Re-evaluating DaVita
The Ozempic/etc (Semaglutide) threat, and some long-term thoughts
First, happy holidays! Merry Christmas to those who celebrate or who just enjoy the time off with friends and family. Happy new year too, as I may not have time to write until then!
Sorry this post has nothing to do with Christmas, by the way, and is a continuation of our first post about DaVita. We first talked about it here. Read that first if you haven’t, or you need a refresher. I go over the business’ basics there.
Now, here’s why I’ll be putting this Semaglutide problem to bed unless some really bad incremental information comes out.
Some new, new information
On its earnings call a few weeks ago, DaVita spent time on how Semaglutide impacts it.
Management did their own research and analysis to decide if any change to the company’s multi-year planning was needed. They’re rational and thorough, and looked at a variety of downside cases after speaking with industry consultants, medical practitioners, epidemiologists, and reviewing literature.
(Maybe you can see why I like this company’s management; plenty of firms I know of would just dismiss it and not even think carefully through the potential consequences, only to then be forced to try and play “catch-up” to avoid dying… like the cable networks’ CEOs in the face of Netflix. DaVita’s management is rational and is candid with the shareholder.)
Because it takes decades for chronic kidney disease (CKD) to progress to kidney failure (end-stage renal disease), anyone with early-stage CKD who takes Semaglutide won’t impact DVA’s ESRD patient volumes for 10-20 years or more. This part won’t matter much. I did a similar exercise back in 2019 when I originally bought the stock. I assumed diabetes was somehow magically cured and new ESRD patients stopped flowing into the system, causing the business to go into terminal decline. The impact to value was something like -50% (and ~30% downside vs. the price I paid), but there was no real evidence it would happen or play out this way, so the odds were low. DaVita’s time horizon and thinking about this threat matches that.
It’s the later-stage CKD people who matter, since some of them are going to get ESRD and go onto kidney dialysis in the next 5-10 years. Here, DaVita pointed out something I neglected to think about enough. I know most people with CKD pass away before their disease progresses to ESRD, meaning we die of something else before our kidneys fail, even if we have kidney disease. The main one is cardiovascular disease.1 Obesity (which drives all this stuff and which Semaglutide addresses) is also a contributing risk factor for cardiovascular disease. So I didn’t think about: reducing obesity in the population would also mean cardiovascular disease declines. Plenty of individuals have (or acquire) both CKD and cardiovascular disease. A bit morbid, but by not dying of something else, they will end up with ESRD and on dialysis. According to the company’s analysis, it fully offsets any volume decline. This is partly because a lot more people have and die of cardiovascular disease than of ESRD. Even excluding this and assuming high adoption of Semaglutide, DaVita calculates a 0.5%/yr volume growth headwind from the drug as fewer late-stage CKD patients end up on dialysis with ESRD. If the business’ volumes grow at 1% or 0% or 2%, it doesn’t matter to our investment returns, since we paid paid ~6-8x free cash flow for the stock.
Lastly, there’s the issue of treatment adherence. This one I found. Prime Therapeutics, a US pharmacy benefit manager2 (which has internal data on the flows of prescriptions) found that two-thirds of people on Semaglutide or similar quit within a year. Semaglutide’s adherence might be low because the drug is expensive, often not reimbursed by insurers, and even artificially low because there is a shortage limiting access. Nonetheless, although there’s hype now and many people want to try Semaglutide today, the “base rates” are already that many, if not most, of these people will quit (Prime Therapeutics found it was ~65% who quit in a year). Because Semaglutide acts like a synthetic hormone, stopping it and going back to your previous lifestyle means you also lose all the ground you gained. Even if somehow 30-50% of Americans took the drug, which would be huge, and ignoring the benefits from lower cardiovascular mortality above, because most of them will quit, the ultimate market penetration will be much lower. That might be a mid-single-digit headwind for DaVita over 10-15 years, for a business already likely to grow volumes in the 2% range, sending volumes to -3% or so. That won’t mean an investment loss for a business we bought at an 11-15% free cash flow yield — i.e., we paid $7 billion for a company that does up to $1 billion annually. It would just mean there’s no upside.
All this, once again, and no matter which way I cut it, points to Semaglutide not being much of an issue for our DaVita investment.
