LIVE SYSTEM · GOVERNED SPINE | Julia | Research platform | Full corpus | Citation
A governed, systematic risk-premium harvesting platform — in Julia.
Blaque Baux does not try to predict returns. After rigorous out-of-sample testing found no exploitable predictive edge at the horizons and instruments studied, it was built around the edge that does survive: risk structure. It harvests diversification and trend premia, times risk with a stateful daily process, and routes every order through a governed execution layer that enforces hard, tested invariants.
Not investment advice. Educational/research software. All performance figures are historical backtests or paper-trading results, net of modeled costs — not a promise of future results. Trading carries substantial risk, including loss of principal. See LICENSE.
Blaque Baux is a quantitative research initiative and a subsidiary of Carter Warrens. BlaqueBaux.com is the home for the work; the code lives here on GitHub — open to study, test, and build bespoke strategies on top of.
Anyone can point an AI at a market. The edge is understanding what the data actually says — and turning it into something you can act on. We test relentlessly and put most of it on the record as rejected, with the reason; what survives is built, governed, and validated before it is ever called real. That combination — honest research, reproducible evidence, and execution you can trust — is why Carter Warrens leads on strategy and implementation, not merely uses the tools everyone now has.
→ The capstone white paper — docs/whitepaper.html
Blaque Baux — An Honest Architecture for Systematic Risk. The whole of it in one document: the philosophy, the engine, the four books that trade, the derivatives layer, the graveyard of rejected ideas, and the ten durable laws the failures taught. Start here.
- The spine — a two-sleeve strategy: an inverse-vol / equal-risk-contribution base that harvests the diversification premium by risk structure, plus a 12-month time-series-momentum trend sleeve that hedges crises (it made money in 2022, short bonds / long energy). Each sleeve is volatility-targeted on its own realized P&L, blended, and scaled by a drawdown-based regime brake. Rebalanced once daily.
- Governed execution — a venue-agnostic controller enforcing idempotency, per-pool budget and loss limits, data-staleness, position reconciliation, a kill switch, and full fill lineage. No language model is anywhere near the order path; the strategy is reproducible code.
- A live-money safety gate — pre-trade drawdown/loss halts, gross-leverage and per-name caps, account/data checks, and alerting, on top of the controller's invariants.
- Honest validation — the methodology and the results, including what didn't work, are documented in full (see below). This repo tries to be a trustworthy quant reference, not a pitch.
| Sharpe | Sortino | Max drawdown | 2022 | Cadence | |
|---|---|---|---|---|---|
| Spine (production, regime brake) | ~0.97 | ~1.30 | ~−11% | −1.4% | daily |
Honest context: this is a single-digit-CAGR, ~1.0-Sharpe strategy — institutional-quality
risk-adjusted performance, not headline returns. Leverage to reach double digits is possible but
carries proportional (double-digit) drawdown, and is net-negative at today's margin rates — the
trade-off is quantified in docs/leverage_decision.html. There is
no double-digit-return-at-low-risk configuration, and this repo says so.
The spine is the survivor of a much larger program. This repo is a research corpus: two dozen strategies were built and tested across return-prediction, convexity / tail-hedging, correlation structure, and leverage — and most were rejected, on the record, with the reason. The complete white paper is the map:
→ docs/compendium.html · PDF — The Complete Method
| Status | # | Examples |
|---|---|---|
| Live / production | 2 | the spine; DBA agriculture sleeve |
| In live A/B | 2 | multi-horizon trend; split-universe spine |
| Research (kept) | 6 | Gamma-ARMA regime framework; Taleb barbell; the curveball; inverse-carry tail hedge; diversified tail hedge; the optimizer library |
| Tested & rejected | 10 | Bayesian & cross-sectional alpha; blue-chip / mid-cap prediction; 15-min alpha; earnings lead-lag; pairs stat-arb; vol-overlay hedge; convex-response function; carry sleeve; carry-as-base; leverage-to-double-digits |
Ten durable laws came out of it — e.g. convexity is free (trend) or paid (long-vol), never both cheap
and fast; timing the tail removes the tail; correlation is priced instantly — a risk tool, not
alpha; aggression multiplies edge, and with none it multiplies only ruin; you can't short
momentum-driven strength (it's ruin) — and being long it survives on drift, not selection: winner-picking
is beta, the tradeable residue is trend convexity. Every result is a reproducible sketch in
scripts/research/ (34 scripts), consolidated in
docs/research_thread_summary.pdf.
That most of the scorecard is red is the point — a strategy is only as trustworthy as the ideas it was willing to kill.
Everything — covariance estimation, the sleeve construction, risk metrics (Sharpe/Sortino/Calmar/ VaR/CVaR), the governed-execution invariants, the validation methodology (Information Coefficient), and the empirical findings (momentum-vs-reversal by horizon, daily-vs-intraday, the leverage trade-off) — is specified with formulas and code references in:
Architecture and design decisions: docs/CANONICAL_ARCHITECTURE.md.
The spine is one strategy built on a general-purpose optimization library (src/module_13_portfolio/,
module PortfolioOpt) that is useful on its own:
- Optimizers — mean-variance (min-variance, max-Sharpe, efficient frontier), Black–Litterman, risk parity / HRP / max-diversification, tail-risk (min-CVaR / min-CDaR via Rockafellar–Uryasev LPs), and cost-aware mean-variance (trades from current holdings with linear + impact penalties).
- Monte-Carlo robustification (
robust.jl) — Gaussian / IID / block / stationary bootstrap and Student-t data-generating processes feeding a Michaud resampled frontier and a feasible-set Monte-Carlo cloud. This is MC used to defend against estimation error, not to forecast price. - A REST service (
scripts/portfolio_server.jl, JSON on:8766) with a Python Dash dashboard (scripts/dashboard.py):/optimize,/frontier,/resampled_frontier,/backtest,/metrics,/random_portfolios.
