Corcept Therapeutics Incorporated (NASDAQ: CORT) — The Molecule Went Generic, the Story Went Up 117%

Report date: 2026-07-25 · Price (2026-07-24 close): $95.36 · Market cap: ~$10.4bn · EV: ~$9.9bn Sector: Health Care — Biopharmaceuticals (Endocrinology & Oncology) · CIK: 0001088856

This article carries no investment recommendation and no price target. The analysis sections discuss valuation solely as embedded expectations and scenarios. The single exception is the clearly-labelled Claude's Take block immediately below, which is the author’s own subjective opinion.

⚡ Claude’s Take

The author’s own independent opinion, offered as general information only — not investment advice. The analysis that follows carries no position and no price target.

Verdict: AVOID-here. HOLD if owned, with a strong bias to trim into strength. Accumulation zone ~$48–62; genuine value below ~$45.

Tag: “The molecule went generic, the story went up 117%.”

Corcept has spent 2026 doing something remarkable: it lost its patent case, lost its lead approval, converted 78% of its own franchise to a 30%-discounted generic, swung to an operating loss — and the stock has more than doubled off the March low to sit 16% below its all-time high. Both halves of that sentence are true, and the market is currently only reading the second. What actually happened is that a cash-generative orphan monopoly became a price-taking generic business with a well-funded pipeline attached, and the equity is now priced as though only the pipeline exists. At ~$9.9bn of enterprise value against ~$769m of trailing revenue (12.8x, above every full-year close in eight years) and a 91st-percentile price/sales versus its own decade of history, the price already capitalises roughly $2.2–2.8bn of eventual revenue — which is close to management’s own end-of-decade ambition of $3bn+, taken at nearly face value and discounted only lightly for the fact that the FDA has already said no once.

The framing is neither momentum nor falling knife: the factor model assigns CORT zero loading to Momentum, Value, Quality and Growth, an R² of 12.8%, and 74.7% annualised idiosyncratic volatility. Roughly 93% of this stock’s variance is its own regulatory calendar. What looks like a trend is a run of binary outcomes landing favourably — ROSELLA’s overall-survival hit, an early Lifyorli approval, an NDA resubmission — and coin-flips, unlike momentum, do not persist. The lifetime record is the tell: 22 years listed, a 0.195 Sharpe, and an 85% maximum drawdown. This security has given everything back before. Against that, two things genuinely argue the other way and keep this a HOLD rather than an outright short-and-forget: director G. Leonard Baker bought 100,000 shares for $3.31m at ~$33 on 2026-03-17 — the only open-market insider purchase in five years, executed four days after the low — and Lifyorli’s ROSELLA data (OS hazard ratio 0.65, p=0.0004, all-comers, no biomarker test) is the real thing in a disease with almost nothing else. But note who is not buying: the CFO, the Endocrinology President and the Chief Development Officer have realised ~$78m between them and hold 16,130, 9,755 and 3,985 shares respectively.

Conviction: medium. Flips bullish: relacorilant approved in Cushing’s at the ~December 2026 PDUFA and Q2/Q3 revenue printing above ~$250m, proving the guidance hockey-stick is fulfilment recovery rather than hope. Flips bearish: a second Complete Response Letter in December — the resubmission contains re-analysed old data, not a new trial, and it is the same evidence base the FDA already declined — or a nine-figure settlement in the Teva/payor antitrust actions, against which zero is currently accrued.

📈 Stock Price Action — Five-Year Event Map

Over five years CORT has traded from a low of $16.53 (2022-01-27) to a high of $114.22 (2025-03-31) and sits today at $95.36, 16.5% below that high, with a 52-week range of $32.15–$96.66. The stock is above its 21-, 50- and 200-day EMAs ($89.04 / $79.20 / $64.12). The shape of the last eighteen months is unusual even for biotech: a one-day double, a two-day 56% collapse, and a four-month 197% recovery — three of the largest moves in the company’s listed history compressed into sixteen months.

Cycle narrative. (1) For two and a half years CORT was a slow-compounding orphan cash generator with an unresolved patent case hanging over it, and the market paid it 4–7x sales accordingly. (2) The single day the patent risk crystallised — the first trading session of 2024, after Judge Bumb’s December 2023 non-infringement ruling — took a quarter off the equity. (3) It recovered anyway, because Corcept pre-empted Teva by launching its own authorized generic in June 2024 and because CATALYST reframed hypercortisolism from a rare disease into a large undiagnosed one. (4) 2025-03-31 is the most consequential day in the stock’s history: ROSELLA’s progression-free-survival hit converted a single-product endocrinology company into an oncology story, and the stock doubled in one session on 14x normal volume. (5) It then faded for nine months as the specialty-pharmacy channel broke down and the company failed to convert a 61% increase in new prescriptions into more than a 37% increase in tablets shipped. (6) On 2025-12-30–31 the FDA declined to approve relacorilant in Cushing’s — the asset the whole endocrinology bull case rested on — and the equity halved in a single session; the 2025-12-31 move was a −49.9% idiosyncratic return in the factor model, i.e. essentially none of it was market or sector. (7) Seven weeks later the Federal Circuit affirmed that Teva’s generic does not infringe, removing the last legal defence of the base business, and the stock bottomed at $32.15. (8) Since late March the sequence has reversed: Lifyorli approved three and a half months early, two-year ALS survival data, and a Cushing’s NDA resubmission — a 197% advance off the low. (Price moves are FACT; attributed causes are INTERPRETATION. No price target, no recommendation.)

1. Executive Summary

Corcept Therapeutics sells one molecule. Mifepristone 300mg — branded Korlym, approved in 2012 for hyperglycaemia secondary to hypercortisolism, and since June 2024 also sold by Corcept as its own authorized generic — generated 100% of the company’s $761.4m of FY2025 revenue. Around that single product the company has built a pipeline of more than 1,000 proprietary selective cortisol modulators across four chemical series, of which relacorilant is the lead and now, as Lifyorli, a second commercial product.

The last eighteen months have been the most consequential in the company’s history, in both directions. On the positive side, the pivotal ROSELLA trial of relacorilant plus nab-paclitaxel in platinum-resistant ovarian cancer met both of its dual primary endpoints — progression-free survival (March 2025) and overall survival (January 2026, hazard ratio 0.65, p=0.0004, median OS 16.0 versus 11.9 months) — and the FDA approved Lifyorli on 2026-03-25, three and a half months ahead of its PDUFA date. Corcept’s own prevalence trials, CATALYST and MOMENTUM, established that 24% of patients with difficult-to-control type 2 diabetes and 27% of patients with resistant hypertension have hypercortisolism, materially expanding the diagnosable population.

On the negative side, three things broke. First, on 2025-12-30 the FDA issued a Complete Response Letter declining to approve relacorilant in Cushing’s syndrome despite the pivotal GRACE trial having met its primary endpoint; the stock fell 56% in two sessions. Second, on 2026-02-19 the Federal Circuit affirmed that Teva’s generic Korlym does not infringe Corcept’s method-of-use patents, terminating the last legal barrier around the base business — Sun and Hikma hold settlements permitting them to follow. Third, and least discussed, the economics inverted: selling, general and administrative expense rose from $280.3m (FY2024) to $448.7m (FY2025) to a $145.4m quarterly run-rate in Q1-2026, at which point SG&A equalled 88% of revenue and the company posted a $49.6m operating loss and negative operating cash flow.

The competitive position is weaker than the 98.3% gross margin suggests. The composition-of-matter patent on mifepristone has expired; the method-of-use patents have been held not to be infringed; Corcept’s authorized generic now represents ~78% of volume at roughly a 30% discount to Korlym’s list price. FY2025 revenue growth decomposes into volume +37.0% and realised price −17.7%. The company is holding units by cannibalising its own price — the signature of a franchise without pricing power.

What remains is a genuine but non-excludable asset: Corcept spends heavily to expand the diagnosed hypercortisolism population, and every patient it diagnoses is then addressable by Teva’s generic, by Recordati’s Isturisa, by Xeris’ Recorlev, and prospectively by Crinetics’ atumelnant — now backed by Vertex’s balance sheet following the July 2026 acquisition agreement. Corcept creates the market; competitors free-ride on it.

Capital allocation is mixed and tilting negative. The company has never raised dilutive equity, carries no debt, and self-funds a deep pipeline — genuinely admirable. But it exhausted 94% of a $200m buyback authorisation at an average of $66.71 in FY2025 and repurchased nothing in Q1-2026 with the stock between $32 and $47. CEO compensation rose 54.7% to $15.3m in a year when operating income fell 67%; the largest component, option awards of $12.9m, vests on the passage of time with no performance condition and no return-on-capital metric. Over five years insiders sold $147.9m and bought $3.3m — the single purchase being director G. Leonard Baker’s 100,000 shares at ~$33 on 2026-03-17.

Five adversarial proceedings and one federal criminal/civil investigation are outstanding, including two treble-damages antitrust actions (Teva, and a payor group led by Aetna) predicated on the same exclusive-specialty-pharmacy conduct, and a False Claims Act qui tam in which the DOJ declined to intervene in May 2026. Nothing is accrued against any of them.

At ~$9.9bn of enterprise value the market is capitalising roughly $2.2–2.8bn of eventual revenue — close to management’s own $3bn+ end-of-decade ambition taken near face value. The next binary is the relacorilant Cushing’s PDUFA expected around December 2026, on a resubmission containing re-analyses of the same dataset the FDA already declined once. Q2-2026 results are due 2026-07-29.

2. Business Overview

2.1 What the company actually sells

Corcept is a commercial-stage biopharmaceutical company whose entire scientific thesis is a single mechanism: modulating the effects of cortisol at the glucocorticoid receptor (GR). The insight, pursued since the company’s founding in 1998, is that excess cortisol activity — rather than excess cortisol production — drives a wide range of pathology, and that a drug which competes with cortisol at the receptor can blunt that pathology without destroying the body’s ability to make cortisol at all. This is the distinction Corcept draws between its approach and the cortisol-synthesis inhibitors (ketoconazole, osilodrostat, levoketoconazole) that dominate the treatment landscape.

Commercially, until March 2026 the company sold exactly one thing:

• Korlym® (mifepristone) 300mg tablets, FDA-approved in 2012 for the treatment of hyperglycaemia secondary to hypercortisolism in adults with endogenous Cushing’s syndrome who have type 2 diabetes mellitus or glucose intolerance and have failed surgery or are not surgical candidates.
• An authorized generic of Korlym, made available by Corcept itself in June 2024 — the same molecule, same indication, sold at approximately a 30% discount to Korlym’s wholesale acquisition cost.

