Hatch-Waxman Act

Start with the founding arbitrage. A new drug costs its innovator roughly $1–2.5bn of risk-adjusted R&D and a decade of trials, and until 1984 a would-be copier in the US had to repeat much of that. The Hatch-Waxman Act changed the deal. A generic maker files an ANDA (Abbreviated New Drug Application) that proves only bioequivalence, meaning the same active ingredient and the same absorption profile in a small human study. Then it rides on the innovator's original safety and efficacy data. Development cost falls from billions to $1–5m for a simple oral solid. The complex forms cost more. In exchange, innovators got patent-term restoration and defined exclusivity periods.

That statute created the modern generics industry, and Indian companies became its largest suppliers. They brought process-chemistry skill honed under India's pre-1995 process-patent regime, US-inspected plants, and engineers who cost a fraction of New Jersey's. Roughly 9 of 10 US prescriptions are generic, and Indian plants supply on the order of 4 of those 10 (order of magnitude, so check the current share).

How does a generic drug get through the ANDA pathway?

Now follow one product through its life. Once you understand a single ANDA end to end, the whole model follows:

1. Select the target. You want a brand with expiring patents, attractive US sales and, ideally, some technical or legal difficulty that will keep the crowd small. Portfolio teams rank candidates by "$ sales per expected competitor."
2. Develop and file. Reverse-engineer the formulation and run bioequivalence, then file the ANDA with a GDUFA user fee. That costs $1–5m for a simple product and $10–50m+ for a complex one, and it takes 2–4 years before filing.
3. Certify against the patents. Every ANDA certifies to each Orange Book patent under Paragraph I/II (no patent, or expired), Paragraph III ("I'll wait until expiry") or Paragraph IV ("your patent is invalid or I don't infringe, and I'm coming now"). Para IV is a declaration of war. The innovator has 45 days to sue, and a suit triggers an automatic 30-month stay on FDA approval while the case runs.
4. Approval. FDA review runs roughly 1–3 years. GDUFA has compressed the medians, so check the current figure. Approval is site-specific. The ANDA names the plant, and the plant must be in good inspection standing, so the whole machine is hostage to that plant's record. The compliance ladder shows what happens when the record goes bad.
5. Launch and erode. On day one of a multi-player launch, price falls 40–60% below brand, and each later entrant cuts again. At 8–10 players an oral solid settles 80–95% below the original brand price. From then on, your revenue on this molecule shrinks every year.

What is the generics treadmill equation?

That gives you the treadmill equation. A generics company's US revenue is a decaying base plus a launch stream:

US revenue(t+1) = US revenue(t) × (1 − erosion rate) + new-launch revenue