julia --project=. scripts/portfolio_server.jl # optimizer backend on :8766
python scripts/dashboard.py # dashboard UI on :8050A worked cross-family demo — scripts/research/multi_sleeve_portfolio.jl —
runs risk-parity / min-variance / max-diversification / HRP / min-CVaR over the full
keeper set: the five asset-class spine plus ten reconstructed keepers — crude→refiner
(CRACK), beta-hedged market-neutral (BORE), vol-scaled multi-horizon trend (TREND),
brown/blue camp rotation (CAMPROT), drawdown-bounce (DDBOUNCE), Block's four-block cross-asset
trend (BLOCK), the live drawdown-regime brake (REGIME), the actual Gamma-ARMA crisis detector
wired in (GAMMA_REG — module 4 ARMA+GARCH tail-index/vol + module 5 detect_crisis_regime),
and two paid-convexity tail hedges, the Taleb barbell (BARBELL) and vol-gated curveball
(CURVEBALL). It surfaces three honest lessons. First, the return-earning keepers push the
diversified book past the best single sleeve, and the risk budget flows to the genuinely
uncorrelated fragments (TREND corr −0.14, BORE −0.06), not the high-Sharpe beta sleeves —
high standalone Sharpe doesn't earn weight, low correlation does; and folding in BLOCK shows the
converse — it's 0.84-correlated to TREND, so the optimizer just splits the trend budget between
them and the book lifts only +1.60→+1.66 (the FX/dollar block, Block's one new axis) — redundant
sleeves don't earn new weight either. Second, the two regime timers differ sharply: the live drawdown
brake catches slow drawdowns but whipsaws, while the wired-in Gamma-ARMA detector flags only
~2% of days yet catches ~67% of the COVID crash and is the best single ingredient (+1.13,
COVID −8% vs SPY −33%) — it times the tail cheaply; but it's 0.88-correlated to SPY (so
still down-weighted as beta) and its thresholds fit COVID in-sample, which is exactly why the
framework stays research and the live spine trusts the simpler brake. Third, the paid-convexity
hedges are negative-carry insurance (CURVEBALL alone a −86% ruin) that a variance objective
misprices — risk-parity hands BARBELL ~31% because it's low-vol and anti-correlated.
Convexity must be budgeted, not optimized in: free convexity (trend) the book earns, paid
(long-vol) you size, timed (the crisis detector) is seductive in-sample — the spine's "harvest
risk structure" thesis with the tail-hedge and regime-timing caveats made visible.
On the Gamma-ARMA framework: modules 1–6 (ARMA+GARCH, the Gamma-hyperprior DPM regime model,
detect_crisis_regime) are present in-repo and importable, and unit-tested undertest/runtests.jl. "Quarantined" means only that they sit outside the production validation gate and the live path — a separate research lineage that showed no edge over the simpler drawdown brake, not code that was removed.
The keeper book has a governed dry-run/paper driver — scripts/keeper_book_live.jl
(wrapper run_keeper_book_daily.sh, launchd
com.blaquebaux.keeper_book.plist). It rebuilds the
book daily (risk-parity over the 8 ingredients), expands each sleeve into its current instrument weights,
nets them per symbol, and routes the targets through the same Layer-3 safety gate + governed execution
controller as the spine (preflight → execute_rebalance! → reconcile). It defaults to dry-run
(computes the book, runs the gate, logs the netted targets — places nothing) and graduates to Alpaca
paper only with its own isolated keys/ledger.
Before graduating it, scripts/keeper_book_validation.jl is the
validate-before-live gate: a fully causal walk-forward that recomputes the book from data strictly
before each rebalance, nets it to the instrument level (so it pays the sleeves' real internal
turnover), net of costs, against a stated pass/fail bar (+ a purged-K-fold cross-check via
module_11_cv). The keeper book clears it: OOS net Sharpe +1.28 (5 bps/side) / CAGR +6.4% /
maxDD −6%, positive in 9 of 10 years — an honest haircut from the demo's +1.66 gross-in-sample, but
still well ahead of SPY (+0.79 / −34%), with all five checks passing. That is the research earning its
graduation to the paper path — still not validated to the spine's full production bar.
The non-keepers are not discarded — they are governed as tactical regime sleeves. Several near-miss
sleeves (each real and mechanism-grounded, but below the standalone keeper bar) are run the way they are
meant to be used: small, deployed only in their favorable regime, time-boxed to a quarter or two, and
combined so no one sleeve carries the book. scripts/tactical_book_validation.jl
is their gate: a causal, net-of-cost walk-forward of the combined book (cost-push / beige / bulgar / pead),
which comes out +0.46 net Sharpe, beta ≈ 0, uncorrelated to the keeper book (+0.05) — so as an overlay it
lifts the keeper book +1.28 → +1.35 at half weight (~+5%), where any one of them alone adds nothing. The
fourth sleeve, PEAD (post-earnings drift: long top-third / short bottom-third surprise among names still in
their drift window), is event-driven — fed by an earnings-calendar pipeline
(scripts/pead_calendar.py → pead_earnings_calendar.json)
and exempt from the time-box (its positions self-limit as the drift window rolls off); it qualifies
(market-neutral, +0.11, uncorrelated) but adds only marginally — diversification is bounded by own Sharpe. The
governed driver is scripts/tactical_book_live.jl (wrapper
run_tactical_book_daily.sh, launchd
com.blaquebaux.tactical_book.plist): it checks each
sleeve's regime, applies the three rules (10% cap / regime gate / a persisted time-box clock that
forces a stand-down + cooldown after a quarter or two of continuous deployment), nets the combined
market-neutral book, and routes it through the same safety gate + governed execution as the spine. It
defaults to dry-run (and dry-run never advances the time-box clock), with its own fully isolated
keys/ledger/state so it can never touch the spine or keeper accounts.
Two companion analyses build on the same keeper set (via the shared keeper_ingredients.jl
builder). negentropy_ranking.jl asks — in Schrödinger's
negentropy language — what the optimizer pays for: not standalone Sharpe (it avoids it, −0.43) and
not fat tails (marginal non-Gaussianity earns nothing), but inverse volatility (+0.88), with
independence only weakly rewarded; a book built to harvest independence + low vol still reproduces the
engine's risk-controlled character. hedge_saturation.jl draws
the convexity-budget curve: barbell drawdown-protection saturates by ~10% weight, after which only
the negative carry compounds — yet a naive risk-parity assigns the barbell ~29%, deep past the knee. Both
make the demo's "budget convexity, don't optimize it in" concrete.
Full math for all of the above: §9 of docs/FINANCIAL_METHODS.md.
(Crypto note: a Deribit BTC volatility signal is available as a risk input via
module_1_data/data_feeds_production.jl; the spine trades ETFs, not crypto assets.)
