The open book.
A technical record of what the engines read, what each sleeve believes, how the book is assembled, what failed, and what can be reproduced. This is a living publication: evidence updates from the same artifacts that drive the live paper dashboard.
- State generated
- 23 Aug 2026, 08:25 UTC
- Current composition
- 4 sleeves / equal quarters
- Research grade
- C+ / not cleared
- Evidence hash
- ac88931664effa…0ba4ced6
One evidence path, end to end.
ALPHAC is not one prediction model. It is a controlled path from time-aware data to four economically different return streams, then through portfolio construction, risk controls, broker execution, and a public record guarded against impossible output.
Four sleeves. Four reasons to exist.
Equal capital is not equal confidence. Each sleeve has a separate economic premise, clock, evidence history, broker path, and disclosed weakness. The quarter weights are portfolio construction; they are not a claim that every edge is equally strong.
AlphaForge
Crypto funding carry
25% OF NEUTRAL COREFunding-rate carry on Binance USDT-M perpetuals, market-neutral.
- Input
- Realized perpetual-funding settlements and point-in-time liquidity
- Transform
- Cross-sectional funding carry, constrained to a market-neutral book
- Clock
- Hourly monitor / weekly target refresh
Evidence note. 0.68 is the FULL-HISTORY figure and hides the tail: -1.63 Sharpe / -19.6% drawdown in 2022 (FTX). Carry compresses with size/crowding. CORRECTION (2026-07-05): from go-live to 07-05 this sleeve was flat and we published 'funding carry compressed, holding cash' — that explanation was WRONG. The real cause was a signal-wiring bug: the live blend included equity-fundamental alphas that are undefined on crypto, which invalidated every signal; the validated carry-only configuration was not what was deployed. Found in an internal audit and fixed the same day — the loop now runs the exact blessed walk-forward configuration (carry_fund_21, weekly rebalance). No money was misreported (the sleeve genuinely held cash at $100k); the published explanation was wrong, and we correct it here rather than rewrite it.
AlphaMax
US-equity 12-1 momentum
25% OF NEUTRAL CORE12-1 cross-sectional momentum, dollar-neutral long/short, split-adjusted, survivorship-free.
- Input
- Split-adjusted, survivorship-free US-equity prices
- Transform
- 12-1 cross-sectional momentum, dollar-neutral long/short
- Clock
- Daily monitor / quarterly portfolio refresh
Evidence note. Genuinely broker-executed as of 2026-08-07 on its own Alpaca account, but the live record is only days old. 0.91 is the TOP of a 0.07-1.17 construction band; the deep-history forward is ~0.08 with DSR 0.34 (does NOT clear the deflation gate). The honest central estimate is ~0.5-0.6, not 0.91 — the live book will settle it. A few unshortable / non-fractionable small-caps are skipped, so the live book is a close tradable proxy of the research strategy, not an exact replica.
AlphaTrend
Managed-futures trend
25% OF NEUTRAL CORETime-series momentum across a 17-market basket (equity-index, rates, commodities, FX via liquid ETFs), long/short on each market's own trend, inverse-vol weighted. THE RE-DERIVATION IS DONE (2026-08-07) AND IT WENT AGAINST US: at the honest trial count (N=133) and the pooled trial variance, its DSR is 0.000 — not the 0.83 we published, and the WORST of the 4 sleeves rather than the best. The old 0.83 rested on a V[SR] input ~80x too small and n_trials=5 while AlphaMax was graded at N=101; it was marked on an easier exam and then called the soundest. Net Sharpe 0.25 over its full 5,179-day history. Positive skew, two-decade stable, near-uncorrelated to equities. We keep it because removing it measurably worsens book drawdown, not because its edge is established.