The 10-K refers to these two collectively as “our Products.” They are not two products in any economic sense; they are one product sold at two prices, and the mix between them is the single most important driver of realised revenue per tablet.

Since 2026-03-25 there is a second, genuinely distinct commercial asset:

• Lifyorli™ (relacorilant), approved in combination with nab-paclitaxel for adults with platinum-resistant epithelial ovarian, fallopian tube or primary peritoneal cancer who have received one to three prior systemic regimens, at least one of which included bevacizumab. List price $37,900 per 28-day cycle (nine doses).

2.2 Revenue segmentation and business model

Corcept reports a single operating segment; the Chief Executive Officer, Dr Joseph Belanoff, is the Chief Operating Decision Maker. There is no geographic segmentation of consequence — sales are effectively all United States, with a Marketing Authorization Application for Lifyorli pending with the EMA and a likely decision in Q4-2026.

The model is a classic orphan-drug specialty-pharmacy model:

1. A field sales force calls on physicians treating hypercortisolism, supported by medical science liaisons.
2. Product ships through a specialty distributor and, until 2026, a single exclusive specialty pharmacy.
3. Patient support programmes and donations to independent charitable foundations underwrite co-pays and deductibles. Corcept’s stated policy is that “no patient with hypercortisolism will be denied access to our Products for financial reasons,” and it supplies product free to uninsured patients who do not qualify for charitable support.
4. Net product revenue is gross revenue less government rebates and chargebacks, co-pay assistance, prompt-payment discounts to the distributor, and returns reserves. Government rebates are material and growing — the provision related to current-period sales rose from $52.8m (2023) to $86.3m (2024).

Charitable-foundation donations are not netted against revenue; they are recorded as a deduction to SG&A. This is disclosed and conventional, but it means the SG&A line contains an item that is economically closer to a revenue contra than to selling expense.

2.3 The specialty-pharmacy rupture

From 2017 until 2025, Optime Care, Inc. was the exclusive specialty pharmacy for both Korlym and the authorized generic. This arrangement is now simultaneously a historical operating problem and the factual predicate for two antitrust lawsuits.

The operating problem: Optime could not keep up. Management disclosed on the Q4-2025 call that FY2025 produced a 61% increase in new prescriptions but only a 37% increase in tablets sold — the gap, in the President of Endocrinology’s words, “an illustration of the lack of capacity at our pharmacy vendor.” The 10-K states flatly that FY2025 “net product revenue would have been materially higher had our primary specialty pharmacy vendor been able to fully meet demand.”

The remedy created a second problem. Corcept notified Optime in June 2025 that it would cease to be exclusive from September 2025, terminated the agreement in October 2025 effective 2026-01-08, and transferred substantially all pharmacy services to Curant Health Georgia, LLC in Q4-2025. Transferring thousands of patients’ prescriptions and medical files “disrupted our business in November, December and January,” and left a prior-authorisation backlog that management said was still being worked through as of the Q1-2026 call in May.

Management now says it will expand beyond a single pharmacy in Q4-2026 — an admission that single-vendor exclusivity was an operating vulnerability as well as a legal one.

2.4 Recurring versus non-recurring

Revenue is recurring in the useful sense: hypercortisolism is a chronic condition and Korlym is a chronic daily therapy, so the revenue base is a patient book rather than a transaction stream. Patients titrate to dose over months, which is why management argues that new patient starts understate near-term revenue and overstate near-term cost. Lifyorli, by contrast, is a finite-duration oncology therapy — median overall survival in ROSELLA was 16 months and treatment duration is shorter still — so its revenue is fundamentally a flow of new patient starts rather than an accumulating book. The two businesses have genuinely different revenue physics and should not be modelled the same way.

Verdict. A single-molecule chronic-therapy franchise with a real second product newly attached, run through a channel that has just been rebuilt under duress and is the subject of litigation. The description “commercial-stage biopharmaceutical company” flatters it: for revenue purposes this is a one-drug company that has just added a second drug, with everything else a call option.

3. Industry Dynamics

Corcept now operates in two structurally different markets, and conflating them is the most common analytical error made about this company.

3.1 Hypercortisolism — an orphan market that Corcept is expanding and losing simultaneously

Structure. Endogenous Cushing’s syndrome was historically treated as a genuine rare disease — a few thousand diagnosed US patients, high-priced therapy, minimal competition, high barriers. The named approved competitors are few: Signifor (pasireotide) and Isturisa (osilodrostat), both marketed by Italy’s Recordati S.p.A., and Recorlev (levoketoconazole) from Xeris Biopharma. Off-label use of ketoconazole (an antifungal) and metyrapone (a diagnostic agent) is common and cheap. Since January 2024, Teva’s generic mifepristone competes directly.

Market size — and the reframing. The economically interesting fact about this industry is that Corcept has, at its own expense, redefined its size. The CATALYST trial screened 1,000 patients whose type 2 diabetes remained uncontrolled despite best care including GLP-1 agonists, and found 24% had hypercortisolism. Its randomised treatment phase (136 patients) showed a 1.47% mean HbA1c reduction on Korlym versus 0.17% on placebo at 24 weeks, alongside reductions in weight and waist circumference, with many patients reducing or discontinuing other glucose-lowering agents. Results were published in Diabetes Care in December 2025 and referenced in the March 2026 AACE diabetes-management guidance. The MOMENTUM trial then found 27% of resistant-hypertension patients had hypercortisolism, presented at the American College of Cardiology in March 2026.

If those prevalence figures translate into clinical practice, the addressable population is not a few thousand patients but a subset of the millions with resistant diabetes and resistant hypertension. That is a genuine, evidence-backed TAM expansion, and it is Corcept’s most valuable non-molecular asset.

Profit pool and competitive intensity. Here is the problem. The profit pool Corcept is expanding is not defensible by Corcept. The composition-of-matter patent on mifepristone has expired. Teva sells the same molecule. Sun Pharmaceutical and Hikma hold litigation settlements permitting them to sell mifepristone once FDA-approved, conditional on Teva’s product remaining commercially available — so the structural expectation is three or four generic mifepristone sellers, not one. Every newly diagnosed hypercortisolism patient that Corcept’s trials create is a patient any of them can serve at a lower price.

Corcept’s response — launching its own authorized generic in June 2024 — is the correct defensive move and is also an admission of what has happened. By end-2025 the authorized generic was ~75% of volume; by 2026 it is ~78%, at roughly a 30% discount to Korlym’s WAC. The company has voluntarily repriced three-quarters of its franchise downwards to keep the units.

Regulatory landscape — an idiosyncratic overhang. Mifepristone’s active ingredient is approved in another product for the termination of early pregnancy. The 10-K devotes substantial space to the consequences: post-Dobbs state legislation restricting mifepristone, the June 2024 Supreme Court standing ruling that preserved access on procedural grounds only, and — most concretely — that in September 2025 HHS announced it will re-examine the safety of mifepristone. The filing states there “can be no assurance this re-examination will not result in restrictions on the distribution of mifepristone for any use, including the treatment of patients with hypercortisolism.” This is a political risk to a revenue line, wholly outside management’s control, with no obvious hedge.

3.2 Platinum-resistant ovarian cancer — a small, ugly, under-served market

Structure. Platinum-resistant epithelial ovarian cancer — disease that recurs or progresses within roughly six months of platinum-based chemotherapy — is one of oncology’s genuinely bad neighbourhoods. Prognosis is poor, single-agent chemotherapy response rates are low, and the standard of care (weekly paclitaxel or nab-paclitaxel, sometimes with bevacizumab) has been static for years. Mirvetuximab addresses only folate-receptor-α-high patients; KEYTRUDA (pembrolizumab) is limited to PD-L1 CPS≥1, which management sizes at 50–60% of patients by biology but only 35–40% after accounting for actual testing rates.

Where Lifyorli sits. ROSELLA is an unusually clean dataset for this setting: both dual primary endpoints met, an overall-survival hazard ratio of 0.65 (p=0.0004), median OS 16.0 versus 11.9 months, safety essentially unchanged versus nab-paclitaxel alone (management reports half the ALT elevations in the combination arm), oral administration, and — commercially decisive — no biomarker test required. Inclusion in NCCN guidelines as a preferred regimen 15 days after approval is a meaningful access signal.

Profit pool. The population is small. Management targets 5,000 US physicians who account for ~90% of volume and expects to become market leader “in a relatively short time frame,” guiding to “more than $1 billion in annual revenue in the United States by the end of the decade.” At $37,900 per 28-day cycle and a realistic treatment duration, that implies a large share of a modest patient population — achievable, but it requires near-dominance rather than participation.

Capital-cycle read. Applying the Marathon lens: Corcept’s endocrinology business sits past the peak of its capital cycle — high historical returns attracted a generic entrant, price is falling, and the incumbent is spending heavily to defend units. Its oncology business sits at the start of one — a newly opened profit pool with a first mover, no direct mechanistic competition, and returns that will attract capital only after others replicate the biology. The two are moving in opposite directions inside one P&L, which is precisely why the consolidated numbers look confusing.

Verdict: a structurally deteriorating core industry and a structurally attractive but small new one. Orphan endocrinology was a good industry and, for Corcept specifically, has stopped being one — the molecule is off-patent, the method-of-use defence has failed in court, and three additional generic sellers are contractually queued behind Teva. Platinum-resistant ovarian cancer is a hard, small market in which Corcept has an unusually strong and genuinely differentiated position. The honest summary is that the good industry is the small one.

4. Competitive Position

4.1 Naming the moat — and finding it mostly gone

Applying the Greenwald taxonomy rigorously, there are exactly three sources of genuine competitive advantage: supply-side cost advantage, demand-side customer captivity, and economies of scale combined with captivity. Corcept’s position must be tested against each.

Supply-side cost advantage: absent. Mifepristone is a contract-manufactured small molecule. Cost of sales was $13.0m on $761.4m of revenue in FY2025 — a 98.3% gross margin. That margin is a pricing phenomenon (orphan-drug pricing in a channel with charitable co-pay support), not a cost phenomenon. Teva’s manufacturing cost is not materially higher than Corcept’s; indeed Teva, as one of the world’s largest generic manufacturers, plausibly has the advantage. There is no cost moat here and there never was.

