PillPack founder TJ Parker: drug pricing debate is distorted by list vs. net price confusion, EO's most promising element is direct consumer pricing reform

Summary

TJ Parker, the PillPack founder turned Matrix Partners investor, argues that the central confusion distorting the drug pricing debate is the gap between list prices and net prices — and that the executive order's most promising element is a mechanism that could close that gap for consumers directly.

The core problem Parker describes is straightforward: when an uninsured consumer fills a branded prescription at the pharmacy counter, they pay the list price. For Ozempic, that is roughly $1,000. The actual net price — what the PBM negotiated — is around $250. The $750 difference flows to the PBM, the insurer, the employer, or some combination. Consumers bear the full gross price while everyone else in the chain transacts at the net.

The press has made this worse by consistently citing list prices in cross-country comparisons. The Wall Street Journal recently quoted Jardiance at $611 for a 30-day supply in the US versus $70 in Switzerland and $35 in Japan. Parker says that comparison is misleading — the US net price for that drug is probably closer to $100–$150. The real gap versus other countries is two to three times, not the ten-to-twenty times figure that shows up in political rhetoric.

The EO's most interesting provision

The element Parker is most focused on is a specific provision requiring pharma companies to make comparable direct-to-consumer prices available at something close to list prices in other developed markets. For Ozempic, the UK list price is $150–$250 — close to the US net price already. If pharma were required to offer that cash price domestically, Parker argues the entire rebate architecture becomes unnecessary. Consumers transacting at rational prices removes the commercial logic for backdoor rebate negotiations and the fees that create the current complexity.

Parker is careful about the implementation spectrum. If the EO amounts to government price controls pegged to net prices in other countries, that is a meaningful threat to pharma economics and R&D incentives. If it instead gives consumers the option to buy at prices comparable to foreign list prices — or import from abroad — that is more like opening competition than imposing controls. Which version emerges from implementation is what he says investors should watch.

On the market's muted reaction — biotech rose roughly 3% on the day rather than selling off — Parker reads it as two things: pharma's net economics are largely unaffected if the real target is rebate margin captured by middlemen, and there is genuine skepticism about implementation given the track record. The prior administration attempted a rebate repeal, got it far along, and was ultimately blocked by the CBO because rebates do reduce premiums. An MFN approach on Medicare Part B also failed. The PBMs and insurers sold off, which Parker says makes sense given what the EO implies for their margin.

Drug compounding

Parker is vocally against drug compounding outside of genuine shortage situations. His objections are both safety-based — a compounding center producing a commercially available drug at scale previously caused a meningitis outbreak that killed roughly 100 people — and structural. Compounding outside shortages effectively violates IP law and eliminates the R&D incentive that patent protection is designed to create, which he argues is worse for innovation than price controls.

The market dynamic driving compounding demand, he says, is the same list-versus-net price distortion. GLP-1 compounders charge roughly $100–$150 against a pharmacy list price of $1,000. If direct consumer prices come down to $200, the price delta that makes compounding attractive largely disappears.

Where Parker sees startup opportunity

The investment thesis Parker is building around is not AI diagnostics or vertical medical models — he is skeptical that dedicated health AI products outperform foundation models on diagnosis, and doubts consumers pay $20 a month for something ChatGPT handles 80% as well for free.

The gap he finds more compelling is the transactional infrastructure. Healthcare has no product catalog, no visible pricing, no checkout flow. ChatGPT now runs Shopify checkout inline; nothing equivalent exists in healthcare. Once a patient has a probable diagnosis and needs a prescription, labs, or a referral, there is no digitized path to actually transact. Parker frames that as the infrastructure problem worth solving — building the equivalent of a product catalog and pricing layer for healthcare, independent of wherever the consumer front door ends up sitting.

On AI administrative efficiency more broadly, he thinks the value will accrue to service businesses that implement it well rather than to pure-play software vendors, with one exception: companies that use AI to compound differentiation in consumer-facing experiences — like dropping the cost of booking a specialist appointment from $40 per human-assisted booking to $2 via AI — have a clearer path to venture-scale returns.

Parker does two to three deals a year at Matrix, concentrated early-stage. His watch on the EO comes down to one question: is the final implementation about fixing the consumer's gross-versus-net problem, or about setting true net MFNs against other countries? The first is structurally benign for pharma and interesting for consumer health startups. The second makes him nervous.