- Input
- Seventeen liquid ETF proxies across equities, rates, commodities and FX
- Transform
- Per-market time-series trend, inverse-volatility weighted
- Clock
- Daily monitor / biweekly target refresh
Evidence note. We called this 'the soundest sleeve'. On the re-derived numbers that was wrong: its DSR is 0.000 at honest N=133, the lowest of the three. It IS genuinely broker-executed on Alpaca paper and it does survive 2008/2020/2022 with ~0 equity correlation. As of the v2 restart (2026-08-07) it is FOLDED INTO the flagship ALPHAC book at a 20% weight. The case for keeping it is measured drawdown reduction, plus the fact that its recent 3 years (-0.23) are a small and unrepresentative slice of a 20.6-year record (+0.25 full, +0.28 over 15y) — judging a trend sleeve on its worst window is the most documented allocator error there is. The case against it is that 0.000 is 0.000. Both are stated because both are true.
AlphaVintage
PIT macro-surprise size spread (IWM/SPY)
25% OF NEUTRAL COREPoint-in-time CPI surprise — the standardized AR(3) residual of headline (PCPI) and core (PCPIX) inflation, differenced WITHIN a single ALFRED vintage — traded as a dollar-neutral IWM-minus-SPY size spread, monthly.
- Input
- ALFRED point-in-time headline and core CPI vintages
- Transform
- Standardized AR(3) surprise mapped to an IWM-minus-SPY spread
- Clock
- Monthly, only after the release becomes knowable
Evidence note. CORRECTION 2026-08-19 — we published this sleeve at net Sharpe 0.3403 with Newey-West t 1.82 over 5,996 days. Those numbers are WITHDRAWN. Re-run on 2026-08-16 with a calendar correction that retains zero-exposure sessions, its net Sharpe is 0.2298 and its Newey-West t is 1.267 over 6,296 portfolio days (5,998 active). That t FAILS the 1.5 bar this candidate pre-registered, so its own result artifact records verdict KILLED — and we deployed it on 2026-08-10, six days before the corrected run existed, then left the old figures on this page. It is still carrying a quarter of the book while we decide whether to withdraw it; that decision is open and we are not going to describe a killed candidate as a validated one in the meantime. It never cleared the 0.95 deflation gate on any version of the numbers. Two further disclosures stand: it runs at HALF the researched notional (gross 1.0x not 2.0x) because 2.0x breaches our own runaway brake and this account's Reg T limit — Sharpe is unchanged by scaling, dollar return is halved; and the research modelled NO short-borrow cost while the live sleeve is short SPY ~95% of days, a drag we charge at 50bp/yr.
Correctness is encoded, not implied.
The engineering claim is deliberately narrower than “the backtest is true.” These are executable invariants: the specific classes of leakage, arithmetic drift, and publication error the system is designed to make difficult.
Every weight and leverage applied to a forward day is a pure function of returns strictly before that day, enforced by the PIT reader and a truncation-invariance test.
Every scripted fill, fee, funding payment and final equity is asserted.
Historical BTC funding is reproduced at rel=1e-12.
Engine cost inputs match the registered feature library instead of forking formulas.
The null is part of the product.
A strategy is not promoted because its chart looks persuasive. Trial count, deflated Sharpe, overfit probability, costs, and pre-written gates decide what the research can honestly claim. The current verdict remains no-deploy on backtest evidence alone.
Deployed-path and prototype-screen failures remain in the public ledger.
Survival does not mean statistical proof; each caveat remains attached.
Required: 0.95 or higher.
Required: below 0.20.
NO-DEPLOY (honest null). Best config DSR 0.2112 < 0.95 gate AND PBO 0.8818 > 0.20 gate. No variant beats baseline and clears deflation simultaneously. Crypto-perps alone is a thin cross-sectional signal that does not survive honest deflation net of cost; the equity sleeve is required as ballast.
Search debt, surfaced.
Every measurement stays on record. DSR selection N removes only rolling window remeasurements and duplicate identities across ledger profiles.
92 identities remain before mandatory review.
