Humana Inc. (HUM) Earnings Call Transcript & Summary

Awesome. Thanks so much, everyone. So I'm Steve Baxter, the health care services analyst here at Wells Fargo. Really pleased to have Humana with us this morning. As you likely know, Humana is one of the largest MA companies and also is growing footprint in Medicaid and services. From the company, we have CFO, Susan Diamond. Thanks again for being here. Do you want to make any opening remarks or we just hop right into it?

Nothing. We will go right in.

Great. Okay. So yesterday, you put out an 8-K reaffirming guidance. There's obviously been a lot of focus on utilization dynamics that you discussed on the second quarter call. What update can you provide us on utilization and where things stand as you've had a chance to further study the second quarter and the trends that you've seen since?

Sure. So as we mentioned on our second quarter call, within the quarter, starting in the back half of May, we did see a higher level of inpatient activity relative to our initial expectations. Now when we study overall hospitalization, hospital utilization broadly, the overall number of events in the hospital was consistent with what we expected. Just a higher share of those events were billed as an inpatient event versus an observation stay. So we saw a higher level of inpatient, but then a lower level of observations relative to what we had expected given what we saw coming out of the first quarter, recognizing we're continuing to adapt to the human record changes. As we stated that, our belief is that, that is just an ongoing sort of learning and reaction to the two-midnight rule as providers submit initial authorization requests, some of those get denied. They have the opportunity to appeal. And as they learn through that process, I think they're probably improving some documentation and learning how to better appropriate document the level of care that they believe is needed, that's resulting in a higher level of authorizations being previously an inpatient stay. So we saw really in the back half of May and then through the second quarter with variance that we experienced very consistent. And so a lot of the questions we were getting in the second quarter is how do you know that that's stabilized or are you going to continue to see some increasing elevated levels. As you guys know, we get real-time authorization data for inpatient events through the month of August. We continue see a variance that was in line with expectations that we set as at the second quarter. So it has stabilized from our perspective since the back half of May. We continue to launch overall inpatient authorization levels, observation levels, appeal rates, overturn rates and all, I would say, very stable and consistent with what we had expected coming out of the second quarter and contemplated in our guidance. As you saw yesterday, we reaffirmed guidance for the full year. So continue to feel good about that. While the geography is a little bit different, some of that elevated cost pressure that we'll see as a result of the inpatient event. We have other offsetting good guidance over the course of the year that, again, continue to still get about '24 as well as '25.

Got it. And if we -- I'll come back to the two-midnight rule dynamic in a second. But if you think about everything else going on, cost-wise, I guess, anything to call out with how the second quarter reserves have developed utilization beyond inpatients in the second quarter? I know your visibility is a little bit slower there.

Yes. So we will open claims for unit cost and then non inpatient. And I would say with July claim development and then preliminary runs that we do through the month of August, nothing of concern. And I would say generally in line with expectations, a little bit of positivity that we continue to see on inpatient unit costs. I think you mentioned last time, we've seen some sort of higher trends on things like chemo and then on inpatient space. But some of the orthopedic pressure that you saw last year that has not been problematic this year. That seems to be stabilized. So I would say, all in, feel good about the July reserve development and what we've seen is sort of preliminary for audit.

Okay. And it's good to hear that the two-midnight rule dynamic feels like it's stabilized a bit for you. I guess as you step back and studied it and maybe tried to look at it at a either geographical level or system level or individual hospital level. I guess what trends have you found that maybe give you confidence that there's not further pressure that could emerge in the back half of the year. I think one thing that's come out is that hospitals seem to have a certain point of view that there's more opportunity here. I guess how do you guys get comfortable with that dynamic?

Yes. So we've been watching the returns closely, as you can imagine, as we see reports like the provider systems to the degree they're saying something different, we always study that additionally. I'd say we've been reviewing the -- again, the absolute levels of inpatient authorization requests, the appeal rates, the denial rates, and overturn rates. By not only in the aggregate, but also geographically, by plan, by facility and we're not seeing anything that gives us any concern that we see pockets of sort of continued elevated trend. It's relatively stable across any of those cuts. One reason we think there may be a difference is in the commentary, you might remember in the first quarter we mentioned that we were seeing a higher level of appeal rates than we had seen historically and might have expected, especially when you think about more absolute authorization requests being approved, so to see higher appeal was not what we had initially expected. So there was some question about, is that just a pull forward of what you might have seen over the tail period because there is a fairly long period of time that providers can appeal a denial. We took the position that, that was just going to be an absolute higher level of appeals in our thinking and that we see overturn rates comparable to historical. So we allowed for that today, we would say that, that is likely what we're seeing. We haven't seen that it suggested it was just a pull forward. So some of the hospital commentary, maybe just a reflection of them not having anticipated that as early as we did. But our estimates to contemplate an absolute higher level of appeals, which then will result in comparable overturn rate as we've seen historically. And it just may be a timing difference in terms of when we contemplated that versus when providers may have.

Do you think you have opportunities to improve your avoidance rate in the back half of the year or into next year? Like if this emerges as something that over time is a gradual source of pressure? Or do you think it kind of moves into the dynamic of hospitals are always going to be trying to maybe push things incrementally and you guys have a responsibility to be able to deal with that?

Yes. As far as two-midnight rule itself, as we mentioned initially, at the start of the year, our avoidance rates were lower than we expected, but over the course of the quarter, every week we saw improvement and we continue to train our clinicians based on what we were seeing, frankly, do some provider training as well to make sure they were interpreting it correctly. So when we exited the first quarter, our avoidance rates were in line with what we had expected and that has been very stable ever since very much in line. We've continued to do detailed chart audits. We actually recently went through a CMS audit on the two-midnight rule changes and everything continues to validate that our clinicians are making appropriate determinations on whether inpatient level of care is, in fact, needed. So I would say on the margins, we may see some changes. We're continuing to look at things like high-complexity cases and routing those to specific clinicians that have deep expertise in those areas, but those are not high volume sort of events. So we'll continue to work to improve it. As you said, I think hospitals will certainly continue work. But I'd say the bolus of sort of the implementation, we've each addressed. I think over time, we'll continue to more broadly try to impact trend. I would say that going in this time last year as we're negotiating with providers, they largely believe two-midnight rule would have no positive impact in terms of their results. So that was hard to use as a point of leverage. Now that we've actually very clearly seen the results, it does give, I think, our contractors more information, as they more broadly work with hospital systems around the various components that hopefully will allow us to take some ground in terms of overall medical cost management. But I'd say that the avoidance rate itself, I think, are sort of largely in line with what we would expect.

Got it. Okay. And then as we think about from the second half -- or excuse me, the second quarter MLR the progression that you guys have in the insurance business, to the third quarter, I think gravitating to somewhere around like 90.5%. I think the progression from Q2 to Q3 to Q4 looks a little bit different than maybe it has in previous years. What are the key factors that are impacting the progression from Q2 to the balance of the year that we should think about?

Yes. So -- and there have been a lot of questions trying to assess just the seasonality and some people trying to go all the way to pre-COVID, which we would say is not terribly productive just because of the business mix changes and other things that have happened since then. In terms of this year's seasonality, it does look different than you would see typically. And some of that's just a function of the overall contribution of insurance pretax to the overall enterprise is going to create a different seasonality to EPS pattern. Normally, you do see some increased MLR in the back half of the year in the individual MA book just because the seasonality of revenues associated with MA, where over the course of the year as you continue to enroll, say, agents and see the impact of deceased members, which have much higher than average risk scores. You tend to see negative seasonality on the revenue side, which drives some of that increase to MLR in the back half. You've got the obvious respiratory season in the fourth quarter, that's going to drive naturally higher MLRs. And then there's a couple of more unique things the last couple of years. One is some changes that were made by CMS on D-SNPs. They changed the way cost share applies to the maximum out-of-pocket calculation, we knew this in terms of pricing. So it's contemplated, but it does create a different seasonality pattern where duals will now largely hit the loop where in the past, they did not. So it creates some more planned liability in the back half of the year. And then one of the other larger things is the increase in supplemental benefits. And so those do natural see a higher utilization level in the back half of the year, particularly the fourth quarter. Those benefits cannot be rolled over to the next year. So you typically do see a higher level of utilization of things like OTC cards, fee benefit cards and even some things like dental services, we'll see in the fourth quarter. We were anticipating an even higher level of utilization in some of those services in the fourth quarter of '24, just recognizing the benefit changes we've made for '25. And if people get visibility to that, knowing that those benefits will be reduced, we do anticipate, again, an even further elevated use with some of those benefits recognizing that they won't have the same capacity sort of free utilization in 2025.

