Sandoz's settlement will not cure the generic drug cartel

A $450 million settlement is a tidy way to close a chapter. It does not change the underlying business.

On August 3rd Sandoz, the Swiss generic-drug maker, agreed to pay $450 million to 43 American states to resolve their antitrust claims. A second deal worth $28.5 million settles parallel class-action suits from indirect purchasers such as pharmacies and hospitals. Together, the two settlements bring to a close every remaining US federal and state government claim against the company in what has become one of the longest-running cartel prosecutions in American pharmaceutical history. Sandoz denies any wrongdoing.

The money is not trivial, but it is not ruinous either. Sandoz's American subsidiary will pay $400 million over seven years beginning in 2027, with an additional $50 million to states that settled earlier, according to the company's statement. Sandoz reported net sales of $11.1 billion in 2025. Today's two settlements total $478.5 million, roughly 4% of that figure. Stretched across seven years, it is an annoyance, not an injury. The company says the deals have no impact on its 2026 guidance or its medium-term outlook.

The trouble is that this is not the first time Sandoz - or its former parent Novartis - has settled over the same alleged conduct. In 2020, while still part of Novartis, Sandoz paid a $195 million criminal penalty to the Justice Department, the largest criminal antitrust fine ever levied on a US domestic corporation at the time, plus $185 million to resolve the parallel civil case. In February 2024, after spinning off as an independent company in October 2023, it paid $265 million to settle direct-purchaser class actions. In December of that year it paid another $275 million to end-purchaser claimants. Added to today's $478.5 million, the cumulative bill totals about $1.4 billion.

The alleged misconduct dates from 2009 to 2019, when Sandoz and dozens of other generic manufacturers supposedly conspired to allocate markets, rig bids and fix prices for around 165 generic drugs, according to court filings and the Department of Justice's criminal case. In 2020 Sandoz admitted to four criminal antitrust counts covering conduct between 2013 and 2015, including conspiracies involving drugs such as clobetasol cream, benazepril HCTZ and tobramycin inhalation solution. The settlements cover conduct over a wider period and broader range of products, but contain no new admission of guilt.

The real question is not whether the money adds up. It is why the system keeps producing the same outcome. The answer lies in the market structure of generic pharmaceuticals, which is less a competitive bazaar than a collection of narrow oligopolies.

Generic drugs account for roughly 90% of prescriptions dispensed in America. The regulatory pathway that brings them to market - the abbreviated new-drug application, or ANDA - is designed to reward the first company to challenge a brand-name patent. That first mover gets 180 days of market exclusivity. The result, unintended but predictable, is that many individual generics are supplied by only two or three manufacturers. That is not how fierce competition looks. When a market has three sellers, a phone call is an easier path to higher prices than a price war.

The incentive to coordinate is clear. Generic margins are thin, and the product is indistinguishable from a rival's. The natural competitive response to a rival cutting price is to cut it too, which leaves everyone worse off. The natural collusive response is to agree not to, which leaves everyone better off. For the individual firm, the expected cost of a settlement - even a large one, paid out over years and without an admission - is often less than the expected gain from a decade of inflated prices.

The settlements tell a story about how antitrust enforcement works in practice. The Department of Justice prosecuted the criminal side with some vigour in 2020. States then litigated their own claims over several years, each settling on its own timetable. Class-action plaintiffs followed. The result is a slow drip of recoveries that does not alter behaviour because the financial hit arrives years after the conduct, is spread across years of payment, and comes without the deterrent of individual criminal liability for the people who made the decisions. The executives implicated in the 2013-2015 conduct were no longer employed by Sandoz when the company pleaded, according to Novartis's statement at the time.

To be sure, the states' efforts are not pointless. Generic drugs are supposed to be the mechanism by which the American health-care system keeps medication costs down. When that mechanism is manipulated, patients, insurers and public programmes such as Medicaid all overpay. A $67 million recovery from smaller settlements with other manufacturers, as announced in April by a 48-state coalition led by Connecticut's attorney general, is a drop in the ocean compared with the total harm, but it is a drop nonetheless. And the cooperation requirements embedded in some settlements - forcing settling firms to assist investigations of non-settling ones - may yet produce evidence against the companies that have not yet paid.

Yet settlements are not a substitute for structural reform. The trouble is that no one seems to know what that reform looks like. The Federal Trade Commission has floated proposals to increase competition in generics, including limiting exclusivity periods and making it easier for a second or third manufacturer to enter a market. Those ideas are sound in theory. In practice they run into the usual difficulty: the people whose job it is to approve generic drugs sit in the same agency, the FDA, that oversees the ANDA process. And the companies that would lose from more competition are already well-connected to the politicians who oversee the regulators.

A wiser approach would focus on the bottleneck, not the symptom. The ANDA exclusivity system creates artificial scarcity on purpose. Narrowing or eliminating the 180-day exclusivity window for the first generic filer would reduce the incentive to coordinate on drugs where two or three firms dominate. Expanding the use of reference pricing, under which insurers and public programmes pay what a drug costs in other countries, would reduce the willingness of states and patients to absorb inflated prices. Neither idea is popular with industry lobbyists. Both would be more effective than another round of settlements.

Sandoz's stock, which trades on Switzerland's SIX exchange, is valued at around $28 billion, according to market data as of July 31, 2026. The company is growing its biosimilars business - cheaper versions of complex biologic drugs, the next frontier in generic-style competition - and its 2025 results showed accelerated sales growth. Investors should not expect this settlement to dent the business model. The company will pay its money, move on and carry on making the same drugs in the same market structure that made the alleged conduct profitable in the first place.

That is the lesson. Fines are a tax on bad behaviour. When the market structure makes the bad behaviour rational, the tax is just a cost of doing business. Consumers pay first.