This ledger proves what was measured and how selection N is counted. It does not prove that any identity is profitable, independent, or admissible. Inspect the machine-readable ledger
Read the lineage, including the nulls.
These papers bind shared mechanism, literature, authorship, and trial accounting to entire research families. They remain incomplete packets until every identity-level section is proved.
A target is not a result.
The governing objective is an honest forward portfolio Sharpe of 1.5 across up to 14 evidence-backed sleeves, including at least 10 new admissions, with expected maximum drawdown near 11%. The book's measured backtest-to-forward haircut means that objective requires a research Sharpe band of 2.25–3.00. These are different units, and neither is a promised result. The book remains smaller if point-in-time lineage, costs, capacity, deflation, execution, or correlation gates do not clear.
Across 20 asset groups.
Two universes × three horizons per family.
No documented identity match; not performance evidence.
Key-free evidence work currently authorized.
Historical aliases cannot re-enter as new.
Blocked pending an identity-level distinction.
0 family failures; current book exact-match.
Taxonomy and governance only.
Targets cannot manufacture evidence.
Shared fail-closed admission contract.
Each tested or justified as inapplicable.
240 coverage cells. Zero sleeves claimed. Passing this audit is not return evidence and does not create a sleeve. It only permits a separately sealed key-free feasibility review. 60 cells may proceed only to feasibility; 96 remain stopped for literature review; 42 require overlap resolution; 24 are retired/killed and 6 monitor a forward-only continuation. Returns opened: 0. Return identities spent: 0.
Inspect all 240 identities / Inspect all 2,880 gate evaluations / Inspect source-bound lineage / Inspect the admission contract
earnings narrative change
Not an admission. Passing this contract makes a result technically eligible for an explicit admission decision. It does not guarantee a sleeve, future performance, the target Sharpe, or the target drawdown. Point estimates cannot override confidence-bound, crisis-conditional, tail-loss, execution-evidence, or capacity-reconciliation failures. v6 additionally refuses to load if its own significance floors cannot all be satisfied at once, gates the average pairwise correlation on its upper confidence bound as well as its point estimate, and publishes the arithmetic relating the correlation gate to the portfolio objective -- including the distance the gate does not close.
The machine waits for a human.
These source-only packets are frozen before scoring. An independent reviewer must complete every label and attest that parser outputs, prices, and returns were not consulted. Until then, no downstream accuracy or sleeve claim exists.
Repurchase / issuance Item 703
Determine whether each filing contains the frozen Item 703 table structure.
- Frozen sourceHash-bound
- Independent reviewWaiting
- Seal and scoreLocked
This packet enables independent source labeling only. It contains no parser prediction, return, price, Sharpe, drawdown, correlation, capacity, or admission claim and spends zero return identities.
Active ownership Item 4 v3
Classify source language and exact aggregate ownership under the frozen protocol.
- Frozen sourceHash-bound
- Independent reviewWaiting
- Seal and scoreLocked
This packet enables an independent blind source audit only. It contains no machine predictions and makes no accuracy, return, Sharpe, drawdown, or sleeve claim.
earnings narrative change
US equities / one to three months
Point-in-time change in issuer language and disclosed operating risks, residualized against earnings surprise, momentum and sector
83,070 official 10-K records mapped; 82,491 pair-eligible filings; 0 metadata failures. The locked runner cannot load returns until the corpus is complete.
- Required evidence
- Point-in-time filing text, earnings-call transcripts, speaker roles, revisions and delistings
- Potential sources
- SEC EDGAR / LSEG StreetEvents / FactSet CallStreet
- Kill rule
- Kill if section lineage cannot be reconstructed point-in-time, language changes are unstable across filing formats, or residual correlation to AlphaMax exceeds 0.35.