And when we think about the emergence of some of the incremental costs this year, some of the offsets, I think you guys have talked about higher levels of risk adjustment, higher levels of claims recovery. I guess how should we think about how you're approaching booking those things this year and whether this is a durable offset into next year to potentially offset this higher level of cost as we think about the margins that you're trying to get to in MA next year?

Sure. And it's creating the seasonality of MLR for '24 as well, where we have seen higher levels of prior year development to claims and then favorable MRA, most of which was related to final 2023 as well. When we report prior year development for you guys externally, it's just the claims side, the revenue doesn't get incorporated into that, but we share that in our actual results. So on the MRA, if you remember, we had a very large enrollment cohort in 2023. We don't have visibility to the 2022 claims, which will determine their '23 reimbursement. So we make assumptions based on historical patterns. And ultimately, what we saw as the final '23 payment reflected a higher risk score for that new bolus of '23 members than we had anticipated, which was favorable. That had not been contemplated in our bids and so it was incrementally positive for '25 as well. And for '23, it gets recorded sort of in the first quarter when that payment is received. When we think about '24, now that we have visibility to those final restores, we would anticipate that in our normal course sort of booking across the length. So the seasonality will look a little bit different for '25 than it did in '24 because we didn't have visibility to it. On the more normal course prior year development, some of that is just arguably claims conservatism in our year-end reserves, but ultimately unwound. But there is a component that we can isolate that is related to activity that we do. So think of postpaid claim audits. The benefit that we've seen from some of those audits has been favorable relative to what we expected, and that also was not something contemplated in the '25 bids and so it's something that you'll see recur again next year. That will look more like typical PPD and seasonality in terms of more front half of year weighted, and that's a function of our reserving practices where we never assume claims are more than 100% complete. When in reality, they are at the end of the year and so you plan for that positive development numbers in the first half of the following year.

Okay. And then when I think about the pressure that you're offsetting this year, the inpatient side, I think, has taken a little bit of time to ramp and have the true tubing level activity kind of emerge. How do we think about kind of looking at like a partial year of this inpatient effect versus annualizing that into a full year impact next year?

Yes. So some people assume that it's back half of May, it's not avoided, when if you remember that first quarter, we saw some pressure as well, just because those avoidance rates were not on the levels that we had expected until we were exiting the first quarter. So there was some inherent pressure in the first quarter just because it took a period of time for the avoidance rates to get to the expected levels. So it's still not a full 12 months, which is fair, but it's not sort of 7.5 months either. So we would say that pressure as we look at sort of what we expect at the time of bids and what we are now anticipating, we've accounted for that higher level of ultimate net cost for inpatient activity. But then also looking at risk scores, the prior year development and all of the other variables that would have been contemplating at the time have been all in, when we consider all of those variables, we continue to feel good about the MLR that we were targeting for 2025 in our bids and then the earnings progression that would have supported. So while the geography is a little bit different, continue to feel good about the totality of assumptions that went into our bids and that based on what we're seeing in emerging experience that there's sufficient offsets such should we continue to feel good about what we're targeting in pricing.

Okay. That makes sense. And I noted on the second quarter call, I think it was still probably early, so maybe you guys weren't are as specific as you might able to be now or in the coming months, but you're talking about exiting a few hundred thousand members in MA. Can you talk a little bit more if you can specifically about maybe the sizing of that? How should we think about maybe the gross number of members that you'd exit, what you might hope to retain and then kind of the financial implications of either lost members or retained members and how to think about that in terms of your 2025 earnings growth?