Requires Julia (1.10+). From the repo root:
julia --project=. -e 'using Pkg; Pkg.instantiate()'# one-time
julia --project=. test/runtests.jl # gate suites (should be green)Backtest / inspect the spine (uses the bundled scripts/data/sector_panel.csv):
julia --project=. scripts/leverage_decision_data.jl # regenerates the leverage-analysis dataPaper trading through Alpaca (paper keys need no account approval):
export ALPACA_KEY_ID=PK_your_key ALPACA_SECRET_KEY=your_secret
julia --project=. scripts/spine_live.jl # PAPER by default; safety gate always onLive money is deliberately gated: it requires an explicit BB_LIVE_CONFIRM sentinel, a funded /
approved brokerage account, and the safety gate green. Do not flip it lightly.
src/
module_1_data/ data adapters: CSV, Alpaca, IBKR panel providers
module_7_execution/ governed ExecutionController + venue adapters (Alpaca / IBKR)
module_8_governance/ Layer-3 live-money safety gate
module_11_cv/ purged / combinatorial cross-validation
module_13_portfolio/ PortfolioOpt: moments, risk-based weights, the spine, metrics
scripts/
spine_live.jl production daily driver (safety-gated)
run_spine_daily.sh launchd wrapper (scheduled pre-open run)
spine_end_to_end.jl full pipeline on cached data (integration demo)
docs/
compendium.html THE COMPLETE METHOD — every strategy, scored (start here for breadth)
FINANCIAL_METHODS.md the validated math (start here for depth)
CANONICAL_ARCHITECTURE.md architecture & decisions
research_thread_summary.pdf consolidated convexity/correlation/leverage findings
leverage_decision.html interactive leverage trade-off visual
scripts/research/ 34 reproducible research sketches (the scorecard's evidence);
keeper_ingredients.jl is the shared keeper-set builder they reuse
test/ gate + (quarantined legacy) suites
- Paper-tested end-to-end (data → strategy → governed orders → ledger with lineage → reconciliation) against a real broker paper account.
- The strategy is validated out-of-sample; the live path is verified on paper.
- Real capital has not been deployed. The legacy Gamma-ARMA base modules (a separate research
lineage — modules 1–6) remain in-repo and unit-tested under
test/runtests.jl, but sit outside the production validation gate and the live path; the live spine uses the simpler drawdown regime brake (§3.4). "Quarantined" means gate/live exclusion, not removal.
This repository is the canonical, validated core. A larger body of earlier and exploratory work
is preserved outside the repo (a local Archive/ tree) for provenance — it is not published here
and not wired into the live path. It's catalogued so the lineage is clear and so the honest
"what's next" is on the record:
- Crypto-Quant MVP(archived) — a Streamlit app with a natural-language (LLM) interface for
cryptocurrency portfolio risk analysis via Monte-Carlo simulation. A plausible future front-end /
product direction; it is a separate prototype, never connected to the spine. Today crypto appears
in the core only as a risk input (Deribit BTC volatility) — see
data_feeds_production.jl. - Alpha research track(archived; see
FINANCIAL_METHODS.md§10) — a Bayesian return-estimation engine and a Monte-Carlo / "Six-Sigma Oracle" risk engine. The math is sound but measured no predictive edge at the horizons tested, which is why the live strategy harvests risk premia instead. This is "Path A" — kept as funded-research material, not production. - Earlier prototypes(archived) — Python builds (
v1/v2/polyglot) and a standalone optimizer service, all superseded by this Julia core (the in-repoportfolio_server.jlreplaces the old optimizer service).
Direction of travel: the near-term roadmap is depth on the validated core (broader instrument universe, live-money graduation off paper, more governance coverage), not re-adopting archived components. Anything from the archive returns only if it clears the same out-of-sample edge bar the core was held to (§7). No archived component has cleared it yet — and this README will say so until one does.
This repo is the base/blueprint. Each family repo consumes this engine as a git submodule and steers it at a different market — one platform, many directions:
Evaluation standard (family-wide). Returns are fat-tailed, so a bare Sharpe misleads. Every sleeve's research toolkit now tests normality with Jarque-Bera before leaning on any mean-variance number, and reports risk-adjusted performance with Jensen's alpha (return beyond what beta earns) and M² (return re-scaled to the benchmark's own volatility) — not Sharpe alone. Where JB rejects normality (it usually does for daily bars), the left tail / max-drawdown carries as much weight as the mean. This is also the honest core of the conditional-keeper idea: a book "fine until the crash" is only a keeper if a validated regime gate makes the gated version clear the bar on the full sample, black swans in. The toolkit also carries
portable_alpha(α, β)— it ports an uncorrelated stream onto beta (β+α) and scores the combination vs the bogle hurdle with a crisis-correlation check (the 2008 tell: does the alpha stay uncorrelated in the tail?). It reclassifies "diversifier, not a market-beater" nulls that are genuine alpha sources — bore clears the hurdle ported (crisis-corr −0.39) — while correctly refusing hedges with negative carry (BTAL, and bear): hedges don't port, alpha sources do.
| Repo | Focus | State |
|---|---|---|
| blaquebaux | base engine + validated risk-premium spine | live path (paper) |
| blaquebaux-blunt | short-horizon tactical (crude→refiner sleeve) | live driver built — validation PASS |
| blaquebaux-breakthrough | the meta-allocator (capstone) — assemble the keepers into one portfolio | research validated (the capstone KEEPER) — risk-budgeting the family's keepers (spine/trend/tail/gulf/growth/PE) into one book compounds: Sharpe +1.16 (above the best single keeper +1.02 and the Bogle hurdle +0.87), CAGR +11%, maxDD −17% (SPY −34%), Jensen α +4.4% / M² +5.1%, avg keeper pairwise corr +0.37. Not a new premium — the disciplined assembly of what survived; the one sleeve that turns 40 ingredients into a product. Governed live allocator BUILT — allocator emits the target (risk parity over the brigade ingredient set + regime-conditional bastion bear insurance), Julia rail routes it through the Layer-3 safety gate; dry-run PASS (19 names, gross 1.0x). Defaults to breakthrough, with brilliant (min-var) & bossy (leverage) as modes. The sleeve that turns 40 ingredients into one governed product |
| blaquebaux-brilliant | 2nd-place capstone — the optimizer allocator | research validated (2nd of 3 capstones) — max-Sharpe / min-variance over the same keepers clears the Bogle hurdle but can't beat humble risk parity: max-Sharpe overfits to +0.90 (uses return estimates), min-variance ties-not-beats (+1.09, lowest DD −15%) vs breakthrough +1.16. DeMiguel's 1/N is hard to beat, on the family's own keepers. Humility > cleverness · governed live rail built (dry-run PASS) |
| blaquebaux-bossy | 3rd-place capstone — conviction / leverage allocator | research validated (3rd of 3 capstones) — concentration under-performs (Sharpe-wt +0.68, top-2 +0.72 — sheds the diversification that made breakthrough work); only 1.5× levered risk-parity pays: +16% CAGR (beats the index headline) but same Sharpe +1.16 / M² +5.1% at a deeper −25% DD. Leverage buys return, not edge. For the bold, not the best · governed live rail built (dry-run PASS) |
| blaquebaux-believer | benchmark capstone — faith (buy & hold, never rebalance) | research validated (the faith benchmark) — hold the keepers, never touch them: +0.92 Sharpe but drifts into concentration (growth 17%→28%, PE 17%→34%), −25% DD. Disciplined risk parity beats it (+1.15, −17%): rebalancing earns its keep — discipline holds the line where faith drifts. Meant to be beaten, and is · governed live rail built (dry-run PASS) |
| blaquebaux-bemused | benchmark capstone — ignorance (random allocation) | research validated (the null hypothesis) — 5,000 random keeper-weightings: breakthrough sits at the 96th percentile (allocation skill is real but modest), yet 99% of random keeper-portfolios beat SPY — so the edge is the KEEPERS, not the weighting (ingredients: SPY +0.87→+1.05; allocation: +1.05→+1.16). The family's honest rigor check |
| blaquebaux-brigade | expanded capstone — keepers + curated void-fillers | research validated (breadth beats purity) — risk parity over the 6 keepers plus the uncorrelated near-keepers that fill genuine voids (crypto/bitdollar, defensive/buffett, reversal/bounce) lifts the book +1.49→+1.54 (2021–26 window, crypto/BITO history). Lift from the genuinely uncorrelated adds (crypto corr +0.35, reversal +0.14); defensive (+0.93) rides along. Fill the voids that actually diversify · governed live rail built (dry-run PASS) |
| blaquebaux-boundless | kitchen-sink capstone — keepers + ALL near-keepers | research validated (more is not better) — the uncurated set (incl. redundant allweather≈spine, momentum≈growth, latam≈gulf) lands +1.51 < brigade +1.54: risk parity can't diversify a correlated cluster, so the redundant adds water it down. The deepest capstone lesson — ingredient quality & diversity, not quantity. brigade keeps; boundless is the on-record proof curation matters |
| blaquebaux-bastion | insurance capstone — keepers + a standalone-null hedge (bear) | research validated (a null rehabilitated as a portfolio keeper) — bear loses alone (Sharpe −0.34, negative carry) but at −0.82 corr-SPY / −0.63 to the book it lifts the whole portfolio: a 20% overlay takes Sharpe +1.16→+1.26, halves the drawdown (−17%→−11%), and improves Calmar/Sortino/M² monotonically; min-variance wants it (maxDD −15%→−8%). Portfolio value ≠ standalone value — the bleed lesson generalized, the deepest capstone finding · governed live rail built (dry-run PASS) |
| blaquebaux-balsamic | core + trend satellite — a mega-capstone enriched with an uncorrelated managed-futures trend | research validated + governed live rail built (dry-run PASS) — a breakthrough core blended 70/30 by risk with a managed-futures long/short trend satellite: the one outside sleeve that lifts the mega-capstones (the diversification law's other half — a satellite helps iff it is uncorrelated and positive-return; another capstone fails it at +0.92 corr, a tail hedge fails it on negative return). At corr +0.15 the blend beats the core on every axis — Sharpe +1.17→+1.26, maxDD −17%→−13%, skew −0.29→−0.19, and the satellite's +7.6% in 2022 cushions the core's −11.3% to −6.8%. Diversify across uncorrelated ingredients, not correlated allocators |
| blaquebaux-bloated | NVDA mixed into the capstones — an honest null | research done (honest null — deliberately not traded) — bolting NVDA (the market's engine) onto the diversified capstones lifts every Sharpe it touches (ΔSharpe +0.08–0.16) yet fails the outside-sleeve law's spirit: NOT NEW (corr +0.77 to the QQQ growth keeper — the book already owns it), NOT SAFE (single-name blow-up the Sharpe can't price; NVDA maxDD −66%). The counterweight that fixes it — gold + managed-futures trend — is just a capstone rebuilt (bloated_2); inverse ETFs (PSQ/SQQQ/YQQQ) give a perfect −1.0 corr and worthless return (bloated_3); and mixing all of it (bloated_4) resolves on Sharpe, not on wealth — the equal-risk kitchen sink posts the best Sharpe (+1.49) yet makes you poorer than the plain capstone (2.5x vs 2.7x). Sharpe ≠ wealth; add the garnish, keep the capstone, and stop. Joins bear as the family's cautionary sleeve |
| blaquebaux-bull | naive long-growth book (QQQ/VUG/XLK), governed | live driver built — the naive counterpart to broad (no trend/vol-target). Raw book is growth beta, a regime bet not alpha (bogle #3); made ownable by balanced's rate_regime overlay, validated ON — on this unmanaged book de-risking when rates rise cuts maxDD −35%→−30% (13%), lifts Sharpe +0.96→+1.04 / M² +1.0%→+2.3%, improves skew −0.15→−0.05, keeps 90%. The one book in the family that earns the rate overlay on-by-default (broad, managed, declined it — benchmark #4's law from the earning side) |
| blaquebaux-bear | regime-gated short — the mirror of bull | research done (honest null) — perma-short bleeds the risk premium (Sharpe −0.84, −80% DD), so the only honest bear is gated: short only when market_regime is risk-off. That halves the bleed but still loses (Sharpe −0.45, CAGR −7.7%, M² −23.7%, skew +0.31) — a hedge, not a book (positive Jensen α only from −0.83 corr-SPY). Fatal flaw: market_regime coincides, can't lead, so it's short into only 65% of crash onsets while paying carry. Costly, incomplete tail insurance — redundant with bleed. No driver; timing the short side needs a leading signal the family lacks. Joins backsliders/brute-force |
| blaquebaux-boom | mega-cap blue chips (momentum tilt) | live driver built — validation PASS. Two regime overlays wired, both OFF by default.bonds: 0% DD cut / −0.05 Sharpe (the earlier −22% DD was a 2022-window IEX artifact). market_regime (the conditional-keeper test): cuts maxDD 41% (−19%→−11%), nudges Sharpe +1.22→**+1.26**, lifts M², and flips skew −0.31→+0.23 (removes the momentum-crash left tail) — but gives back ~20% of return, failing the retain-80% bar, because BOOM already vol-targets (self-de-risks, cf. broad/bridgewater). Ships opt-in as drawdown/left-tail insurance, not a default keeper |
