The 15% Trap: How a Single Number Broke Healthcare Pricing

Mark Cuban's Cost Plus Drug Company sells a cancer drug for $47 a month. That same drug often costs well over $2,000 at your pharmacy. Both include a 15% markup. The markup is the same — the price is dozens of times higher, and nobody's asking why.

In this episode, I break down why percentage-based pricing is the single most inflationary structural design choice in American healthcare. Not because people are corrupt, but because a math decision made decades ago created a system where every participant — insurers, PBMs, brokers, hospitals, pharmacies — gets richer when costs go up.

I identify three distinct "pricing diseases" in healthcare:

• Percentage Parasitism — when compensation scales with cost, not work
• Chargemaster Fiction — fake list prices with negotiated discounts off fictional numbers
• Opacity Arbitrage — profiting from the inability of other parties to see the real price

We're only treating one of them. And I make the case that the AI industry has already solved this problem with per-unit token pricing — they just don't know they solved it for healthcare too.

Watch the full video: https://youtu.be/px1eRptDHeg

Full sources and the deep dive: danmccoymd.substack.com

Transcript

SPEAKER_00 0:07

Mark Cuban's Cost Plus Drug Company will sell you a cancer drug for$47 a month. That same drug often costs well over$2,000 a month at your pharmacy. Both of these include a 15% markup. The markup is the same, the price is dozens of times higher, and nobody's asking why we use the same percentage for both. That 15% number, it sounds so reasonable, modest even. It has been quietly baked into healthcare at every level: insurance premiums, drug prices, broker commissions, and hospital contracts. And it is the single most inflationary structural design choice in the American healthcare system. It is the thing that makes everyone in the supply chain richer when costs go up. A couple of weeks ago, I did a video on tokens as the new RVU, how AI is forcing a completely different measurement system for clinical work. If you haven't seen that one, I'd encourage you to watch it because this video is the companion piece. That video is about how we measure output. This video is something I think is more fundamental. This video is how we measure price. And I think the way we measure price in healthcare is structurally and mathematically broken in its current form. Not because people are corrupt, not because regulators are asleep, but because a design decision made decades ago, basing compensation on a percentage of total spend created an incentive structure where every single participant in the supply chain benefits when costs go up. And I want to be clear, this isn't just about insurance companies. This is about PBMs, it's about brokers, it's about hospital contract structures, it's about the transparency programs that are getting celebrated right now, but are still using the same broken math. And what makes this so interesting to me is that the AI industry has already solved the problem. They just don't know they solved it for healthcare too. Let's get into it. Let me show you how the number 15 keeps showing up. The Affordable Care Act established something called the medical loss ratio, the 80-20 rule. Health insurers must spend at least 80% of premiums on medical claims. The remaining 20%, which in practice averages about 15%, is what they keep for administration and profit. When the ACCA passed in 2010, this was celebrated as a consumer protection. It was supposed to cap insurer greed, and in a sense it has. Since 2012, insurers have rebated roughly$12 billion to consumers when they exceeded that 20% threshold. Sounds great, right? Now look at Mark Cuban's Cost Plus Drug Company. Their model is beautifully transparent. You pay the manufacturer's acquisition cost plus a 15% markup, plus a$5 pharmacy fee, plus a$5 shipping, and that transparency has made Imatanib a leukemia drug available for$47 a month, instead of often around$2,000 to$2,500 at retail pharmacies. And then there's Trump RX, a new federal program built around something called most favored nation-style drug pricing. Now, early announcements have highlighted drugs like Ozimpic dropping from list prices over$1,000 toward publicly touted targets as low as$200 a month in some scenarios. Both of these programs are being celebrated. Both of them are real progress. And I don't want to take anything away from that. But I want you to notice something. Cost Plus uses a 15% markup. Insurance keeps roughly 15% of premiums. PBMs take a percentage of drug spend. Hospital contracts are often structured as a percentage of charge master prices. And broker commissions, they're 2% to 8% of premiums. Every single one of these is a percentage of the total cost. And that is the design flaw that nobody's talking about. Because when your compensation is a percentage of the total, your compensation goes up when costs go up. You don't have to do anything differently. You don't have to work harder, you don't have to deliver better outcomes. Costs rise and your slice gets bigger automatically. And here's where it gets counterintuitive. The ACA's MLR was designed to constrain and sure profits. It was a cap. You can only keep 20%. Sounds like it should work, right? But researchers at the University of Chicago published a peer-reviewed study in the American Economic Journal and found the opposite. After