Demand-side captivity: weak and weakening. Switching costs for a once-daily oral tablet are close to nil at the molecule level — a pharmacist can substitute a generic. What captivity exists is procedural, not pharmacological: patients are enrolled in a specific specialty-pharmacy programme with a specific patient-support apparatus, prior authorisations are tied to that channel, and physicians default to the pathway they know. That is real friction, and Corcept has clearly exploited it — Teva’s antitrust complaint is essentially the allegation that Corcept exploited it unlawfully. But procedural captivity is fragile, is now under legal attack from two directions, and was materially disrupted by Corcept’s own forced pharmacy migration.

Economies of scale plus captivity: no. With 730 employees and $761m of revenue there is no scale barrier that a Teva, a Recordati or a Vertex cannot clear.

4.2 What Corcept actually has

Stripping away what is not a moat leaves two things that are real.

First, a diagnosis-generation franchise. Corcept has spent years and hundreds of millions of R&D dollars proving that hypercortisolism is far more prevalent than believed — CATALYST, MOMENTUM, the Diabetes Care publication, the AACE guidance citation, the ACC oral presentation. No competitor has done this work. Corcept has the field force, the medical science liaisons and the relationships to convert that evidence into screening behaviour. This is a genuine, expensively-built asset.

It is also, critically, non-excludable. Corcept cannot patent the finding that 24% of resistant diabetics have hypercortisolism. Once a physician screens and diagnoses, the prescription can be filled with Teva’s mifepristone at a lower price, or with Isturisa, or with Recorlev. Corcept bears the full cost of market creation and captures only a share of the resulting value — and a declining share, at a declining price. The FY2025 numbers are the arithmetic proof: volume +37.0%, realised price −17.7%, and SG&A up 60% to fund it. That is an expensive way to grow a market for one’s competitors.

Second, a proprietary chemical library. More than 1,000 compounds across four structurally distinct series, with composition-of-matter patents on relacorilant and its successors running to 2033–2042 in oncology and 2033–2040 in hypercortisolism. This is the durable asset. Crucially, it is not the asset generating today’s revenue — it is the asset that would generate tomorrow’s, if relacorilant is ever approved in Cushing’s and if Lifyorli’s label expands. The moat, such as it is, sits entirely in the future.

4.3 Head-to-head

The Crinetics/Vertex item deserves emphasis because it is invisible in Corcept’s own disclosure. Prior /Claude work on Crinetics documented atumelnant’s Phase 2 data in congenital adrenal hyperplasia — near-complete androgen normalisation with glucocorticoid tapering to physiologic doses — and its parallel Cushing’s programme. Vertex is paying $85/share, roughly 1.5x the combined peak-sales estimate for the two assets, precisely because it believes that biology. Corcept’s Cushing’s franchise, already genericised, will in three to five years plausibly face a differentiated, mechanistically distinct, Vertex-funded competitor.

4.4 The ROIC test

Greenwald’s test for whether a moat exists is whether returns on invested capital persistently exceed the cost of capital. Corcept’s history says the moat was real: ROIC of 22.3% (2022), 18.1% (2023), 20.0% (2024), with ROE of 41.3% (2022), 30.4% (2023) and 29.8% (2024). Those are unambiguous excess returns.

They are also gone. ROE fell to 16.8% in FY2025 — and that figure flatters, because FY2025 net income of $99.7m exceeds pre-tax income of $66.5m thanks to a $33.2m tax benefit. In Q1-2026 the company generated a −30.1% operating margin. Trailing-twelve-month operating income is negative $8.2m. On the one test Greenwald says matters, the moat has stopped showing up in the financials — exactly at the point when the patent defence failed and the generic mix passed three-quarters.

Verdict: no durable competitive advantage in the current revenue base; a real but unproven one in the pipeline. The 98% gross margin is a price, not a moat. The diagnosis-generation franchise is genuine and valuable but cannot be fenced. The chemical library is the only asset that would deteriorate a competitor’s position if Corcept lost it — and it produces almost no revenue today. An investor buying CORT at $95 is buying the library, not the business.

5. Growth History and Forward Opportunities

5.1 The historical record

Two things stand out. First, the seven-year revenue CAGR to 2025 is 17.1% — genuinely good for a single-product company. Second, operating margin has fallen in every single year since 2019, from 36.4% to 5.9%. Revenue tripled and operating profit fell by half. Growth has been bought, progressively more expensively, every year.

5.2 Organic versus acquired, and the quality of the growth

All of it is organic. Corcept has made no material acquisitions in its history. That is a genuine credit — this is not a roll-up dressing up purchased revenue as growth.

But the composition of recent growth is poor. FY2025’s +12.8% decomposes into volume +37.0% and average realised price −17.7%. Q1-2026’s +4.9% decomposes into volume +12.4% and price −6.7%. Corcept is selling substantially more tablets for progressively less money per tablet, because it is deliberately migrating its book to a 30%-discounted authorized generic to keep Teva out. Management confirmed the AG reached ~75% of volume at end-2025 and ~78% in 2026, and described the mix as “essentially stabilised” — which, if true, means the price headwind largely annualises out from here. That is the single most important thing to verify in the next two prints.

There is also the fulfilment gap. FY2025 saw new prescriptions +61% against tablets sold +37% — roughly 24 percentage points of demand that the specialty pharmacy could not convert. The 10-K states revenue “would have been materially higher” absent the constraint. This cuts both ways: it is evidence of genuine underlying demand, and it is evidence that the company’s own operating infrastructure destroyed a material amount of revenue for at least a year.

5.3 Forward opportunities

Endocrinology — the CATALYST/MOMENTUM conversion. Management expects the Cushing’s syndrome business to reach “at least $2 billion in annual revenue by the end of this decade.” The mechanism is physician behaviour change: if AACE guidance and ACC presentations move screening practice, the diagnosed population expands materially. This is the largest single line in any bull case and it depends on something Corcept does not control — clinical practice adoption — and on Corcept capturing that expansion despite selling a genericised molecule.

Relacorilant in Cushing’s. The NDA was resubmitted 2026-06-17 with “additional data analyses”; management expects a PDUFA date approximately six months later, i.e. ~December 2026. Approval would matter enormously: it would give Corcept a patented Cushing’s drug (composition-of-matter to 2033–2040) plus seven years of US orphan exclusivity, restoring branded pricing to a franchise that has just been reset to generic economics. This is the pivotal question for the equity.

Lifyorli label expansion. Corcept is running relacorilant plus nab-paclitaxel in platinum-sensitive ovarian (with bevacizumab, the BELLA trial’s first arm), endometrial, cervical (a Phase 2 collaboration with Arcagy Research) and pancreatic cancer. Management states successful results “would immediately increase the number of patients that relacorilant might potentially help by fivefold.” BELLA’s first arm reads out by end-2026; the remainder by end-2027.

Other franchises. Three genuine shots on goal, all binary:

• dazucorilant in ALS — the Phase 2 DAZALS trial (249 patients, double-blind, placebo-controlled) showed an 84% reduction in risk of death at one year (p=0.0009) and 87% at two years (p<0.0001) at the 300mg dose. Management is explicit that the drug “does not appear to prevent functional decline. It prevents early death.” A Phase 3 with overall survival as the primary endpoint is planned to start in 2026. Caveat: the DAZALS survival findings were exploratory analyses, not the trial’s primary endpoint, and gastrointestinal tolerability drove most discontinuations.
• miricorilant in MASH — the 175-patient Phase 2b MONARCH study is fully enrolled with results by end-2026.
• nenocorilant plus nivolumab — Phase 1b in solid tumours, testing whether GR antagonism augments immunotherapy.

Verdict: high-volume, low-quality growth in the base business; high-quality but entirely unproven growth in the pipeline. Seven years of 17% compound revenue growth is real, but it has been achieved with monotonically deteriorating margins and, most recently, by trading price for volume. The forward opportunities are numerous and genuinely interesting — this is not a company short of shots on goal — but essentially every one of them is a binary event with a date attached rather than an extrapolation of anything currently happening. Investors should be clear that they are underwriting a schedule of catalysts, not a growth rate.

6. Financial Quality

6.1 The income statement has inverted

The trajectory is unambiguous. Over nine quarters revenue rose 12.3% while SG&A rose 158.2%. Operating income went from $29.5m to negative $49.6m. In Q1-2026, SG&A alone was 88.1% of revenue.

Management’s explanation is straightforward and is probably true as far as it goes: the company built two commercial organisations at once — an expanded endocrinology field force ahead of an expected relacorilant approval, and an entirely new oncology division ahead of Lifyorli. The oncology build was, in the event, well-timed (approval came 3.5 months early and the team was ready). The endocrinology build was spent against an approval that did not arrive. The 10-K guides that SG&A will be “higher in 2026 than in 2025,” and R&D likewise.

6.2 Quality of earnings — read the tax line

FY2025 requires care. Reported net income was $99.7m. Pre-tax income was $66.5m. The difference is an income tax benefit of $33.2m, versus a $20.3m tax expense in FY2024. MD&A attributes it “primarily due to increased stock compensation deductions and decreased pretax income.”

So FY2025 net income exceeded pre-tax income by 50%, and the year-on-year EPS decline (diluted $1.24 → $0.83, −33%) materially understates the operating deterioration (operating income $137.0m → $44.8m, −67%). Anyone anchoring on reported EPS is looking at a number flattered by option-exercise tax deductions generated by insiders selling stock. The same dynamic recurs in Q1-2026: a pre-tax loss of $44.7m becomes a net loss of $31.8m after a $13.0m benefit.

Stock-based compensation is large and growing: $85.0m in FY2025 (11.2% of revenue), up from $61.7m and $49.1m. Of that, $61.6m sat in SG&A. Diluted shares (119.99m in FY2025) exceed basic shares (103.86m) by 15.5% — a substantial and persistent option overhang. When the company posts a loss, as in Q1-2026, that overhang is anti-dilutive and disappears from the share count, making loss-per-share look better than the economics warrant.

6.3 Cash flow

The business is genuinely capital-light — cumulative five-year capex is under $4m. Cash conversion has historically been excellent, with cash flow to net income above 1.4x in most years, helped by non-cash SBC. But FY2025 free cash flow fell 27.7% despite revenue growth, and Q1-2026 turned cash-negative. Q1 is seasonally the weakest quarter (payer reauthorisation, free bridging drug), so one negative quarter is not a trend — but it is the first negative operating quarter in many years and it arrived alongside the highest SG&A quarter in the company’s history.

