Bridgewater's two engines — All-Weather (replicable) and Pure Alpha (a data gap) — examined honestly.
Bridgewater is a member of the Blaque Baux family. The core repo is the engine and blueprint — a governed, systematic platform (Julia) with a venue-agnostic execution controller and a Layer-3 live-money safety gate. Bridgewater points that engine at the world's largest hedge fund's two signature strategies and inherits the governance wholesale.
Not investment advice. Educational/research software. Nothing here is validated. See LICENSE.
git clone --recursive https://github.com/blaquebaux/bridgewater.git
julia --project=engine -e 'using Pkg; Pkg.instantiate()' # one-time engine setupBridgewater runs two very different engines, and they sit on opposite sides of what a price feed can see:
- All-Weather is a public, replicable idea: balance a portfolio by risk contribution across the four macro environments (growth up/down × inflation up/down) using stocks, bonds, gold, and commodities, so no single environment dominates. It is essentially risk parity — and, notably, it is close to what the Blaque Baux spine already does (inverse-vol / equal-risk-contribution across SPY/IEF/GLD/DBC/DBA). That makes it fully testable, and the honest question is what it is: a low-vol diversified compounder, its real failure mode (2022, when stocks and bonds fell together), and whether it beats the family's own spine.
- Pure Alpha is a discretionary global-macro fund — leveraged, uncorrelated-to-markets, and entirely private: no public NAV, no ticker, no replicable rulebook. It cannot be reconstructed from price bars. We say so plainly rather than fake it (the same honesty as basket's private funds).
- Reconstruct All-Weather. Risk-parity across SPY/IEF/TLT/GLD/DBC/DBA (+
RPAR, the listed Risk Parity ETF, as a direct proxy). Characterize it: Sharpe, drawdown, and the stock-bond-correlation regime that breaks it (2022) — feeding straight into bonds' finding that the diversification is regime-conditional. - All-Weather vs the spine. Does Bridgewater's flagship beat the family's own risk-parity spine, or is it the same idea? Honest either way — if it ties, that validates the spine.
- Pure Alpha — the gap, stated. Document why discretionary macro is unobservable from prices, and what (if anything) a liquid proxy could stand in for — likely nothing faithful.
Full detail in research/README.md. The scorecard (Alpaca SIP, 2016–2026):
| # | Question | Verdict |
|---|---|---|
| 1 | What does All-Weather deliver? | ✅ a real low-vol risk-parity compounder — inverse-vol +0.95 Sharpe, ~⅓ the market's vol / half the DD, but +7% vs SPY +15% return; RPAR (the listed ETF) is a poor leveraged implementation (+0.35, −30% in 2022) |
| 2 | Any different from the family's spine? | ✅ no — and the spine beats it: correlation +0.96; spine +0.97 vs All-Weather +0.84, because the spine omits the long-bond TLT sleeve |
| 3 | The failure mode? |
The synthesis: Bridgewater's flagship is a great idea the family already runs — and runs better. All-Weather is a genuine low-vol risk-parity compounder (higher Sharpe than the market at a third of the vol), but it's the same book as the Blaque Baux spine (0.96 correlated), and the spine posts a higher Sharpe (+0.97 vs +0.84) by holding intermediate rather than long bonds — the exact choice that spared it the 2022 rate shock. There's no secret sauce a disciplined in-house inverse-vol book lacks (and the listed RPAR ETF proves the concept is only as good as the execution). All-Weather's one weakness is the positive stock-bond-correlation regime that breaks the "bonds diversify stocks" premise — the regime bonds found is partly knowable a quarter ahead, so it's best paired with the bonds regime read. Pure Alpha stays a documented gap (discretionary, private, unobservable). The through-line: diversification is regime-conditional.
The static book is a legitimate, distinct sleeve even though the spine beats it: the recognizable public
Dalio All-Seasons allocation — 30% SPY / 15% IEF / 40% TLT / 7.5% GLD / 7.5% DBC — run on the
same engine + Layer-3 safety gate as the spine. It is the family's no-forecast, maximally-diversified,
low-vol beta sleeve: +0.77 Sharpe at 8.9% vol and −23% maxDD (vs SPY's 18% / −34%). live/bridgewater_live.jl.
The interesting part is what it teaches about wiring. All-Weather's one known failure is the positive
stock-bond regime (2022) — and the sibling bonds sleeve publishes
exactly that (63d SPY-IEF correlation). So this is the right signal for the book's actual weakness.
Yet live/bridgewater_bonds_regime_validation.jl FAILS the
family bar, and the overlay ships OFF:
| All-Seasons book (full 2016–2026 SIP, net 5bps) | Sharpe | CAGR | vol | maxDD |
|---|---|---|---|---|
| FULL static (shipped) | +0.77 | 6.7% | 8.9% | −23% |
| + bonds-regime overlay | +0.74 | 5.2% | 7.3% | −19% |
| SPY (reference) | +0.88 | 15.3% | 18.0% | −34% |
The overlay cuts drawdown (−23% → −19%, a 20% cut) but costs Sharpe (+0.77 → +0.74) and ~22% of
return — so it fails "not worse on Sharpe" and "retains ≥80% of return." Why, when it's the right
signal? The 63d correlation flags "hedge dead" on ~33% of days, while All-Weather only truly breaks
in the acute 2022-type episodes; de-risking the other pos-corr stretches costs more return than it saves.
This is benchmark #4's monotonic law from the other side:
All-Weather already self-manages risk (risk-parity across five asset classes, only −23% maxDD), so a
blunt de-risking overlay can't earn its keep — the mirror image of blackstone,
a naive high-beta book with no risk control, where the market_regime overlay earned a full Sharpe point.
Right signal, wrong book. Opt in with BB_BONDS_OVERLAY=1 if you specifically want the ~20% drawdown cut.
BB_DRYRUN=1 bash live/run_bridgewater_daily.sh # dry-run: logs the target, places nothingDry-run by default; graduates to paper once ~/.config/blaquebaux/alpaca_bridgewater.env exists. Real
money additionally requires BB_LIVE_CONFIRM. Kill switch: ~/.config/blaquebaux/HALT.
Live driver built — a governed All-Weather book; the regime overlay tested and declined. All-Weather validates the family's own spine (corr 0.96) and is mildly beaten by it (less duration). The driver ships the recognizable static All-Seasons allocation as a distinct low-vol/low-drawdown diversification sleeve, but the bonds-regime overlay — the right signal for its 2022 weakness — fails the family bar (cuts DD 20% but costs Sharpe and return) because the book already self-diversifies, so it ships OFF (opt-in). Pure Alpha stays a documented data gap. Ships dry-run/paper; not yet run as real money.
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.
This repo is one sleeve of the Blaque Baux family — a single governed engine steered in many directions. The core repo is the base/blueprint and holds the full family roster.
engine/ the Blaque Baux platform (git submodule -> blaquebaux/base)
research/ three sketches (reconstruct All-Weather, vs the spine, 2022 failure + Pure Alpha gap) + scorecard
live/ bridgewater_live.jl (governed All-Weather book) + bonds-regime validation (declined) + wrapper/plist
MIT. (c) 2026 Carter Warrens.