of the key factors you've talked about, obviously, the place where you've done nothing but kind of outperform has been on the membership side and the revenue side. You've got a strong class this year. Talk to us about the class this year. You've got a pretty wide range of expected performance, right? I think the medical margin is $30 to $60, typically in year 1. So before we talk about 2024, how is that 2023 class looked relative to that $30 to $60.

Yes. It's performing -- sorry. It's performing pretty well. it's definitely going to be above the $45, so we'll be better than breakeven from an EBITDA contribution in this class. That's a nice jumping off point for us until next year. Justin, as you kind of mentioned, it's one of the -- it is actually the largest class we've ever brought in at 130,000 members. We'll get some same geography growth out of that next year as well. But that big of a class with a very decent starting point, kind of above the midpoint of where we would normally operate moving into next year, which is the biggest year that you get of a step-up in medical margin year-over-year is really just a positive, a very nice tailwind for us going into next year without even talking about the class of 2024 yet.

Justin, what I would add on the class of '23 huge 130,000 members but also first health system, main health, performing very well. Class of '24, two health systems coming on Premier in Dayton, Holland and Western Michigan. So I think we are very encouraged that health systems that come in maybe with a higher cost basis. Our ability to start to get after that cost and do the normal work that we do in terms of setup around getting patients in and getting an appropriate burden of illness for that population. So class in '23 is starting in a good place for this year, biggest class we've ever had. And then you know in our model, the move from year 1 to year 2 is always a very substantial step up. And so we see that not just in terms of the revenue PMPM but in terms of the magnitude of 130,000.

Right. And your point is that $45 plus typically goes to $100 plus in year 2. So we're all just sitting here doing math around a bit. I think you guys saw the earnings bridge that I put together the EBITDA bridge. You're just saying, look, 130,000 members are going to go from $45, $50 to $100. Then you look at the class of 2024 and you're saying that's going to do north of $60, right? So walk us through your visibility on that. What are the 2 or 3 kind of key drivers of why that -- why you feel so confident that this is going to outperform your normal kind of trajectory?

So it's a few things. One is we have a longer implementation cycle, which gives you the ability to get patients in for their initial assessment visits. Two is we have better clinical data with this new technology platform that we bought. This is the first class in which we're able to integrate with those EMRs, which helps you in terms of identification complex patients enrollment. And the third is this class will have clinical programs available to them far earlier than we expected. And I guess the last is we can look at their underlying cost performance today and that is probably in line or maybe a little bit better than what we've seen in other classes on an adjusted basis.

Got it. Are those areas -- if we talk quickly about the class of '25 and how that's shaping up? And also, is that something where you think is unique to the class of '24 or do you feel like that implementation that earlier implementation, getting the clinical programs rolled out early, is that something that we can kind of pencil forward as being at the high end or above going forward because.

Your question is what's sustainable.

Class of '24 because it has every one of those 4 components, including the fourth one, which it was pretty good across these markets, pretty strong performing group [indiscernible] going to be, I think, especially positive for us. It's a good year for that happening with a [indiscernible] class of 2024. But I think generally speaking, all of our classes, we've always said historically, and if you look at the cohort data where we give you what how each class started. We've said 30 to 60 because that's essentially the range that we've experienced. The idea would be because of the other elements that are drivers, as Steve talked about, longer implementation, which we're going to have for the class of '25, we're very well advanced in terms of contracting in those providers set up things like that our ability to use the incremental data sources that we're going to have now that we'll even start to be able to use during the implementation year 0. And then you're just the idea that if you have longer implementation, you can do things like already start implementing clinical programs that we never would do that before in a year 0 market. I think all that together means, yes, we would expect that over time, our average is going to be more [indiscernible]. We'll see. The wildcard is always what's the starting point for those markets, but I wouldn't expect that we're going to see starting classes below that breakeven point.

So my headline would be PMPM on the higher end of our range, larger classes and the forward impact of that substantially.

And so talk about the -- any early read on the class of 2025? Do you think it's going to look more like 2023 and '24 or kind of more like '21 and '22?

