Three Alzheimer's new molecular entities were approved between June 2021 and July 2024, a rate of about 0.75 a year. 1.216 years remain before 2027-12-31, and openFDA lists no pending original application the model can see. A Poisson rate over that window gives 60%.
Every trial has odds. Nobody wrote them.
One market per Alzheimer's trial, on one primary outcome named word for word: the treatment arm against the placebo arm. An agent prices each one every night and hashes its forecast on chain before the data exists.
0.0004ETH in the contract9 markets open
replay Brier 0.19081 readouts replayedfee 2% of every settled pool
Stake on a readoutsettles on the sponsor's own number
AMfcast 15% XanaMIA a6526bAfcast 21% Polaris-AD 5afb86BMfcast 43% ADEPT-2 2b3bf6APfcast 53% ACP-204 72329dSLfcast 38% Masupirdine 6dd678HRfcast 31% TRONTIER 1 3d722cHRfcast 30% TRONTIER 2 511614PMfcast 48% Translation Challenge 47c542FDAfcast 60% FDA new entity cb7999AMfcast 15% XanaMIA a6526bAfcast 21% Polaris-AD 5afb86BMfcast 43% ADEPT-2 2b3bf6APfcast 53% ACP-204 72329dSLfcast 38% Masupirdine 6dd678HRfcast 31% TRONTIER 1 3d722cHRfcast 30% TRONTIER 2 511614PMfcast 48% Translation Challenge 47c542FDAfcast 60% FDA new entity cb7999
Closest to readout
CalendarPrimary completion from the registry, soonest first.
1APolaris-AD
2026-05 estimated · n 1,535
21%agent2AMXanaMIA
2026-09-22 actual · n 247
15%agent3SLMasupirdine
2026-12 estimated · n 375
38%agent4BMADEPT-2
2026-12-09 estimated · n 500
43%pool 100%5APACP-204
2028-01 estimated · n 1,074
53%agentAgent's highest odds
The modelTonight's probability that the named endpoint hits.
1FDAFDA new entity
U.S. FDA · Any new molecular entity
60%0.75 a year2APACP-204
Acadia Pharmaceuticals · ACP-204
53%prior 30%3PMTranslation Challenge
Prima Mente · Wet-lab finalist
48%prior 50%4BMADEPT-2
Bristol Myers Squibb · KarXT
43%prior 30%5SLMasupirdine
Suven Life Sciences · Masupirdine
38%prior 30%Replayed readouts
ReplayPosted results, priced by the same model the night before.
1UOMIND
NCT02720445 · p 0.530
20%NO2ATADVANCE-2
NCT05557409 · p 0.407
36%NO3AIAPOLLOE4
NCT04770220 · p 0.6058
12%NO4AIINVOKE-2
NCT04592874 · p 0.7975
19%NO5AAbroad
NCT05771428 · p 0.9819
16%NOA
Polaris-AD NCT05531526 · P3
Clinical Dementia Rating Scale - Sum of Boxes (CDR-SB)
Agent YES rung 2 expected
21%
pool empty+ $PLCB · 2% of the pool, 70% to stakers
0.000ETH stakedcloses2027-12-31
AM
XanaMIA NCT06125951 · P2/3
Effects of 10 mg Xanamem on integrated cognitive and functional abilities
Agent YES rung 2 expected
15%
pool empty+ $PLCB · 2% of the pool, 70% to stakers
0.000ETH stakedcloses2027-12-31
SL
Masupirdine NCT05397639 · P3
Cohen-Mansfield Agitation Inventory (CMAI)
Agent YES rung 2 expected
38%
pool empty+ $PLCB · 2% of the pool, 70% to stakers
0.000ETH stakedcloses2027-12-31
BM
ADEPT-2 NCT06126224 · P3
Change from Baseline to End of Treatment in the Neuropsychiatric Inventory-Clinician: Hallucinations and Delusions (NPI-C: H+D) score
Agent YES rung 2 expected
43%
pool 100% · gap -57+ $PLCB · 2% of the pool, 70% to stakers
0.0002ETH stakedcloses2027-12-31
PM
Translation Challenge ORGANISER · P-
A statement on primamente.com, or a joint release by an organiser, naming a finalist hypothesis as tested in the lab
Agent YES rung 2 expected
48%
pool empty+ $PLCB · 2% of the pool, 70% to stakers
0.000ETH stakedcloses2027-12-31
FDA
FDA new entity OPENFDA · P-
An original application (submission type ORIG, status AP) for a new Alzheimer's molecular entity in openFDA drugsfda
Agent YES rung 3 expected
60%
pool empty+ $PLCB · 2% of the pool, 70% to stakers
0.000ETH stakedcloses2027-12-31
AP
ACP-204 NCT06159673 · P2/3
Scale for the Assessment of Positive Symptoms-Hallucinations and Delusions subscales (SAPS-H+D) total score change from baseline (Substudies 1, 2A, 2B)
Agent YES rung 2 expected
53%
pool empty+ $PLCB · 2% of the pool, 70% to stakers
0.000ETH stakedcloses2028-12-31
HR
TRONTIER 1 NCT07169578 · P3
Change From Baseline to Week 72 in Clinical Dementia Rating, Sum of Boxes (CDR-SB)
Agent YES rung 2 expected
31%
pool empty+ $PLCB · 2% of the pool, 70% to stakers
0.000ETH stakedcloses2029-06-30
HR
TRONTIER 2 NCT07170150 · P3
Change From Baseline to Week 72 in Clinical Dementia Rating, Sum of Boxes (CDR-SB)
Agent YES rung 2 expected
30%
pool empty+ $PLCB · 2% of the pool, 70% to stakers
0.000ETH stakedcloses2029-06-30
Forecasts 9
Written by the model from five public inputs. The hash goes on chain before the readout, so the reasoning cannot be edited afterwards.