| blaquebaux-brash | aggressive: crypto, alternatives | research done (growth-vs-ruin lab; fractional-Kelly rule) + live driver built — ETF-proxy gate MIXED; thesis PASSES on the real crypto rail (+0.72 aggressive), crypto execution now wired — trades real BTC/ETH (aggressive, governed) |
| blaquebaux-bleed | contrarian; positioned for the tails | research done (regime-spanning tail basket) + live driver built — validation PASS (as insurance: +79% vs SPY -82% on crash days) |
| blaquebaux-brace | vol-risk-premium harvest — bleed's counterparty | research done (honest null) — selling vol is the insurer's side (negative skew, mirror of bleed): put-write/covered-call lag SPY (skew −1.1 to −1.8), DIVO ~matches, SVXY −95% landmine. Gating on market_regime cuts maxDD −29%→−13% but only by surrendering return (M² −2.3%→−2.6%, no risk-adjusted gain). Not a portable-alpha source (corr +0.85, crisis-corr +0.83). No driver; joins bear/backsliders/brute-force. (ETF-risk: PUTW delisted ~Apr-2025) |
| blaquebaux-bottom | sub-small-cap / penny names | research done (cap-ladder bounce rejected; a large-cap play) + live driver built — validation PASS |
| blaquebaux-brittle | near-expiry far-OTM options/ETFs | research done (short-vol premium is a trap; naked rejected) |
| blaquebaux-broad | broad-market & thematic ETFs (IVES, GRNY, QQQ, TQQQ) | research done (leverage law; managed-exposure keeper) + live driver built — validation PASS; consumes blaquebaux-bonds' regime signal (overlay ON, marginal: full-cycle Sharpe +0.94→+0.96, 95% of return kept). The only equity consumer where the overlay survives a full-cycle re-test — it already self-de-risks via trend + vol-target. Declined benchmark's market_regime (redundant: gating cuts Sharpe +0.84→+0.80, its vol-target already does the vol-timing). Consumes balanced's rate_regime opt-in (rising rates = growth headwind): lifts Sharpe +0.96→+1.03 / M² +0.8%→+2.0% but 0% DD cut → fails the DD bar, ships opt-in (self-manages, same as market_regime) |
| blaquebaux-bore | market-neutral, indifferent to bull/bear | research done (beta-hedged keeper) + live driver built — validation PASS. Declined benchmark's market_regime overlay (measured beta −0.07; gating a neutral book hurts, Sharpe +0.42→+0.31) — wrong signal for a market-neutral book. Reclassified as the family's portable-alpha source: neutral (not a market-beater) standalone, but ported β+α onto SPY it CLEARS the Bogle hurdle — Sharpe +0.83→**+0.97**, Jensen α +4.9%, M² +2.4%, and crisis-corr −0.39 (helps more in crashes — the anti-2008 profile). DBMF ports too; BTAL (a hedge) doesn't. Governed β+α driver built (bore_portable_live.jl: full S&P beta via IVV/VOO + bore's alpha, ~2.2× gross, own paper leg) — the first sleeve to turn a neutral null into a hurdle-clearing live book |
| blaquebaux-bridge | statistical arbitrage / pairs | research done (honest null — the mirror of bore) — a rules-based pairs book is the cleanest neutral, tail-safe stream in the family (beta-SPY +0.01, crisis-corr +0.09, skew +0.96) — the ideal portable-alpha vehicle — but earns ~0 after cost (Sharpe +0.15, CAGR ~0%): the classic mean-reversion edge has been arbitraged away since the 2000s. Jensen α +0.1% / M² −0.0% → nothing to port. Profile is not edge. A few pairs live (GLD/GDX +0.45) but the basket washes out. No driver; joins bear/brace/backsliders/brute-force |
| blaquebaux-bounty | cross-asset carry (getting paid to hold) | research done (honest null — the brace pattern) — ETF-harvestable carry (credit HYG−IEF, term IEF−BIL, EM EMB−IEF) is a crash-risk premium, not clean alpha: diversified basket Sharpe +0.39 (vs SPY +0.88), corr-SPY +0.80, skew −2.02, and it fails the port (Jensen −1.7%, crisis-corr +0.84 — crashes with equities). Term carry has the right profile (corr −0.12) but no return in the rising-rate regime. FX carry (DBV) delisted ~2023; the pure futures carry factor is out of reach via cash ETFs (data gap). No driver; joins brace/bridge |
| blaquebaux-bite | event-driven / merger arbitrage | research done (honest null — the last uncorrelated-alpha hope, diluted) — merger arb MNA Sharpe +0.47 but excess-over-cash only +0.7%/yr, skew −2.36, crisis-corr +0.70 (deals break when markets fall) → not the uncorrelated alpha it looked like; ported onto SPY it fails (Jensen +0.1%, M² −0.4%). Spin-offs CSD are just small-cap beta (β1.15, −59% DD). Real edge (deal selection/leverage/access) unreplicable via ETF — confirms bind. No driver; joins brace/bridge/bounty |
| blaquebaux-bulk | defense / military & adjacent | research done (moderate factor; systematic null) |
| blaquebaux-brown | conservative-leaning sectors (energy, mining, ag, firearms, prisons) | research done (Brown/Blue rotation keeper) + live driver built — validation PASS |
| blaquebaux-blue | entertainment/film, green energy, tech | research done (Brown/Blue rotation keeper) + live driver built — validation PASS |
| blaquebaux-beyond | short-horizon growth (CAGR over weeks, not years) | research done (growth-momentum keeper) + live driver built — validation MIXED (stays dry-run) |
| blaquebaux-bubble | the AI complex viewed as one | research done (crowded factor; bubble not fadeable) |
| blaquebaux-basel | Basel-regulated banks (one regulated factor) | research done (one-factor; macro sleeve) |
| blaquebaux-bio | biotech; idiosyncratic FDA events (the anti-Basel) | research done (systematic null) |
| blaquebaux-bounce | range-bound "kangaroo" market (mean-reversion) | research done (gated reversal keeper) + live driver built — validation MIXED (stays dry-run) |
| blaquebaux-emea | Europe, the Middle East & Africa | research done (null — US beta wearing a flag, 11 ETFs → 1.8 bets; FX drag; no rotation edge). Dollar overlay tested: only trades return for drawdown (Sharpe flat, 12% DD cut) — does not rescue the null |
| blaquebaux-apac | Asia-Pacific | research done (US beta + severe FX drag, Japan −229%; the one region with a rotation pulse, long-short +0.30) + live driver built — validation PASS. Dollar overlay declined: the live L/S book is dollar-neutral (measured beta −0.01 to UUP) — wrong signal, correctly not consumed |
| blaquebaux-latam | Latin America | research done (null — US/commodity beta, worst tail −55%, unhedgeable in-wrapper FX, rotation hurts). Dollar overlay tested: meaningfully de-risks (DD −55%→−40%, 28% cut) at ~flat Sharpe — a better ingredient, still below SPY, not a keeper |
| blaquebaux-bitdollar | crypto / dollar-crypto axis | research done (trend+vol-target keeper; dollar axis rejected) + live driver built — ETF-proxy gate MIXED; thesis PASSES on the real BTC/ETH rail (+0.72), crypto execution now wired — trades real BTC/ETH (fractional, governed) |