I think in our earlier post, the only important mistake I made was neglecting to think about how rivalry and internal competition among the insurers might cause them to reimburse the drugs even if it cost them more and wasn’t worth it initially, since it would mean a market share opportunity. Maybe they’d include it in higher-tier health plans to corporate customers (though that doesn’t seem widespread). At low market penetration, it would go unnoticed in the overall inflation in the cost of health insurance plans, since maybe 5% of those insured take the drug, and that cost is spread over the 100% of insured individuals and corporates paying for the plans.
Regardless, most of the evidence points to me (and anyone else) seeing ghosts with this one, and that it should be left alone. It’s probably why the stock fell to $75 on the fears, but has since ripped to $106.
If you want to think in terms of Bayes’ Theorem and “Bayesian updating”, hopefully the above (and the previous post on DaVita) is a good example of how. You just re-underwrite the investment thesis given the new information. New information comes in and you think about what it tells you, what its magnitude and probability is, and add that to the magnitude and probability of what you already know.
Longer-term stuff
The call had a couple other interesting points which reinforce the idea that DaVita has 3 major drivers over the next several years and which I have been twiddling my thumbs waiting for over the last 2 years as the company exits COVID.
A shift from fee-for-service to value-based payments
DaVita reduces the payers’ (insurers and the US government via Medicare) costs via better patient outcomes. For example, they end up in the hospital for acute care less often than competitors’ patients, which means the insurers are reimbursing less for those hospital visits.
The value DaVita should capture from this differential performance is getting more and more explicit every year. That’s because the overall US healthcare system is shifting from from fee-for-service to payments-for-value. Value-based contract agreements are being made bilaterally (one on one) between big healthcare providers and the insurers. This benefits payers, and the best-performing providers, at the cost of the worst-performing ones in both camps. The industry couldn’t do this before because both sides were fragmented: in the 2000s, there were lots of smaller insurers and lots of small dialysis providers. Not many big ones who made up most of the market share. It wasn’t feasible for each to negotiate hundreds of contracts. Now that both industries are much more consolidated, you can. So they are.
The contracts are loaded with legalese, but the main thing is: if the healthcare provider saves the insurer money compared to some benchmark, then the insurer pays the provider a portion of the savings and they share in it. That’s it. It doesn’t matter where the savings come from. If, for example, the average ESRD patient is costing an insurer $50,000 in hospital visits annually (dialysis patients end up in hospital ~4x/yr for several days at a time such as due to infection), but the average DaVita patient is going to the hospital only 3.5 times, then DaVita is saving the insurers $6,250. Under a value-based care contract, it gets paid some percentage of $6,250. It doesn’t matter that DaVita isn’t the hospital — it just matters that based on how they treat the patients, their patients end up in the hospital less, and hence saved the insurer some hospital costs. That extra payment for saving money comes on top of the regular dialysis treatment prices DaVita and any insurers already negotiated.
The second way things might work is that the contract downloads the full cost and risk of a group of patients off the insurer and onto the provider like DaVita. For taking on the risk, the insurer pays DaVita a fixed amount per patient per year (called “capitated payments”3). DaVita takes on the whole responsibility of telling the patient and the patient’s other doctors and healthcare providers what best to do, based on DaVita’s data. Now that they receive a fixed revenue but have all the control (and risk) of the patient’s cost, they’re incentivized to go about trying to save money. There’re loads of levers to pull. DaVita will negotiate better pricing for ESRD-specific drugs because they have 40% of the US ESRD market and hence their patients are some of the biggest end-users of some drugs. DaVita will also steer patients to certain doctors, healthcare facilities, and treatments. It also researches and changes its own dialysis treatment methods vs. what competitors do. Etc. Healthier people cost the healthcare system less money, so the whole thing becomes win-win for DaVita, the payers, and the patients.
Outside the dialysis industry, it will probably remain the responsibility of the insurers to steer the patients around to the best providers and recommend changes in treatments to doctors and such. They might download less risk/reward to others. That’s because the insurer has the deepest relationship with most patients (maybe other than your GP) and because it has the most data about them.
