VETOCHARTS
VETO.CHARTS
MARGARYAN LABSFINANCIAL FRONTIER
PUBLIC RESEARCH LEDGER / 2026
PROJECT ZERO SCIENCE BEFORE SIGNALS

FINANCIAL
FRONTIER.

Markets may hide regularities. Our job is to test them—not to promise returns.

An independent Margaryan Labs research track. Mathematics, reproducible experiments, negative results, and explicit limits. Only validated methods may reach VETO.

Explore the evidenceOur method
No guaranteed returns. No invented discoveries.
01RESEARCH02MATHEMATICS03EXPERIMENT04FALSIFICATION05EVIDENCE06VETO
STUDIES IN LEDGER—documented research records
CONFIRMED EDGE00none established so far
MAIN EXPERIMENTE1L2 — awaiting verifiable data
LAST PUBLICATION—Editorial timestamp, not market feed
VETO / RESEARCH MISSION CONTROL

Beyond signals.
Testable questions.

Connect market evidence, falsifiable hypotheses and an experimental workbench. Research never receives live trading authority automatically.

Open Research LabMR-05 · Future EvidenceDiscovery Engine · MR-04Alpha Factory · 3 experiments
FINANCIAL FRONTIER / ALPHA RESEARCHWORKFLOW · CONCEPT
01
OBSERVEMarket state
02
QUESTIONFalsifiable idea
03
TESTObserved transitions
04
VERIFYLimits & review
05
DECIDEManual governance
WORKER JOURNAL → RESEARCH BENCHNO AUTO TRADING
RESEARCH / EVIDENCE DISTRIBUTION

What does the research actually show?

Shares of published study statuses, not a strategy win rate or investment return.

—studies
No verified published ledger available — chart withheld.
I. RESEARCH TERRITORIES

Questions we don't yet know how to answer.

MARGARYAN LABS / 001–003
Loading the published evidence ledger…
II. THE EVIDENCE LEDGER

Every hypothesis leaves a trace.

Loading
Loading actual research records…
Load the research ledger to inspect results.

This is an editorial evidence ledger. It updates when reviewed records are published, not from simulated AI activity. Dates and statuses reflect documented studies; a hypothesis is not promoted without new evidence.

III. OUR METHOD / NON-NEGOTIABLE

We try to prove ourselves wrong.

01

Define a falsifiable question

Null hypothesis, alternatives and mathematics.

02

Prove the data provenance

Source, timestamps, integrity, data rights and leakage.

03

Beat strong baselines

Chronological holdouts, costs, regime stability.

04

Stop when evidence fails

A failed test is published. Production requires a separate gate.

H017 · PRE-REGISTERED FORWARD SCIENCE

New predictions. Not retrospective explanations.

One precommitted, falsifiable question on BTC source marks: does an eight-observation absolute-return estimate outperform the frozen 32-observation baseline at forecasting the next recorded movement? No real trades.

Open H017 protocol

H022 · PRECOMMITTED EXTERNAL HOLDOUT · 11 OCT — 10 DEC 2026

A new level: test on two external venues.

A fixed 60-day future window, OKX and Coinbase BTC source feeds and a frozen 8-versus-32 volatility question. The experiment is not completed and results are not invented.

Replication Lab 022
IV. AI RESEARCH LABORATORY / CROSS-LAB PROTOCOL

One research ecosystem. Distinct scientific disciplines.

Waiting for a verified laboratory source.

DRAFT RESEARCH INTEGRATION
Open educational Discovery Bench — local, not published