Catalyst Pharmaceuticals [CPRX] — Valuation

Two horizons, two instruments, both reported. Framework: Criteria, 2026-07-29. No position verdict is issued.

0. The governing fact

The 12-month price is not a forecast. It is a realised, settled, known number: $31.50 per share in cash, received 2026-07-15, on completion of the Angelini Pharma merger. The security stopped trading on 2026-07-14 at $31.49 and was delisted the following day.

Both instruments below are therefore run twice: once as the framework requires them, and once as a no-deal counterfactual — what the analysis would have concluded had the acquisition not happened. The counterfactual is the part with methodological content, and it is where the screen's error is located.

1. Implied-path test — the Valuation Criteria

assets/reverse_dcf.py. Solved for the required revenue CAGR. Terminal value is ~100% of the modelled EV (the script's construction), so the reverse DCF is the mandatory primary long-horizon output.

Parameters held fixed, named explicitly

Results — the required parameter under six parameterisations

Run A reproduces the screen exactly (9.3%), which confirms the screen's arithmetic is correct and localises the error entirely in its parameter choices.

The two parameter errors, and why they do not cancel

Error 1 — the terminal margin was capped downward by 20.9 percentage points. The screen recorded terminal_margin_capped_from: 0.438 and set terminal_margin: 0.229, basis "max(own, industry median) — CAPPED at industry p75". Catalyst's actual operating margin is 43.8% (FY2025) and 44.8% (TTM), sustained above 40% for three consecutive years and structurally supported by a 2.2%-of-revenue R&D budget and a virtual-manufacturer model. Capping a demonstrated, durable 43.8% at a SIC-28 industry p75 of 22.9% is not conservatism; it is substituting a peer statistic for an observed fact about the subject — the same defect the framework's own anchoring rule prohibits for exit multiples. This error pushes the required CAGR up.

Error 2 — the exit multiple was anchored above the name's own all-time p90. The screen used 23.4x EBIT. Measured over CPRX's own six-year trading history (n = 1,498 sessions, 2020-07-27 → 2026-07-14, EV computed point-in-time from lagged TTM fundamentals):

23.4x sits above the 90th percentile of every multiple this stock ever traded at. The framework's rule is that an exit multiple may only be drawn from a comparator set whose growth brackets the subject's — but there is a stronger anchor available here that requires no comparator set at all: the multiple an informed strategic acquirer actually paid, in cash, with full diligence and control.

The realised transaction multiple is 12.49x. The screen used 23.4x — 87% higher. This error pushes the required CAGR down, and it dominates.

Net effect. Run F isolates it: correcting the terminal margin alone (leaving the 23.4x multiple) would have produced a required CAGR of 26.5%, not 9.3%. Correcting the multiple alone (run B/C) produces 10.6–11.1%. The two errors are of opposite sign but very different magnitude, and the reported 9.3% is roughly the product of both being wrong.

The margin — demonstrated − required

This is the number the strategy ranks on, and everything turns on which "demonstrated" is correct.

The 40.1% denominator is invalid for this test. 61.5% of the revenue added over that window was acquired or in-licensed (CPRX_Research.md §2.2). A CAGR computed across two business combinations demonstrates the company's acquisition capability, not its ability to grow revenue — and the implied path being solved for is a five-year organic revenue path.

On the correct denominators the margin is between +8.2pp and −3.6pp, not +30.8pp. Against the company's own written FY2026 guidance — issued 2026-02-25, three months before the screen ran — the price required more growth than management said it would deliver.

Result: INDETERMINATE. Not FAIL. The price input is a frozen tick from an extinguished security, so no implied-path statement about it is meaningful. On the merits and pre-deal, this is at best a PASS WITH ARGUMENT — and the argument would have to be that a mid-single-digit guided grower with $755.9m of undeployed cash re-accelerates through acquisition. That argument is available (§3 of Research documents the mechanism) but it is a capital-allocation argument, not a product-cycle one, and the 2025 record is over 100 targets reviewed and zero deals done.