analyst revision drift
US equities / one to three months
Point-in-time consensus estimate revisions after controlling for earnings surprise and momentum
- Required evidence
- As-of analyst estimates, contributor counts, announcement timestamps and delistings
- Potential sources
- LSEG I/B/E/S / FactSet Estimates / Visible Alpha
- Kill rule
- Kill if revision timestamps cannot be reconstructed point-in-time or residual correlation to AlphaMax exceeds 0.35.
options dispersion
equity options / weekly to monthly
Index-versus-single-stock implied correlation dispersion with defined crash loss
- Required evidence
- Historical SPX and constituent option chains, index membership, bid/ask, greeks and corporate actions
- Potential sources
- OptionMetrics / Cboe DataShop / ORATS
- Kill rule
- Kill if the edge is ordinary short index variance, crisis expected shortfall breaches budget, or constituent costs consume half the gross premium.
merger arbitrage
US equities / event driven
Announced cash-deal spread convergence conditioned on financing, regulatory and termination risk
1,965 official target anchors found. Later outcome coverage reached 91.8%, but prior deal-state coverage reached only 67.0%; the pre-written 80% gate failed. No return test was run.
A separately locked tender-only branch downloaded 100 target filings and extracted Item 4 from 94.0%. It still failed: only 10.6% yielded one safe cash price, while 74.5% were ambiguous and board posture resolved in only 22.3%. The parser was not retuned and no return test was run.
- Required evidence
- Point-in-time deal terms, revisions, breaks, votes, close dates and borrow costs
- Potential sources
- SEC EDGAR / LSEG Deals / FactSet Mergerstat
- Kill rule
- Kill if deal terms cannot be reconstructed without hindsight or realized break losses dominate the costed spread.
index reconstitution flow
global equities / days to weeks
Announcement-to-effective-date demand from transparent index additions and deletions
- Required evidence
- Point-in-time announcement archive, index weights, float, ADV and closing-auction data
- Potential sources
- S&P Global / FTSE Russell / MSCI
- Kill rule
- Kill if announcement history is survivorship-biased or auction slippage consumes the forecast edge.
electricity load weather spread
power and natural-gas futures / day-ahead to weekly
Forecast errors in weather-driven regional electricity load relative to marginal fuel and power prices, neutralized to commodity trend
- Required evidence
- Point-in-time weather forecasts, ISO load forecasts and revisions, nodal or hub power prices, fuel prices, listed contract rolls and executable bid/ask
- Potential sources
- EIA / NOAA / PJM / ERCOT / CME DataMine / ICE Data Services
- Kill rule
- Kill if forecast vintages cannot be reconstructed, listed liquidity is insufficient, or returns reduce to natural-gas trend or generic seasonality.
Read locked feasibility protocol / Inspect sealed feasibility result
securities lending supply
US equities / days to weeks
Point-in-time borrow-fee and utilization changes as a constrained-supply return source
- Required evidence
- Historical borrow fees, utilization, recalls and availability by security
- Potential sources
- S&P Global Securities Finance / DataLend / Orbisa
- Kill rule
- Kill if locate and recall costs erase the signal or historical availability cannot be established point-in-time.
credit equity relative value
corporate credit and equities / weeks to months
Issuer-level equity-versus-credit dislocation after neutralizing market, sector, duration and quality
- Required evidence
- Point-in-time bond prices, reference data, TRACE liquidity and equity corporate actions
- Potential sources
- FINRA TRACE / LSEG / Bloomberg
- Kill rule
- Kill if issuer mapping is hindsight-contaminated, liquidity makes fills implausible, or the result is merely quality or momentum.
active ownership escalation
US equities / one to six months
New Schedule 13D control-intent ownership followed by specific governance, board, capital-allocation or strategic action
Schema-aware SEC discovery passed: 22,353 initial accessions, 800 immutable headers, 100% target/filer lineage, and 47.3% contemporaneous ticker coverage with a 43.8% Wilson lower bound.
The locked Item 4 stage then failed. Parser v2 resolved 159/160 submissions and extracted 139/160 sections ( 86.9% versus 90% required). No parser v3 or return test was opened.