the MLR was introduced, medical claims actually rose 7% in the individual market and 2% in the group market. Premiums were unaffected. Let me explain why, because the math is really important. If you're an insurer and you're allowed to keep 20% of premiums, then 20% of a$1 billion premium pool is$200 million in gross margin. But 20% of a$2 billion premium pool is$400 million. You doubled your profit without changing a single thing about your operations. Now think about this from the insurer's perspective. You have two options. Option one, invest it heavily in care management programs that reduce overall medical spending. This costs money, it eats into their 15 to 20% overhead. And if it works, if it actually works, it will reduce claims. Their total premium pull shrinks, which means that 20% of a smaller number is, well, less money. Option two, don't fight rising provider rates. Let costs just drift naturally upward. Your 20% of a bigger number is more money, and it didn't cost you anything. Rand published a review and concluded, and I'm going to try to quote them directly, there is little evidence that the MLR requirement has promoted premium affordability or constrained insurer profits as intended. Instead, the MLR drives premiums and higher medical care spending. When I first read that, I had to read it three times. A consumer protection designed to limit insurer profit actually incentivized higher cost. That is a structural design failure, not a corruption problem. And it gets worse when you add vertical integration. United Health Group generated on the order of$400 billion in revenue in 2024. They are simultaneously the insurer, the PBM through OptimRX, the pharmacy, and the healthcare provider through Optim Health, which alone generated over$100 billion. Well, over half of OptimRX revenue comes from affiliated United Health businesses. So the insurer can pay inflated prices to hospitals it owns, count those payments to medical claims or whatever they want to call it, had to help satisfy MLR rules, and the money never leaves the corporate family. The percentage-based cap that was supposed to protect consumers became the mechanism through which vertically integrated conglomerates like this increase profits by increasing costs. And I want to be clear, I'm not just picking on United. This is a structural incentive. CVS owns Aetna and CAREMARC, Cigna owns Express Scripts, the big three PBMs manage 79% of all prescription claims. The FTC found that these three companies generated about$7.3 billion in excess revenue on specialty degeneric drugs over roughly five years on just two cancer drugs alone. The report showed well over a billion dollars in excess revenue in under three years. The Ohio Medicaid audit in 2018 found PBMs charging about a 31% spread on generic drugs. Think about that. Collecting around$208 million in spread on generics in a single year. This is what happens when every participant's compensation is a percentage of total cost. The incentive to reduce cost is structurally absent. Now, here's where most analysis of this topic stop. They say, okay, the incentives are broken. We need transparency. I hear that all the time. And we're getting transparency. That's real. The CMS hospital price transparency rule went into effect in 2021. The No Surprises Act in 2022, a Trump RX launched recently, cost plus drug shows you every penny. But I want to push on this because I think there's something underneath the transparency conversation that we're not really examining. Hospital price transparency compliance is a useful case study. CMS has reviewed about half of American hospitals. 65% received at least one warning. Only 20 hospitals out of thousands have actually been penalized. And here's my favorite statistic. CMS reviewed a sample of hospital files and found that over 90% of estimated allowed amounts were encoded as 99s. Okay, 999, 999, 999. These are placeholders that technically counted as compliant but were meaningless to patients. CMS has told hospitals to stop doing this. But even when hospitals do publish real prices, what are we actually seeing? We're seeing a charge master, a list of 50,000 prices that hospitals set with essentially no connection to actual cost. The national average charge-to-cost ratio is 3.4. Some departments charge 28 times their actual cost. Rand found that private insurers pay 254% of Medicare rates on average. And in some states, it's over 300%. In the transparent programs, cost plus drugs is genuinely transformational for generic drugs, but it still uses a percentage markup. And this is where I want to slow down because this is the question I think nobody is asking. If Cost Plus charges 15% on a$6 drug, that's 90 cents, fair enough. But if cost plus charges 15% on a$6,000 drug, that's$900. Is the work of dispensing a$6,000 drug really$900? Is it really$1,000 more effort than dispensing the$6 drug? Of course not. The pill goes in the same bottle. The pharmacist does the same verification. The shipping box is the same size. But the margin is a thousand times larger because it's a percentage. And this is the charade. Even the programs we celebrate for their transparency, and they genuinely deserve celebration, are still using a pricing model where the margin scales with the price of the product, not the cost of the service. And here's where my analysis gets different from what you're going to see anywhere else, because I think we've been treating