6.4 Balance sheet

As of 2026-03-31:

Net cash of roughly $506m, no financial debt, a 2.86x current ratio and a tangible common equity ratio of 78.3%. This is a genuinely strong balance sheet and it is the reason the December 2025 CRL was survivable rather than existential. Cash and investments did decline from $603.2m (end-2024) to $532.4m (end-2025) to $515.4m — but that reflects $245.9m spent acquiring shares in 2025, not operating burn.

The balance sheet’s one soft spot is what is not on it. Two treble-damages antitrust actions and a False Claims Act qui tam are outstanding, and the 10-K confirms: “No such amounts are accrued as of December 31, 2025.” A settlement in the low hundreds of millions — not an unusual outcome for pharmaceutical antitrust and FCA matters of this shape — would consume a meaningful fraction of the net cash position.

6.5 Returns on capital

Through 2024 these were the returns of a business with a real franchise. FY2025’s 16.8% ROE is inflated by the tax benefit; on a pre-tax basis the return on equity would be roughly 10%. On trailing-twelve-month operating income of negative $8.2m, ROIC is not meaningfully computable. The economics that justified the historical multiple have, for now, stopped existing.

Verdict: economics have deteriorated with scale, not improved. Incremental operating margin in FY2025 was −106.7% — every incremental revenue dollar came with roughly $1.95 of incremental SG&A. The gross margin is spectacular and irrelevant; what matters is that the operating margin has fallen every year for six years and is now negative. The balance sheet is strong and unencumbered, which buys time and optionality. But the honest characterisation of FY2025–Q1-2026 is that Corcept spent its entire accumulated operating profitability building commercial infrastructure for two products, and the FDA approved one of them.

7. Capital Allocation

7.1 The record

Corcept has never paid a dividend, never made a material acquisition, never issued debt, and never raised dilutive equity — the company has funded itself from product revenue since 2015. In an industry defined by serial dilution, that is a genuine and unusual credit, and it deserves to be stated plainly before the criticisms.

Uses of capital have been three: R&D, the commercial build, and share repurchase.

R&D. Spending rose from $113.9m (2021) to $254.9m (2025), funding four structurally distinct compound series across hypercortisolism, oncology, MASH, ALS and neuropsychiatry — more than 30 studies by management’s count. The strategic logic is coherent: one mechanism, many indications, each a call option on the same underlying science. The results are mixed but not embarrassing — one approval (Lifyorli), one CRL (relacorilant in Cushing’s), one strong Phase 2 survival signal (dazucorilant), one fully-enrolled Phase 2b (miricorilant). For a self-funded company this is a defensible allocation of a quarter-billion dollars a year.

Commercial build. This is where judgement is more questionable. SG&A rose 60% in FY2025 and a further 60% year-on-year in Q1-2026, to a run-rate approaching $600m annually against $761m of FY2025 revenue. A large portion was spent standing up a Cushing’s commercial organisation sized for a relacorilant launch that has not happened and may not happen until 2027 at the earliest. The oncology build, by contrast, was vindicated — the team was trained and ready when approval arrived early, product was available within five days, and 200+ physicians prescribed within the first month. One of the two bets paid; both were funded at full size simultaneously.

7.2 The buyback: pro-cyclical

The Board authorised $200m in January 2024. Corcept spent $188.6m of it — 94% — before the stock fell by half, at an average of $66.71. In Q1-2026, with the stock at its lowest level in two years, roughly $11m of authorisation remained and the Board did not re-authorise. The company held $515m of cash and investments throughout.

In pure hindsight, $66.71 was a fine price — the stock is $95 today. But the process was pro-cyclical in the textbook sense: maximum buying when the pipeline narrative was strongest and the multiple richest; nothing when the price was lowest and the balance sheet was still intact. Capital allocation is judged on the decision rule, not the outcome, and the decision rule here was “buy when things feel good.” A counter-cyclical rule would have left dry powder for March 2026.

The gross figure is larger still: FY2025 total cash spent acquiring shares was $245.9m, of which $56.8m related to net option exercises and $16.1m to RSU tax withholding — i.e. roughly $73m of the buyback budget was, in substance, funding insider option monetisation rather than reducing the share count for continuing holders. Shares outstanding nonetheless rose from 105.1m (end-2024) to 107.3m (Q1-2026).

7.3 Compensation and incentive alignment

Three observations.

First, the quantum. Every named executive’s total compensation rose 55–59% in FY2025, a year in which operating income fell 67%, the FDA issued a Complete Response Letter on the lead asset, and the stock closed the year down 31%. Option awards to the four non-CEO officers rose 84% each.

Second, the structure. The proxy asserts that “except for their salaries, compensation for our named executive officers is performance-based.” The option grants “vest over a four-year period, subject to the recipient’s continued employment.” There is no disclosed performance condition, no relative-TSR modifier, no revenue or margin hurdle, and no return-on-capital metric anywhere in the programme. Options do have a real performance link — they are worthless if the stock falls below strike — but time-vesting options are a directional incentive, not a capital-efficiency one, and they reward volatility. For a company whose central capital-allocation question is whether to keep spending SG&A at 88% of revenue, the absence of any return metric in the incentive scheme is a meaningful governance gap.

Third, the framing. The proxy’s “Executive Summary” of 2025 accomplishments lists the ovarian-cancer NDA, ROSELLA, the EMA filing, trial initiations and “strengthened our global patent portfolio.” It does not mention the 2025-12-30 Complete Response Letter or the Federal Circuit loss. Both occurred within the compensation year. Selective framing in a proxy is common; it is nonetheless the document in which the Board explains why it raised the CEO’s pay 55%.

Say-on-pay received 94% support at the 2025 meeting (on FY2024 pay). The FY2025 vote is at the 2026 meeting.

7.4 Insider behaviour

The full five-year Form 3/4/5 corpus — 309 of 309 filings parsed — yields a stark picture:

• Total open-market purchases: 2 transactions, $3,313,809. Both by director G. Leonard Baker, Jr. on 2026-03-17: 100,000 shares at a weighted average of ~$33.14, four days after the 52-week low and eight days before the Lifyorli approval. Baker reported 1,146,631 shares held afterwards.
• Total sales: 225 transactions, $147.9m. By year: 2022 $0.8m · 2023 $7.3m · 2024 $9.8m · 2025 $96.3m · 2026-YTD $33.8m.
• Largest sellers: Maduck $42.1m, Belanoff $29.3m, Lyon $26.7m, Guyer $26.2m, Mokari $9.6m.
• Resulting holdings are the striking part. After realising those sums, the CFO holds 16,130 shares, the President of Endocrinology 9,755, and the Chief Development Officer 3,985. The pattern is cashless option exercise followed by immediate and near-total disposal.

Virtually all 2026 officer sales carry 10b5-1 designations, which is proper compliance and mitigates the timing inference. It does not mitigate the level inference: executives who exercise and sell essentially everything, retaining four-figure share counts against eight-figure realised gains, are not building an ownership stake. The CEO (2.57m shares) and two long-tenured directors (Baker 1.15m, Wilson 1.08m) are the exceptions, and Baker is the only person in five years who has bought.

Verdict: capital allocation is mixed, tilting negative. The refusal to dilute, the absence of debt, the avoidance of value-destroying M&A and the self-funding of a genuinely broad pipeline are real, durable credits that distinguish Corcept from most of its peer group. Against them: a pro-cyclical buyback that exhausted its authorisation at $66.71 and bought nothing at $32; a compensation programme with no capital-returns metric whose payouts rose 55%+ in a year of a 67% operating-income decline and a regulatory rejection; and an executive team that has sold $148m and bought $3m in five years. The Board is not allocating badly so much as it is not allocating counter-cyclically, and it is paying for effort rather than for returns.

8. Changes and Headwinds — Last Two Years

8.1 The chronology that matters

January 2024 — Teva launches generic Korlym. Following Judge Bumb’s 2023-12-29 D.N.J. ruling of non-infringement, Teva launched. The stock fell 25.6% on the first trading day of 2024.

June 2024 — Corcept launches its own authorized generic. The correct defensive response, priced at roughly 30% below Korlym’s WAC. It preserved units and permanently reset realised price. By end-2025 the AG was ~75% of volume; by 2026, ~78%.

December 2024 – December 2025 — CATALYST and MOMENTUM. CATALYST’s prevalence phase (24% of resistant diabetics) and treatment phase (1.47% HbA1c reduction versus 0.17%) were published in Diabetes Care in December 2025 and referenced in AACE guidance in March 2026. MOMENTUM (27% of resistant hypertensives) was presented at ACC in March 2026. These are the strongest evidence in the bull case.

2025-03-31 — ROSELLA meets its PFS endpoint. The stock doubled in a session.

June – October 2025 — the specialty-pharmacy rupture. Optime de-exclusivised in June, terminated in October effective January 2026, replaced by Curant. FY2025 saw new prescriptions +61% against tablets +37%. The transition itself “disrupted our business in November, December and January.”

2025-12-30 — the Complete Response Letter. The FDA declined to approve relacorilant for hypertension secondary to hypercortisolism. Per the 10-K, “the letter acknowledged that our GRACE trial had met its primary endpoint and that our GRADIENT trial had provided confirmatory evidence,” but the FDA “stated that additional evidence of efficacy would be required.” Management’s Q4-2025 account adds that the CRL cited liver-enzyme elevation, and that it contained “an analysis at the end that they performed that we had never seen and really can’t replicate.”

2026-01-22 — ROSELLA meets its OS endpoint. Hazard ratio 0.65, p=0.0004; median OS 16.0 versus 11.9 months.

2026-02-19 — Federal Circuit affirms. No infringement of either the '214 or '800 patent. Management said it “plans to appeal.”

2026-03-25 — Lifyorli approved, 3.5 months early. NCCN preferred-regimen inclusion 15 days later; 200+ prescribers in the first month.

2026-05-19 / 06-03 — DOJ declines to intervene in the qui tam; the relator serves an amended complaint 2026-06-02.

2026-06-17 — relacorilant NDA resubmitted for Cushing’s, with additional analyses of existing data. PDUFA expected ~six months later.