| blaquebaux-blurred | deliberately uncorrelated names, traded as one | research done (null — uncorrelated equities are a +0.17 floor & unstable; diversify across asset classes) |
| blaquebaux-backsliders | broken decliners, 25%+ off high, no path back (short) | research done (short-the-fallen null; the long bounce is the edge) |
| blaquebaux-brute-force | names propped up by options/squeeze/flow | research done (fade rejected; needs positioning data) |
| blaquebaux-block | a basket of derivative strategies | research done (the 4 blocks interlock but stay ~4.6/8 diversified; linkages real yet regime-dependent & priced-in — a risk map, diversification is the edge) |
| blaquebaux-burry | Michael Burry's book, re-examined (the mechanism, not the man) | research done (cautionary null — the style is mostly ~1.0-beta equity (+0.6 corr-to-momentum), buying the hated (3y reversal) nulls out (L/S Sharpe −0.00), shorting froth is regime/ruin (−98% DD, +86% only in 2022), concentration is a ruin machine; residue (gold, Bleed) already in the family. Joins brute-force/backsliders) |
| blaquebaux-buffett | Buffett's strategy in fragments (quality/value/safety/leverage) | research done + live driver built — validation MIXED (stays dry-run). On the full 2016–2026 SIP history the cheap-safe-quality QUAL/USMV/VLUE/MOAT blend is not genuinely defensive (DD −35% ≈ SPY −34%, Sharpe +0.82 < +0.89) — marginally lower-vol but market-like, so it does not clear the defensive bar. Bonds overlay OFF (adds nothing). Defensive beta, not alpha; the real moat (cheap float) is un-buyable |
| blaquebaux-beltway | Democratic-era darlings (Biden/Obama/Clinton) — do they hold up? | research done (emphatic null — inverted; 2016+ only, Clinton/Obama out of sample) — the party-aligned basket loses under its own party 4/4; DEM darlings −14.8%/yr under D vs +39.8%/yr under R; under Biden clean energy −21%/yr while fossil +26%/yr — macro dominates, and darlings revert (rank persistence −0.95). Joins bubble/burry) |
| blaquebaux-brics | the tradable core of BRICS (best emerging growth engines) | research done + live driver built — the keeper (Gulf KSA/UAE/QAT as a low-corr EM diversifier); validation PASS as a diversifier (corr-SPY 0.56 vs EEM 0.74, vol 15% vs 20.7%, standalone Sharpe +0.46, full-SIP 2016-2026). 'Best of BRICS' (momentum) & the full basket (EM beta) rejected; Russia excluded on the data. A diversifier, not a market-beater. Dollar-regime overlay wired, validated & ON (its right macro signal, not the bonds one): de-risk ×0.5 when the dollar trends up → Sharpe +0.48→**+0.51**, maxDD −35%→−20%; brics also publishesdollar_regime.txt for the family. Paper/dry-run |
| blaquebaux-bonds | the bond–equity relationship (macro overlay for sizing/hedging) | research done + live driver built — regime overlay + regime-signal emitter; validation PASS as an overlay (full-cycle: cuts equity DD 41% −34%→−20%, keeps 61% of return, 11% vs 18.7% vol; timing ~static 60/40 as research found — the one overlay that strengthens on the full cycle). The corr-regime read (72% persistent a quarter out; hedge works only in neg-corr) is published for the family to size against. Paper/dry-run — not a live-money endorsement |
| blaquebaux-basket | exchange / swap funds — hidden private-wealth vehicles | analytical study done (4 sketches, grounded in Fidelity / Morgan Stanley GIC / Concentrated-Position-Toolkit docs). All Reg D private placements (democratized $1M/QP → $100k/accredited via Cache, still private); two structures — 721 (≥20% leveraged real estate → rate risk, K-1s, ~80% index) vs 351 (diversified-contribution, cleaner). No alpha (fund = index − fee); the product is the capital-gains deferral, conditional (+5–22% after-tax hold-to-death via the step-up, break-even fee ~1.8%). Its niche vs the toolkit (sell+TLH / completion / collar / CRT): a very-low-basis position you can't afford to sell or dilute and intend to hold into your estate. A tax/estate vehicle, not a strategy |
| blaquebaux-blank | SPACs / blank-check shells (trust carry, deSPAC shorts, busts) | research done (diagnostic null — de-SPAC decay real (−9pp/yr, 60% below the $10 trust) but untradeable: naive short −26%/yr on −89% DD as ASTS +643%/RKLB +429% run it over; broken-subset short buried by 20-100% borrow; sound trust carry needs SPAC-level data — parked. Joins bubble/brute-force) |
| blaquebaux-benchmark | market internals as a risk-regime read (VIX, credit, breadth, Dow Theory, defensives) | research done + regime emitter built — internals coincide, they don't lead (peak cross-corr k=0), but the composite gates SPY net of cost (Sharpe +0.87→**+1.11**, DD −34%→−16%). Publishes market_regime.txt (family's 3rd regime signal) labeled honestly as vol-timing, not breadth alpha (strip vol and breadth/credit/DT barely beat B&H). Internals confirm, don't forecast |
| blaquebaux-bridgewater | Bridgewater's All-Weather & Pure Alpha | research done — All-Weather is a real low-vol risk-parity compounder (+0.95 Sharpe, ~⅓ market vol) but the same book as the spine (corr 0.96), and the spine beats it (+0.97 vs +0.84) by omitting long-bond TLT — an external validation of the spine. Pure Alpha = documented gap. Live driver built: the recognizable static All-Seasons allocation as a distinct low-vol/−23%-DD diversification sleeve — but its bonds_regime overlay, the right signal for its 2022 failure, fails the bar (cuts DD 20% yet costs Sharpe +0.77→+0.74 and 22% of return) because the book already self-diversifies → overlay OFF. The mirror image of blackstone: right signal, but a book that doesn't need it |
| blaquebaux-bind | the rest of the hedge-fund landscape (Millennium, Citadel, pod shops) | research done (diagnostic null) — the liquid replicators are watered-down positive-beta (equal-weight book +3.5%/yr, not the pods' 10-15%); the only genuinely uncorrelated slices (DBMF managed-futures, BTAL anti-beta crisis hedge) = the family's trend + Bleed already. Real pod-shop edge (leverage × PMs × execution × access) is private/unreplicable. Nothing to add |
| blaquebaux-blackstone | listed private-equity managers (BX/KKR/Apollo/Carlyle/Ares) | research done — the "buy, prop, exit" model is a levered pro-cyclical market/credit factor, not diversification: beta-SPY 1.39 / beta-XLF 0.99, 1.8/6 eff-bets, beta-credit +1.55, up/down beta 1.34/1.41 (falls harder), coupled to private-credit BDCs. Huge in the easy-money bull (+784% vs SPY +337%) but 31% vol / −45→−57% DD → Sharpe +0.83 < SPY +0.88 (extra return is leverage, not skill), the opposite of a hedge — so the live driver ships it governed: equal-weight PE book + benchmark's market_regime overlay ON (validation PASS). The first consumer book to earn that overlay — on this uniquely unmanaged high-beta book it lifts Sharpe +0.76→+0.95 (above SPY), holds CAGR flat, cuts DD −45%→−31% for free |