health care costs are too high like just it's one problem. It's not one problem. It's three distinct pricing diseases, and each one requires a different treatment. I spent a lot of time thinking about this and I want to name them because I think having precise language for what's broken helps us fix it. The first pricing disease is percentage parasitism. That is what we've been talking about. When your compensation is a percentage of the total cost, you are structurally parasitic on the cost itself. The higher the cost, the more you earn. The insurer's 15 to 20%, the PBM spread on drugs, whether they're generic or name brand, the broker's commission on premiums, the hospital's percentage of charge master contracts. So here's a specific scenario you'll recognize if you work in benefits. A health insurer broker earns, say, 5% of your company's total premiums. Your premium is a million dollars a year, the broker makes$50,000. Next year, premiums go up 12% because of two catastrophic claims. The broker's commission now$56,000. They made$6,000 more dollars because your employees got sicker. Let me be clear. The broker didn't do anything wrong. The incentive structure did something wrong. And here's the compounding effect. When one participant in the supply chain benefits from higher cost, they have no incentive to push back on the other participants who are also inflating the cost. The insurer doesn't push back on the hospital. The PBM doesn't push back on the drug manufacturer. The broker doesn't push back on the insurer because every percentage-based participant benefits from the same rising tide. The second pricing disease is charge master fiction. This is different from percentage parasitism. This is the practice of setting a fake list price for a service that bears no relationship to the actual cost of delivering it. And then negotiating discounts off that fictional number. Stephen Brill's Bitter Pill investigation in Time Magazine, we all remember that, found hospitals charging something like$50 for a single Tylenol that costs$1.5 cents, a hundred to one markup, and a bacterial ointment for$108 that Amazon sells for$3. These aren't percentages. These are fictions. The chargebaster exists because nobody, not the hospital, not the insurer, not the patient, is pricing the actual cost of the service. Instead, the hospital sets an absurdly high list price. The insurer negotiates a 40% discount. The insurer tells the employer, we saved you 40%. The employer feels good, but 40% off a fictional number can still be 300% above the actual cost. The only people who pay the charge master rate, that's usually the uninsured, the people with the least ability to pay, the people with no negotiating power. This is not a percentage problem. This is an accounting fiction problem. Let that sink in. The third pricing disease is opacity arbitrage. This is the most subtle and I think really the most damaging. Opacity arbitrage is when an intermediary profits specifically from the inability of the other parties to see the actual price. PBMs are of the textbook example. The FTZ's investigation found that PBMs are marking up specialty generic drugs by over 1,000%. And on some HIV and cancer drugs, the market was literally incomprehensible. One generic antipsychotic drug that cost about$20 a month in the open market. Ohio Medicaid was paying over$140 for PBM spread pricing. Why? Because the plan sponsor couldn't see what the PBM paid the pharmacy, and the pharmacy couldn't see what the PBM charged the plan sponsor. The PBM sat in the middle, in the dark, and extracted the spread. This is not a percentage problem. This is not a fiction problem. This is a visibility problem. And it is one that transparency laws are actually starting to fix. The FTC is investigating states like Ohio are banning spread pricing, and Trump's RX direct to consumer model bypasses PBMs entirely for cover drugs. Now I want you to notice the pattern in what I just described. Percentage parasitism is a math problem. It requires changing the compensation model from percentages to flat fees or a per unit pricing. Charge master fiction is an accounting problem. It requires pricing services based on actual costs, not made up list prices. Opacity arbitrage is a visibility problem. It requires transparency, which we're actually making progress on. We're attacking opacity arbitrage. We're barely talking about charge master fiction, and we're not talking about percentage parasitism at all. In fact, the programs we celebrate the most, including Cost Plus Drugs, still use percentage-based pricing. And this is where I think the AI industry has something genuinely important to teach healthcare. In a previous video, I talked about tokens as the new RVU, how AI companies price their services on a per unit basis. Anthropic charges you per million tokens, not a percentage of your revenue, not a percentage of the value the output creates, a fixed price per unit of work performed. If Claude helps you write a contract worth$10 million, you pay the same per token rate as if it were to help you write a grocery list. The price reflects the cost of the computational work, not the value of the output. Now imagine applying that principle to healthcare. What if a pharmacy charged a flat dispensing fee, say$15 per prescription, regardless of whether the drug costs$6 or$6,000? The pharmacist does the same work. Why should the margin be a thousand times larger for the expensive