Well, it's going to look more like '23 and '24 because it's going to be large. It will have health systems in it, which we only started in '23 and '24. It will have large independent groups who've been in their communities for decades. So I think it has all of those things, and it will be a mix of new states and existing states. One of the things in Class '24 is you've got substantial growth in existing geographies in which you've already got payer contracts, you got clinical programs set up infrastructure team, et cetera. And so the class of '25 will have a mix of that included as well as new states where we're setting up.

Got it. And just going back to the -- you mentioned the 3 markets that you were seeing the higher cost. I assume that's the higher medical cost. I assume the OTC cards and all that is more widespread. Or is that both?

Well, it is more widespread, but it's an issue in those markets as well. But the lion's share is more in core medical it's more in those areas that I talked about. Hence, the work that we're doing around things like ASC, better visibility.

Any specific, anything about those 3 regions that's kind of correlated specific regions of the country or types of groups or years of implementation payers, anything you can kind of point to?

Yes. So I mean it's -- I wouldn't say it's a geographic focus around those. I do think -- they are markets that have seen some large growth from a same geography standpoint and this variability that I talked about. They have -- their top quartile physicians are performing extremely well. And so I think there's an opportunity for us to really get after that variability as we talk about it and bring those bottom -- those lower-performing PCPs up. And it's a little bit more of a hands-on work with them, leveraging things like our regional pod structures, et cetera and just getting them better data.

Got it. So one of the things that I think has been an area of focus for investors has been the relative consistency of negative development in the business, right? I think everyone understands, I think most people understand that your business is going to have a lot more variability than a managed care company, right? Just the structure of it, right? Again, you're one step removed you've got revenue true-ups, right, membership true-ups that are going to happen, right, in addition to cost true-ups. So I think everyone understands that you're going to have more volatility. I think everyone would like to at least see 2-way volatility rather than be consistently negative. So you talked about this in the slides trying to set up a year-end kind of reserve that or kind of medical claims liability that is going to be more reflective and less likely to generate negative development in '24 or like it has in '22 and '23. Can you expand on that?

Yes, absolutely. It's a really important question. And you're right on. I mean, historically, particularly this year, the issues that we have are driven by that very high complexity that we have with payers and the extent to whether we're getting really good, timely, accurate information from them or not. That obviously was a problem coming into this year where, of course, we came into this year thinking that we were appropriately reserved and we're -- and just really uncovering as we came into this year, a number of data gaps, line spots, just issues that we had to go back and work with payers. So one of the things that we've done this year is we've done a couple of things. One is, we've made a very big concerted effort both investing in some very knowledgeable expert resources that we've brought in and then organizing better to work more closely with the payers to say, we need -- where we can, we need to get better information and we need to identify to you where those gaps have been in the past and try to close those gaps. And we just need to be having really good conversations with the payers about where they think costs are going. So we've tried to implement all those things. And what that results in, I believe, is our ability to account for all of that complexity and hopefully get to an actuarial assessment of where our costs are going that we're going to essentially reserve to that hopefully is more. Then when you do that, of course, when you do an actuarial assessment, you also have a range of potential outcome that you're going to have. It's not going to be an endpoint. And then the second idea is and what we're doing is, hey, we should always be reserving in the higher end of that range. So hopefully, we have better data to begin with, so we're getting to a more accurate range, more accurate assessment. And then by always taking a velocity of reserving in the high end of that range we ought to put it in a much better position [indiscernible]. Because you're right, I mean, the complexity that we have just this year before we even expand into next year around 100 different payer provider contracts out there and data comes in at that level. That's a feature right of our model that we're able to do that, that we're able to grow and have multipayer risk in every one of these markets. We have to put these other things in place, obviously, to overcome the potential downside of that, which [indiscernible].

And Justin, on this was in your question, but I do think it is important for us to look at the net development of our business because we do have true-ups that will come with both membership and revenue in claims. That's why we try to get you and others to look at the net development.