Sensitivity — over the exit multiple, never over scenario probabilities

Terminal margin held at the actual 43.8%. Sensitivity is run on the parameter that determines the answer.

The flip point sits between the own-history median (11.26x) and p75 (15.19x) — i.e. the price was justifiable only on the assumption that CPRX would exit at a multiple in the top quartile of its own six-year range. That is the honest statement of where the judgement lived, and it is stated on the parameter that can change the answer.

Implied compression from today's trading multiple, stated as a number: trading at 11.58x TTM EBIT (56th percentile of own history) against a solved exit of 11.26x — implied compression of 0.32x, or 2.8%, essentially flat. Against the screen's 23.4x exit the implied expansion is +11.8x, or +102%, which no part of the screen's output flagged.

2. Twelve-month target

2.1 The actual result — realised, not forecast

There is no distribution around this number. It is contractual, it has been paid, and no residual claim on the asset exists. Any expected return computed off the $31.49 screen price is +0.03%, not the +30.8pp margin the screen reported.

2.2 The no-deal counterfactual — built as valuation.md requires

This is what the target would have been on the framework's own method, and it is the more instructive number.

Step 1 — near-term base. Consensus was unobtainable (Alpha Vantage quota exhausted; documented). Company guidance is used instead, which for a 12-month horizon is the better instrument: FY2026 total revenue of $615–645m, issued 2026-02-25, alongside "promoted product revenue expected to grow by 20% or more".

Step 2 — named product-cycle events inside 12 months, each dated and each carried in the Catalyst Calendar: - FIRDAPSE upstream royalty expiry, January 2026 — blended rate 18.5% → 6.0%, already in the Q1-26 print (gross margin 87.3% → 90.3%). - FYCOMPA generic erosion, continuing — three generic tablets and one generic suspension on market; the '497 patent expired 2026-07-01. - AGAMREE Phase 1 immunosuppression-dose readout, 2026-06-30 (delivered). - AGAMREE Canadian launch by sub-licensee KYE, following Health Canada approval.

Operating margin is held flat at the TTM 44.8% — the royalty tailwind and the AGAMREE royalty drag are assumed to offset. This is an assumption, not a forecast, and it is the single largest judgement in the counterfactual.

Step 3 — anchor the multiple on the name's OWN trading range, with the percentile stated. Six-year own EV/EBIT history as tabulated above. Percentile at the frozen spot: 55th (11.55x). At the unaffected pre-deal close of $25.94: 41st percentile (10.26x). No peer median is used anywhere.

Step 4 — the target.

No-deal 12-month target: $32.15 — own-history median EV/EBIT of 11.26x (50th percentile, stated) on guided FY2026 revenue at the midpoint. +24.0% to the unaffected close of $25.94. Range across the guidance band and the p25–p75 multiple band: $25.55 – $42.03.

2.3 The sanity band — and a genuinely useful cross-check

valuation.md requires an external target as a check on the output, never as a calibration target. No sell-side target was obtainable (quota). But something better is available: an informed acquirer's actual, binding, diligenced valuation of the whole company.

The framework's own-history method lands within 2% of the price a strategic buyer paid in cash. That is the strongest single validation of the 12-month instrument in the record to date, and it is worth logging precisely because it was not tuned to reach that answer — the multiple came from the name's own median, mechanically, and the revenue came from published guidance.

It also reframes the transaction. Angelini's "21% premium" was a premium to a depressed price: CPRX was trading at the 41st percentile of its own EV/EBIT history on 2026-04-22. Against its own median multiple on guided earnings, the $31.50 deal price was a 2% discount to fair value. Shareholders were paid a control premium relative to where the stock had drifted, and roughly nothing relative to what the business was worth on its own historical terms.

2.4 Targets above spot

Per item B16 — a process whose every target sits below spot is expressing a house view rather than valuing companies. This name's targets are above both the unaffected price (+24.0%) and, marginally, the deal price. Recorded for the calibration tally.

3. Summary of the two outputs

Screen output vs this memo: