← Back to scans

When will the FDA approve COMP360 Psilocybin for Treatment-Resistant Depression? — Before July 2027

KXFDAAPPROVALDATECMPS-360-27JUL01 · Science and Technology · 2026-05-03
30%
Agent
73%
Market Price
-43.0%
Edge
50%
Confidence
Volume: 5,489
Spread: 5.0c
Days to resolution: 423
Markets in event: 3
Final Rationale
My independent forecast of 17% suffered from two genuine reasoning flaws the critic identified: sq3 partially double-counted execution risks already captured in sq2, and I likely underestimated CNPV's actual design target (FDA public statements describe ~1-2 month total review for selected drugs, not just shaving that off Priority Review). Accepting Challenges 1, 2, and 5 partially while rejecting full deference to the market: with sq1≈0.75, sq2≈0.40 (more aggressive CNPV), and sq3≈0.75 (CNPV selection is revealed FDA preference for psychedelics), I get ~0.22, and I add a small further upward correction for the possibility that rolling modules are already substantially filed. I deviate sharply from the 73% market because it is extremely thin (22 days, 249 contracts/day) and just spiked +43% on CNPV news without any change to the binding timeline constraint: COMPASS's own Q4 2026 NDA target leaves a punishingly short window for filing acceptance, REMS finalization, and DEA scheduling before July 1, 2027. The market is pricing news sentiment, not the calendar.
Price vs. prediction
How the market moved after the agent's call. Solid blue line is the actual market price; dashed line is the agent's static forecast at $t{=}0$.
$n = 58$ follow-ups
Re-scan Context
This market has been scanned before. Previous predictions:
DatePredictedMarket PriceConfidence
2026-04-23 22% 49% 50%
1. Planner Opus · Fermi decomposition + tool plan
Tools chosen: kalshi_data article_search web_search wikipedia_lookup code_execution
Sub-questions (Fermi decomposition)
  1. Has COMP360 (psilocybin for TRD) submitted or is it on track to submit an NDA/BLA to the FDA before January 2027? (w=0.35)
  2. If an NDA/BLA is submitted, will the FDA grant Priority Review or Breakthrough Therapy designation enabling a decision before July 2027? (w=0.35)
  3. Has the FDA shown willingness to approve psychedelic-assisted therapies (e.g., psilocybin or MDMA) for psychiatric indications by mid-2027? (w=0.30)
Planner reasoning
This question requires assessing the probability that COMPASS Pathways' COMP360 psilocybin therapy receives FDA marketing approval for treatment-resistant depression (TRD) within approximately 14 months (by July 1, 2027). The key causal chain is: (1) Phase 3 trial completion and positive results, (2) NDA/BLA submission to FDA, (3) FDA acceptance and review, (4) Advisory Committee meeting, (5) Approval decision. The temporal bottleneck is severe: even with Priority Review (6-month clock) or Breakthrough Therapy Designation, an NDA would need to be filed no later than approximately January 2027 for approval by July 2027. As of May 2026, COMPASS Pathways was still conducting Phase 3 trials (COMP005 study). The MDMA/PTSD precedent (Lykos Therapeutics rejection in August 2024) is critically important — it demonstrated FDA's heightened skepticism toward psychedelic-assisted therapies, particularly around functional unblinding, abuse potential, and REMS requirements. Reference classes: (a) novel CNS breakthrough therapies with expedited designation — approval rate ~40-50% within 2 years of NDA; (b) psychedelic-assisted therapies specifically — 0 approved as of 2026, 1 rejection (MDMA); (c) first-in-class drugs with complex delivery requirements. The political environment under a more conservative FDA administration may further complicate approval. The realistic scenario for YES requires: Phase 3 data already complete and positive → NDA submitted by ~Q3 2026 → Priority Review granted → No Complete Response Letter → Approval by July 2027. Each step is uncertain, making this a low-probability outcome.
2. Research Parallel tool calls · legacy v2
ToolStatusTimeItemsSummary
kalshi_data OK 1.7s -
article_search OK 0.8s -
web_search OK 55.1s -
wikipedia_lookup OK 1.4s -
code_execution OK 0.0s -
3. Evidence Synthesis Sonnet · 17 items · legacy v2
#StrengthCredibilityDirectionSourceClaimPriced In
1 STRONG 88 ↑ UP web_search COMPASS Pathways achieved two positive Phase 3 trials (COMP005 in June 2025, COMP006 in February 2026), both hitting primary endpoints with high statistical significance across 1,000+ participants with TRD. Yes
2 STRONG 85 ↓ DOWN web_search COMPASS indicated it expects to complete its NDA submission in Q4 2026, with the final COMP006 26-week Part B data expected in early Q3 2026 — meaning full submission is not anticipated until after that data is available. Yes
3 STRONG 90 ↑ UP web_search FDA granted COMPASS a rolling NDA review on April 24, 2026, allowing submission of portions of the NDA as they become ready rather than requiring a complete package at once. No
4 MODERATE 82 ↑ UP web_search COMPASS raised ~$341 million in February 2026, extending cash runway into 2028, indicating strong financial position to complete the NDA process. Yes
5 STRONG 75 ↓ DOWN code_execution Timeline feasibility calculation suggests that for approval before July 1, 2027, a complete NDA submission would need to occur by approximately July 2026 (standard review) or January 2027 (priority review); with Q4 2026 NDA submission planned, only priority/accelerated review pathways remain viable. Yes
6 STRONG 90 ↑ UP web_search COMP360 already holds FDA Breakthrough Therapy Designation for TRD, which provides intensive FDA guidance and can support faster review timelines. Yes
7 STRONG 88 ↑ UP article_search On April 24, 2026, the FDA selected COMP360 for the Commissioner's National Priority Voucher (CNPV) program, which shortens the post-filing review by 1–2 months while maintaining full safety and efficacy standards. No
8 STRONG 82 ↓ DOWN web_search The CNPV program shortens review time by only 1–2 months after NDA filing — a modest acceleration compared to Priority Review's 6-month vs. 12-month standard clock; this alone is insufficient to bridge the timeline gap if NDA is filed in Q4 2026. No
9 MODERATE 70 ↓ DOWN article_search The FDA's top drug regulator (Richard Pazdur, installed Nov 2025) raised alarms about the legality and safety risks of expedited approval initiatives pushed by FDA Commissioner Makary, suggesting internal friction over accelerated reviews. Yes
10 MODERATE 72 ↑ UP article_search Trump-era FDA leadership has moved away from advisory committee meetings for individual drugs, reducing one procedural step that typically adds months to review timelines — a potential modest accelerant. Yes
11 STRONG 92 ↓ DOWN article_search FDA rejected Lykos Therapeutics' MDMA-assisted therapy for PTSD in August 2024, citing concerns about functional unblinding, abuse potential, and inadequate REMS — establishing a precedent of skepticism toward psychedelic-assisted therapies. Yes
12 STRONG 88 ↑ UP article_search The FDA granted National Priority Vouchers to three psychedelic drug companies on April 24, 2026 (psilocybin for TRD, psilocybin for MDD, and methylone for PTSD), signaling an active shift toward facilitating psychedelic therapy reviews under the current administration. No
13 MODERATE 72 ↓ DOWN article_search The FDA under the Trump administration has shown a pattern of reversals and multiple drug rejections (8+ drugs denied/discouraged per RTW Investments), adding uncertainty about approval outcomes for novel CNS therapies. Yes
14 STRONG 78 ↓ DOWN wikipedia_lookup As of the Wikipedia entry, no psychedelic-assisted therapy has received FDA marketing approval; psilocybin remains Schedule I and COMP360 is the most advanced psilocybin candidate in development. Yes
15 MODERATE 75 ↑ UP web_search COMP360's safety profile shows the vast majority of TEAEs are mild-to-moderate and resolve within 24 hours — potentially addressing one key FDA concern (safety) that contributed to the MDMA rejection. Yes
16 MODERATE 55 ↑ UP kalshi_data The Kalshi market for COMP360 FDA approval before July 2027 is priced at 73%, with a dramatic +43% rise in 7 days and +63% rise in 30 days, suggesting very recent market-moving news (likely the April 24, 2026 CNPV announcement and rolling review grant). No
17 WEAK 40 NEUTRAL kalshi_data The Kalshi market has very limited history (22 data points) and low average daily volume (249 contracts), suggesting thin liquidity that may cause price to overreact to recent news rather than reflect deep fundamental analysis. No
Information Gaps
  • No clarity on whether COMPASS intends to accelerate NDA submission ahead of its Q4 2026 target given the rolling review grant — could the NDA complete filing by Q3 2026?
  • No detailed information on exactly what the Commissioner's National Priority Voucher review timeline looks like in practice — the 1-2 month shortening is mentioned but total post-filing review duration is unclear
  • No information on FDA's specific feedback during rolling review on REMS requirements, functional unblinding methodology, or scheduling rescheduling for psilocybin (Schedule I barrier)
  • No information on whether DEA scheduling action (required for any psilocybin approval) has begun or is anticipated — this could be an independent bottleneck
  • Lack of detail on COMP006's 26-week Part B data quality/readout (expected early Q3 2026) — this is critical for NDA completion
  • No information on whether the FDA has accepted any rolling NDA modules already submitted by COMPASS
  • No data on comparable timelines for drugs that used both rolling review and CNPV together — the combination is novel
  • No information on whether the current FDA leadership under Makary/Tidmarsh has expressed specific views on psilocybin-assisted therapy approval
Key Uncertainties
  • Whether the NDA can be fully submitted before Q4 2026 (earlier submission critical for approval before July 2027)
  • Whether the CNPV program combined with rolling review could effectively compress the total review period enough to enable approval by July 2027 from a Q4 2026 filing
  • DEA scheduling reclassification requirement — psilocybin remains Schedule I and approval would require DEA action that is not solely within FDA's control
  • Whether COMP006 26-week data (expected early Q3 2026) will meet the high bar required for NDA completion and FDA acceptance
  • Political/regulatory uncertainty: whether RFK Jr./Trump administration's FDA would view psilocybin favorably or unfavorably given the administration's heterodox positions on both psychedelics (potentially supportive) and drug oversight (unpredictable)
  • Whether FDA will raise unblinding or functional placebo control concerns analogous to those raised against MDMA, which could trigger a Complete Response Letter
  • The Kalshi market's 73% price appears inconsistent with the fundamental timeline analysis — whether this reflects genuinely new information about accelerated timelines or market mispricing of thin volume
Market Context
Price: 0.73 · 7d: +0.43 · 30d: +0.63
Volume: LOW — 249 contracts average daily volume · Depth: SHALLOW — only 22 days of data, price range 4%-80% suggests extreme volatility
Reliability: LOW — very thin market with dramatic recent price movement (likely reacting to April 24, 2026 CNPV/rolling review news), insufficient liquidity for reliable price discovery; 73% may significantly overstate true probability relative to fundamental timeline constraints
4. Independent Forecaster Opus · No market price · legacy v2 Did NOT see market price
Independent estimate: 17% (55% confidence)
base rate
18%
55% conf
evidence driven
13%
55% conf
contrarian
20%
55% conf
Spread
7pp
Low disagreement
sq1: Has COMP360 submitted or is on track to submit an NDA before January 2027?
75%
Base rate for sponsors meeting their stated NDA timelines after positive Phase 3 results and rolling review grant is roughly 60-70%. Strong positive evidence: two positive Phase 3 trials completed, rolling NDA review granted April 2026, strong cash position. The stated Q4 2026 target with rolling review means partial submissions are likely already underway. The risk is COMP006 26-week data delays pushing past January 2027. With rolling review, 'submission' is gradual — I estimate ~75% probability the rolling submission is substantially complete (or accepted for filing) by January 2027.
sq2: Conditional on submission, will FDA review be fast enough to enable approval before July 2027?
25%
Even with Breakthrough Therapy + CNPV + rolling review, the timeline is brutally tight. CNPV shortens review by only 1-2 months. If full NDA is accepted for filing in Q4 2026/early 2027, standard Priority Review (6 months) lands around mid-2027 at earliest, and CNPV could shave to 4-5 months — putting approval in roughly Q1-Q2 2027. However, this assumes no Complete Response Letter, no major REMS negotiation delays, and DEA scheduling action in parallel. Base rate for first-cycle approval of novel CNS drugs with complex REMS is ~50%, but timing precision is the bigger issue. The Lykos REMS/unblinding precedent and DEA scheduling requirement are major drags. Estimate ~25%.
sq3: Has FDA shown willingness to approve psychedelic-assisted therapies by mid-2027?
