Terns Pharmaceuticals

Phase A — Understand the business

Company Overview

Terns Pharmaceuticals was a clinical-stage biopharma founded in 2017 by Weidong Zhong (ex-Global Head of Antiviral Research at Novartis Institutes for BioMedical Research; before that Gilead HCV/HBV discovery), seeded by Lilly Asia Ventures, and built on a "hub-and-spoke" model of in-licensing small molecules from Eli Lilly. It IPO'd on Nasdaq in Feb 2021 at $17.00/share (~$128M raised).

The business never had a P&L — it was pre-revenue its entire life (hence the +clinical battery). Its "products" were a rotating pipeline of clinical assets, and it pivoted its therapeutic identity three times in five years:

1. NASH/MASH (2017–2023) — the founding mission. Lead assets TERN-101 (FXR agonist, originally Lilly's LY2562175), TERN-201 (SSAO/VAP-1 inhibitor, in-licensed from Lilly Mar-2018), TERN-501 (THR-β agonist).
2. Obesity (2024–Oct 2025) — pivoted to metabolic/obesity behind TERN-601, its first internally discovered asset, an oral small-molecule GLP-1 receptor agonist.
3. Pure-play oncology (Aug 2025 → acquisition) — radical refocus onto TERN-701, an oral allosteric BCR-ABL1 inhibitor for chronic myeloid leukemia (CML); legacy metabolic assets pushed to a "partner-funded" model.

Customers / payment structure: none — no commercial product, no revenue, no take-or-pay. The only "customer" that ever mattered was the acquirer. our figures is empty `` and correctly so.

The one-sentence business model in hindsight: Terns was an asset-incubation vehicle — license/discover small molecules, generate Phase-1/2 proof-of-concept, and sell to a large-cap. It executed exactly that, terminating in the Merck buyout. The value creation was entirely in TERN-701; everything else was a write-off or a sunk cost.

Supply Chain (→ CDMO / manufacturing, +clinical swap)

A pre-commercial biotech has no manufacturing-scale supply chain; the relevant map is discovery → CDMO → clinical sites → (would-be) payers:

• Upstream IP origin: Eli Lilly (NASH assets, in-licensed) and Terns' own medicinal-chemistry org in Shanghai/Foster City (TERN-601, TERN-701 were internally discovered). The China-rooted discovery engine (Zhong's heritage; LAV backing) was a genuine structural feature — cheaper, fast med-chem.
• Manufacturing: outsourced to contract development & manufacturing organizations (CDMOs) — specific CDMO names not disclosed publicly → n/a. Single-source CDMO dependency is the standard chokepoint for a company at this stage but was never the binding constraint here.
• Clinical operations: CARDINAL (TERN-701) was a global, multicenter Phase 1 across US + ex-US sites; FALCON (TERN-601) was US-based multicenter.
• Downstream "end customer": for TERN-701, hematologists/oncologists treating CML — but commercialization rights, manufacturing scale-up and payer access all transferred to Merck on close. The chokepoint that actually mattered was capital + a commercial partner, and Terns resolved it by selling itself.

This lens is structurally thin by design for a pre-revenue name — names where they exist, honestly marked n/a where they don't.

Competitive Advantages (→ asset/data/IP moat, +clinical swap)

The moat was never corporate (no brand, no scale, no network effects, no switching costs — it was a 100-person clinical biotech). The moat was molecule-level, and it was real for exactly one asset:

• TERN-701's differentiation vs. the only approved allosteric peer (Novartis' asciminib / Scemblix): preclinical data presented at EHA 2025 showed greater potency than asciminib against several BCR-ABL1 resistance mutations in both the active-site and allosteric domains. Clinically, CARDINAL enrolled a heavily pretreated 3L+ population — 38% had received prior asciminib, of whom 75% had discontinued asciminib for lack of efficacy — and TERN-701 still drove deep responses in that group. That is the crux of the "best-in-disease" claim: working after the incumbent allosteric drug fails.
• FDA Breakthrough Therapy designation for TERN-701 in CML — a regulatory moat signal (priority interactions, expedited path).
• IP estate: internally discovered → cleaner composition-of-matter ownership than the in-licensed NASH assets. Patent life specifics not sourced → n/a.
• Bargaining power: essentially none over suppliers; the only leverage Terns held was scarcity of a differentiated late-Phase-1 CML asset in a market facing the Keytruda cliff — which is precisely what let it command a $6.7B exit on Phase 1/2 data.

