TPG's $5 Billion Gamble on a Company That Exists Because of an Antitrust Divestiture

A company called Lyric, which checks whether medical insurance claims are paid correctly, might be sold for $5 billion. That is the headline. The weirder part is where Lyric comes from: it exists because a federal antitrust judge told the largest health insurer in America to sell off a piece of its business.

TPG — the private equity firm — bought that piece for $2.2 billion in 2022. It was the claims-editing unit of Change Healthcare, called ClaimsXten, which UnitedHealth had to divest to get approval for its $13 billion acquisition of Change. The judge ordered the sale. TPG was the buyer. In 2023, TPG rebranded ClaimsXten to Lyric. Now, four years later, TPG is working with JPMorgan to sell it again, at a price close to $5 billion.

That is a 2.3x return on purchase price for a business generating roughly $250 million of annual EBITDA, implying a 20x multiple. For a private healthcare software company, that is an ambitious ask. And it lands in a sector where the public-market peer, Claritev, saw its stock drop roughly 80% over roughly eight months in 2025 and early 2026 as investors worried about AI disrupting payment-integrity software.

So the question for any potential buyer — and for an observer trying to understand the structure of this deal — is not whether Lyric makes money. It is whether the business model at the center of that $5 billion is something AI makes stronger or something AI can render obsolete.

What Lyric actually does

Health plans — the insurance companies that pay doctors and hospitals — lose money on claims in small increments. A few dollars here, a coding error there, a duplicate payment that nobody noticed. Across tens of millions of claims, it adds up to billions.

Lyric sits at the beginning of the claims payment workflow and checks whether the claim should be paid the way it was submitted. It catches things like unbundled charges (when a provider bills separately for services that should be bundled), frequency violations (getting paid more often than a plan allows), and duplicate billing. The company says nine of the top 10 U.S. health plans use it. Major clients include UnitedHealth, CVS, and Humana.

In other words, Lyric is embedded at the gate. When a health plan decides to pay a claim, Lyric has already looked at it first. That kind of position — sitting between the claim and the payment for nearly every major health plan in the country — is not something a startup builds in three years. It took Lyric, as ClaimsXten, decades to accumulate that footprint, starting well before 2000.

The revenue model is typically some combination of per-claim fees and contingency arrangements, where the company gets paid a share of what it saves the health plan. Either way, Lyric's incentives are aligned with catching errors that cost the payer money. The bigger the plan, the more claims flow through, and the more revenue.

The antitrust origin is the structure

This is the part that matters for understanding the valuation. Lyric didn't grow organically to $5 billion. It was carved out of one of the largest healthcare data companies in the country and sold to a private equity firm under court order.

The divestiture created an odd dynamic. UnitedHealth was the original owner of the ClaimsXten business, and UnitedHealth (through its UnitedHealthcare insurance arm) is also one of Lyric's biggest customers. After the sale, the vendor and its largest client became independent — which is exactly what antitrust is supposed to produce. But it also means Lyric's entire customer base, including UnitedHealth, could in theory build in-house what Lyric now provides as a standalone service.

TPG seems to have counted on that not happening. Since buying the company, TPG has added acquisitions — ClaimShark in 2025, Concert earlier this year — expanding the platform, which it now calls Lyric42. It has invested in AI, arguing that the company's dataset gives it an advantage over competitors. TPG has said Lyric experienced "significant acceleration in revenue growth" since the acquisition, though the specific growth rate has not been disclosed.

The $5 billion asking price asks buyers to believe TPG is right: that the business grew materially, that the customer base deepened, and that the moat is wider, not narrower, than it was in 2022.

The AI question is not a marketing problem, it is a structural one

Here is where the story splits into two readings.

The case for the $5 billion price: Lyric is already the most data-rich player in payment integrity. If AI's advantage in this space comes from training on real claims data at massive scale, Lyric has a head start. It processes claims for nine of the top 10 health plans, covering 190 million lives. AI models trained on that volume of real adjudication data could be genuinely hard to replicate. An AI-native competitor might sound impressive in a demo, but building a ruleset that actually works across every major health plan's benefit design, provider network, and coding structure requires institutional knowledge that Lyric has accumulated for decades. The moat is the data and the relationships, and AI might actually widen it.

The case against it: AI is exactly the kind of technology that could bypass the ruleset companies. The industry literature is full of warnings that AI can synthesize unstructured clinical notes, detect patterns humans miss, and process claims in under 200 milliseconds. If health plans can deploy their own AI — or buy from a newer, cheaper vendor — to do the same checks that Lyric does, the question becomes why they keep paying Lyric. The Claritev stock collapse suggests that at least some investors think the answer is "they might not."

A few dollars of leakage per claim, caught by an expensive proprietary platform, starts to look like a problem you can solve with a general-purpose model once that model is good enough. The margin for payment integrity companies like Lyric and Claritev is already high — Claritev reported adjusted EBITDA margins around 60% — which means there is real economics at stake if a cheaper alternative emerges.

The truth is probably somewhere in between. Lyric's position at the gate gives it time to adapt. But the $5 billion price doesn't hedge. It requires a buyer to commit capital to a single reading of where this goes.

What a 20x EBITDA multiple means in this market

TPG bought ClaimsXten for $2.2 billion in 2022, when software multiples were still elevated from the prior cycle and private equity had cheap leverage. Now, two years into a period of higher rates and compressed valuations, asking 20x EBITDA for a company in a sector that has taken public-market punishment is a bold move.

The broader PE software exit environment tells you why TPG might still try. Private equity dealmaking began recovering in 2025 after three years of drought, though Q2 2026 saw deal values drop again. The exit conveyor belt has been jammed, which means fewer comparable transactions to anchor pricing on. TPG may be testing whether a buyer — perhaps a strategic acquirer who sees Lyric's customer access as worth a premium — will pay up.

Or TPG may simply be setting an anchor. The asking price is the starting point of a negotiation, not a final offer. If the $5 billion number pushes buyers into the $3.5 to $4 billion range, that still represents a 1.6x to 1.8x return on TPG's purchase price, on top of the cash flows Lyric generated during its ownership.

What this means for you

Lyric is not a publicly traded company. You can't buy shares in it. But this deal tells you something about a sector that matters to anyone invested in healthcare companies.

The payment integrity market — the entire market for software and services that catch incorrect healthcare claims — was estimated at roughly $16 to $17 billion in 2026, growing toward $30 billion by 2031. That is a real, expanding business. The question is who captures the value: the established players with embedded relationships and proprietary data, or the next generation of AI-native companies that promise to do the same job cheaper and faster.

UnitedHealth, CVS, Humana, and the other giants that buy healthcare software are the ones who will decide. They are also the ones who would feel the most competitive pressure if AI makes payment integrity commoditized. A cheaper, better alternative to Lyric could shift billions of dollars in cost savings from vendor profits to plan margins.

TPG is essentially asking a buyer to bet that the old order holds. The $5 billion price is a claim that Lyric's moat is still the strongest one in the room. Whether that claim survives buyer due diligence — and whether AI proves the buyer wrong three years from now — is the question the market hasn't answered yet.