TonightReplay
Read the model ACP-204 starts from a 30% prior for one neuropsychiatric endpoint in a Phase 2/3 trial. HTR2A scores 0.42 for association with Alzheimer's in Open Targets, moderate genetic support, and 1,074 participants give the trial room to detect a modest effect. Acadia Pharmaceuticals has 1 posted readout in the replay record with 1 positive, PubMed lists 13 papers on the drug in 24 months, and the sum comes to 53%.
The organiser's own calendar puts hypothesis validation in summer 2027, six months before the deadline. It is the challenge's first edition, so there is no record of it keeping a schedule, and the market only counts a public statement naming a tested hypothesis. Even odds, nudged both ways, give 48%.
ADEPT-2 starts from a 30% prior for one neuropsychiatric endpoint in a Phase 3 trial. CHRM1/CHRM4 scores 0.43 for association with Alzheimer's in Open Targets, moderate genetic support, and 500 participants leave little room for a small effect. Bristol Myers Squibb has 0 posted readouts in the replay record with 0 positive, PubMed lists 17 papers on the drug in 24 months, and the sum comes to 43%.
Masupirdine starts from a 30% prior for one neuropsychiatric endpoint in a Phase 3 trial. HTR6 scores 0.42 for association with Alzheimer's in Open Targets, moderate genetic support, and 375 participants leave little room for a small effect. Suven Life Sciences has 1 posted readout in the replay record with 0 positive, PubMed lists 3 papers on the drug in 24 months, and the sum comes to 38%.
TRONTIER 1 starts from a 12% prior for one cognitive and functional endpoint in a Phase 3 trial. APP scores 0.87 for association with Alzheimer's in Open Targets, strong genetic support, and 1,101 participants give the trial room to detect a modest effect. Hoffmann-La Roche has 4 posted readouts in the replay record with 0 positive, PubMed lists 28 papers on the drug in 24 months, and the sum comes to 31%.
TRONTIER 2 starts from a 12% prior for one cognitive and functional endpoint in a Phase 3 trial. APP scores 0.87 for association with Alzheimer's in Open Targets, strong genetic support, and 1,060 participants give the trial room to detect a modest effect. Hoffmann-La Roche has 4 posted readouts in the replay record with 0 positive, PubMed lists 28 papers on the drug in 24 months, and the sum comes to 30%.
Polaris-AD starts from a 12% prior for one cognitive and functional endpoint in a Phase 3 trial. PDE5A scores 0.29 for association with Alzheimer's in Open Targets, moderate genetic support, and 1,535 participants give the trial room to detect a modest effect. AriBio has 0 posted readouts in the replay record with 0 positive, PubMed lists 9 papers on the drug in 24 months, and the sum comes to 21%.
XanaMIA starts from a 16% prior for one cognitive and functional endpoint in a Phase 2/3 trial. HSD11B1 scores 0.10 for association with Alzheimer's in Open Targets, little genetic support, and 247 participants leave little room for a small effect. Actinogen Medical has 0 posted readouts in the replay record with 0 positive, PubMed lists 3 papers on the drug in 24 months, and the sum comes to 15%.
The public test
Replay81 posted Alzheimer's readouts, priced on what existed the night before. Published once, sha256 227ee012.
Brier0.190base rate0.186hits20/81
How a market settles
LadderThe first source that speaks decides the market. Nobody votes on it.
- 01Posted registry result
ClinicalTrials.gov
- 02Sponsor's own statement
Sponsor domain
- 03FDA action
openFDA drugsfda
- 04Termination for futility
ClinicalTrials.gov whyStopped
- 05Void at the deadline
The clock
72 hours to dispute, with a bond, before anyone can finalize and claim.
$PLCB
The rule2% of every settled pool, in ETH. 70% goes to the $PLCB distributor for stakers, 30% to the forecasters who were right.
2%
of every pool, taken at finalization
70% to $PLCB stakers30% to forecasters
Forecasters bond 1,000 $PLCB per market, the house agent 2,000. Half of a bond is burned when the forecast was wrong and confident.
ClinicalTrials.gov
PubMed
Robinhood Chain
ETH
Pons