But… dialysis is sort of unique because the sub-population of ESRD patients is always at the dialysis clinic. You have to get treatment 2-4 times a week, for about 4 hours at a time, which gives a huge opportunity for the dialysis provider to build a relationship with the patient. The clinics’ nephrologists (kidney doctors) also have to co-ordinate a lot with patients’ other doctors for other treatments and such. DaVita ends up knowing more about its patients anyone else does (like the insurers), and also has a deeper relationship and more human touch-points. Think about it: they can talk to you like a human being, in the clinic, rather than trying to change your behavior with an email or text message; is a nurse you’ve been seeing 2 years or a text message more likely to get you to change your behavior for the better? Another reason I know DaVita knows a lot and has a lot of data on patients is that the insurers willingly share some of their internal data with DaVita, which it can integrate with its own data in an effort to try and optimize the patients’ treatments. The insurer wouldn’t do this if there was no value-add: it’s a big operational and competitive risk for them.
So in the dialysis business, it’s the clinical providers that are the best guys to bear the cost risk of the patients and to be rewarded for reducing costs by changing treatment for the better.
But first… you have to bake a pie before you make a contract about sharing the pie.
When I first bought the stock in 2019, I was aware this pie could exist in the future. I saw that (a) there was a cost advantage any payer had for working with DaVita and only DaVita, so DaVita should be able to get some of that, (b) all the incentives were in play for the insurers to want to make these contracts with guys like DaVita. For example, Medicare Advantage in the US is actually administered by the private insurers. The government pays them a fixed amount per person per year in exchange for offloading all the cost risk of those government-insured patients onto the commercial insurers (like UnitedHealth and Cigna). That makes it easy for the insurer to then turn around and re-contract all these costs and risks down onto the providers like DaVita and look for the ones who can save the most money. Then the insurer just gets a profit on the difference — on any money saved — between the two contracts. And I saw that (c) CMS, the government entity that administers Medicare, had already started testing this economic model with the providers with a small sample of patients.
This is the pie:
It doesn’t matter what the words like ESCO mean. The ESCO was the trial the providers did with CMS/Medicare. The results were that DaVita, in the orange bars, saved more money for the payers than the rest of the industry in the gray bars did. If the patients who go to DaVita clinics are costing the insurers $10,000 less annually than the insurers are forking over across the rest of the industry, there is now a $10,000/patient/yr pie for DaVita to go after.
Now the contracts are happening. Already, ~25% of DaVita’s patients are treated under some form of value-based or risk-sharing contract, in which DaVita is paid more for saving more (by making the patient healthier). The pie then gets split something like this:
So say the average patient costs $100,000 annually. (Yeah, like we said when we introduced DaVita, dialysis patients are not cheap). DaVita takes on the full risk of the patient and becomes responsible for all their healthcare costs — hospital visits, treatments for other diseases, drugs, family doctor visits, etc. They have to pay for all of it and tell the patient where to go and what to do. DaVita’s doing a bunch of analytics etc. to make it happen. If DaVita can get some savings, like the ~10% in trials, then they get paid $100,000 but the actual cost they’re paying to other healthcare and pharmaceutical providers is $90,000. Now we have our $10,000 pie. The payer keeps some of it, say $4,000, since the payer is partially investing in this, providing data to DaVita, and giving them permission to do this stuff. Of the remaining $6,000, some DaVita shares with all the doctors and such in the value chain as an incentive to improve patient health and cost. Then it costs DaVita some amount to run the program, given all the analytics and tracking they’re doing. What’s left is DaVita’s profit per patient under these programs, say ~$3,000, or ~$2,250K after tax. Win-win-win.
DaVita’s got over 200,000 patients. If you’re netting $2.25K/patient/year, you’re pulling in an incremental $400 million annually. DaVita usually gives lower numbers on the order of $150-300 million pre-tax longer-term. I don’t think anyone knows, or can know, exactly what the long-term profit outcome is going to be, but any outcome is material profit growth for a business doing $1 billion a year in free cash flow (excess cash profit) today. It’s a boon one way or another.
Today, DaVita is losing money in this business and expects to break even in 2025-2026. Yet, all the actors’ incentives are in place, DaVita has already shown it can add (then capture) value, and it’s now starting. In the dialysis niche, DaVita sits at the epicenter of this shift to payment-for-value. This was my “upside” case back in 2019 as I thought it looked very plausible. Today, the evidence says things are going that way, although it will be less profitable than the the math sitting in my original 2019 report on the company.