TERN-601 (obesity) had no durable moat — it was a me-too oral GLP-1 in the most crowded field in pharma (Lilly orforglipron, Novo, Pfizer, Structure, Viking, Roche), and its data was sub-competitive (see Lens 5). The moat thesis lived and died with the oncology asset.

Segments

Not applicable in the operating sense — Terns was single-reportable-segment, pre-revenue, no product or geographic revenue split. our figures is empty `` and correct. The meaningful "segmentation" is by pipeline asset, captured as the Pipeline-by-phase table in Lens 5 (the +clinical swap). R&D spend by program was not broken out granularly in public sources → n/a at the dollar level.

---

Phase B — Measure performance (+clinical overlay: pipeline, not earnings)

Pipeline by phase

The asset table is the company. State at the point of acquisition:

TERN-701 / CARDINAL — the data that got the company bought:

• MMR (major molecular response) at the RP2D (≥320 mg QD): 80% at 24 weeks (95% CI 61.4–92.3) among efficacy-evaluable patients with >24 weeks follow-up (n=30). Earlier/cumulative cohorts read ~50% MMR at 3 months and a pooled MMR/DMR of 75% / 36% as characterized by William Blair.
• Population: predominantly 3L+, with 38% prior asciminib (75% of those had failed asciminib for lack of efficacy) — i.e. responses in the hardest-to-treat patients.
• Safety (data cutoff 13-Sep-2025, n=63): 87% (55/63) remained on treatment; no dose-limiting toxicities, MTD not reached; Grade ≥3 TEAEs all <10% (neutropenia 8%, thrombocytopenia 8%); most common any-grade TEAEs diarrhea 22% / headache 18% / nausea 16%, Gr 1–2; no AE-related discontinuations at the December update. Clean for a TKI.

TERN-601 / FALCON — the failure that killed the obesity story:

• Phase 2, ~30/cohort, doses 250/500/500-slow-titration/750 mg vs placebo, 12-week primary endpoint.
• Maximum placebo-adjusted weight loss ~4.6% at 12 weeks — well below the ~6–8%+ bar set by oral competitors at comparable timepoints, and a long way from injectable-class efficacy.
• ~12% discontinuation for adverse events; three grade-3 (asymptomatic, reversible) liver-enzyme elevations post-treatment, two deemed drug-related — a hepatotoxicity flag in an indication that demands a pristine safety profile.
• Outcome: Terns announced it would not advance TERN-601 and would not invest in further metabolic assets — a clean kill, redirecting to TERN-701.

The pre-revenue financial frame: revenue $0; the relevant metric is cash runway — see Lens 11. Net loss / opex specifics by quarter were not pulled from the empty shelf → n/a at line-item granularity, but the company carried only operating losses.

Earnings Calls → mechanism/data-event sentiment (+clinical swap)

With no revenue, "calls" = data-update calls + conference presentations; the sentiment arc is one of escalating, then redirected, conviction:

• 2021–2023 (NASH era): measured, target-engagement-focused ("best-in-class THR-β," SHBG biomarker) — proof-of-concept language, no commercial confidence.
• 2024 (obesity pivot): bullish — Sep-2024 Phase 1 TERN-601 showed "statistically significant" ~4.9% weight loss in 28 days; management leaned hard into obesity as the value driver and raised capital against it.
• Aug–Oct 2025 (the hinge): decisive pivot to "pure-play oncology" before the FALCON miss, then the 21-Oct kill of TERN-601 — management's tone shifts from "broad metabolic platform" to "TERN-701 is the asset; we are concentrating all capital here." The recurring phrase becomes "best-in-disease" for TERN-701; the thing they stopped saying was anything about obesity/GLP-1.
• Dec-2025 → Jan-2026: maximal confidence — ASH oral presentation, $1B+ cash runway "into 2031," "2026 priorities" framed entirely around TERN-701 registration path. That confidence is exactly what a buyer pays a premium to take off the table.

The sentiment trend is textbook for a single-asset oncology story that got acquired: rising conviction on the winner, abrupt silence on the losers.

Catalyst calendar + mechanism comps (+clinical swap — comps by target, not P/E)

The terminal "catalyst" already fired (the buyout). For completeness, the mechanism-comp landscape Merck bought into:

Pricing/sales comps are mechanism-relevant, not multiple-relevant: asciminib peak-sales and 2025 actuals were not cleanly sourced → n/a (Novartis does not isolate Scemblix in the snippets retrieved). The be-early signal is that Wall Street modeled TERN-701 peak sales above the net price Merck paid (Leerink: >$1B by 2032, peak ~$6.2B by 2040 ) — see Lens 12.