- Required evidence
- Point-in-time Schedule 13D/13D-A state, reporting groups, Item 4 intent changes, ownership percentages, exits, delistings and borrow
- Potential sources
- SEC EDGAR / LSEG Ownership / FactSet Ownership
- Kill rule
- Kill if specific active intent cannot be separated from boilerplate point-in-time, target/reporting-person lineage is unstable, or residual correlation to existing equity sleeves exceeds 0.35.
treasury auction concession
US Treasury futures and liquid duration ETFs / one to three sessions
Temporary duration concession before repeated US Treasury coupon auctions followed by post-auction normalization, conditioned on issue size and dealer-balance-sheet pressure
2,037 official coupon-auction events passed the locked metadata gates, including 1,377 from 2013 onward. Core fields and event identities were complete; 99.56% had at least one day of notice. No prices were opened and zero return identities were spent.
A second timing audit found 319 two-year auctions, a median 5 calendar days of formal notice, and 0 events with the 10 calendar days needed even before counting trading sessions. The published 10-session identity remains closed until archived tentative schedules are sealed.
- Required evidence
- Auction announcements and results, reopening status, issue size, when-issued or futures prices, contract rolls, bid/ask and primary-dealer positioning
- Potential sources
- US Treasury Fiscal Data / Federal Reserve Bank of New York / CME DataMine
- Kill rule
- Kill if the effect is absent after the 2013 publication date, reduces to unhedged duration beta, cannot survive one-tick-plus-fee costs, or correlates above 0.35 with AlphaTrend.
Read locked feasibility protocol / Inspect sealed feasibility result
Read identity-timing protocol / Inspect identity-timing result
cftc hedging pressure
commodity futures / weekly to monthly
Cross-sectional commodity risk transfer from commercial hedging pressure using only CFTC reports available after their Friday release, neutralized to trend and curve carry
183,676 official metadata rows across 21 years passed schema and identity checks, but exact historical release lineage is 0% and no tradable-contract map is sealed. The candidate is data-gated. No positions, prices, or returns were opened.
- Required evidence
- Point-in-time COT disaggregated positions, release timestamps, futures settlements, rolls, limits, bid/ask and contract multipliers
- Potential sources
- CFTC / CME DataMine / ICE Data Services
- Kill rule
- Kill if release-lagged positions do not add beyond trend and curve carry, the tradable map is unstable, crisis liquidity breaks the cost model, or correlation to AlphaTrend exceeds 0.35.
Read locked feasibility protocol / Inspect sealed feasibility result
pre fomc announcement drift
US equity-index futures or SPY / intraday event window
A publication-aware replication of the equity premium reported in the hours before scheduled FOMC announcements, isolated from overnight and ordinary market beta
79 official regular FOMC statement events span 2016–2025. Release-time coverage and the 2:00 p.m. ET convention both measured 100%, but point-in-time schedule-revision lineage is 0%. Returns remain unopened and zero identities were spent.
- Required evidence
- Point-in-time FOMC schedule changes, intraday index futures or SPY quotes, spreads, holidays and announcement timestamps
- Potential sources
- Federal Reserve / Polygon / CME DataMine
- Kill rule
- Kill if the post-publication sample is non-positive net of costs, the result is ordinary overnight beta, schedule revisions cannot be reconstructed, or left-tail behavior violates the event-risk budget.
Read locked feasibility protocol / Inspect sealed feasibility result
repurchase issuance flow
US equities / one quarter to one year
Filing-known completed corporate equity demand: completed repurchases versus contamination-resolved completed issuance, never authorization headlines
SEC COLLECTION AND SEMANTICS COMPLETE AWAITING BLIND LABELS. The 48-test no-return pipeline is implemented. Every stage binds exact manifest identities, immutable part hashes and zero return access, and exits nonzero when its required seal or gate is absent. Blind labels must be sealed against the parser source before parser output exists. Company Facts or Item 703 extraction alone can never pass full feasibility; combined coverage, amendment replay, quarterization and contamination gates remain mandatory.
