Skip to content

Site under construction, formerly SigmaX has begun the process of transitioning to AIgentXS.

SYNTHETIC DATA · READ-ONLY BACKTEST · NOT_CAPITAL_READY — research evidence only. This run executed no order and moved no capital. SigmaX proposes; humans authorize. No capital authority, no allocation recommendation.

← Proof Foundry
Evidence dossierrun · scenario_oos_synthmaturity L4 · Replicated patternReplicated
claim_trend_oos_synthTrend-following survives synthetic OOS
Verdict

OOS candidate

Closest-to-proof on synthetic data — NOT real out-of-sample proof. The only admissible next step is replication on real data.

SYNTHETICShadowNOT_CAPITAL_READYNOT_TRUSTEDNOT_REVIEW_READY
Archived record — recorded 93 days ago, not a current observationrecorded_at 2026-06-10T00:05:00Z

Provenance on this page is DERIVED, not taken from what the record says about itself: the declared data mode is checked against the dataset fingerprint the run hashed, and a self-declared trusted / review_ready is shown as a claim because nothing on this public surface verified it. data_mode "SYNTHETIC" is corroborated by a SYNTHETIC data_fingerprint => shown as declared

Rationale · Walk-forward passed for "Trend-following survives synthetic OOS" but on SYNTHETIC data — this is not real out-of-sample proof. Per doctrine synthetic results are never trusted, review-ready, or capital-ready. Next admissible test: replicate on REAL data.

Failure reasons on record
synthetic_data_not_trusted
Proof objective

What this run is trying to prove

Establish, on a fair and pre-registered basis, whether "A well-separated MA-crossover trend follower (fast=10/slow=50), selected on validation and evaluated out-of-sample, retains a positive walk-forward OOS edge across repeated synthetic trend swings — the closest a synthetic run can get to proof, but it is still synthetic and needs real-market replication." survives real-market out-of-sample testing net of fees and across regimes — and is therefore worth advancing toward human review.

Standard of proofshadow_forward

Why it matters · Capital is never authorized by the Foundry (capital_authority="none"). The objective exists to separate a genuine, durable strategy edge from an in-sample artifact BEFORE any operator review, so that only trusted evidence — never an unsupported performance claim — reaches the decision record.

Minimum evidence required
  • real-market (non-synthetic) data over a strictly held-out OOS window
  • walk-forward OOS edge that is positive after fees across a majority of folds
  • evidence the edge is not concentrated in a single regime (regime replication)
  • fee/slippage resilience: net-of-cost edge survives realistic execution assumptions
  • forward shadow persistence on inert/paper data, unseen at fit/validation/OOS
Disqualifiers (any one sinks the claim)
  • synthetic-only evidence (synthetic data can never be capital_ready or review-ready)
  • in-sample / validation-only support with no out-of-sample test
  • failed real-data out-of-sample (OOS Sharpe <= 0 after fees) => rejected or research_only
  • missing fee or regime evidence (incomplete => proof_gap, not proven)
  • single-fold dependence or look-ahead/leakage in the evaluation
  • a negative control that reproduces the 'edge' (the result is an artifact)
  • forward decay: an OOS edge that does not persist in shadow testing
The fair case in favour

Best evidence for

The strongest fair case in favour of the claim. None of it is proof on its own — each item carries its limitations, and any in-sample / validation number is shown as supports early pattern — not OOS proof.

mechanismsupportsL1

Proposed mechanism for "A well-separated MA-crossover trend follower (fast=10/slow=50), selected on validation and evaluated out-of-sample, retains a positive walk-forward OOS edge across repeated synthetic trend swings — the closest a synthetic run can get to proof, but it is still synthetic and needs real-market replication." is structurally plausible (non-degenerate parameterization).

claim_type=strategy_edge · strategy_degenerate=false · genome_fast=10 · genome_slow=50 · genome_thresh=0

  • plausible mechanism only — not yet causally verified
  • parameterization could still be an overfit artifact
  • synthetic data only — not real-market, never trusted proof
in_samplesupportsL3

Strong in-sample fit (Sharpe 80.57). Preserved as supporting evidence — in-sample only, not trusted proof.

supports early pattern — not OOS proof

in_sample_sharpe=80.5717

  • in-sample only — overfit-prone, never trusted proof
  • no held-out generalization established by this receipt
  • synthetic data only — not real-market, never trusted proof
validationsupportsL4

Survives validation (Sharpe 83.43). Preserved as supporting evidence — validation only, not trusted proof.

supports early pattern — not OOS proof

validation_sharpe=83.4269 · degradation_in_to_validation=-2.8552

  • validation only — tuned-window result, never trusted proof
  • validation window may share regime structure with the fit window
  • synthetic data only — not real-market, never trusted proof
walk_forwardsupportsL5

Survives held-out walk-forward OOS (Sharpe 85.03, return 30.46%).

oos_sharpe=85.0325 · oos_return=0.3046 · oos_max_drawdown=-0.0469 · positive_fold_frac=1 · degradation=-0.0554 · n_folds=5 · passed=true