Stock-Price Catalysts (the >5% movers, 2021→2026)

The tape tells the whole boom-bust-rescue arc:

• Feb 2021: IPO at $17.00; rode metabolic/NASH enthusiasm.
• 2022–2023 "biotech winter": broad de-rate; NASH disappointments and the FXR-agonist class falling out of favor → multi-year grind lower.
• Sep 2024: TERN-601 Phase 1 obesity data (+) — ~4.9% weight loss in 28 days; obesity hype + a $172.7M raise; sharp re-rate up.
• 21 Oct 2025: FALCON Phase 2 miss (−) — obesity program killed; Citizens cut PT to $15 from $20; major leg down. This is the single most important negative catalyst — it should have been near-terminal for the equity.
• Aug–Dec 2025: oncology pivot + CARDINAL build (+) — Breakthrough designation, "best-in-disease" data narrative, $1B+ runway raise; the stock re-rates up on TERN-701 alone, climbing back toward the low-$50s on M&A speculation.
• 25 Mar 2026: Merck deal announced (+) — $53.00/sh; stock to ~$52.80 (+5.6% on the day), pinned to the offer thereafter.
• 5 May 2026: deal closes — TERN delisted.

What the pattern reveals: TERN was a pure binary-catalyst stock — it reacted only to clinical readouts and the takeout, not to macro or earnings. The market's entire valuation flipped from "obesity also-ran" to "scarce best-in-class CML asset" inside ~5 months, and a strategic buyer monetized that re-rate.

---

Phase C — Judge people & books

Management

• Founder: Weidong Zhong (CSO/founder archetype — deep antiviral/med-chem scientist, Novartis + Gilead pedigree, China-discovery DNA). Built the asset engine that eventually produced TERN-701 and TERN-601 internally — the source of the only real value.
• CEO at exit: Amy Burroughs (appointed 7-Feb-2024 — a professional-manager/dealmaker archetype, not a scientist). Background: commercial/strategy at Genentech, P&G brand management, prior CEO of Cleave Therapeutics where she ran financings and licensing/collaboration deals. Her hire ~13 months before the buyout is telling: a board installs a financing-and-BD CEO when the endgame is a sale, not a build. Capital-allocation grade: A on the exit. She presided over (a) the brutal-but-correct kill of the failed obesity asset, (b) concentration onto the winner, (c) a large runway raise that maximized the company's negotiating position, and (d) a $6.7B all-cash sale on Phase 1/2 data — a near-ideal outcome for a single-asset biotech.
• Other execs: Mark Vignola (CFO), Melita Sun Jung (CBO) — BD/finance-heavy bench, consistent with a sale-oriented org.
• Skin in the game / insider ownership: our figures absent → n/a. LAV/Deerfield were anchor holders historically.
• Red flags: the serial therapeutic pivots (NASH → obesity → oncology) read as either admirable discipline (kill losers fast) or strategic incoherence (no durable franchise). In hindsight it was the former — they followed the data and the value — but a holder living through three identity changes in five years bore real whiplash.

Verdict on the operators: they ran the right playbook for what this asset was — incubate, kill ruthlessly, sell at the top. Judged by P&L-equivalent (shareholder cash returned), this is a management success.

Forensic Red Flags (+clinical re-point: trial integrity, going-concern, dilution)

For a pre-revenue, soon-acquired biotech, classic accounting forensics (revenue recognition, channel stuffing) don't apply — there's no revenue to manipulate. The +clinical risk set:

• Trial-design integrity: CARDINAL was small (n=63), single-arm, open-label, early-phase — the standard caveat. The 80% MMR is impressive but not a randomized comparison to asciminib; cross-trial comparison risk is real. The bull case rests on Phase 1/2 data that Merck must still convert through registration trials. This is the central scientific risk, not an accounting one.
• The obesity data was a genuine miss, disclosed cleanly — Terns reported the FALCON shortfall and the liver-enzyme signal transparently and killed the program (no evidence of burying it). That is a positive governance signal.
• Dilution / going-concern: as a cash-burning pre-revenue name it diluted repeatedly (IPO, 2024 raise, late-2025 upsized offering to ~$1B). SBC flattering non-GAAP is moot (no non-GAAP earnings). Going-concern was resolved permanently by the acquisition.
• Cash-vs-earnings divergence: moot — pure cash burnCash burnHow much more cash goes out than comes in, per period. The clock on a company with no profits., no earnings to diverge from. Line-item burn n/a from the empty shelf.

Regulatory findings (required sub-section):

• SEC enforcement (EDGAR LR + AAER): None possible to verify via the research layer — regulatory/regulatory-findings.md records cik: null, so our model ran no EFTS search. (Terns' real CIK is 0001831363; a future ingest could close this gap. No SEC re-fetch this run per scope.)
• Non-SEC (FTC/DOJ/FDA/CFPB): web search for "Terns Pharmaceuticals" (FTC OR DOJ OR FDA OR... ) enforcement surfaced no enforcement actions, consent decrees, fines, or penalties — only normal FDA interactions (Breakthrough Therapy designation for TERN-701, IND clearances), which are favorable, not adverse.
• 10-K Item 3 (Legal Proceedings): not read — filings shelf empty, SEC re-fetch out of scope → n/a from research layer. Standard pre-deal merger-objection / 14D-9 disclosure lawsuits are routine for any take-private and, where they occur, are immaterial; none surfaced as material in web search.
• Net: No material regulatory or legal findings surfaced via web search and the (empty) research-layer regulatory file as of 2026-06-29. SEC EDGAR LR/AAER was not machine-verified due to the missing CIK — flagged as a known coverage gap rather than a clean bill.

---

Phase D — Project & stress-test

rNPV / path-to-tradeable (+clinical swap — the question that actually matters)

There is nothing to project — the value was crystallized in cash. The forward question for a +clinical name is normally "does runway reach the next inflection?" Here it's already answered terminally:

• Runway: post the late-2025 raise, ~$1.0B cash with runway "into 2031" — i.e. funded well past the next CML inflections. Runway was a strength Merck paid to capture, not a risk.
• Terminal value realized: $53.00/share in cash, $6.7B equity ($5.7B net of cash). That is the rNPV the market clearing price assigned — no DCF of mine improves on a closed cash transaction.
• Implied valuation of TERN-701 alone ``: net consideration ~$5.7B minus negligible value for the abandoned metabolic/NASH assets ⇒ essentially the entire ~$5.7B net price is the market's risk-adjusted value of TERN-701 on Phase 1/2 data. Against Leerink's >$6.2B peak-sales-by-2040 model, that is <1× peak sales — the crux of the "Merck got it cheap" debate (Lens 12).
• EPS projection: n/a — no revenue, no EPS, company acquired. No our model create (correctly skipped — unattended --watchlist, and there is no scoreable forward EPS or binary readout left; the binary already resolved YES via takeout).

Bull vs Bear (adversarial — re-cast as "did Merck pay right?", the only live question)

The investable question on TERN closed on 5-May; the residual analytical question is whether Merck got value — which is the post-mortem's real payoff.

Bull (Merck got a bargain / the asset is special):

• TERN-701's 80% MMR at 24wk RP2D in a 3L+, asciminib-failed population is potentially best-in-disease — exactly the profile that displaces incumbents.
• Street peak-sales > price paid: Leerink modeled >$1B by 2032, ~$6.2B peak by 2040 — Merck paid ~$5.7B net for an asset analysts value above the price. William Blair and others called the ~6% premium-to-last-close among the lowest in years and floated a competing bid that never materialized.
• Strategic fit / timing: Merck faces the Keytruda patent cliff (~2028) and needs oncology growth assets; a differentiated oral CML drug with Breakthrough status is a clean, ownable franchise.
• Clean safety + oral convenience + FDA Breakthrough = a credible registration path.

Bear (Merck overpaid / the asset disappoints):

• Phase 1/2 data only — small, single-arm, open-label. CML response rates frequently regress when randomized and scaled; cross-trial superiority over asciminib is not yet proven in a head-to-head.
• Asciminib is entrenched and moved to 1L (NCCN Cat-1 preferred); a 3L+ resistant niche is real but smaller than the headline TAM, and displacing a Novartis incumbent with a dominant 1L position is hard.
• CML is a chronic, well-served, increasingly generic market (imatinib/dasatinib generic) — payers will resist a premium 3L+ oral unless the benefit is unambiguous.
• Merck took a ~$5.8B IPRD charge in Q2/FY2026 — a real near-term earnings hit for an asset that could still fail Phase 3.