Yes. So as you said, we commented we expect to lose a few hundred thousand members. That is a net number. So the plan exit numbers we'll share today. So what you'll see when the landscape does come out is the number of members impacted by a plane exit is about 560,000. So think of that as roughly 10% of our individual MA membership base. That's going to come in 2 forms. One -- and wanted to be clear is there are very few places, I think it's literally 3 counties where we will have no presence at all. So completely insignificant. So our coverage and footprint will remain large the same. When we did an exit, it was either a full plan exit where we exit a plan, but there are other Humana plans available in the market or we took a plan that may be covered x number of counties and we're now going to cover x-minus 2. So we trim the number of counties that they cover. But again, nearly all of those members have other options. So as we think about those members impacted and how the plans that are left behind are positioned, the assumption is that we'll retain, call it, roughly half of those. If you do the math you get to this sort of a few hundred thousand loss of members. That is not atypical from what we've seen historically when we do plan exits. And as we said before, for the rest of the book, not impacted by plan exits, they're going to seem roughly flat in terms of what we would expect. And that was something that we couldn't see better results than that and we couldn't grow. It's just we wanted to make sure with the earnings progression that we are trying to achieve that it was not dependent on driving industry average growth on the rest of the book. And if we are able to drive membership growth, and that's incrementally positive. And so you can think of those members impacted by exit. If we chose to exit, you can assume it was unprofitable from a contribution margin standpoint. And the team had a perspective that within a reasonable period of time, think of that as 2-ish years, they couldn't reasonably get it to a contributing level of performance. And so in those cases, they would decide to exit, they could have launched a new plan in response to that, which they felt like could better optimize and have more flexibility around or they feel comfortable with what other plans already existed in the market. So the exit itself is positive in the sense that those claims were not contributing. And so just exiting even if we don't retain the members, it's positive. If we do then ultimately retain more of those members that's incrementally positive because the plan choice is left behind are priced in such a way that they will be positively contributing. So we kind of get it on both sides. So we do acknowledge that membership growth in '25 is, there is a lot of range of potential outcomes. It also is going to have a disproportionate impact on the absolute level of earnings that will drive growth next year. Just we feel good about the margin percentage and the pricing decisions we've made, but how we ultimately see the membership play out in terms of how many members we retain, what plan choices do they make in terms of the plan options available will be an important consideration. And to appropriately assess that, we'll need to see this whole landscape that they must release in early October, so that on the third call, we'll be able to give you guys some color or commentary in terms of how we feel like the plans are positioned, some of the early reads from AEP, mostly on the enrollment side, broker feedback, and some of those things as respect to ultimate positioning that we see.

Got it. I mean that makes sense. I appreciate the color. So we think about the difference between the membership that you feel like might just be lost to the market versus what you might ultimately keep. It feels like there should be a pretty wide range and profitability assumptions related to that. Like, obviously, I think like you're dealing with all the negative loss operating leverage of lives that leave the plan completely versus transitioning into plans that might not be overwhelmingly profitable. But from a contribution margin basis, it would seem to be a much better alternative than losing the members entirely. I guess how should we think about the spread between those populations?

Yes. I mean when we think about every member that we are able to retain is positive because we haven't left any -- there are a few plans that may still be negative contributions. That's more a function of just maturity and they may have a disproportionate number of agents because there are newer plans and those things. But structurally, we feel like are well positioned. So every member that we retained is positive relative to the performance that we've seen historically. And I would say, across the industry, while the overall performance of the book might vary, I would say all of the large competitors have a similar dynamic where we all have launched sort of higher benefit value, PPOs, typically plans that have been very pressured given V28 and the trend we've seen. So you've seen all of us talk about sort of exiting more plans in typical and disproportionate benefit changes. I do think those will be concentrated in those plans that we all have. They are low margin, if not negative margin. And so that will create some disruption. We also have plans that are very well performing. D-SNPs is an example in general, are higher performing and have weathered some of this better. So I think all of us are going to work hard to address the underperforming plans and work to protect the higher-performing plans. And so exactly how people did that is what we need to see, what exact benefits and changes that people make. But I think you'll see some probably even investment level in those higher-performing plans where we weren't going to take MAX PDC and might actually bring the margin down a little bit because those plans are meeting, if not exceeding, sort of the target margin and have the potential to grow. So we want to retain those members and continue to grow while addressing some of these sort of financially troubled plans. And again, we all have those. And so it's those members as they start to move around, how do the resultant plans sort of fits well with one another and how do the benefits resonate with consumers and brokers. We did a ton of research that inform the benefit changes that we were going to make, recognizing it was a higher level than the industry had been managed in the last couple of years. So we feel really good about the decisions we made, but we're going to have to give the full detail to fully assess it.

Got it. Okay. So in terms of the kind of the flat expectation for the rest of the book, I guess how are you thinking about that across the different populations? I think you alluded to a little bit maybe D-SNP versus non D-SNP. I guess I'd love to understand what informs that for you. And then to the extent, I believe you started to get some feedback from brokers maybe during the month of August as you introduced what your plan designs look like for 2025. I guess what have you heard from them? And any kind of competitive intelligence that starting to emerge as well would be great.