8.2 The three structural headwinds

Permanent price reset. The move to a 78%-authorized-generic mix at a 30% discount is not cyclical. Management describes the mix as “essentially stabilised,” and if that holds the rate of price decline should moderate sharply — but the level is reset. Two further generic entrants (Sun, Hikma) are contractually queued.

A cost base built for an approval that did not arrive. SG&A of $145.4m in Q1-2026 against $164.9m of revenue was sized for two launches. One happened. Management guides SG&A higher again in 2026. This is a fixed-cost problem that only revenue growth solves — and the guidance implies precisely that.

A litigation stack with nothing accrued. Five proceedings plus a federal investigation. The Teva antitrust action goes to trial in March 2027; the payor action (Aetna, HCSC, Humana, Molina) alleges “substantially similar” conduct in California state court; the qui tam proceeds without DOJ but with FCA treble damages and per-claim penalties available; the NJ USAO investigation into Korlym promotion and payments to prescribers remains open with Corcept “within the scope”; and a securities class action covers 2024-10-31 to 2025-12-30. The mifepristone political overhang — HHS’s September 2025 safety re-examination — sits behind all of it.

Verdict: on balance these developments weaken the thesis, though not uniformly. The genuine strengthening is Lifyorli — a real approval, a real differentiated dataset, and a second revenue stream from a patented molecule. Everything else has gone the wrong way: the patent defence is exhausted, the base business is permanently repriced, the cost structure was built for an approval that was refused, and the legal exposure has broadened. The single most important forward item is the ~December 2026 PDUFA, and it is worth stating plainly that the resubmission contains re-analyses of the same dataset the FDA already reviewed and declined — not a new trial. That is a materially lower-probability path than a fresh randomised study, and it is what the market is currently pricing as likely to succeed.

9. Risk Analysis

The two risks that dominate. First, the December 2026 PDUFA: a second CRL would remove the entire premise for the Cushing’s SG&A build, strand a franchise that has already been repriced to generic economics, and — on the evidence of 2025-12-31 — could take half the equity. Second, the litigation stack: two treble-damages antitrust actions and an FCA qui tam, all predicated on the same exclusive-specialty-pharmacy conduct, all carrying multiplied damages, and all carrying a zero accrual. The March 2027 Teva trial date makes this a 2027 event, not a 2026 one, which is precisely why it is currently underpriced.

The risk that is overstated. Financial distress. With $506m of net cash, no debt and a capital-light model, Corcept can absorb several years of the current burn without financing. The equity has valuation risk, not solvency risk.

10. Valuation Discussion

No price target and no recommendation. This section describes what the current price requires.

10.1 Where the price sits

10.2 Against its own history

A valuation-percentile dataset — which ranks each multiple against roughly a decade of the stock’s own range — puts CORT in the 90.4th percentile on the composite, the 91.2nd on price/sales and the 82.8th on price/book. The 97.3rd-percentile P/E should be discounted per standing practice, since GAAP EPS is distorted by the collapse in operating income and the tax benefit; the price/sales percentile is the reliable read.

Year-end EV/Sales over eight years:

The current multiple exceeds every full-year close in eight years, by a wide margin. The only comparable level was intra-2025, when the stock briefly traded at 15.9x EV/Sales in the weeks after the ROSELLA PFS readout — a level it gave back entirely within nine months.

The critical point is what is being multiplied. In 2019–2024, 5–7x sales was applied to a business earning 20–36% operating margins. Today 12.8x is applied to a business with a negative trailing operating margin. On any earnings- or cash-based measure the re-rating is far larger than the sales multiple alone conveys.

10.3 Embedded expectations

Work backwards from $9.88bn of enterprise value.

Assume Corcept eventually stabilises at a 15–20x EV/EBIT multiple — reasonable for a specialty pharma with patent-protected products and modest growth — and a mid-teens to 20% sustainable operating margin. That is a generous margin assumption given SG&A is currently 88% of revenue and guided higher, and that the historical peak was 36% when the company had one product and a fraction of today’s cost base.

At 17.5x EV/EBIT and an 18% operating margin, $9.88bn of EV implies roughly $565m of sustainable EBIT, i.e. ~$3.1bn of revenue. Flex the assumptions — 20x and 22% margin — and the requirement falls to ~$2.25bn. Tighten them — 15x and 15% — and it rises to ~$4.4bn.

So the current price requires, on a discounted-to-today basis, something in the region of $2.2–3.1bn of eventual annual revenue at healthy specialty-pharma margins.

Management’s own stated ambition is “at least $2 billion” from Cushing’s syndrome plus “more than $1 billion” from US Lifyorli by the end of the decade — approximately $3bn+.

The market is therefore paying, today, close to the full present value of management’s own end-of-decade plan. It is discounting that plan only lightly for the four years of execution risk, for the fact that the FDA has already declined relacorilant once, for the possibility that the Cushing’s franchise is permanently generic, and for a litigation stack against which nothing is reserved.

10.4 The near-term arithmetic problem

FY2026 guidance is $950m–$1,050m, raised from $900m–$1,000m at Q1. Q1-2026 actual revenue was $164.9m.

Reaching the $1.0bn midpoint requires ~$835m across Q2–Q4 — about $278m per quarter. That is 69% above the Q1 run-rate and 34% above the highest quarterly revenue the company has ever recorded ($207.6m in Q3-2025). Management’s explanation is coherent — Q1 seasonality from payer reauthorisation, the prior-authorisation backlog from the pharmacy migration, new patients starting below full dose, and a Lifyorli launch that only began on 2026-03-25 with “almost all” of the guidance range still coming from endocrinology. But it is a very large step-up to deliver in three quarters, and the Q2 print on 2026-07-29 is the first genuine test.

10.5 Scenarios

Illustrative enterprise values, not price targets.

Bear (~$2.5–3.5bn EV). The December 2026 PDUFA produces a second CRL. Relacorilant never reaches the Cushing’s market, or does so only after a new randomised trial completing in 2029+. Sun and/or Hikma enter, compressing realised price further. Lifyorli plateaus at $300–400m as physician adoption proves slower than the first-month enthusiasm implied. An antitrust or FCA settlement lands in the $150–400m range. Sustainable revenue $700–900m at a mid-single-digit operating margin, plus ~$0.4bn of residual net cash.

Base (~$6.5–8.5bn EV). Relacorilant is approved in Cushing’s in 2026–27, restoring branded pricing and seven years of orphan exclusivity to a meaningful share of the endocrinology book. Lifyorli reaches $500–800m in the US with modest EU contribution. CATALYST/MOMENTUM-driven screening lifts endocrinology to $1.2–1.5bn. SG&A stops scaling and operating margin recovers to 20–25%. Revenue $1.8–2.3bn by 2029–30. Litigation settles for a manageable sum.

Bull (~$14–18bn+ EV). Management’s $3bn+ arrives on schedule. Lifyorli’s label expands into platinum-sensitive ovarian, endometrial and cervical disease on the 2026–27 BELLA and Phase 2 readouts — a fivefold increase in addressable patients on management’s own arithmetic. MONARCH succeeds in MASH and/or the dazucorilant Phase 3 replicates the DAZALS survival benefit, adding a third franchise. Corcept becomes a multi-franchise specialty pharma.

10.6 What the market is pricing correctly, and incorrectly

Correctly: the quality of the ROSELLA dataset and the value of an all-comer, no-biomarker oral agent in platinum-resistant ovarian cancer; the genuine TAM expansion from CATALYST and MOMENTUM; the strength of the balance sheet; the breadth of the pipeline; and the fact that Q1-2026’s weakness contained a real, recoverable fulfilment component.

Incorrectly, on this reading: the probability that a resubmission of re-analysed data clears an FDA that has already declined it; the permanence of the price reset in a franchise that is now 78% generic with two further entrants queued; the size of the fixed cost base relative to the revenue that must service it; and — most of all — an unreserved litigation stack containing two treble-damages antitrust actions and an FCA qui tam, with a March 2027 trial date that is currently beyond most investors’ horizon.

Verdict: the price embeds the plan working. At 12.8x trailing sales — above every year-end multiple in eight years, applied to a business with negative trailing operating income and a 91st-percentile price/sales versus its own decade — CORT offers little compensation for the bear case and prices the bull case close to par. The distribution of outcomes here is unusually wide, and the current price is not near the middle of it.

11. Variant Perception

11.1 The consensus

The prevailing view, visible in sell-side commentary and in the 197% advance off the March low, runs approximately as follows: the December CRL was an aberration by an FDA that “missed the mark”; Lifyorli’s early approval validates both the science and the commercial organisation; CATALYST and MOMENTUM have permanently expanded the Cushing’s market; the resubmission will be approved around December 2026; and management’s $3bn end-of-decade target is credible. On that view, guidance of $950m–$1.05bn for 2026 represents 25–38% growth and the stock is a growth compounder that briefly went on sale.

11.2 The strongest bull case

It is genuinely strong, and worth stating at full force.

Corcept owns a mechanism, not a molecule. Cortisol modulation at the glucocorticoid receptor has now produced: a marketed drug in Cushing’s for fourteen years; a positive pivotal trial in Cushing’s (GRACE); a positive dual-endpoint Phase 3 in platinum-resistant ovarian cancer with an OS hazard ratio of 0.65; an 87% two-year reduction in risk of death in ALS; and early liver-fat and fibrosis improvement in MASH. That is an extraordinary hit rate for one biological hypothesis, and Corcept owns more than 1,000 proprietary compounds across four chemical series to exploit it, with composition-of-matter protection into the 2040s.

The Cushing’s TAM has been genuinely and permanently redefined by Corcept’s own prevalence work — 24% of resistant diabetics, 27% of resistant hypertensives — and that evidence is now in AACE guidance and ACC presentations. Physician behaviour changes slowly but it does change. The fulfilment gap (new prescriptions +61% versus tablets +37%) proves the demand is real and was artificially suppressed; a working pharmacy converts that backlog into revenue. Lifyorli launched into NCCN preferred status with 200+ prescribers in month one. The balance sheet has $506m of net cash and no debt. And director Baker bought $3.3m of stock at the low.

11.3 The strongest bear case

The bear case is not “the science is wrong.” It is that the equity has been repriced from a cash-generating business to a pipeline option without the fundamentals having made that transition.