40%
Base rate: zero psychedelic-assisted therapies approved historically. Lykos rejection in 2024 establishes skepticism precedent. However, the April 2026 CNPV grant to three psychedelic companies is a strong signal of regime change in FDA receptivity. COMP360's safety profile (mild/transient TEAEs) addresses one Lykos concern. Functional unblinding remains a structural concern for any psychedelic trial. The current FDA leadership signals openness but is unpredictable. Weighted: ~40% — meaningfully higher than base rate due to CNPV signal, but still below 50% given no precedent and structural concerns remain unresolved.
Show full reasoning trace
base rate: Novel first-in-class CNS drugs with breakthrough designation have ~40-50% approval within 2 years of NDA filing; psychedelic-assisted therapies have 0/1 approval rate (Lykos rejected). For drugs filing NDA in late 2026 to be approved by July 2027 (sub-8 month window), historical base rate is roughly 15-25% even with priority review.
evidence updates: UP: Two positive Phase 3 trials (+significant), rolling review granted (+5%), CNPV granted (+3-5%), strong safety profile (+2%), FDA shifting toward fewer AdComs (+2%). DOWN: Q4 2026 target submission leaves minimal review window (-10%), Lykos precedent on unblinding/REMS (-5%), DEA Schedule I barrier requiring parallel scheduling action (-5%), no psychedelic ever approved (-3%), Pazdur internal friction on expedited reviews (-2%).
combination method: Conditional chain with correlation adjustment between sq2 and sq3 since CNPV evidence appears in both.
final: The fundamental timeline math remains punishing: COMPASS targets Q4 2026 NDA, leaving roughly 6-8 months for FDA review, REMS finalization, and DEA scheduling action. Even with every accelerant (BTD + CNPV + rolling review + no AdCom), approval by July 1, 2027 requires near-flawless execution. The CNPV grant is a meaningful positive signal but only shaves 1-2 months. I estimate ~18%, slightly above my prior 14% due to the new rolling review and CNPV evidence, but well below the implied market pricing.
ensemble: {'ensemble_count': 3, 'perspectives': {'base_rate': {'p_yes': 0.18, 'confidence': 0.55}, 'evidence_driven': {'p_yes': 0.13, 'confidence': 0.55}, 'contrarian': {'p_yes': 0.2, 'confidence': 0.55}}, 'spread': 0.07, 'individual_p_yes': {'base_rate': 0.18, 'evidence_driven': 0.13, 'contrarian': 0.2}}
5. Devil's Advocate Sonnet · legacy v2 Market price revealed
Math Error Reasoning Flaw Missing Info
Challenges
  1. The conditional chain multiplication (0.75 × 0.25 × 0.40 / correlation adjustment) gives ~0.075, but you arrived at 0.17. The math doesn't transparently support the final number — sq3 (FDA willingness to approve psychedelics by mid-2027) appears to double-count concerns already in sq2 (REMS, unblinding, DEA scheduling are all timing/execution risks captured in sq2). If sq3 is meant to capture *substantive* willingness, conditional on sq1 and sq2 succeeding mechanically, it should be much higher than 40% — arguably 70-85%, since CNPV selection itself is strong revealed preference. Reconsidering: P = 0.75 × 0.30 × 0.75 ≈ 0.17 works, but your stated sq3=0.40 with sq2=0.25 implies ~0.075.
  2. sq2 may be too pessimistic on CNPV. The CNPV program is explicitly designed to compress review to 1-2 months in pilot cases, not just shave 1-2 months off Priority Review. Public FDA statements about CNPV describe target review times of ~1-2 months total post-filing for selected drugs. If true, a Q3-Q4 2026 rolling completion could yield approval by Q1-Q2 2027 comfortably. The forecaster's interpretation of CNPV as 'shaving 1-2 months off standard priority review' may understate its acceleration.
