← Analysis page  ·  David Hay hub  ·  Research hub

Actionable insights — Friday POW!: AstraZeneca (AZN)

The repeatable analysis behind the pick: not what was bought, but how to buy a pipeline on its hit rate instead of a single catalyst, how to tell a setback cluster from a broken franchise, how to screen for a quality compounder at a two-yardstick decade low, how to find the asset the market has not priced, and how to substitute a correction history for a missing breakout — written to be rerun on the next de-rated large-cap with a deep late-stage pipeline.
2026-SEP-18 · Haymaker (Substack newsletter, paid) · The Haymaker Team / David Hay · ↗ Read · full analysis · article text
How to read this page: each insight is a method — the reasoning that took Haymaker from "three specific setbacks in three months" to "that repricing is the opportunity." The boxed line shows how it played out here. (Written newsletter — no timestamps; the Read link opens the source post.) Contrast with Sep-04's INCY: that pick was one binary catalyst taken as a free option; this one deliberately diversifies away the binary — the thesis is the ratio of wins to losses across 20+ readouts.

1. Buy a pipeline on its hit rate, not on any one readout

The repeatable method
  1. Count the late-stage readouts scheduled over the next 12–18 months and the company's recent pivotal record (positives vs total last year).
  2. Score each week's news as a tally — wins vs losses — rather than reacting to the headline failure.
  3. Check that the failures hit non-core programs: does the miss touch the franchise that carries the revenue and growth?
  4. Require breadth across therapy areas so "no single setback blows up the investment case."
  5. Buy when a cluster of misses has de-rated the stock while the tally and the core franchise remain intact.
Here: "two wins and one loss in a pipeline that has 20+ trials reading out over the next eighteen months, and that ratio is the investment thesis in a nutshell." Enhertu (DESTINY-Lung04) and Tagrisso won; camizestrant failed — but it "does not touch the oncology franchise that is 48% of total product revenue and growing 16%." Record: 16 positive pivotal trials in 2025. AZN.
Watch for

2. Diagnose the discount: name the setbacks and size them

The repeatable method
  1. When a quality large-cap is 20%+ off its high, list the specific events that created the discount — don't accept "sentiment."
  2. Size each one in revenue (third-party estimate cuts are a usable yardstick) against the company's total base.
  3. Set the losses against what the business is still delivering: organic growth, EPS growth, margins, reaffirmed guidance.
  4. If the named setbacks are small relative to the base and guidance holds, the discount is a repricing of uncertainty — buyable.
Here: "The discount has been created by three specific setbacks in three months, all manageable" — e.g. Morningstar's $700M cut on camizestrant against a $61B revenue base still growing 6% constant-currency into 11% core EPS growth, guidance reaffirmed. "The stock is at $164 because three programs missed in three months and the market repriced the uncertainty; that repricing is the opportunity."
Watch for

3. Screen for a two-yardstick decade low in a stock that was never cheap

The repeatable method
  1. Plot 10 years of forward P/E and price-to-sales together (the house's preferred yardstick is P/S — sales are harder to fake).
  2. Look for both near their decade lows at the same time, and measure the dip from prior peaks.
  3. Confirm the stock historically commanded a premium — a structurally cheap stock at its usual multiple is not a signal.
  4. Find the last time both lows coincided and check what followed.
  5. Compare the forward multiple with the market's to frame relative cheapness.
Here: forward P/E 14.9× "near its lowest level in a decade," ~40% below the 23–25× peaks, vs a market P/E of 21; P/S 4.16× vs >6× (2020–21) and ~5.5× (2025). "The last time both metrics were simultaneously near these lows was 2016 to 2017, before a multi-year re-rating drove the stock from approximately £40 to over £130."
Watch for

4. Find the late-stage asset the market is not yet paying for

The repeatable method
  1. Scan the pipeline for a drug with replicated positive Phase 3 data (not a single trial) that has drawn little coverage.
  2. Ask whether it expands the treatable population beyond what existing drugs reach — the commercial breakthrough is often who it works in, not how well.
  3. Check management's peak-sales number and whether it was raised after the data; note regulatory speed (Priority Review) and the decision date.
  4. Tie it back to the long-range target: is this the asset that bridges the gap?
Here: tozorakimab for COPD (~400M patients) — three positive Phase 3s (OBERON, TITANIA, MIRANDA) working "across all eosinophil levels," where existing biologics help only ~30%; peak sales raised to >$5B, Priority Review, decision Q1 2027 — "exactly the kind of asset that could help propel AZN to its $80 billion revenue target."
Watch for

5. When the preferred breakout is missing, substitute the stock's correction history — and say so

The repeatable method
  1. Check the chart against your preferred entry pattern (for Haymaker, a breakout from a tight trading range).
  2. If it isn't there, state it plainly rather than bending the chart.
  3. Look instead at how prior corrections within the long-term up-trend resolved; if each was a buying opportunity, the current pullback can qualify on the same terms.
  4. Let valuation carry the entry case when the technical signal is absent.
Here: "there isn't the clear breakout from a tight trading range that is our preferred price action pattern. On the positive side… its prior three corrections over this timeframe were excellent buying opportunities." Price ~$165, below its moving averages (~$174 / ~$183).
Watch for

Methods distilled from the paid Haymaker newsletter post for personal study. Not investment advice.