The FTC Borrowed CVS's Own Savings Number

Introduction

The FTC settled its insulin case against CVS Caremark on July 14, and the number that ran in the headlines was $13 billion in savings. FTC Chairman Andrew Ferguson never used it. What he said was that the settlement "brings billions in real savings to consumers", and reporters got to $13 billion on their own, by adding two of the FTC's separate projections together. One of those two projections, the $4.5 billion piece, is a figure CVS had already written down for its own investors months earlier. So the regulator picked up the regulated company's sales forecast and reissued it as a federal win, pinned to the drug costs of the roughly 88 million people on a Caremark plan. If you're one of them, that forecast is now being described to you as your savings.

The Second Middleman to Settle

CVS Caremark is the second of the country's three dominant pharmacy benefit managers to close out the FTC's September 2024 insulin case, and on July 14 the enforcer and the company each put out a statement selling the same deal. Express Scripts settled first, in February; OptumRx, the third, has a tentative deal whose terms still aren't public. All three middlemen that sit between your insurer and your pharmacy are now in the same pipeline, and Reuters called the Caremark deal a mirror of the Express Scripts one.

The machinery underneath is the part I walked through last month: these middlemen pocket the spread between a drug's inflated sticker price and the rebated price they actually pay, and your out-of-pocket cost gets pegged to the sticker. It's how a vial of insulin can carry a $274 list price against a net price closer to $43, and why a JAMA study still found 1 in 5 adults under 65 with diabetes rationing insulin over cost (via AJMC).

On the money, the FTC's release says the order "locks in place up to $8.5 billion in consumer savings over the next 10 years and unlocks up to $4.5 billion in additional savings for patients over the same 10-year period from point-of-sale rebates." Two separate ten-year estimates. The FTC's own release never adds them together; Benzinga and other outlets did, and $13 billion became the headline. The order itself carries no financial penalty and no admission that CVS did anything wrong.

Where the $4.5 Billion Comes From

Set the two July 14 releases next to each other and the $4.5 billion stops looking like a government estimate. CVS's own release tells its clients it expects to "help our clients drive estimated savings of $450 million per year for each of the next 10 years" through point-of-sale rebates. Run the multiplication. $450 million across ten years is $4.5 billion, matching the FTC's "$4.5 billion... from point-of-sale rebates" to the dollar. That match is the tell: it doesn't look like anyone at the FTC ran the math independently. They took the company's projection and stamped it.

The other half of the headline is shakier. Neither release explains how the FTC got to $8.5 billion "locked in." No methodology, no model anyone outside the FTC can check, just a decade-long projection presented as if it were money already in your pocket. And the vote to send the deal out for public comment was 1-0-1: one commissioner in favor, one recused, because the agency currently has almost no sitting commissioners. The deal now runs through 30 days of public comment before it becomes final.

Caremark's headline is nearly double the $7 billion Express Scripts settlement, but it also covers more members, about 88 million, and neither release says how many actually see savings, or whether any of it narrows that $274-versus-$43 gap for a real patient.

The Provision That Does Nothing Yet

The part of the settlement that drew the political spotlight is the part that currently changes nothing. The consent order tells Caremark to count purchases made through TrumpRx.gov toward members' deductibles and out-of-pocket maximums, but only, in its own words, "in the event of certain legislative and regulatory changes." Those changes haven't happened.

TrumpRx, launched in February, is the White House's cash-pay drug marketplace built on "Most-Favored-Nation" pricing. Because it runs outside insurance, what you spend there doesn't touch your deductible today. The provision is written to close that gap once Congress passes an enabling law, which as of now is still a proposal the White House sent over in January. On paper the credit is real; in practice, it's switched off.

Ferguson led with it anyway, saying the settlement builds on "previous wins for President Trump's healthcare agenda, including innovations like TrumpRx." A federal antitrust enforcer used a settlement announcement to plug a White House program by name, and it's the same program whose central claim a New York Times investigation had already knocked down in March. Working with German public broadcasters, the Times found TrumpRx's "world's lowest prices" pitch simply wasn't true: for Novo Nordisk's Wegovy, the TrumpRx price beat just one of eight benchmark countries.

One more wrinkle sits under all of it. The "legislative and regulatory changes" the deductible provision waits on are PBM-reform bills, the kind CVS spends heavily to shape. CVS Health put $9.93 million into federal lobbying in 2025, including on PBM bills moving through Congress. None of that breaks a rule, but it leaves CVS on both sides of the trigger: bound by a provision that fires only if Congress acts, and paying to influence whether Congress acts.

Who Benefits

Two winners here, and the patient isn't obviously either one.

CVS Health leaves without a fine and without conceding it did anything wrong. Better than that, it recasts a federal antitrust case as a press release about its own affordability work, down to the nearly $80 billion it says it saved clients last year. Its executive Ed DeVaney called the deal an advance on "changes we have already put in place." A J.P. Morgan analyst had already judged the terms manageable before the ink was dry.

The FTC and the administration behind it get the number. "$13 billion" is politically useful, bigger than the Express Scripts figure, and it arrives with a TrumpRx tie-in that lets an enforcement action double as an ad for the President's drug-pricing platform. That a third of the number is CVS's own forecast, and that the TrumpRx piece stays dark until a law passes, doesn't fit in the headline. The payoff here is cover, not cash: a round number that makes a modest, conditional deal read like a rescue.

The Enforcer Helped Write the Number

To be fair, the deal isn't hollow. It carries a provision the Express Scripts settlement didn't: Caremark is barred from interfering with independent pharmacies that use third-party "hub" services, with an appointed monitor to enforce it. That term answers a House Judiciary Committee report from January that accused CVS of using its network rules to squeeze out rivals, a report CVS called "misguided." There's also a new $25-a-month insulin cap for members whose plans opt in, and the advocacy group Patients for Affordable Drugs welcomed the settlement. If your plan takes the cap, that's real money off a real bill.

Still, the shape of the whole thing is what stays with me. This is the pattern the June piece pointed at, one settlement further down the road: the enforcer and the enforced now share a number. When the FTC's headline figure is CVS's own investor projection, and the most-promoted feature is dark until a law that hasn't passed does, the settlement reads less like a check on CVS than a joint production with it. The structural fixes can matter and the headline can still be doing political work the evidence won't support.

The Bottom Line

Strip the $13 billion down and what's left is a real but ordinary deal: new structural rules and an opt-in insulin cap, sitting under a savings number the company wrote about itself. The regulator's job was to check CVS, and on the headline number, it mostly transcribed CVS's own figure instead.

The open question is OptumRx. It's the last of the three big middlemen still at the table, and its terms aren't public. If its settlement shows up with the same architecture, a big ten-year number that the company's own figures quietly produced, plus another nod to TrumpRx, then all three deals start to look less like the FTC settling cases than the FTC and the industry agreeing, three times over, on how to describe the same status quo. Watch whether Optum's number, when it lands, is one anybody outside the company actually calculated.