Pre-mortem (it's 2028 and the deal looks bad): TERN-701's Phase 3 either (a) fails to show superiority over asciminib in a controlled setting, (b) surfaces a late safety signal at scale, or (c) lands in a 3L+ niche too small to justify $5.7B as CML therapy continues commoditizing — and Merck's IPRD write-off proves prescient rather than conservative.

Contrarian view (what the market refused to see at the time): through Oct-2025 the market priced TERN as a failed obesity also-ran (Citizens PT $15) and almost entirely missed that the oncology asset was the crown jewel. The re-rate from ~$15-implied to $53 in ~5 months is the cost of that mispricing. The enduring lesson: a single-asset biotech's value can invert overnight when the market is anchored on the wrong asset.

Devil's Advocate (short-seller — moot on TERN, sharp on the asset)

You cannot short TERN — it's delisted and cash-settled at $53. The only place skepticism still bites is Merck's thesis:

• Where the value is concentrated: 100% in one Phase-1/2 asset. If TERN-701 fails Phase 3, Merck's $5.7B is largely impaired (the metabolic/NASH assets were already written off by Terns).
• The most dangerous competitor bulls underestimate: Novartis/asciminib itself — entrenched, now 1L-approved, with real-world data and a sales machine. "Best-in-class on Phase 1 data" has a long graveyard of drugs that never beat the incumbent in the clinic that mattered.
• Worst capital-allocation read: none egregious at Terns (the pivots were value-accretive); the risk sits with Merck paying up on early data under cliff-driven pressure — exactly when buyers overpay.
• What must hold for the price paid: TERN-701 must (a) replicate its response rates at Phase 3 scale, (b) carry a clean safety profile in a larger, longer-exposed population, and (c) win meaningful share against a 1L-entrenched asciminib in a commoditizing market. If CML efficacy disappoints 20–30%, the asset re-rates from "$5–6B franchise" to "modest 3L+ niche," and the deal looks like a cliff-panic overpay.
• Single scenario that permanently impairs: a Phase 3 readout showing no superiority + a meaningful safety signal simultaneously — plausibility moderate; it is the whole bet.

Management Questions (15, ordered by information value)

(Now effectively questions for Merck oncology / the program team, since Terns no longer has independent management — ordered so the answer that would most change the view is first.)

1. What is the Phase 3 / registration design for TERN-701, the comparator (asciminib head-to-head, or SoC?), the line of therapy (3L+, 2L, or a push to 1L?), and the timeline to a pivotal readout?
2. In the CARDINAL data, what is the durability of the 80% MMR beyond 24 weeks, and the deep molecular response (MR4/MR4.5) rate that actually predicts treatment-free remission?
3. What exactly differentiates TERN-701 from asciminib mechanistically and clinically in a head-to-head — and what evidence beyond preclinical potency supports superiority?
4. What were the full TERN-601 FALCON safety data (the grade-3 LFT elevations), and does that hepatotoxicity signal implicate Terns' broader internal med-chem (relevant to other internally discovered assets)?
5. How large is the addressable 3L+/asciminib-failed CML population, and what is the realistic peak-sales path given a commoditizing, generic-heavy market?
6. What price/reimbursement assumptions underwrite the deal model, and how do they survive a 1L-entrenched asciminib?
7. What is the manufacturing/CMC readiness (CDMO, scale-up, COGS) for a commercial oral oncology drug?
8. What resistance-mutation coverage does TERN-701 have (incl. T315I, compound mutations) versus ponatinib/olverembatinib?
9. What is the IP/exclusivity runway (composition-of-matter expiry, formulation patents) on TERN-701?
10. What retention / integration plan exists for the Terns discovery team that produced the asset — does the China-rooted med-chem engine survive inside Merck?
11. How does TERN-701 fit Merck's post-Keytruda oncology strategy beyond a single indication (Ph+ ALL? other BCR-ABL settings?)?
12. What milestones in the next 12–18 months de-risk the IPRD charge, and what triggers a further write-down?
13. What is the plan, if any, for the shelved metabolic/NASH assets (TERN-501/601/101/201) — partnered, divested, or dead?
14. What competitive readouts (Novartis, Ascentage, others) could change TERN-701's positioning before its own pivotal data?
15. What is the go/no-go threshold at the next interim that would stop the program?

---