Skip to content

Repository files navigation

Blaque Baux Benchmark

LIVE PATH · REGIME EMITTER  |  Python / Julia  |  Interactive study  |  Research corpus  |  Citation

Market internals as a risk-regime read — do the confirming/diverging signals predict, or just coincide?

Benchmark 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. Benchmark points that engine at the market's own internals and inherits the governance wholesale.

Not investment advice. Educational/research software. Nothing here is validated. See LICENSE.

git clone --recursive https://github.com/blaquebaux/benchmark.git
julia --project=engine -e 'using Pkg; Pkg.instantiate()'   # one-time engine setup

The thesis

Traders watch a wall of "internals" to judge whether a market move is healthy or hollow: VIX (fear), the advance-decline line / $TICK (breadth), HYG vs LQD (credit risk appetite), the MOVE index (bond vol), SPY vs RSP (cap- vs equal-weight participation), Dow Theory (DJIA confirmed by DJT, with DJU as the defensive tell), and the put/call ratio $CPC (sentiment). The folk belief is that divergences here warn — breadth rolling over, transports failing to confirm, credit widening — before price does. Benchmark asks the honest question: do these internals actually lead the market (a tradable/sizing regime), or do they merely coincide with it (a dashboard)? If they lead, benchmark becomes the family's third regime publisher (after bonds' stock-bond regime and brics' dollar regime); if not, it's an honest diagnostic.

Data note: several of these are index/breadth tickers a daily-bars vendor doesn't serve, so they're proxied (or flagged) with tradable instruments:

requested proxy
VIX VIXY VIX-futures ETF
HYG vs LQD (credit) HYG / LQD direct
$ADD / $TICK (breadth) RSP / SPY equal- vs cap-weight ($ADD & intraday $TICK not on daily bars)
DJIA vs DJT (Dow Theory) DIA vs IYT industrials vs transports
DJU (defensive) XLU utilities
MOVE index (bond vol) TLT realized vol the MOVE index isn't on Alpaca
put/call $CPC CBOE sentiment index not available (gap)

Research — first pass done

Full detail in research/README.md. The scorecard (Alpaca SIP, 2016–2026):

# Question Verdict
1 Are the internals a coherent dashboard? ✅ correctly signed (VIXY corr −0.77, credit +0.47, defensives −0.43; risk-off internals spike on the worst SPY days)
2 Do they lead SPY or coincide? coincident — fwd-20d spreads tiny/wrong-signed; peak cross-corr at k=0 (they mirror the tape)
3 Is the composite a tradable regime? ⚠️ passes net of cost (gated SPY +0.87→+1.11 Sharpe, DD −34%→−16%) — but it's vol-timing: non-vol internals barely beat B&H (+0.89)
4 Where does it earn its keep? ✅ on a naive long book (SPY B&H +0.87→+1.11) — value shrinks monotonically with the book's own risk management; redundant on managed sleeves (broad +0.84→+0.80)

The synthesis: benchmark is a coherent risk dashboard that resolves into a genuine risk-off regime — which, honestly, is vol-timing, not the breadth/credit edge it looks like. The internals coincide with the market (they confirm, they don't forecast); the composite still gates profitably because risk-off regimes persist (high-vol clusters), and that benefit comes almost entirely from the VIX / bond-vol components — strip them and breadth/credit/Dow-Theory barely clear buy-and-hold. So it's managed-vol wearing an internals label, overlapping broad's vol-target and bonds' crash behavior. Publishable as a family risk-off regime (market_regime.txt) with that caveat — a risk map to READ and size with, not novel alpha. Same honest lesson as Bubble: diagnostics flag, they don't time.

Live driver — regime emitter (paper/dry-run)

The composite is now a governed driver on the engine (live/benchmark_live.jl). Each run it:

  1. Publishes the market regime to ~/.config/blaquebaux/market_regime.txt (composite score + a risk_on flag) — the family's third regime signal after bonds' stock-bond and brics' dollar. The file itself carries the honest label: mostly vol-timing, not breadth forecasting.
  2. Trades the regime's own expression — hold 80% SPY when risk-on (the 85% single-name safety-gate cap; the rest cash), flat/cash when risk-off — through the same Layer-3 gate, ledger, reconcile, kill switch and HWM as the spine.
BB_DRYRUN=1 bash live/run_benchmark_daily.sh          # compute + publish the regime, place nothing

Validation of record is research/benchmark_3_composite_regime.py — gated SPY beats buy-&-hold net of cost (Sharpe +0.87→+1.11, maxDD −34%→−16%), with the honest caveat that the edge is vol-timing (strip the vol signals and the non-vol internals barely clear buy-&-hold). Dry-run verified (today: risk-on, composite +0.20). Not a live-money endorsement; paper by default.

About Blaque Baux

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 Blaque Baux family

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.

Layout

engine/     the Blaque Baux platform (git submodule -> blaquebaux/base)
research/   three sketches (internals dashboard, lead-vs-coincide, composite regime) + scorecard
live/       benchmark_live.jl (regime emitter + gated-SPY book) + run wrapper + plist

License

MIT. (c) 2026 Carter Warrens.

About

Market internals as a risk-regime read — internals coincide (don't lead), but the composite gates SPY net of cost (Sharpe +0.87→+1.11). Regime emitter built: publishes market_regime.txt, honestly labeled vol-timing not breadth alpha. [Live path]

Topics

Resources

Code of conduct

Contributing

Security policy

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages