Neurocrine Biosciences [NBIX] — Valuation

Analysis date: 2026-07-29 · Framework: Criteria, 2026-07-29 · Method: references/valuation.md Spot: $182.13 (Alpaca IEX, 2026-07-29 intraday)

Two outputs are produced, over two horizons. Neither replaces the other.

1. Verified inputs

Both bases are carried through the whole document. The pro-forma basis is the primary one — it is what an investor buys today — and it is labelled as an estimate everywhere it appears. Operating lease liabilities of ~$470m (current + non-current) are not in either EV; a book that treats leases as debt should add ~$4.67 per share of EV.

2. Implied-path test — the Valuation Criteria

2.1 Anchoring the exit multiple — GROWTH_MATCHED

The 129-name Tier-1 universe in reports/scan_v2 contains no commercial-stage biopharma peer — the only pharma-adjacent name is JNJ, a conglomerate. Anchoring NBIX to MSFT, AVGO or ANET because they happen to grow at a similar rate would satisfy the letter of the growth-match rule and violate its purpose. A comparator set was therefore built from EDGAR + Alpaca specifically for this memo.

Comparator set: US-listed commercial-stage biopharma, growth bracketing 5–15%, positive EBIT. Growth is three-year revenue CAGR to the latest fiscal year; EV/EBIT on trailing twelve months, net cash from the latest balance sheet.

Reference points outside the growth bracket, shown for range but not used to set the base: United Therapeutics 18.0% / 13.0x; BioMarin 15.4% / 35.6x; Halozyme 28.4% / 22.6x; Alnylam 53.0% / 72.8x; ACADIA 27.5% / 44.5x.

The bracket is set to the exit-year growth rate, not the historical one. NBIX's demonstrated three-year CAGR is 24.3%, but by year 5 (2031) INGREZZA is a mature ~$3.5bn franchise, CRENESSITY and VYKAT XR are past their launch curves, and growth depends on whether osavampator or direclidine reach market. A 9–12% exit-year growth rate is the reasonable central expectation, and the comparator set spans 5.6%–14.8%, which brackets it. Basis: GROWTH_MATCHED.

Base exit multiple: 18.0x EV/EBIT. It sits above the set median of 14.5x and below Vertex at 23.8x, i.e. inside the anchor range and above the lowest anchor — the rule that the base may not sit below every stated anchor is satisfied without needing a separate argument. Alkermes at 51.3x is an EBIT-denominator artifact (11.3% margin) and does not pull the base up; Regeneron and Incyte at 13.5x/13.9x are names with visible LOE overhang, which NBIX does not have inside the five-year window (§ Research 4). Implied compression from today's 19.9x traded EV/EBIT: −1.9x, or a 9.5% multiple compression. Stated as a number, as the spec requires.

2.2 Terminal margin

30.0%, held fixed. Basis, stated with the argument because it does not simply take the peer median: - Own, demonstrated: TTM GAAP EBIT margin 25.4%; FY2025 non-GAAP operating margin 29.6% ($845.5m / $2,860.5m); Q1 2026 non-GAAP operating margin 27.8%. - Pharma comparator median EBIT margin: 30.0% (set above) — and 34% across the wider set including UTHR, HALO and VRTX. - 30.0% is the peer median and roughly the company's own current non-GAAP operating margin. It is not the max(own, peer median) convention taken to the wider-set 34%, and the reason is stated: NBIX runs R&D at 35.5% of revenue against two simultaneous Phase 3 programmes. Assuming that normalises all the way to the wider peer level would be assuming the pipeline stops, which contradicts the mechanism the memo credits.

2.3 The test

python3 reverse_dcf.py --spot 182.13 --shares 100.582 --net-cash 226.0 \
    --revenue 3387.4 --years 5 --wacc 0.10 \
    --terminal-margin 0.30 --exit-multiple 18.0

Parameters held fixed: horizon 5 years; WACC 10.0%; terminal EBIT margin 30.0%; exit multiple 18.0x EV/EBIT; shares 100.582m; net cash $226m (pro forma, estimated); starting revenue $3,387.4m (NBIX TTM + VYKAT TTM). Solved for: revenue CAGR.

Terminal value is 100% of modelled EV by construction, so the reverse DCF is the primary long-horizon output and the sensitivity runs on the multiple.

The margin at other defensible definitions of "demonstrated":

This is the honest frame and it should not be buried. Total-company growth clears the required path comfortably. INGREZZA alone — 84.6% of trailing revenue — does not. The +14.5pp margin is entirely attributable to CRENESSITY and VYKAT XR growing into the mix. That is exactly the mechanism the Research document credits, and it is the correct thing to underwrite — but the test is a bet on the de-concentration working, not on the base franchise.