Yes. Yes. Of course, in our quarterly filings, where we, of course, require them, we should give the visibility to what just the claims side of that development is, but the volatility works in both directions. And we have actually had positive, if you want to call it, that volatility on the revenue side. So our methodology -- the information you're getting to forecast revenue is different. It's largely around how you're assessing members, which we get pretty good information about and then how you're forecasting your risk adjustment or essentially that revenue top off that's going to come in. And we are -- that is much more within our control and we have much better information. So we're probably, as it turns out a little bit conservative on the revenue [indiscernible] obviously, not enough reserve on the claim side. And net of that is that $22 million that we talked about [indiscernible].

So I think that's -- the focus I would agree should be on the net. Honestly, I think it has been on the net. And so the in terms of your ability, like if I think about this, if trend remains at July levels and you've got $50 million set up, you should come out with a decent amount of conservatism, right? If you just kind of leave it there.

Yes. I mean our objective is to say, hey, let's put enough into our outlook for the year to make sure that we're covering those -- that potential that we could see actually higher utilization. So that's why we picked that right now, we picked that original 60 up to 90 when we saw what was actually coming through Q2. So that's what gets you that kind of $10 million incremental versus the asset that we would have given to go. But the idea is, yes, that we're going to end the year with that's going to be -- put us in a very good position to be appropriately accrued for the year. You can look at it on a net basis, but there's no reason why we would expect to have. We shouldn't be going into a year with a significant underaccrual of our cost stand-alone either.

So that's where I kind of wanted to kind of finish up there is just the -- my concern is that there is your focus on trend, right, and it's kind of out of control a bit right now. We just -- the visibility, I should say, is limited, right, for the plans and for you. So you've got this $10 million, which does feel like it's conservative. But if you need it, are you willing to just like -- I think what investors would like to see is to just have you clear the decks on this reserve issue. So even if it's -- if trend continues at June levels, which is what you've kind of assumed.

Higher than that, but yes.

Theoretically, you would need to top off reserves. Additionally, to make sure that you're not going to have negative development next year?

Yes. So we've actually -- so that right now, the -- what we put into our balance of the year forecast and for balance of the year, let's just call it, the months that we don't have like 90% plus [indiscernible]. We've actually said we're going to put -- we put about $10 million on to original asset. That's a little bit higher than what we experienced.

2 million a month.

Yes, exactly. So it's kind of the average of the 3 months of May June, July, I guess the way to think about that. So when we get to the end of the year, we'll obviously watch exactly how we do these months develop. And when we close at the end of the year, we're going to use all of those techniques that I talked about to try to get the most appropriate reserves that we -- the idea is certainly not to go through the end of the year [indiscernible].

Our intent is to leave this year at the high end of our...

The high end of your range of what?

Of the actuarial assessment.

Okay. Versus your 50-50, you're going to pick something that's much more conservative.

Well, yes. So 2 things, just a good way to close that concept out is, first of all, the 50-50, you want to make sure it's 50-50. That's the first thing that I talked about. How do you get better information, how do you understand where your gaps have been in the future? How do you try to close those clients back up. And when we finished last year thinking we had a 50-50 estimate and obviously, we were off for a lot of those reps. So the first part is you've got a fix that. Then when you feel like you have a more accurate 50-50 accrued to the high end of that range, [indiscernible].

Got it. I'd like if I were to just ask for like a relative positioning there. Where were you -- where were you reserved at the end of 2022 relative to your best guess and where do you think you'll be in 2023?

We definitely thought when we ended '22, we were above the midpoint of our actuarial range. The problem was that we just ended up later saying, obviously, now we've identified a bunch of gaps that we had in that. And some of them were just when we talked about that even coming through the first and second quarter call, some of them were -- there's a lot of different reasons and a lot of different payers. But more lately, a pretty good chunk of that issue was around just we didn't have good reporting on supplemental benefits last year, that was the bulk problem.

All right. Well, look, guys, I want to thank you for your time. Any last remarks before we let you go.

Well, I mean I just conclude with, I think we have a really strong business I think when you look at the number of physicians and groups, patients we're bringing on, it's working very well. When you look to '24 and beyond strength this membership and in Reach, I think, is very important. On the medical margin side, we've tried to be very transparent and I think we've got levers to address that.

All right. Thanks, guys. I really appreciate you being here with us.

Thanks, Justin.