(By the way, notice this thing is independent of the Semaglutide thing, so if both (a) the impact from Semaglutide was really bad, and (b) DaVita successfully pulled this lever, the net is that even in most of the worst cases, the two offset each other, and in the better cases, value-based care is actually a bigger positive impact than the Semaglutide headwind. DaVita might lose patients, but the effect is blunted because its profit per patient goes up.)
A COVID echo
DaVita’s patients heavily skew to people who are elderly, lower income, and are very ill with multiple “co-morbidities” (e.g., many have multiple illnesses like being overweight, having diabetes, cardiovascular disease, and/or something else; most are also elderly). They take an average of ~11 medications. Unfortunately, dialysis patients are not at all healthy like most of us are.
This meant they were disproportionately hurt by the pandemic, passing away much more often than the average person who contracted COVID. In 2020, DaVita thinks it lost ~7,000 US patients vs. what it otherwise should have had. At the end of ‘19, it had ~207,000 US patients, meaning there was a ~3.5% volume headwind. During 2019, it was growing volume in the mid-2% range. In 2020 (where the pandemic impacted only part of the year), treatment volume was only up 1%. In 2021, DaVita again estimated it had 6,200 patients fewer than it should have, because of COVID-related mortalities (again, about ~3% of its patient base), and treatment volume was down 1.9%.
With the pandemic now (mostly) behind us, the company’s said COVID-driven mortality is down significantly today (maybe 1,500 or less annually), and so the underlying ~2% growth in patients should again begin to feed through. DaVita itself thinks that growth should be above 2% in the next few years. All this is because patients with ESRD were more likely to die of COVID than those with mid-late-stage-CKD, but some of those with CKD end up with ESRD and on dialysis. Right now, treatment volume growth is turning positive, from -0.2% in 2Q23 to +0.5% in 3Q23.
That’s going to be a boon to profit growth, which had fallen during COVID as (a) volumes fell and (b) the company deleveraged on its fixed costs like the cost of leasing the clinics, so there was also a lower percentage profit margin on lower revenue. This is now going into reverse and fixing itself.
The international business
Over the last 10 years, DaVita has been working on taking its model international. Kidney disease isn’t unique to America, and many other countries’ healthcare systems are both large and lack the incentives to make significant treatment improvements other than what medical practitioners learn from schooling, academic literature, and private sector sales staff coming to their office to pitch some new medical device or drug. DaVita, by contrast, has the scale to do kidney disease and treatment research on its own. DaVita brings incrementally more innovation to the table. DaVita’s international patients now outperform local benchmarks in every country the business operates, just as it does in the US. Mortality rates at its international clinics are down 20% since 2020.4 This business still isn’t material to profit today, but it shows the strength of DaVita’s value proposition to the healthcare system, which is its moat (we talked about how DaVita’s scale lets it invest in reducing cost and improving patient outcomes here). The international business was losing money when I bought the stock and now makes a ~5% margin as it has gained scale, vs. the 15-20% margin the US business makes. It serves ~46,000 patients internationally and is a small piece of a growing patient population. That business may be contributing $100 million annually in a few years from $50 million, adding a percentage point or two to profit growth.
Lastly, it’s worth mentioning DaVita was impacted by the tight US labor market, which is especially tight in healthcare, mainly nurses (many of whom burned out during the pandemic). Wage costs were rising much faster than DVA’s treatment payment rates, so its margins fell. Wage costs are now increasingly under control, and that’s been a margin tailwind for the business, too.
In all, the company’s setup over the next 3-6 years looks attractive to me, and it seems reasonable to me that DVA’s management can hit their 3-7% profit growth targets by pulling the levers above. There is already momentum. Without taking you down an Excel discounted cash flow model rabbit hole, a business growing 3-7% earning 15-25% on equity is worth more like 13-20 times free cash flow, vs. the 10x it trades at now, and the 7x I’d paid to buy a couple times. In the meantime, since DaVita spends all excess cash flows on share repurchases, it’ll continue buying in up to 10% of the company’s shares annually (at what look like attractive levels). In fact, part of the reason DaVita went through its analysis on Semaglutide is because it decided that it had repaid enough debt with its cash flow, and that now was an opportune time to flip back over to buying back stock after it fell. The company’s own management thinks like a shareholder would.