Yes. So as I said, we were very focused on the sort of current profitability of the plan to try to protect higher-performing plans in terms of preserving membership and then positioning them to continue to grow while addressing the lower-performing. You can think of that in some cases, there are certain products like D-SNPs, some of our sort of honor plans do quite well. So those would be some specific areas that you would see us work really hard to protect, certain geographies. Typically those that are higher-risk areas, South Florida to make an example where, again, still continue a very rich benefit value and do well and so we would want to protect those. So broadly, it was a profitability sort of matrix that we would value as well as just the opportunity for further growth. So combining those 2 factors. I would say, though, for risk providers recognize that many of them are under pressure. Some of them saw pressure disproportionate from the supplemental benefit investments where dental in particular, where you saw some of that investment was more concentrated in risk penetrated markets. That was a point of frustration for providers and so certainly worked very hard, and we're very focused on those plans and mindful of the benefit changes we were making, feel confident that the risk providers would be satisfied with the level of changes that we made and the resulting impacts to their surplus. So that is something we're very curious to see. I think all of the plans have been -- were under pressure to do that. And so it will be very interesting to see if someone didn't take appropriate action, how does the risk provider react to that? And are they going to continue to support plans that are not financially viable. And that was certainly a priority for us to make sure that we were designing the plans in a way that would supportive of the risk parameters.

Okay. And then to the extent that brokers are starting to give some feedback on how they're going to deal with the idea that benefits industry-wide seems like it's going to be compressing some, I guess, what's the initial feedback? Like if you're in a plan that has benefits coming down, but it's still the most attractive plan in the market. Like do you think those plans still retain a lot of that membership? Or how are you guys thinking about the potential for switching in 2025?

Yes. And to your earlier question, so I'd say, we've all done sort of our broker meetings. I would say less detail has been shared generally this year than we've seen in the past, just recognizing, I think everybody anticipated there might be more last-minute benefit changes or solving direct subsidy things in historical. So it is later to come out. And generally, I'd say less specific than we've seen historically, which is why the full landscape will be so important. But I would say, basically what we have seen so far in broker reaction has been positive in terms of sort of the changes we made and how we approach the plan design. So that's been good. We'll have to see ultimately how it looks, but so far that's been positive. I think generally people are anticipating a higher level of disruption and trying to prepare for that. It's still going to be a challenge when you think about -- it's a 52-day selling cycle and you're putting almost every Medicare beneficiary through. So I do think it will be challenged. I think the plan exits in particular, brokers will be focused on because in the absence of the plan change, they're going to lose the renewal commission on that. So I think they will definitely be very focused on getting in front of those beneficiaries and having a conversation about what other options are available. To your point, there's going to be a lot of frustration over the absolute level of changes across the board. But they will go through the process. And even if they are frustrated by the level of benefit changes we made, if, in fact, that plan is still the best option they will stay typically in the plan, and that's what we've seen historically in cases where we plan to make some of these types of changes. But we recognize that it still will be a frustration. We try to support the broker as we provide information to them on the members we believe will be most impacted and most susceptible to sort of this enrollment based on the changes we've made as well as their own utilization patterns. So that they can get a front -- in front of that. We do a lot of support where we're willing to take on some of that administrative burden to say, we're happy to sort of facilitate a needs assessment and facilitate a plan change were necessary and still protect you as the writing agent. So we can free up capacity for them to service sort of new enrollments. So there's a variety of things that we'll do. We are staffing for a higher level of inbound activity across our service centers and ourselves, call centers, given the disruption we expect, so we're trying to plan for it, but I do think it will be a bit -- it will be a challenge for the industry.

Okay. And beyond the enrollment dynamics and benefit design, another thing we'll be watching for in October is obviously the updated Star ratings for 2026. There's been a bunch of methodology changes and kind of walk backs over the past couple of years. How are you guys thinking about STARS opportunity? I know you're a top-performing plan, so your ability to improve is obviously limited. But what are you watching on STARS? What are you focused on?