drug? What if an insurance company earned a fixed per member per month fee instead of a percentage of premiums? Their incentive would instantly flip. Instead of benefiting from higher costs, they'd benefit from lower costs because their fee stays the same, while reduced claims mean fewer operational headaches. What if hospital contracts were based on a multiple of Medicare, a fixed reference price, instead of a percentage discount off of a fictional charge master? Rand has literally recommended this. They said private insurers should move away from discounted charge contracting and shift to contracting based on a percentage of Medicare or a similar fixed price arrangement. The AI token model isn't a metaphor, it's a proof of concept. Per unit pricing works. It aligns incentives, it rewards efficiency instead of inflation, and a trillion dollar industry has already adopted it. Here's where this gets directly applicable to your work, whether you're an employer, a benefits consultant, a clinician, or someone who just pays health insurance premiums. First, diagnose which pricing disease you're dealing with. The next time you look at a healthcare bill, a benefits renewal, a drug price, or contract, ask yourself this. Is this a percentage parasitism problem, a charge master fiction problem, or an opacity arbitrage problem? Because the treatment is different for each one. If your broker earns a percentage commission and your premiums just went up, that's percentage parasitism. The fix isn't a different broker. The fix is a flat fee broker and they exist. Firms like Epic, Nava, and a growing list of independent consultancies have publicly moved forward on flat fee or fee-based advisory models. When your broker earns the same fee regardless of your premium, their incentive aligns with yours for the very first time. If your hospital charged you$80,000 for a knee replacement and your insurer tells you, hey, congratulations, we negotiated it down to$40,000. That's Charge Master Fiction. The actual cost of that procedure at an ambulatory surgery center might be$12,000. The fix isn't better negotiation. The fix is reference-based pricing, paying based on a multiple of Medicare or actual cost, not a discount of some made-up number. If your PBM won't tell you what they actually paid the pharmacy, that's opacity arbitrage. The fix is a pass-through contract where you see every dollar. Ohio banned spread pricing in Medicaid after the audit found over$200 million in hidden markups in a single year. You can do the same thing in your plan. Second, start asking the per unit question. Every time you see a percentage-based price in healthcare, ask, what is the actual work being performed? And does a percentage reflect the cost of that work? 15% of a$6 drug is 90 cents. 15% of a$6,000 drug, remember this, is$900. Is the work$900 more? If not, you've identified percentage parasitism and you can start pushing for a flat fee or per unit alternative. Self-insured employers and about 60% of covered workers are in self-insured plans, have the most leverage here. You bear 100% of claims risk. You're the one who benefits the most from per unit pricing. You're the one who should be demanding it. Third, watch how AI pricing becomes a template. I don't have a crystal ball on this, and I'm skeptical of anyone who claims to know exactly how healthcare pricing will evolve, but I do see the direction. AI companies have proven that per unit pricing works for a massively complex service. A million tokens cost$2, regardless of whether those tokens are used to cure cancer or right a limerick. The price reflects the cost of the work, not the value of the output. Healthcare delivery is heading in the same direction, even if it takes a decade. The transparency laws are forcing visibility. The cost plus model is proving that transparent pricing really does attract consumers, and Trump RX is showing that governments can negotiate reference pricing. And the AI analogy gives us a vocabulary for what the end state looks like. Price the unit of work, not a percentage of the total. And so we come back to where we started. Mark Cuban's Cost Plus sells a cancer drug for$47. Your pharmacy charges well over$2,000. Both include a 15% markup. The 50% isn't the problem and it isn't the solution. It's the symptom of a deeper structural choice, basing compensation on a percentage of cost rather than cost of the work itself. That structural choice, repeated at every level of the supply chain, insurer, PBM, broker, hospital, pharmacy, creates a system where every participant gets richer when costs go up. Not because they're corrupt, but because the math rewards it. And we're only treating one of them. Transparency is progress and it's real and it matters. But transparency on top of percentage-based pricing is still percentage-based pricing. We need to go deeper. We need to price the work and not the bill. The AI industry figured this out. The price per token, per unit of work, regardless of the value that the output creates, it's the fairest, most efficient pricing model and technology, and it's the model that healthcare should be moving toward. I'll be going deeper on this on a written piece on my Substack. If this changed how you think about healthcare pricing, well, then subscribe. This is the kind of analysis I want to keep doing. Going underneath the conversation that everybody else is having, finding the structural question that nobody's asking.