| blaquebaux-benefactors | foreign creditors — US Treasury holdings by country vs their economies/markets | research done (honest macro null) — FRED data (Japan 1984, China 2003) supplied by hand (box can't reach FRED). A creditor's Treasury buying has only a weak ~+0.15 coincident link to its own market, swamped by global (SPY) beta, no clean lead — for China the market leads the flow (+0.18), so causality runs backward. Creditor status is a macro read, not a signal. Market test ETF-capped to 2016-2026 |
| blaquebaux-buyouts | post-buyout targets — pump-and-dump on the exit? | [Concept] — the PE targets (cf. blackstone = the managers). Private "prop it up" leg is invisible (no ticker); testable surface = sponsor-backed IPOs/re-IPOs (IPO/FPX) + post-lockup drift → a short/avoid thesis on the sponsor exit. Fat-tail toolkit (JB + Jensen's α + M²) |
| blaquebaux-bigbrother | America's new SOEs — US government equity stakes (Intel, CHIPS, bailouts) | [Concept] — does a government backstop prop the stock (implicit put) or drag it like a China SOE (state ownership destroys value)? Two opposed hypotheses; the tell is the tail. Rare hand-curated stake events + an industrial-policy/defense basket vs SPY |
| blaquebaux-battered | distressed assets that survive to pay out (the MF Global playbook) | [Concept] — fallen angels (ANGL), high-yield; the right-thesis-killed-by-leverage lesson. A prime conditional-keeper case — the fat left tail is the true cost, so verdicts lean on JB + max-drawdown. OTC distressed invisible; ANGL/HYG proxies, survivorship flagged |
| blaquebaux-bets | the gambling complex (sports betting, casinos, prediction markets) | [Concept] — durable "house always wins" premium, or levered consumer discretionary? Kalshi/Polymarket private (gap); listed books/casinos + BETZ testable vs SPY/XLY. Fat-tail toolkit |
| blaquebaux-balanced | value investing / margin of safety — still crucial? | research done — value is a rate-regime bet, and "margin of safety" is mostly a myth at the factor level. Value lagged growth risk-adjusted (RPV α −2.3% / M² −4.9% vs RPG −3.4%; spread −1.9%/yr) — a different beta, not alpha (cf. bogle). It crashed harder than growth (pure value maxDD −51% vs −37%, worse skew) — the safety folklore inverted; its only real safety is lower downside vol + cushioning growth-specific selloffs. Its one strong, clean signal is rate-conditional (value−growth +12.3%/yr rising rates, −12.3%/yr falling), but a coarse rate-timed rotation (α +1.8%) doesn't clear the SPY hurdle net of cost (M² −0.5%). A conditional-keeper ingredient, not a standalone edge — publishes rate_regime.txt (the family's 4th regime signal) for value-sensitive consumers |
| blaquebaux-bogle | Vanguard / low-cost indexing | research done (meta-sleeve — the hurdle) — the VOO line Sharpe +0.89 / Jensen-α 0 / M² 0 is the bar the family is measured against. Broader diversification cost return this US-led decade (VTI α −0.7%, VT α −1.6%); 0/7 Vanguard tilts beat plain VOO risk-adjusted (tilting = different beta, not alpha); only 2/10 popular active/thematic funds cleared it (QQQ/XLK — pure tech-beta, a regime bet), and ARKK hid a −81% DD behind +13% CAGR. Bogle stands; the toolkit makes it precise. Publishes bogle_hurdle.txt (the plain-VOO bar) + a reusable scorer — the family yardstick. (Also corrected M² to the Sharpe-difference form → benchmark-vs-itself = 0.) |
| blaquebaux-bureaucrats | congressional trading under the STOCK Act (Pelosi et al.) | [Concept] — do lawmakers beat the market, or is it a mega-cap-tech tilt? NANC/KRUZ tracker ETFs priceable from 2023; granular STOCK Act data hand-supplied (egress wall). Distinct from beltway (party-era winners) — this is the legislators' personal book |
| blaquebaux-belt-tighteners | the inflation domino across sectors, globally | [Concept] — belt-tightening cascade XLY→XLP→XLB→XLU→XLRE→XLE, then contagion when US + China sneeze together. A correlation-tightening study; inflation via breakevens or hand-loaded CPI; XLC only from 2018, global capped ~2016 |
Cross-family paper A/B is monitored by scripts/family_summary.py (each leg's keys live in
~/.config/blaquebaux/, so it snapshots whatever sleeves are active).
The family grew leaf-first (funds, sectors, themes); mapping the tree surfaced the trunk primitives it skipped — fundamental style/asset branches, several of them prime portable-alpha sources. Names reserved, not yet built:
| # | name | branch | notes |
|---|---|---|---|
| 1 | volatility / vol-risk-premium | ✅ built — honest null. Harvesting the VRP (bleed's counterparty) is the insurer's side (negative skew); via listed ETFs it lags the index gated or not, SVXY is a −95% landmine, and it's not a portable-alpha source. On the shelf | |
| 2 | carry (cross-asset) | ✅ built — honest null. ETF-harvestable carry (credit/term/EM spread) is a crash-risk premium, not clean alpha: +0.39 Sharpe, corr-SPY +0.80, skew −2.02, crisis-corr +0.84. FX carry (DBV) delisted; the pure futures carry factor is out of reach. On the shelf | |
| 3 | event-driven / merger-arb | ✅ built — honest null. Merger arb (MNA) is watered-down (excess-over-cash +0.7%/yr, skew −2.36, crisis-corr +0.70 — deals break when markets fall); spin-offs (CSD) are just small-cap beta. The real edge is unreplicable via ETF. On the shelf | |
| 4 | statistical arbitrage / pairs | ✅ built — honest null. The ideal portable-alpha vehicle (beta +0.01, crisis-corr +0.09) but no alpha to carry — the classic pairs edge has decayed (book Sharpe +0.15 after cost). Profile is not edge. On the shelf | |
| 5 | blaquebaux-bourse | rates / FX / commodities, directly | the curve, FX pairs, gold/oil/ags — the asset classes we only touch via ETFs |
| 6 | blaquebaux-bloom | seasonality / calendar anomalies | turn-of-month, sell-in-May, tax-loss January |
| 7 | blaquebaux-buzz | alt-data / sentiment | NLP news/social, earnings-call tone (egress-walled, à la bureaucrats) |
| 8 | the meta-allocator (capstone) | ✅ built — the capstone KEEPER (first trunk keeper). Risk-budgeting the keepers compounds: Sharpe +1.16 (best single +1.02, SPY +0.87), half the market's drawdown, M² +5.1%, avg keeper corr +0.37. Not a new premium — the disciplined assembly of what survived. Governed regime-conditional allocator = the graduation |
Note brace / bounty / bridge / bite are exactly the market-neutral, positive-Sharpe, doesn't-beat-the-index-standalone
streams the new portable_alpha toolkit is built to port — the bore pattern, repeatable. And several of
these are implemented through derivatives, so block's catalog is
where brace (options) and bounty (futures roll) actually get built.