  • OOS pass on this dataset only — needs real-market and forward confirmation
  • fold dispersion not yet stress-tested
  • synthetic data only — not real-market, never trusted proof
fee_stresssupportsL6

Result is fee-resilient (not fee-fragile): OOS return stable from 30.46% (1x) to 29.68% (5x fees).

base_oos_return=0.3046 · x2_oos_return=0.3026 · x5_oos_return=0.2968 · fee_fragile=false

  • fee robustness measured on summary OOS return only
  • slippage/impact beyond fee multiples not modeled here
  • synthetic data only — not real-market, never trusted proof
safety_gatesupportsL5

Safety gates held: can_go_live=false, live_broker_inert=true — no live-capital pathway was reachable for this run.

can_go_live=false · live_broker_inert=true · promoted_live=false · capital_mode=none · capital_authority=none · missing_live_keys_count=3

  • safety gate proves the kill-switch held, NOT that the strategy has edge
  • live broker remains inert (missing keys: AETHERFORGE_BROKER_API_KEY, AETHERFORGE_BROKER_API_SECRET, AETHERFORGE_EXCHANGE_API_KEY)
  • synthetic data only — not real-market, never trusted proof
The case against

Best evidence against

The strongest case against the claim. These receipts are preserved and shown even when the headline verdict already reflects them — disconfirming evidence is never buried.

regime_testlimitationL6

Regime classifier over-calls (dominant label "trend_down" at 31.7%, accuracy 68.1%); regime context is a limitation, not yet load-bearing.

accuracy=0.681 · crisis_recall=0.8 · dominant_label=trend_down · dominant_frac=0.317 · n_regimes=5

  • regime classifier over-concentrates / under-accurate — context unreliable
  • regime-conditional claims cannot lean on this calibration yet
  • synthetic data only — not real-market, never trusted proof
Still missing

Proof gaps

What is still missing before this claim could be trusted. Each gap is a reason it is NOT_CAPITAL_READY.

proof_gapneeded_nextL2

PROOF GAP: no real-market out-of-sample evidence exists — this run is synthetic. Real-data replication is required before any trust.

data_mode=SYNTHETIC · has_real_oos=false · oos_passed=true · run_limitations_count=6

  • no real-market data has been observed — synthetic results are not transferable
  • this gap blocks any trusted / review-ready / capital-ready status until closed
  • synthetic data only — not real-market, never trusted proof
The next step

Next admissible test

walk_forward on REAL shadow data with regime-calibrated folds (data_replication)

Fastest path to evidence · Source REAL-market data for the traded asset(s) and re-run the walk-forward with regime-calibrated folds. The synthetic result — pass or fail — is not trusted proof; promoting the claim to real data is the single highest-leverage step and everything downstream depends on it.

Expected failure modes
  • Synthetic non-transfer: an edge that exists in synthetic regimes vanishes on real-market data.
  • Overfit collapse: strong in-sample/validation Sharpe that degrades to <= 0 out-of-sample.
  • Single-fold dependence: the OOS result is carried by one lucky fold and is not broad.
  • Fee fragility: a gross edge that turns negative once realistic fees and slippage are applied.
  • Regime concentration: the edge exists only in the dominant regime and does not generalize.
  • Look-ahead / leakage: the signal uses information not available at decision time.
  • Artifact mechanism: a negative control reproduces the 'edge', revealing it as a measurement artifact.
  • Forward decay: an OOS-validated edge that fails to persist in forward shadow testing.
Known limits

Run limitations

  • Synthetic data only — no real-market ingestion (real ingestion is on HOLD).
  • SHADOW posture: no live orders, no exchange/broker connection, no capital moved.
  • Walk-forward OOS is on synthetic regimes; results are illustrative, not an alpha claim.
  • Single random seed; not yet replicated across independent synthetic draws.
  • NOT capital-ready: a synthetic run can never be promoted to live.
  • Closest-to-proof, NOT proof: a passed SYNTHETIC walk-forward is not real OOS; the only admissible next step is replication on REAL market data.
On record

Decision record

Research Decision Record

  • Run: scenario_oos_synth
  • Claim: claim_trend_oos_synth
  • Decision: Do not promote strategy.
  • Verdict: oos_candidate
  • Capital ready: NO
  • Labels: SYNTHETIC NOT_CAPITAL_READY NOT_TRUSTED
  • Reviewed by: operator_pending
  • Created at: 2026-06-10T00:05:00Z

Reason

Evidence is SYNTHETIC; synthetic data can never be capital-ready or review-ready. Live trading is disabled (cangolive=false); the kill-switch held — a supporting safety receipt, not a promotion path. Fee stress was non-fragile (supporting receipt preserved) but cannot rescue a negative-OOS result.

The Proof Foundry proposes; the operator disposes. This record carries no capital authority and makes no unsupported alpha claim.

SigmaX proposes; humans authorize. This record carries no capital authority and makes no allocation recommendation.

This dossier is research evidence only. Every figure above is paired with its limitations and its in-sample / out-of-sample framing — no allocation recommendation is made. Data mode: SYNTHETIC. Archived record — recorded 93 days ago, not a current observation. Capital authority: none.