  3. The Lykos precedent is being applied somewhat asymmetrically. Lykos's rejection centered on functional unblinding, trial conduct issues, and therapist misconduct allegations — not on psychedelics being inherently unapprovable. COMP360's trials were larger, more rigorous, and the safety profile is cleaner. Treating Lykos as a -5% drag on COMP360 specifically may overweight a poorly-analogous case.
  4. Information gap on whether COMPASS has already submitted rolling modules is critical. If substantial portions are already filed (rolling review was granted April 2026, a year before resolution), the effective review clock may already be running, materially changing sq1 and sq2.
  5. The 'never approved before' base rate is doing heavy lifting in sq3, but CNPV selection of three psychedelic companies on the same day is exactly the kind of structural regime change that invalidates the historical base rate. The forecaster acknowledges this but may not weight it enough.
Suggested adjustment: +8pp
Two genuine reasoning issues warrant upward adjustment: (1) sq3 likely double-counts execution/timing risks already in sq2, and if recalibrated to capture only substantive 'will FDA approve a psychedelic in principle' it should be 70%+, not 40%. (2) CNPV's actual design target appears more aggressive than 1-2 months off priority review — it's a near-term aspirational fast track. Recombining with sq3 ≈ 0.70 (substantive willingness, given CNPV revealed preference) and sq2 ≈ 0.30 yields ~0.16, but if CNPV is more aggressive sq2 could be 0.35-0.40, yielding ~0.20-0.25. Adjustment is independent of market price.
Market comparison: Forecaster: 17%, Market: 73%. The market is thin (249 contracts/day, 22 days of history) and just spiked +43% in 7 days on the CNPV news, so it likely overshoots. However, the forecaster's 17% may be modestly too low due to internal double-counting in the conditional chain and possible underestimation of CNPV acceleration. A reasonable fundamentals-based estimate is probably 25-35%, with the market at 73% reflecting thin-market overshoot on recent news. The divergence is real but the market's reliability is genuinely low.
6. Aggregator Opus · final calibrated forecast
30% (50% confidence)
My independent forecast of 17% suffered from two genuine reasoning flaws the critic identified: sq3 partially double-counted execution risks already captured in sq2, and I likely underestimated CNPV's actual design target (FDA public statements describe ~1-2 month total review for selected drugs, not just shaving that off Priority Review). Accepting Challenges 1, 2, and 5 partially while rejecting full deference to the market: with sq1≈0.75, sq2≈0.40 (more aggressive CNPV), and sq3≈0.75 (CNPV selection is revealed FDA preference for psychedelics), I get ~0.22, and I add a small further upward correction for the possibility that rolling modules are already substantially filed. I deviate sharply from the 73% market because it is extremely thin (22 days, 249 contracts/day) and just spiked +43% on CNPV news without any change to the binding timeline constraint: COMPASS's own Q4 2026 NDA target leaves a punishingly short window for filing acceptance, REMS finalization, and DEA scheduling before July 1, 2027. The market is pricing news sentiment, not the calendar.
7. Entry Decision Opus Trader
Side
NO
Position Size
$1400
Contracts
5600
Trader Confidence
50%
Trade Rationale
The edge is exceptionally large (-43%), driven by a hard calendar constraint: COMPASS's own Q4 2026 NDA target leaves only ~6 months for filing acceptance, potential AdCom, REMS finalization, and DEA scheduling before July 2027 resolution. The market spiked on CNPV news sentiment but the binding timeline didn't change. Even after the critic's +0.08 adjustment and accepting forecaster confidence is moderate (0.50), the gap from 30% to 73% is enormous. Low liquidity (249 contracts/day) is a real concern but works in our favor here — the 73% print is likely sentiment-driven mispricing rather than informed flow. Sub-question sq2 at 25% is the binding constraint and is well-reasoned around FDA review mechanics.
Allocation Logic
Sized at $1400 — above baseline due to the unusually large edge and clean calendar-based thesis, but capped below max because forecaster confidence is only 0.50, the critic flagged real reasoning flaws (math_error, missing_info), and thin liquidity plus a 423-day horizon add execution/variance risk.
Entry price: $0.25
Current: $0.27
Status: OPEN
P&L: $112.00