Some sleeves feed others rather than trade alone. Four published regime signals now exist (plus bogle's benchmark hurdle), each matched to the sleeves it actually helps — the discipline is match the signal to the sleeve:
- bonds →
bonds_regime.txt(US stock-bond correlation) — for net-long US-equity sleeves. - brics →
dollar_regime.txt(US dollar trend, UUP vs 100d MA) — for USD/international sleeves (a rising dollar is an EM headwind). - balanced →
rate_regime.txt(US rate direction, IEF vs 100d MA) — for value/growth-sensitive sleeves (value beats growth +12%/yr when rates rise, −12% when they fall; the regime is clean, the coarse rotation itself doesn't beat the index — an ingredient). Consumers tested: on managedbroad (QQQ) it lifts Sharpe +0.96→+1.03 / M² +0.8%→+2.0% but cuts 0% off drawdown → opt-in (broad self-manages). On the naive growth book bull (QQQ/VUG/XLK, no trend/vol-target) the same signal earns ON-by-default — DD −35%→−30% (13% cut), Sharpe +0.96→+1.04, keeps 90% — because the naive book has the drawdown gap the managed one already filled. The law holds: a de-risking overlay's value ∝ how unmanaged the book is (and whether its worst DD is rate-driven). bull is the sleeve built to earn it. - bogle →
bogle_hurdle.txt(the plain-VOO bar) + a reusable scorer — not a regime but the yardstick every sleeve is measured against (clear it net of cost/tax: Jensen α > 0 AND M² > 0 vs VOO, or it isn't a keeper). - benchmark →
market_regime.txt(market-internals risk-on/off composite) — a broad de-risking flag, honestly labeled vol-timing, not breadth alpha. Its value shrinks monotonically with how much a book already manages its own risk (benchmark #4): earned by naive high-beta books with no risk control — benchmark's own gated-SPY (SPY +0.87→+1.11) and now blackstone, the first consumer sleeve to earn it (levered listed-PE, Sharpe +0.76→+0.95 / DD −45%→−31% / CAGR flat) — and declined by the managed/neutral sleeves —broad(redundant with its trend+vol-target, +0.84→+0.80),bore(market-neutral → wrong signal), andboom(the conditional-keeper test: cuts DD 41% and flips skew −0.31→+0.23, i.e. removes the momentum-crash left tail, but gives back ~20% return → fails the retain-80% bar since it already vol-targets; ships opt-in insurance, not default). So it's earned where nothing else manages the vol, and refused (or made opt-in) everywhere that already does.
The bonds (stock-bond) overlay, consumed to de-risk gross when the bond hedge is dead. Each consumer is validated independently on the full 2016–2026 SIP history — it ships on only where it earns it:
| consumer | overlay effect (full cycle) | default |
|---|---|---|
| bonds (own overlay vs static 60/40) | −41% drawdown (−34%→−20%), timing ≈ static | ON |
| broad | +0.02 Sharpe, ~flat DD — marginal but consistent | ON |
| bridgewater (All-Weather) | −20% DD (−23%→−19%) but −0.03 Sharpe & keeps only 78% return | OFF |
| boom | 0% DD cut, −0.05 Sharpe | OFF |
| buffett | no DD help, −0.04 Sharpe | OFF |
| keeper book | no change (−6% DD either way) | OFF |
Honest finding (after a full-cycle re-test): the equity-sleeve de-risk overlay is not the broad win it first looked like. An early pass on the engine's default IEX feed (only ~2021+) showed boom gaining a −22% drawdown cut — but that was a 2022-window artifact: boom's worst full-cycle drawdown is the 2020 COVID crash, a negative-correlation episode the overlay correctly leaves alone. On the full cycle the equity overlay barely helps anywhere; only broad keeps it (marginally). The robust, strengthening benefit is bonds' own overlay — the direct hedge-vs-cash switch — which cuts drawdown 41% across the 2020 and 2022 crashes.
The dollar (UUP-trend) overlay is the counter-example that proves the discipline. Pointed at brics (a USD-EM book, −0.66 beta to the dollar), de-risking the Gulf ×0.5 when the dollar trends up improves Sharpe (+0.48→+0.51) and cuts drawdown 43% (−35%→−20%) on the full cycle — a clean win, because it's the right signal for that book. The same book correctly rejects the bonds overlay. The other USD/international sleeves were tested against it too, and the honest results differ by book:
| consumer | dollar overlay effect (full cycle) | outcome |
|---|---|---|
| brics (Gulf) | Sharpe +0.48→**+0.51**, DD −35%→−20% | ON — earns it |
| latam | ~flat Sharpe, DD −55%→−40% (28% cut) | de-risks; a better ingredient, still a null |
| emea | flat Sharpe, DD −40%→−35% (12% cut) | trades return for DD; still a null |
| apac | live L/S book is dollar-neutral (beta −0.01) | declined — wrong signal for a neutral book |
Same lesson as the bonds side: the dollar is the right lever to try for a USD-exposed book, but only brics clears the bar; the rest are de-risked at best, and apac's neutral live book declines it outright.
The lesson: match the signal to the sleeve, validate each pairing on the full cycle, ship on only
what earns it — and beware short-window results. Every consumer keeps a graceful fallback
(missing/stale signal → full gross) and a BB_BONDS_OVERLAY / BB_DOLLAR_OVERLAY toggle.
A deeper law, revealed by the fat-tail toolkit (Jarque-Bera + skew + Jensen's α + M²). Returns are
non-normal (JB rejects normality everywhere), so an overlay's real job is the left tail, which Sharpe
can't see. Re-examining the borderline declines under the toolkit shows an overlay only buys genuine tail
insurance — flips return skew from negative to positive — when it de-risks on a crash-timing signal:
market_regime (vol-timing) did exactly that for boom (skew −0.31→**+0.23**, DD −41%), making it an
opt-in insurance overlay. Overlays keyed on a persistent state — the stock-bond correlation (bridgewater,
~33% of days) or the dollar trend (latam, ~63% of days) — de-risk indiscriminately, so they shed return
and worsen skew (bridgewater −0.32→−0.73; latam −0.94→−1.16) instead of catching the tail. And "defensive"
is itself a tail claim: buffett's blend has a worse skew and negative Jensen's α / M² vs SPY — not
defensive at all. So the toolkit doesn't rescue borderline nulls; it discriminates by mechanism —
crash-timed overlays can insure, state-timed ones only shave — and sharpens every honest rejection.
You're welcome to study, fork, and build on this. If you deploy real capital, validate independently and start on paper. Issues and PRs that improve the math, the tests, or the execution safety are especially welcome.
Explore the production site, interactive LABS, and open research CORPUS.
MIT, plus a not-financial-advice notice. © 2026 Carter Warrens.