Drawdown-Risk Lens (20-day, transparent z-score composite)

A probabilistic drawdown-risk gauge over the seven credit / inflation / rate LEVELS the full-history study flagged for the platform’s multi-week drawdown lane. Renders as the “Drawdown-Risk Lens · 20d” block in the regime card with a 0–100 score, a low / elevated / high band, and the top contributing legs.

This is not a directional predictor, not a timing trigger, and not the validated logistic from the research. It is a transparent equal-weight composite, chosen deliberately for explainability and graceful degradation over fitted accuracy.

What it measures

Each of the seven levels is standardized against its trailing baseline (default ~2 trading years), oriented by economic risk-direction, and averaged into a single number. Higher = more macro stress relative to the recent norm; lower = calmer.

Why this shape, not a fitted model

The full-history study (.research/SYNTHESIS.md) validated the multi-week drawdown signal as a parsimonious L2 logistic over the same seven features. That model is sklearn-based, and the platform’s prod image deliberately ships no machine-learning runtime — scikit-learn and its peers live in requirements-dev.txt and the no-Dockerfile-reference lint (verify-docs.sh:27) blocks them from ever being COPYed into a production container.

The deployable shape is therefore the transparent composite. Trade-offs, on the record:

• Transparent + auditable. Every contribution is a z-score the operator can recompute by hand. The card surfaces the top drivers so it never moves without showing why.
• Equal weights, not fitted. The validated logistic weighted the seven features differently. Equal weighting gives up some accuracy in exchange for not encoding an opaque fit calibrated to a single bear.
• Degrades gracefully. A missing feature is just dropped from the average; below four available legs the gauge suppresses itself (status: insufficient_data) rather than report off a thin set.

The seven inputs + risk-orientation

The orientation table lives in app/signals/drawdown_risk.py:_FEATURE_ORIENTATION and is documented inline with the rationale. Summary:

A +1 feature contributes +1·z to the mean (positive z = up = more risk). A −1 feature contributes −1·z (so a deeply inverted curve, where z is very negative, still pushes risk up).

How the gauge is computed

For each available feature with a non-zero baseline standard deviation:

1. z = (level − trailing_mean) / trailing_std
2. oriented_z = z × orientation

Then:

3. mean_z = average(oriented_z over available legs)
4. score = 100 / (1 + exp(−mean_z)) — a logistic squash with steepness 1.0, so mean_z = 0 maps to 50, +1σ to ~73, +2σ to ~88, −1σ to ~27.
5. Band: low under +0.5σ, elevated between +0.5σ and +1.25σ, high at or above +1.25σ.

If fewer than four legs are populated the gauge returns status: insufficient_data and the dashboard suppresses the block.

Caveats — read these before you act on it

• One bear in-sample. The live daily history covers ~2021–present, so the only validated drawdown regime is 2022. The composite is calibrated to that world. It is unvalidated in any other kind of bear and the bear-onset research wave is meant to escape this constraint by studying long-history macro leaders.
• Equal-weight is a deliberate downgrade in accuracy. The fitted logistic moved the AUC; the composite does not. Treat the gauge as situational awareness, not a quantitative probability.
• The breakeven sign is regime-specific. 2022 was an inflation-driven drawdown; elevated breakevens read as risk here. A deflationary bear would invert that. The orientation is the load-bearing assumption of the gauge — flag the gauge as suspect if the regime story changes.
• It does not contribute to the health score. This is a separate risk-side read; the 19-component health score is unchanged.

Where it lives in the code

• Pure helper: app/signals/drawdown_risk.py (no DB, no network, no ML).
• Route: GET /api/v1/signals/drawdown-risk (computes trailing baselines + current levels from daily_signals, calls the helper). Schema-versioned by DRAWDOWN_RISK_SCHEMA_VERSION.
• UI: rendered in the regime card on / directly below the regime line, above the historical base-rate block.
• Tests: tests/test_drawdown_risk.py pins orientation, banding, the insufficient-data fallback, and the route’s history-driven baseline computation.