Sleeve discovery

Objective

Source

config/owner_goals.json

Source schema

canli.alphac-owner-goals.v1

Recorded on

2026-09-12

In force from

2026-09-14

Honest forward sharpe target

2

Honest forward sharpe comparison

ABOVE

Portfolio max drawdown target

0.100000

Portfolio max drawdown comparison

AT_MOST

Portfolio max drawdown statistic

REALIZED_MAXIMUM_DRAWDOWN_BOUND

Portfolio max drawdown mechanism status

MECHANISM_LIVE_BOUND_ENFORCED_UP_TO_ONE_DAY_OVERSHOOT

Portfolio max drawdown mechanism source

config/drawdown_control_contract.json

Portfolio max drawdown mechanism status at goal recording

DECLARED_LADDER_DERIVED_FROM_BOUND_MEASURED_NOT_LIVE

Expected max drawdown objective

0.110000

Expected max drawdown objective statistic

expected_maximum_drawdown

Target total sleeves

14

Target total sleeves comparison

AT_LEAST

Target sleeve count

14

Average pairwise correlation objective

-0.051161

Targets are admission evidence

no

Current sleeves

3

Minimum new sleeves

11

Candidate atlas minimum

200

Honest forward sharpe basis. Annualized (365-day convention) from consecutive published UTC daily paper marks, net of modeled costs, zero excess-return benchmark, evaluated by scripts/evaluate_forward_evidence_maturity.py under config/forward_evidence_contract.json (252 returns before an estimate, 756 and PSR >= 0.95 before establishment).

Portfolio max drawdown mechanism. config/drawdown_control_contract.json declares half gross at 5% below the high-water mark and flat at 10%, absorbing until an owner rearm. Activation is declared live in that contract. A daily brake can overshoot within a day; it does not guarantee the future bound. This declaration is not independent evidence of runtime enforcement.

Expected max drawdown objective note. The admission contract's sealed expected-maximum-drawdown objective and gate, evaluated on the current-composition drawdown study. Not the owner's bound: the bound is on realized maximum drawdown and is portfolio_max_drawdown_target.

Average pairwise correlation objective note. An objective, not a gate: the average pairwise correlation at which 14 sleeves of the quality the sealed contract measured reach 3 in-sample, which is forward 2 on the optimistic end of the measured haircut. Derived by inverting the published identity; the PSD floor at this sleeve count is -0.0769.

Portfolio sharpe target

• 3
• 4

Portfolio sharpe target unit. IN-SAMPLE, being the band an honest forward 2 implies at the 1.5x and 2x ends of this book's own measured backtest-to-forward haircut. The headline target is the forward figure.

What reaching it requires. At 14 sleeves of the quality the sealed contract measured (0.464 on the four-curve basis), the low end of the band (3 in-sample, 2 forward on the optimistic haircut) needs an average pairwise correlation of -0.0512 against a floor of -0.0769 (inside the floor). At the +0.00 gate exactly it needs a mean standalone Sharpe of 0.802. The distance from today's measured quality has to be bought with per-sleeve quality, genuinely negative correlation, or both; the sleeve count alone buys none of it.

Portfolio requirement. Improve the combined book through independent economic return sources, not leverage or selection of the prettiest backtest.

Admission rule. A target never creates a sleeve. The owner's minimum is reached only if enough new identities independently clear every admission gate; otherwise the book remains smaller.

Admission gates

Point in time data

yes

Net of costs

yes

Walk forward only

yes

Book deflated sharpe must be measured

yes

Deflated sharpe must be measured

yes

Requires capacity curve

yes

Requires kill criteria

yes

Pbo max

0.200000

Candidate average correlation to existing book max

0

Book average pairwise correlation delta max exclusive

0

Average pairwise correlation upper 95 max

0.100000

Pairwise correlation max

0.350000

Stressed pairwise correlation max

0.500000

Minimum oos observations

756

Net sharpe min

0.150000

Capacity usd min

500000

Credential policy

Storage

environment variables or managed secret store only

Request timing

Only after a candidate passes its key-free feasibility screen.

Never

• source code
• research artifacts
• chat transcripts
• client-side bundles

Audit

Performed at

2026-08-15

Public source

glassbox/kill_log.json

Finding. The v1 queue accidentally relisted eight previously tested families. They are preserved below as retired, not presented as fresh research.

Retired admission gates

Deflated sharpe min. Retired as a per-sleeve gate in v6 because it measured the wrong object and silently required about eight times the declared net-Sharpe floor. Per-sleeve DSR remains a mandatory published measurement.

Book deflated sharpe min. Re-scoped prospectively in v7 from an incremental admission gate to a 0.95 portfolio-maturity gate. Full-union book DSR remains mandatory to measure and publish at every admission; known results were not regraded.

Average pairwise correlation max. Replaced prospectively in v7 because applying a final global level to each incremental admission was path-dependent. A candidate must now average no more than zero correlation to the existing book and must strictly reduce the book's global average; the -0.03 portfolio objective remains mandatory to publish.

Frontier arithmetic

Identity

S_book = s_bar * sqrt(N / (1 + (N - 1) * rho_bar))

Source

config/owner_goals.json projected on the sealed contract's identity and inputs

Target sleeve count

14

Honest forward sharpe target

2

Psd floor at target n

-0.076923

Incremental candidate average correlation gate

0

Incremental book average correlation delta gate exclusive

0

Incremental gates alone establish objective floor

no

Incremental gates alone establish objective ceiling

no

In sample support band

• 3
• 4

In sample support band unit. IN-SAMPLE, being the band an honest forward 2 implies at the 1.5x and 2x ends of this book's own measured backtest-to-forward haircut. The headline target is the forward figure.

Psd floor note. Average pairwise correlation cannot fall below -1/(N-1) for any real correlation matrix. At 14 sleeves the floor is -0.0769.

Honest reading. At 14 sleeves of the quality the sealed contract measured (0.464 on the four-curve basis), the low end of the band (3 in-sample, 2 forward on the optimistic haircut) needs an average pairwise correlation of -0.0512 against a floor of -0.0769 (inside the floor). At the +0.00 gate exactly it needs a mean standalone Sharpe of 0.802. The distance from today's measured quality has to be bought with per-sleeve quality, genuinely negative correlation, or both; the sleeve count alone buys none of it.

Superseded admission contract frontier arithmetic

Identity

S_book = s_bar * sqrt(N / (1 + (N - 1) * rho_bar))

Incremental book average correlation delta gate exclusive

0

Incremental candidate average correlation gate

0

Incremental gates alone establish objective ceiling

no

Incremental gates alone establish objective floor

no

Legacy global zero correlation reference

0

Psd floor at target n

-0.076923

Target sleeve count

14

Honest reading. Tightening the correlation ceiling from 0.15 to 0.00 removes a ceiling that sat BELOW the objective at every plausible quality, which is why the v4 gate could not have produced the objective at any sleeve count. It does not on its own deliver the objective. At the gate exactly, fourteen sleeves of today's measured quality reach the figure recorded above, and the remaining distance has to be bought with per-sleeve quality, genuinely negative correlation, or both. Publishing this is the point: a target sitting beside a gate that forbids it is not a stretch goal.

Psd floor note. Average pairwise correlation cannot fall below -1/(N-1) for any real correlation matrix. At fourteen sleeves that floor is far from binding, which is why a negative average is arithmetically available here and is not at 250 sleeves.