Yes. We didn't have some of the same impacts as last year, one to two, and some of those types of things. But I would say just in general, the program is challenging in terms of just the basic structure, right? Straight down the curve. It's not weighted. And so it can be more difficult to predict. So I would just say we continue to be focused on it. It is early right, CMS has not released the information as such so we don't have visibility to the thresholds just yet. So as you said, it will be early October when we get the information the same time as others. And so we'll certainly comment in. But I would say it continues to be a focus. As you said, we continue to be proud of the work that we have done, we are very high performing. So lot of days, it feels like there's only one way to go, right, just because when you're 94%, I don't know that we ever get to 100%. So continue to focus on across the enterprise and work to improve all of those activities, but too early unfortunately to share any details.

And as we think about some of the planned exits and members moving around, how does that impact STARS or not? Is that a watch area?

It will imply to '26, right? So any enrollment activity will be more for '27. Interestingly enough, plan exits in of themselves do not have an impact on the STARS because the plan won't exist anymore. There can be an impact if a member enrolled in a plan, enrolls in another Humana plan under the same contract. So if we do retain that member, then the fact that they plan change does go into some of the calculations like there's a retention calculation. Plan changes have always gone into that calculation which, frankly, we've been a little bit frustrated by it because we said, look, we shouldn't be disadvantaged because a member decides to find another plan that better suits their needs as their needs change. But nonetheless, that's the way the calc works. So we are less where we have seen a high level of plan change activity in normal course. And so I would say the bar is already fairly high. If some of that ends up being disenrollment versus claims change they count the same. And then again, the fact that the STARS scores are weighed on a curve, we would expect broadly, the industry is going to see an impact to the thresholds, would likely move on that measure itself. But specifically the plan exits, if the member disenrolls and doesn't re-enroll then it actually typically has no impact to measure.

Got it. Another big focus area for people is obviously the Part D redesign. The company has historically made some trade-offs in terms of the underwriting margin versus benefits to the pharmacy segment. As you think about the incremental risk and premiums that you're taking on, how are you thinking about, I guess, the range of outcomes related to that business? . And eventually, how you want to think about underwriting those incremental premiums? It seems like they kind of have to carry a reasonable margin to justify the capital and the risk, but how are you guys thinking about the margins on the incremental from here?

Yes. And so historically, we'd say we've taken a more whole case underwriting approach in terms of the health plan as well as the pharmacy benefit across the PBM and then the dispensing pharmacy. And historically, frankly, the margin has all been on the pharmacy side, the health plan like in some cases, like the value plan has even been negative in some cases, but in the entirety positive. I think in light of the IRA and the additional risk that we're taking on, I do think, generally, I would expect the plans priced for higher risk margin and take a more conservative approach just to the underlying claims assumptions, particularly things like the pipeline of emerging drugs and the pace at which they may get approved, the cost at which they may launch the adoption rates, I think you'll see people take a much more conservative approach just because the liability and the exposure is so much larger. We continue to advocate with CMS. I would argue it's the way it's designed is not a great way to do it. When you think about Part B, they've introduced a significant cost policy, which helps mitigate some of the potential exposure, which I think is the right thing to do because otherwise, we're pricing for a lot of risk margin that, by and large, probably hopefully won't be needed. And so just ultimately would result in margin. So I think over time, we're going to have to figure out a way with CMS to make more sense of sort of the exposure and how to manage it. So I would say, in general, I think the industry will price to higher margin because of the risk that they're taking on. I think for '25, given all the significant changes with the move being implemented, we're going to -- you should assume that we will take a very conservative posture in terms of what we expect from that business in terms of expectations, and then we'll just have to see how it ultimately plays out. I would say the demonstration, which only applies to standalone Part D, obviously, signals that in the absence of that, the premium increases would have been quite large because of the IRA. There are some reasons that stand-alone Part D will see more exposure than MA. Some of that's dual concentration, which they do tend to use specialty drugs at a higher rate and some other reasons. But I would say the demo will be interesting just to see how that plays out, not applying to Part D initially, but hopefully, the learnings from that may cause them to think differently about how they structure Part D as well.

Okay. And I guess, how are you thinking about how -- why the range of outcomes is for the first year to have a program like this?