I know that may not be very high-precision for a few of you finance and investment professionals, but over the years I’ve noticed that honestly, your calculations only need to be roughly right. Then, there just needs to be a really big gap between the price and your math, and a really small gap between the price and the downside. Once it is that obvious, you’re done. Then, just focus on knowing the business and your thesis as best you can, and monitor what’s going on.
When Charlie Munger bought Costco, he didn’t conclude “yes, this stock has 31.2532985987% upside if I’m right, and the Excel model tells me it’s a buy!”
He went and thought a lot about why he should be right. He inverted everything and also thought about why he would not be wrong, or how he would be if he was. He concluded a few things. Costco offers the lowest prices for a given product quality. They were a tiny percentage of a huge retail market in the US alone. They had stellar merchandising capabilities and demand forecasting, and an excellent, honest management team. They had a really good employee base they could afford to pay above-market wages because their business model was better. They sold a small number of items (SKUs), but they sold each one of those in more volume than anybody else could in America. They might not do as much dollar sales as Wal-Mart in total across the entire store & distribution footprint, but they sell more units of many of these SKUs. This gives them a big negotiating and scale advantage the suppliers of those units (I think, these days, Costco is close to selling $10 billion of generic chicken breast a year; just chicken breast. You bet your a** they get the best prices). As Costco gained more market share, they gained more scale, so their low-cost sourcing and distribution capabilities only got better, and the moat grew. Then he said, well, I’m paying 13x profit for this business that does 20% returns on capital, has this increasingly-impossible-to-beat scale advantage, and this huge, huge runway ahead where it could end up several times its current size. All while selling everyday stuff that people will still be buying in 20 years. So he bought the stock.
Again: what you’ve really got to get right is the correct thinking and conclusions, and then a big, big gap between where the stock’s at and what the business is going to be worth. It doesn’t really matter if you make 50% or 150%, honestly. It matters that you have sound reasoning that can get you to 50% or 150%, and not to minus 50%. Then, across a portfolio of bets like that, the rest takes care of itself.
Again, think about it: I bought this thing thinking the demand was ultra predictable, then we had a 1-in-100 year pandemic that upset what I thought was the most straightforward, obvious driver for this business. Was my demand forecast ever wrong! Yet I’ve still done alright because the margin of safety I had when buying had already contemplated problems like that.
Is DaVita worth 13x or 20x its profit? I don’t know. Is it worth a lot more than 7x and is its profit going to be larger in 5 years, all while still earning attractive 15-25% returns on incremental equity because they do something no one else can replicate? Very probably yes.
Chris
When we are in old age, we usually have multiple chronic diseases going at the same time as our bodies deteriorate, but only a few of them are really impacting our health, and so one of those kills us before the others do. That’s the unfortunate reality. For example, we all have cardiovascular disease, which starts off when we’re just little children. Plaque, cholesterols, fats, and other stuff sometimes stick to the walls of our arteries. Some of it stays stuck there. That’s called atherosclerosis. That might start accumulating and building up slowly over decades. Eventually, you can end up with a completely blocked artery, or with constricted arteries and blood pressure and cardiovascular issues as a result, or the blockage can dislodge and do bad stuff. You can get complications as a result, like a heart attack or a stroke. So someone might have late stage CKD, but they also have atherosclerosis, and in a population of many people, most of them pass away from the other stuff like this first before their kidneys fail. Don’t let your atherosclerosis keep you down though, we’ve all got plenty of life left ahead to enjoy!
This is an industry that only exists in the United States, because the healthcare and pharmaceutical industry value chain is more capitalistic with less government involvement in negotiating and setting prices. The pharmacy benefit managers (PBMs) sit between the pharmacies, insurers, and drug distributors on one end, and the drug manufacturers on the other. They exist only to negotiate prices, and they do that by aggregating demand (they represent multiple insurers at once) to get more leverage over guys like Pfizer and Merck. A PBM is basically just a bunch of lawyers and negotiators. Nowadays, because a lot of the largest insurers have consolidated, they have their own PBMs internally and do the negotiating themselves. Like UnitedHealth has Optum, and Cigna acquired Express Scripts years ago. Express, by the way, was a phenomenal business and the stock was a 100x over 20 years.
From the Latin word “caput / capita” meaning “head”. Capitated payments are just per-person payments. Not to be confused with being decapitated!
That’s still above the baseline rate that you would expect from declining COVID related mortality.