Corcept today earns negative trailing operating income on a molecule whose composition-of-matter patent expired, whose method-of-use defence failed in two courts, which the company itself now sells as a 30%-discounted generic representing 78% of volume, and behind which two further generic entrants are contractually queued. Every incremental FY2025 revenue dollar arrived with $1.95 of incremental SG&A. The cost base was built for two launches and got one.

The resubmission is re-analysed old data, not a new trial — the same evidence the FDA reviewed and declined, plus argument. Management’s own account concedes the CRL contained an FDA analysis it “had never seen and really can’t replicate,” which is not obviously a good sign about the strength of the agency dialogue.

The 2026 guidance requires quarterly revenue 34% above the all-time record for three consecutive quarters, starting from a quarter that was 21% below it.

And beneath all of it sits an unreserved litigation stack: two antitrust actions seeking treble damages on the same specialty-pharmacy conduct, an FCA qui tam proceeding after DOJ declination, an open federal criminal/civil investigation into Korlym promotion, and a securities class action — with $0 accrued and a March 2027 trial date.

Meanwhile the factor model reports zero momentum, value, quality and growth loadings, an R² of 12.8%, and 74.7% idiosyncratic volatility. This is not a trend; it is a sequence of coin-flips that has recently landed heads. The lifetime record — 0.195 Sharpe, −85.3% maximum drawdown over 22 years — is what happens when that sequence turns.

11.4 The 3–5 assumptions that decide it

1. Does relacorilant get approved in Cushing’s around December 2026? Bulls assume yes on re-analysis; bears note the FDA already said no on this dataset. This single question is worth several billion dollars of enterprise value.
2. Is the endocrinology price reset finished? Management says the AG mix is “essentially stabilised” at ~78%. If true, the price headwind largely annualises out. If Sun or Hikma enters, it does not.
3. Was Q1-2026 a fulfilment air-pocket or the new run-rate? The entire 2026 guidance depends on the answer. Q2 prints on 2026-07-29.
4. Does SG&A stop scaling? At 88% of revenue and guided higher, the operating leverage in every bull model requires this line to flatten while revenue grows. Nothing in the disclosure yet shows it flattening.
5. What do the antitrust and FCA matters cost? Zero is currently in the price and zero is on the balance sheet. Any nine-figure outcome is pure downside.

11.5 Falsification

• Falsifies the bull case: a second CRL in December 2026; or Q2/Q3-2026 revenue below ~$210m per quarter, which would make the guidance range unreachable and expose the cost base; or a Sun/Hikma launch; or an adverse ruling or nine-figure settlement in the Teva antitrust matter.
• Falsifies the bear case: relacorilant approved in Cushing’s with a clean label and Q2/Q3 revenue above ~$250m per quarter demonstrating that the fulfilment recovery is real; or a Lifyorli label expansion into platinum-sensitive/endometrial disease on the end-2026 BELLA readout; or a settlement of the antitrust matters at a nuisance value that removes the tail.

11.6 Where the positioning read points

The factor evidence argues that consensus is offside in a specific and unusual way. Because CORT carries no style-factor identity whatsoever — no momentum, value, quality or growth loading, and 93% of variance idiosyncratic — the stock is owned almost exclusively by investors with a view, not by systematic or factor capital. That means positioning is concentrated among people who have already underwritten the December PDUFA. There is no passive or factor bid to absorb a disappointment. The 2025-12-31 session, in which a −49.9% idiosyncratic return arrived in a single day, is the empirical demonstration of what that concentration does on bad news.

12. Fact vs. Interpretation

13. Open Questions

1. What exactly did the FDA’s CRL require, and does the resubmission satisfy it? Management describes “additional evidence of efficacy” and a liver-enzyme concern, and refers to an FDA analysis it “can’t replicate.” A redacted CRL is public. Without reading it directly against the resubmission’s contents, the probability of approval in December 2026 cannot be estimated with any confidence. This is the single largest open question in the file.
2. How much of the Q1-2026 revenue shortfall is recoverable? Management attributes it to seasonality plus a prior-authorisation backlog. Q2-2026 results on 2026-07-29 are the first clean read.
3. What is the realistic exposure in the Teva and payor antitrust actions? Both allege the same conduct; both carry treble damages; trial is March 2027; nothing is accrued. No public damages estimate exists.
4. Will Sun or Hikma actually launch, and when? Their settlements are conditional on FDA approval and on Teva remaining commercially available. Neither has a disclosed approval status.
5. What is Lifyorli’s realistic peak? Management says “>$1bn US by end of decade.” The platinum-resistant ovarian population is small and treatment duration finite. No independent patient-number estimate has been reconciled to the $37,900-per-cycle price.
6. Does the SG&A run-rate flatten in 2026, and at what level? Guided “higher”; no absolute figure disclosed. Every operating-leverage assumption in the bull case depends on this.
7. Succession. Belanoff co-founded the company in 1998, is CEO, President and CODM, and is the intellectual author of the cortisol thesis. No succession plan is disclosed.
8. What did the HHS mifepristone safety re-examination announced in September 2025 conclude, if anything? Not addressed in the FY2025 10-K beyond the risk-factor language.
9. Why did director Baker buy 100,000 shares on 2026-03-17? The purchase was eight days before the Lifyorli approval. There is no suggestion of impropriety — the approval date was public information as a PDUFA target and the purchase was made in an open window — but the sizing and timing warrant noting as the strongest single insider signal in the file.

14. What Must Be True

Bull case

Bear case

15. Source Appendix

See Appendix B below for the full citation list with URLs and access dates.

Primary sources. Corcept Therapeutics FY2025 Annual Report on Form 10-K (filed 2026-02-24); Q1-2026 Quarterly Report on Form 10-Q (filed 2026-04-30); Definitive Proxy Statement on Schedule 14A (filed 2026-04-17); Current Reports on Form 8-K dated 2025-03-31, 2025-10-14, 2025-12-31, 2026-01-22, 2026-02-19, 2026-03-25, 2026-06-03 and 2026-06-17; and the complete corpus of 309 Forms 3, 4, 4/A and 5 filed between 2021-07-25 and 2026-07-25. All retrieved from SEC EDGAR (CIK 0001088856) and mirrored locally.

Management commentary. Q4-2025 earnings call (2026-02-24) and Q1-2026 earnings call (2026-05-01), from public transcript sources. Treated throughout as hypothesis requiring external validation, as a standing analytical rule.

Quantitative data. ROIC.ai (income statement, balance sheet, cash flow, profitability ratios, enterprise value, valuation multiples — all material figures reconciled to the 10-K and 10-Q); AZI price history CSV (5,603 daily observations, 2004-04-14 to 2026-07-24) and the AZI valuation index own-history percentiles; FactorsToday factor-loading, leaderboard, specific-volatility and related-stock endpoints.

Related prior coverage. Public Crinetics Pharmaceuticals disclosure and the announced Vertex transaction terms, for the atumelnant competitive read.

Secondary sources. Trade and financial press for the Lifyorli approval and pricing (Fierce Pharma, Pharmaceutical Technology, March 2026); securities-litigation announcements (Rosen Law Firm, Hagens Berman, Bronstein Gewirtz & Grossman, Kuehn Law, April–June 2026); Corcept investor-relations press releases.

The analysis sections above contain no recommendation and no price target; the Claude's Take block is a labelled exception. General information only — not investment advice.

APPENDIX A — Standard Diligence Questionnaire

Report date: 2026-07-25 · Supplemental to the main analysis. Fact / Interpretation / Assumption labels are applied where the distinction matters. No recommendation and no price target appears in this appendix.

General

What thoughtful questions have other investors asked about this company?

From the Q4-2025 and Q1-2026 earnings calls, the questions that actually mattered were these.

Piper Sandler (David Amsellem), on the CRL: “It sounds like you believe the agency, for lack of a better term, moved the goalposts… there is a redacted CRL that lays out some very specific concerns. So it just seems that there’s a lot of daylight between how the FDA is thinking about this filing and how you’re thinking about this filing. And I’m just trying to understand how we in the investor and analyst community can somehow bridge that gap.” INTERPRETATION: this is the single best question asked of management in two years, and it was not answered. Charles Robb’s reply — “I can’t really explain how they moved to their decision” — is a candid admission that the company does not understand why it was rejected. An investor cannot handicap a resubmission whose sponsor cannot articulate the objection it is addressing.

Piper Sandler, on price: what are the assumptions for authorized-generic mix and net price erosion in 2026? Management answered precisely and usefully: AG net price is ~30% below Korlym WAC, mix ended 2025 at ~75% and is ~78% in 2026, “essentially stabilised,” with further discounting built into guidance.

H.C. Wainwright (RK), on the channel: are you adding a second specialty pharmacy so last year’s bottleneck does not recur? Management confirmed a plan to expand the network in Q4-2026 — an admission that single-vendor dependency was a genuine operating vulnerability.

Truist and H.C. Wainwright, on KEYTRUDA: how does Lifyorli position against pembrolizumab in ovarian cancer? Management’s answer — biomarker restriction limits KEYTRUDA to 35–40% of patients after testing rates, versus Lifyorli’s all-comer label and a stronger hazard ratio — is the strongest competitive argument the company makes anywhere.

The question not asked on either call, and the one to put first: what is the range of outcomes in the Teva and payor antitrust actions, and why is nothing accrued?

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? FACT: at a low, and the low is partly structural rather than cyclical. Operating income fell from $137.0m (FY2024) to $44.8m (FY2025) to negative $49.6m in Q1-2026. Trailing-twelve-month operating income is negative $8.2m. INTERPRETATION: roughly half of the decline is a genuine investment trough — SG&A built for two simultaneous launches — and should reverse if revenue scales. The other half is a permanent repricing: realised price per tablet fell 17.7% in FY2025 as the authorized-generic mix passed 75%, and that does not come back.

Driven by the external environment or internal actions? Both, and it is worth separating them. External: Teva’s January 2024 generic launch and the February 2026 Federal Circuit affirmance; the FDA’s December 2025 CRL. Internal: the decision to launch an authorized generic (defensive, correct, and the direct cause of the price decline); the decision to scale SG&A 60% in FY2025 ahead of an approval that did not arrive; and the specialty-pharmacy migration that disrupted three months of fulfilment.