2.4 Sensitivity over the exit multiple (mandatory — never over scenario probabilities)

Pro-forma basis. Terminal margin 30.0%, WACC 10.0%, revenue $3,387.4m, horizon 5 years, net cash $226m.

The test does not flip anywhere in the plausible multiple range on total-company growth — even at 12.0x, below every stated anchor, the required +19.0% CAGR still sits below the demonstrated +24.3%. That is a materially more robust pass than GMED's in this same batch, which inverts at 17.5x.

But on INGREZZA's own two-year stack of +13.9%, the flip point is 15.2x (at 15.3x the required CAGR is +13.39%, a margin of just +0.5pp) — between Exelixis and Vertex. Both columns are shown because they answer different questions, and a book should know which one it is underwriting.

Effect of the Soleno deal on the test, isolated: it raises the required CAGR from +8.54% to +9.76% (+1.2pp), because $2.9bn of cash left the balance sheet in exchange for $285m of trailing revenue. Deals that lower a required path add more revenue per dollar of cash than this one did. The deal made the implied-path test modestly harder to pass, not easier.

3. The 12-month target

Built per valuation.md: near-term estimates + named product-cycle events, multiple anchored on NBIX's own trading history with the percentile stated. Not a DCF; not a peer median.

3.1 Near-term revenue base

No Street consensus was obtainable (Alpha Vantage quota exhausted). NBIX guides INGREZZA only — there is no company guidance for CRENESSITY, VYKAT XR or total revenue. The build below is therefore part guidance and part house estimate, and every line is labelled so the assumption can be attacked rather than inherited.

Named product-cycle events inside the 12-month window that move this base — each dated in NBIX_Catalyst_Calendar.md: - Q2 2026 results (~30 July 2026, i.e. imminent) — first quarter of VYKAT XR consolidation, first Soleno purchase accounting, and the test of whether INGREZZA's Q1 re-acceleration was the comp or the trend. - CRENESSITY sequential trajectory — the Q1'26 add halved. Two more quarters define the launch's asymptote, and the FY2027 CRENESSITY line above is the largest single house assumption in the build. - Zydus INGREZZA SPRINKLE litigation — D. Delaware. No trial date is disclosed. Any ruling or settlement inside the window is a step change in either direction. - osavampator and direclidine Phase 3 progression — no topline date has been disclosed for either. No date is asserted here.

3.2 Multiple — anchored on NBIX's own history

Daily EV/Sales series on as-known TTM revenue (each quarter stepped in at its actual filing date), with the current verified share count and the 2026-03-31 net cash held constant so the series measures multiple movement.

Annual mean EV/Sales: 2021 6.50x · 2022 5.91x · 2023 4.99x · 2024 5.39x · 2025 4.27x · 2026 YTD 4.08x.

On the pro-forma basis (EV $18,093m on pro-forma TTM revenue $3,387.4m) the current multiple is 5.34x, which maps to roughly the 60th percentile of the full window and the high-70s of the post-2023 window. NBIX is mid-range on its own five-year history and upper-half on its recent history — not cheap, not extended.

Target multiple: 5.13x EV/Sales — the median of the full available window. This is deliberately the least aggressive anchor available: it assumes the multiple neither expands nor contracts from its own five-year central tendency. The full window rather than the post-2023 window is used because NBIX has not undergone a structural change comparable to GMED's merger; it is the same company across the whole series, and the wider sample is the better estimator.

3.3 The target

Cross-check on earnings, not used to set the target. FY2025 non-GAAP diluted EPS was $6.39; Q1 2026 was $1.94, annualising to $7.76. At $219 that is 34.3x FY2025 non-GAAP EPS and 28.2x the Q1'26 annualised rate; the stock currently trades at 28.5x and 23.5x respectively. A caution belongs with this cross-check: FY2025 non-GAAP EPS grew +0.9% while revenue grew +21.4% (Research §3.2). A P/E anchor on a flat-EPS year is not a reliable instrument here, which is precisely why valuation.md specifies the revenue-multiple anchor as primary and the P/E as a check.

Sanity band. No external professional target exists for NBIX in the reference book, so there is nothing to check the output against. That absence is stated rather than substituted for: the GMED target in this same batch has a $62 external reference to argue with; this one does not, and it is therefore the less-constrained of the two.

The target's honest sensitivity: it is roughly 41% driven by house estimates (CRENESSITY FY2027 $1,000m and VYKAT FY2027 $450m together are $1,450m of the $4,465m FY2027 line). If CRENESSITY plateaus at the Q1'26 annualised $613m and VYKAT at $380m, NTM revenue falls to $3,846m (FY2026E $3,618m, FY2027E $4,008m) and the target to $204 (+11.8%) on the same multiple. That downside-to-the-estimate case still sits above spot, which is the useful thing to know.

4. Summary