- Required evidence
- PIT XBRL facts, accession-bound Item 703 tables, shares and corporate actions, issuance contamination flags, delistings, borrow, spreads and market impact
- Potential sources
- SEC EDGAR / LSEG Fundamentals / FactSet Fundamentals
- Kill rule
- Kill if completed repurchases and issuance cannot be separated point-in-time from authorizations, amendments, stock compensation, splits, acquisitions and custom tags, or if a later residual is subsumed by existing equity sleeves.
Defects stay visible after the fix.
The red team is not a testimonials section. It is a durable record of where the system or the publication flattered itself, the severity of the defect, and the action taken. Open findings remain open.
A future-dated (2026-06-29) paper loss leaked into the public record and was sealed into seq 0 of the signed transparency chain (read_fwd_curve had no <= today guard).
clamp_live() drops any mark dated after today (UTC) before publish + anchor; a fail-closed guard was added; the corrected state was anchored as seq 1. The append-only chain shows BOTH seq 0 and seq 1 — we don't delete mistakes.
The headline -4.7% max drawdown excluded every crisis (the book curve starts 2023-07, no 2020/2022). It understated risk badly.
Crisis-inclusive worst case (-15 to -18%) is now the headline; -4.7% is labeled 'live-overlap only, not a risk estimate'.
Sleeves were mislabeled: crypto carry (46% weight) has been FLAT since go-live but was billed 'live broker-loop'; equity & MF were billed 'realized next-open fills' but are walk-forward SIMULATIONS, not broker-executed.
Relabeled honestly (crypto = live loop currently flat / holding cash; equity & MF = walk-forward simulation, not broker-executed). Making them genuinely live is the next build.
Standalone Sharpes were recent-window-flattered: crypto 0.68 hides a -19.6% 2022 (FTX) drawdown; equity 0.91 is the TOP of a 0.07-1.17 band whose deep-history forward is ~0.08 (DSR 0.17 at the current N=102 ledger; 0.34 when first published at N=27 — fails deflation either way, and deflation tightens as our own experiment count grows).
Per-sleeve caveats attached; the book forward expectation lowered from 0.7-1.0 to 0.3-0.9 with a stated real chance of ~0 in year one.
Deflation N was framed as 'honest trial count = 8' when 35+ strategy families have been searched across the program (screen-stage kills bypassed the ledger).
Renamed to crypto_perp_grid_trials; exposed cumulative_strategy_families_searched (35+). Wiring every screen into the shared ledger is the remaining engineering step.
Sleeve correlations spike from ~0 (calm) to +0.6 to +0.9 in risk-off — the diversification benefit shrinks exactly in the left tail.
Disclosed in the published metrics + transparency; sizing the equal-risk combiner on the stressed matrix is the remaining engineering step.
The cost model uses a flat half-spread regardless of regime (it never widens in stress); the two deployed sleeves lack a published 2-3x cost-survival rerun.
Vol/regime-scaled half-spread + a published 2x/3x cost rerun of the live sleeves. Tracked.
go-live date overstated 1-2 days vs first real activity; the verifier checks the published JSON, not the raw jsonl source-of-truth.
Per-sleeve effective go-live = first real activity; verifier to accept the raw jsonl. Tracked.
Read the sealed AlphaVintage missing-release correction protocol
Read the claim. Then inspect the artifact.
The book is an interface over downloadable evidence, not a substitute for it. Fetch the raw state, inspect the failed research, and run the published verifier. The exact scope of the claim is content-hashed + byte-reproducible backtest; not blockchain-anchored; no chain-of-custody proofs.
This publication is research and live paper trading only. No client funds are managed, simulated and paper results are not real returns, and nothing here is an offer, solicitation, or investment advice. The record can prove what was published; it cannot prove future performance.