How stable are revenues? FACT: the endocrinology base is genuinely stable in character — hypercortisolism is chronic, Korlym is a chronic daily therapy, and revenue is a patient book rather than a transaction stream. But reported revenue has been volatile: $207.6m (Q3-2025) → $202.1m (Q4-2025) → $164.9m (Q1-2026), two consecutive sequential declines. Management attributes this to Q1 payer reauthorisation seasonality plus the pharmacy-transition prior-authorisation backlog. Lifyorli, by contrast, is a finite-duration oncology therapy and its revenue is a flow of new starts, not an accumulating book. The two have different revenue physics.

Outlook for products/services? Endocrinology: unit growth continuing (volume +12.4% in Q1-2026) at a lower and now roughly stabilised price. Oncology: launching from zero into a small, under-served indication with NCCN preferred status. Pipeline: five distinct binary readouts between end-2026 and end-2027 (BELLA first arm, MONARCH in MASH, the relacorilant Cushing’s PDUFA, the dazucorilant Phase 3 start, and the broader oncology Phase 2 programme).

How big will this market be — growing, shrinking, domestic or international? Overwhelmingly domestic. Sales are effectively all United States; an EMA decision on Lifyorli is expected in Q4-2026 and would be the first material international revenue. FACT: Corcept’s own CATALYST and MOMENTUM trials found hypercortisolism in 24% of difficult-to-control type 2 diabetics and 27% of resistant-hypertension patients — a genuine and evidence-backed expansion of the diagnosable population, now cited in AACE guidance. INTERPRETATION: the market is growing and Corcept is the party growing it. Whether Corcept captures that growth is a separate question, and the answer is unfavourable while the molecule is generic.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? More, decisively. Since January 2024 Teva sells generic mifepristone; the Federal Circuit affirmed non-infringement on 2026-02-19; Sun Pharmaceutical and Hikma hold settlements permitting entry once FDA-approved. Recordati (Isturisa, Signifor) and Xeris (Recorlev) compete with different mechanisms. And Vertex agreed on 2026-07-06 to acquire Crinetics, whose atumelnant — an oral MC2R antagonist — is in Phase 1b/2a for ACTH-dependent Cushing’s. A well-capitalised, mechanistically differentiated competitor is arriving on a three-to-five-year view.

How profitable is the business (ROIC, ROE)? FACT: it was highly profitable and currently is not. ROE 81.0% (2021), 41.3% (2022), 30.4% (2023), 29.8% (2024), 16.8% (2025). ROIC 24.8% / 22.3% / 18.1% / 20.0% across 2021–2024; not meaningfully computable for FY2025 given the collapse in operating income. The FY2025 ROE of 16.8% is itself flattered — net income of $99.7m exceeds pre-tax income of $66.5m because of a $33.2m tax benefit driven by option-exercise deductions.

How profitable is the industry — how many competitors, what barriers to entry? Orphan endocrinology was a structurally excellent industry: small populations, high prices, high regulatory barriers, few entrants. For Corcept specifically it has stopped being one, because the barrier that mattered — patent protection on the molecule — has gone. Applying the Greenwald test: no supply-side cost advantage (contract-manufactured small molecule; the 98.3% gross margin is a price phenomenon, not a cost one); weak and legally-contested demand-side captivity (procedural lock-in through a specialty-pharmacy channel, which is precisely what Teva’s antitrust suit attacks); no scale economies that a Teva or a Vertex cannot clear.

Can the business be easily understood? Yes, unusually so. One mechanism (glucocorticoid-receptor antagonism), one marketed molecule until March 2026, one reportable segment, no debt, no acquisitions, no foreign operations of consequence. The complexity is in the regulatory and legal calendar, not the business model.

Can it be undermined by foreign low-cost labour? Not directly relevant. The relevant analogue is low-cost generic manufacturing, and the answer there is yes — it already has been. Teva is one of the world’s largest generic manufacturers and sells the same molecule.

Do brands matter? Barely, and decreasingly. The clearest evidence is Corcept’s own behaviour: it voluntarily migrated ~78% of its volume from the branded Korlym to its own authorized generic at a 30% discount. A company that believed in its brand equity would not do that. What matters is not the brand but the channel — the specialty pharmacy, the patient-support apparatus and the prior-authorisation infrastructure.

What is the nature of competition? Two-sided. In endocrinology it is price and channel: Teva competes on price for the same molecule; Recordati and Xeris compete on mechanism and side-effect profile. In oncology it is clinical data and guideline position: Lifyorli’s all-comer label, oral administration, 0.65 hazard ratio and NCCN preferred status compete against KEYTRUDA’s biomarker-restricted 0.76.

Customers’ switching costs? Low at the molecule level, moderate at the channel level. A pharmacist can substitute generic mifepristone. What creates friction is enrolment in a specific patient-support programme, prior authorisations tied to that channel, and physician familiarity. INTERPRETATION: this friction is real, is the source of Corcept’s residual pricing, and is being attacked simultaneously by Teva’s antitrust suit, the payor antitrust suit, and the DOJ’s investigation into prior authorisations and reimbursement. It is a moat that plaintiffs are calling a conspiracy.

Financial Condition & Balance Sheet

Assets not fully recognised on the balance sheet? Yes, and they are the main event. The proprietary compound library — more than 1,000 selective cortisol modulators across four structurally distinct series, with composition-of-matter and method-of-use patents running to 2033–2042 in oncology and 2033–2040 in hypercortisolism — is carried at essentially nothing, because R&D is expensed. So is the clinical evidence base (CATALYST, MOMENTUM, GRACE, GRADIENT, ROSELLA, DAZALS) that cost hundreds of millions to generate. On a market capitalisation of ~$10.4bn against $638.0m of book equity, roughly 94% of the market value is in assets that do not appear on the balance sheet.

Off-balance-sheet liabilities? Yes, and this is the most important unquantified item in the file. Five adversarial proceedings and one federal investigation are outstanding: the Teva antitrust action (N.D. Cal., trial March 2027, treble damages); the payor antitrust action brought by Aetna, Health Care Service Corporation, Humana and Molina (Alameda County, “substantially similar” allegations); the qui tam False Claims Act action U.S. ex rel. Elliott v. Corcept (D.N.J., DOJ declined to intervene May 2026, amended complaint served 2026-06-02, treble damages plus per-claim penalties); the NJ U.S. Attorney’s criminal and civil investigation into Korlym promotion and payments to prescribers; and a securities class action covering 2024-10-31 to 2025-12-30. The 10-K states: “No such amounts are accrued as of December 31, 2025.” Beyond litigation, the only other off-balance-sheet items are ordinary operating and finance leases ($9.6m capitalised) and customary milestone/royalty obligations to the University of Chicago on licensed oncology patents.

How conservative is the accounting? Broadly conservative, with two items to watch. Revenue recognition involves significant estimation — government rebates and chargebacks were the auditor’s designated critical audit matter for FY2025, with government-rebate provisions rising from $52.8m (2023) to $86.3m (2024) as volume shifted to the authorized generic. Second, charitable-foundation donations that support patient co-pays are recorded as a deduction to SG&A rather than to revenue; this is disclosed and permissible, but it means SG&A contains an item economically closer to a revenue contra. Otherwise: no capitalised development costs, no goodwill, no intangibles, no acquisitions, minimal fixed assets, and stock-based compensation fully expensed at $85.0m.

How CapEx-hungry is the business? Not at all. Cumulative capital expenditure over the five years 2021–2025 was $3.4m — against $756m of cumulative operating cash flow. Manufacturing is entirely contracted out; the 10-K states the company has no infrastructure or plans to acquire any. This is as capital-light as a commercial pharmaceutical business gets. The scarce resource here is R&D and SG&A, not fixed capital.

Capital Allocation & Management

How much FCF does the business generate, how does management use it, what is the philosophy? FACT: free cash flow was $167.4m (2021), $119.9m (2022), $126.5m (2023), $196.1m (2024), $141.8m (2025) — and negative $16.8m in Q1-2026. The philosophy, consistently applied for a decade, is: fund all R&D from operating cash flow, never dilute, never borrow, never acquire, and return surplus through buybacks. That is a coherent and unusually disciplined philosophy for a mid-cap biotech, and it deserves credit.

Significant acquisitions recently? None, ever. Corcept has made no material acquisition in its history. In an industry where value destruction through M&A is endemic, this is a genuine positive.

Buying back shares? FACT: yes, and pro-cyclically. A $200m authorisation was approved in January 2024. Corcept spent $15.7m in FY2024 and $172.9m in FY2025 — 2.6 million shares at an average of $66.71 — exhausting 94% of the authorisation. It then repurchased $0 under the programme in Q1-2026, when the stock traded between $32.15 and $46.52 and the balance sheet held $515m. The Board did not re-authorise into the drawdown. INTERPRETATION: the outcome looks acceptable in hindsight (the stock is $95 today), but the decision rule was “buy when the narrative is strongest,” which is the wrong rule. Note also that total FY2025 cash spent acquiring shares was $245.9m, of which $72.9m funded net option exercises and RSU tax withholding rather than reducing the count for continuing holders — and shares outstanding nonetheless rose from 105.1m to 107.3m.

Issuing large amounts of new shares to insiders? Yes. Stock-based compensation was $85.0m in FY2025 (11.2% of revenue), up from $61.7m and $49.1m. Diluted shares (119.99m) exceed basic (103.86m) by 15.5% — a persistent option overhang.

Compensation policy of directors/management? FACT: CEO Belanoff’s 2025 total compensation was $15,303,551 (salary $1.182m, bonus $1.190m, option awards $12.908m), versus $9,890,081 in 2024 — +54.7%. Each of the four other named executives received $5,163,087 of option awards versus $2,804,219 in 2024 (+84%), with totals up 56–59%. This occurred in a fiscal year in which operating income fell 67%, the FDA issued a Complete Response Letter on the lead asset, the Federal Circuit affirmed the loss of patent protection, and the shares ended the year down 31%.

The proxy states that “except for their salaries, compensation for our named executive officers is performance-based.” The option grants vest over four years “subject to the recipient’s continued employment,” with no disclosed performance condition, no relative-TSR modifier, no revenue or margin hurdle, and no return-on-capital metric anywhere in the programme. Bonuses are discretionary against qualitative goals. Notably, the proxy’s Executive Summary of 2025 accomplishments omits both the CRL and the Federal Circuit loss. Say-on-pay drew 94% support at the 2025 meeting (on FY2024 pay).