I mean, actually, they're wide, particularly because -- I'd say mostly because we strove for stability in our book and that was even before the demo, and is on stand-alone Part D, that was even before the demo. The risk that you are managing is one a lot of movement within your existing speciality. You want to attract a bunch of more specialty users, you want to run off a bunch of lower utilizers because the way that the IRA impacted the results as you used to get sort of for every catastrophic claim you got, you must pay 80% of that. Like you didn't have to guess what it was going to be, they would just pay it. So now you've got to assume how many of those you have, you're going to get all those dollars shifted into direct subsidy. So lower utilizing members become much more profitable and social utilizers become much more unprofitable. So the balanced risk pool is what's most important. So we strove for stability. The demo further helps hopefully create that stability. So I would say you've got that general risk and move, we'll have to see if it leads to some increased utilization, there are some other data points we could use across group MA, which has some benefits and some other things to inform how we thought about that. But we've been mindful of that. And then it's really the pipeline of drugs. Like do you see accelerated approval for some of these things. That is a bigger area of risk. The risk order protection they introduced on PDP will help minimize the downside risk. And as we think over time, some of those things will be important to just manage the absolute level of premium in particular that we will see.

Okay. Got it. And then just to switch to Medicaid for a minute or so. I know you guys have a lot of moving parts, folks very concentrated in Florida, which I think has still been performing reasonably well. You also have a lot of start-ups because you've done really well in terms of RFPs and organic growth there, I guess on the moving parts of Medicaid, and how do we think about Medicaid earnings progression playing out for the next couple of years?

Yes. So -- and I know there's been a lot of discussion about acuity, with redeterminations. And I think we probably took a little bit more conservative approach just at the outset of the year. So as a respect to redeterminations, we would say that the membership in the acuity are largely in line with what we expected. Some of the ratings we called out Florida specifically in our second quarter commentary, if we were going to see pressure, that's where you'd see it given the concentration in that book. And as we said, it was running slightly favorable relative to expectations, which is good. And some of the newer state implementations, we have seen some discrete areas of pressure consistent with others in the space. And so Oklahoma has seen some pharmacy pressure, there's risk orders that kick in, so mitigate some of that, but certainly some rate correction for the state to do. Kentucky has seen some behavioral pressure, which is a function of rating as well as limitations that they put on our ability to manage it. So we've been working very closely with the state on both of those issues. So again, some discrete medical cost pressure, but we have seen responsiveness from the states in terms of rating action even early. So we would say, in general, the totality of the book we feel good about in terms of the '24 performance. Some of the rate action we have seen will sort of improve in the back half of the year relative to what it would have been, and then we would expect some additional rate improvement next year such that you'll continue to see improvement in that book. Just coming off of rate determinations and then some of those states that were recently implemented mature because it usually takes 2 to 3 years to sort of mature the margin profile. So we do expect to see positive sort of improvement across that book in the next couple of years, assuming a continued sort of reasonable pace of new state enrollment.

Okay. Got it. And if we think about like the totality of all of that for 2024, are you guys expecting to make money in Medicaid this year? Or generally, what's the level of performance?

No. So still, given the number of new state exchanges that we've done and then the sort of adjustments you're going to have just from the higher margin contribution during the waiver period to now, still not in a positive contribution to say yet and that just over the next couple of years, as those recently implemented states mature, then we get to the margin positive contribution overall. But individual states at maturity are contributing positively to growth.

And maybe last question just on the CenterWell and the primary care business there. I guess how do we think about the kind of the earnings trajectory of that business? Obviously, you have the MA industry seeming to have the unit going through broad repricing cycle, there's the interaction of V28 and the comments you previously made that you think you offset it over time, but obviously, in the early stage it is tough, what's the latest thinking on CenterWell on the primary care side?

Yes. So we see continue to feel good about the ability to mitigate it over time, as you said. The original assumption was that for V28 that we would see about 1/3 of that mitigated by plan design changes. We would say generally, we do not see that in the first year across the industry. So again, very anxious to see what benefit changes were made in '25 and hopefully some level of catch up. But we continue to think that's a reasonable assumption over the 3-year implementation period. And then all of the other initiatives that we believe will fully mitigate that are on track for us. So it will take the full 3 years. So by the time we get to '27, we feel like we will have fully mitigated that. And back on the original sort of J-curve trajectory that we laid out, the contribution was a little bit different. There's going to be some higher patient panel growth. Lower margin on an absolute dollar basis, still continue to feel good about that. So the team is on track. We'll have to see again what benefit changes were made across the agnostic book for '25. But I think given, again, all the commentary from providers, I think there'll be a lot of pressure to make the needed adjustment.

Okay. Well, thanks so much. Really appreciate the insight today. Thank you.

Yes. Absolutely, thanks for having us.