Motivations of management? FACT, from the complete five-year Form 3/4/5 corpus (309 of 309 filings parsed): insiders executed two open-market purchases totalling $3,313,809 — both by director G. Leonard Baker, Jr. on 2026-03-17, 100,000 shares at ~$33.14, four days after the 52-week low — against $147.9m of sales across 225 transactions. Sales by year: $0.8m (2022), $7.3m (2023), $9.8m (2024), $96.3m (2025), $33.8m (2026 YTD). Largest sellers: Maduck $42.1m, Belanoff $29.3m, Lyon $26.7m, Guyer $26.2m, Mokari $9.6m.

INTERPRETATION: the resulting holdings are the signal. After realising those sums, the CFO holds 16,130 shares, the President of Endocrinology 9,755, and the Chief Development Officer 3,985. The pattern is cashless exercise followed by near-total disposal. Virtually all 2026 sales carry 10b5-1 designations — proper compliance, which mitigates any timing inference but not the level inference. The exceptions are the CEO (2,565,489 shares) and two long-tenured directors (Baker 1,146,631; Wilson 1,084,543). Baker is the only insider who has bought in five years, and he bought at the bottom.

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? No. Corcept is a Delaware corporation, US-domiciled, listed on the Nasdaq Capital Market under CORT. Ordinary common stock; standard Form 1099 reporting; not an ADR, not a partnership, no K-1.

Dividend policy? FACT: none. The 10-K states: “We have never declared or paid cash dividends. We do not anticipate paying cash dividends in the foreseeable future.” All shareholder return has come via repurchase. This is appropriate for a business with this reinvestment opportunity set.

How profitable is the business? Addressed above. In short: gross margin 98.3%, which is impressive and analytically almost irrelevant; operating margin down from 36.4% (2019) to 5.9% (2025) to −30.1% in Q1-2026; ROE down from 81% (2021) to 16.8% (2025) on a flattered basis. The 98% gross margin is a price, not a moat.

Is net income diverging from cash from operations? Yes, in both directions, and both merit attention. Historically operating cash flow exceeded net income by 1.2–1.5x, driven by non-cash stock compensation — normal and healthy. In FY2025 the more important divergence is within net income: reported net income of $99.7m exceeded pre-tax income of $66.5m because of a $33.2m tax benefit from increased option-exercise deductions and reduced pre-tax income. Diluted EPS fell 33% while operating income fell 67%. Anyone anchoring on reported EPS is looking at a figure flattered by the tax consequences of insiders selling stock. In Q1-2026 the same mechanism converts a $44.7m pre-tax loss into a $31.8m net loss, while operating cash flow turned negative.

Risks & Downside

What factors would cause the stock to decline? In rough order of expected impact: (1) a second Complete Response Letter at the ~December 2026 relacorilant PDUFA — the resubmission contains re-analyses of the dataset the FDA already declined, not a new trial; (2) FY2026 guidance miss — reaching the $1.0bn midpoint requires ~$278m per quarter for Q2–Q4, 69% above the Q1 run-rate and 34% above the all-time quarterly record; (3) an adverse antitrust or FCA outcome, with treble damages available and nothing accrued; (4) Sun or Hikma generic entry compressing realised price further; (5) a disappointing Lifyorli launch trajectory; (6) multiple compression from 12.8x trailing sales — above every year-end multiple in eight years — toward the historical 4–7x range; (7) adverse HHS action on mifepristone following the September 2025 safety re-examination, which would affect 100% of current revenue.

Risk of a catastrophic loss? Real but bounded, and the bound is the balance sheet. With $515.4m of cash and investments, $9.6m of finance-lease debt, ~$506m of net cash and essentially no capital commitments, Corcept can absorb several years of the current burn without external financing. The catastrophic scenarios are therefore valuation rather than solvency events. The empirical precedent is instructive: the stock fell 56.4% over two sessions on 2025-12-30/31, of which the 12-31 move alone was a −49.9% idiosyncratic return. With 74.7% annualised idiosyncratic volatility and 93% of variance company-specific, single-day halvings are within this security’s normal distribution. Lifetime maximum drawdown is −85.3%.

Chance of a total loss? Very low. A total loss would require the simultaneous destruction of a $761m revenue base, the entire compound library, and the net cash position. Even the harshest realistic bear case — a second CRL, three generic entrants, and a large antitrust settlement — leaves a revenue-generating business with a newly approved oncology drug and residual net cash. The bear risk here is a large permanent capital impairment from the current price, not a zero.

Recent News & Events

Has the business environment changed recently? Yes, more than in any prior two-year period in the company’s history. The condensed chronology: Teva’s generic launched (January 2024); Corcept launched its own authorized generic (June 2024); ROSELLA met its PFS endpoint and the stock doubled in a day (2025-03-31); the exclusive specialty pharmacy was de-exclusivised and then terminated amid a fulfilment crisis (June–October 2025); the FDA issued a Complete Response Letter on relacorilant in Cushing’s and the stock fell 56% in two sessions (2025-12-30/31); ROSELLA met its overall-survival endpoint (2026-01-22); the Federal Circuit affirmed Teva’s non-infringement (2026-02-19); Lifyorli was approved 3.5 months early (2026-03-25); the DOJ declined to intervene in the qui tam and the relator filed an amended complaint (May–June 2026); and the relacorilant NDA was resubmitted (2026-06-17) with a PDUFA expected around December 2026. Q2-2026 results are due 2026-07-29.

Significant acquisitions? None by Corcept. The relevant transaction is by a competitor: Vertex Pharmaceuticals agreed on 2026-07-06 to acquire Crinetics Pharmaceuticals at $85.00 per share in cash. Crinetics’ atumelnant is an oral MC2R (ACTH-receptor) antagonist in Phase 1b/2a for ACTH-dependent Cushing’s syndrome with best-in-class Phase 2 data in adjacent congenital adrenal hyperplasia. INTERPRETATION: this places a balance sheet of roughly $13bn of cash behind a mechanistically differentiated Cushing’s asset. It does not threaten Corcept’s 2026 or 2027 revenue, but it materially changes the competitive picture for the end-of-decade $2bn Cushing’s ambition on which the current valuation depends.

Change in accounting policies? No material change. The Q1-2026 10-Q discloses three pending FASB standards to be adopted in future periods — ASU 2024-03 (expense disaggregation, FY2027), ASU 2025-06 (internal-use software, FY2028) and ASU 2025-11 (interim reporting, FY2028). ASU 2024-03 is the one worth anticipating: it will require disaggregation of expense categories in the notes, which should for the first time reveal the composition of the $448.7m SG&A line.

Recent changes — new markets, facilities, management? New markets: oncology, entered commercially on 2026-03-25 with Lifyorli, supported by a purpose-built division under President Roberto Vieira; an EMA decision on the same indication is expected in Q4-2026, which would be the first material non-US revenue. New facilities: none of consequence — the business remains capital-light with contract manufacturing and $0.2m of FY2025 capex. New channel: specialty pharmacy services transferred from Optime Care to Curant Health Georgia in Q4-2025, with an intention to add further pharmacies from Q4-2026. Management: the senior team is notably stable — Belanoff (CEO/President, co-founder since 1998), Mokari (CFO), Guyer (Chief Development Officer), Robb (Chief Business Officer), Maduck (President, Endocrinology), Vieira (President, Oncology). OPEN QUESTION: no succession plan is disclosed for Belanoff, who is simultaneously CEO, President, Chief Operating Decision Maker and the intellectual author of the company’s entire scientific thesis.

This questionnaire supplements the Corcept analysis dated 2026-07-25. It contains no recommendation and no price target.

APPENDIX B — Source Appendix

Report date: 2026-07-25 · All sources accessed 2026-07-25 unless otherwise stated. Primary sources are listed first. Management commentary is labelled and is treated throughout as hypothesis requiring external validation.

A. SEC filings (primary — authoritative)

Filer: CORCEPT THERAPEUTICS INC, CIK 0001088856. The complete trailing-60-month corpus (2021-07-25 to 2026-07-25 — 628 filings) was enumerated and mirrored locally for review.

Methodological note on the insider analysis. EDGAR serves Forms 4 under several different xslF345X0* stylesheet paths. A naive mirror captured only 146 of the 309 filings. Stripping the stylesheet segment from each manifest URL recovered the raw XML for 309 of 309 filings (100%), and every figure quoted in the memo and diligence appendix is drawn from that complete parse. Partial parses of this corpus produce materially wrong conclusions about insider activity.

B. Management commentary (labelled; hypothesis, not evidence)

Both retrieved from public transcript sources. As a standing analytical rule, all management assertions were cross-checked against the 10-K, the 10-Q and external evidence before use; where management’s account and the filings diverge, the filings govern and the divergence is flagged in the memo.

C. Quantitative data services

D. Related prior coverage

E. Secondary sources

F. Analytical frameworks

• Greenwald & Kahn, Competition Demystified (barriers-to-entry taxonomy; the three genuine advantage types; the market-share-stability and ROIC tests; the distinction between a price phenomenon and a cost advantage) and Chancellor/Marathon, Capital Returns (supply-side capital-cycle analysis; the observation that high returns attract capital and mean-revert). Applied in Industry Dynamics (industry structure and capital-cycle positioning), Competitive Position (moat taxonomy and the ROIC test), Growth (growth quality) and Capital Allocation.

G. Data limitations and caveats

1. The Complete Response Letter itself was not read. A redacted version is public and was referenced by an analyst on the Q4-2025 call, but this analysis relies on the 10-K’s characterisation and management’s description. The probability assigned to the December 2026 resubmission therefore rests on structural reasoning — that re-analysed data is a weaker path than a new trial — rather than on the specific deficiencies cited.
2. No independent patient-number estimate for platinum-resistant ovarian cancer was reconciled to the $37,900 cycle price. Lifyorli peak-revenue estimates here are management’s, labelled as such.
3. Litigation exposure is unquantified because no public damages estimate exists in any of the antitrust or FCA matters, and Corcept has accrued nothing. This analysis treats the exposure qualitatively and flags the zero accrual rather than assuming a number.
4. Factor-model loadings are in-sample statistical estimates. Betas, R² and volatility figures are reported as facts; any statement about persistence or mean reversion is labelled interpretation and caveated for regime.
5. The aggregated data services used are third-party sources, not primary. Every figure driving a verdict in this article was reproduced from the 10-K or 10-Q.

Sources compiled 2026-07-25. All primary documents are publicly available from SEC EDGAR and the company’s investor-relations site.