# CrashTestYourStrategy

> Stress-regime intelligence layer for trading strategies.
> Tests strategy behavior against a curated catalog of historical and synthetic stress cases,
> classified along an explicit 10-axis failure-mode taxonomy.
> Synthetic cross-asset portfolio + strategy stress, available on request via a form, with a
> machine-readable realism/value trust layer — see "Portfolio & Strategy Stress" below.

## Overview

CrashTestYourStrategy is a research-grade backtest framework that evaluates how trading
strategies behave under specific market stress conditions. Rather than producing a single
backtest number conditional on one historical path, it runs each strategy against a curated
catalog of stress cases — real historical OHLC slices combined with controllable synthetic
probes — and reports performance by failure mode.

This is NOT a financial advisory tool. It does not make forecasts, recommendations, or
predictions. All outputs are model-based simulations under simplified assumptions.

## Investment-Thesis Catalog (home page)

The home page (`/`) is a curated catalog of investment theses. Each entry is a three-layer
case study: (1) an economic **framework** stated as falsifiable claims with evidence grades,
(2) the **portfolio** that framework implies (assets × risk tiers), and (3) the **stress
evidence** — how that allocation holds up across synthetic regimes and real historical
episodes, with per-claim verdicts. Descriptive, never advisory.

- `/catalog/{slug}` — a single thesis in full. Live:
  - `/catalog/regime-balanced-debasement` (the Systemic Money Model → a regime-balanced
    allocation with a debasement tilt)
  - `/catalog/free-rider-portfolio` (the Free-Rider Portfolio Framework → cap-weighted
    aggregate ownership, VT/BND/BIL, tested against a concentrated winner-pick benchmark)

## Portfolio & Strategy Stress (on request)

Cross-asset portfolio and strategy stress diagnostics — descriptive, never advisory. Separate
from the single-asset strategy reports below.

- **Access:** on request via the form at `https://crashtestyourstrategy.com/contact`. Describe a
  strategy or a portfolio; it is run internally and the result returned by email. Direct agent
  access (REST + MCP) exists but is token-gated — tokens are issued individually after contact;
  there is no open, unauthenticated compute endpoint.

Each diagnostic returns a uniform result: a factual `grounding_summary` (the omitted tail an
optimistic view ignores), a gate-capable `revision_required` flag, and
`methodological_limitations`. The user decides what the outcome means.

- `portfolio_stress_test` — stress a multi-asset
  portfolio (SPY / TLT / GOLD / BTC) across cross-asset regimes (baseline / risk-off /
  rate-shock). Reports per-regime drawdown, tail (VaR + Expected Shortfall), a per-leg
  decomposition, and a `cross_asset_finding` (diversification_intact / hedge_holds /
  hedge_breaks / shared_drawdown) — including the bond hedge that can BREAK under a rate shock
  (the 2022 pattern). Plus the full per-path drawdown DISTRIBUTION
  (`path_drawdown_quantiles` p50–p95/worst, `n_paths`), `probability_weighting` (scenario
  weights: unconditional substrate shares + a nowcast tilt from the validated regime layer —
  model-conditional frequencies, not real-world probabilities), `derisk_comparison`:
  buy-and-hold vs a −10% drawdown-stop on the same paths, and `realism_basis` self-certifying
  the result's realism per asset. Optional `costs` (rebalancing policy + expense ratios +
  per-trade cost) adds a `cost_impact` block via a path-loop engine with real unit accounting —
  the rebalancing POLICY effect and the pure COST drag deliberately separated.
- `challenge_strategy` — surface a strategy's structural vulnerabilities (no vulnerability-free
  claim survives).
- `backtest_integrity` — deflated-Sharpe + regime-coverage check on a claimed backtest (catches
  selection-effect over-optimism).
- `ips_gate` — check a portfolio against investment-policy constraints (max drawdown tolerance,
  time horizon, liquidity need). Checks the full drawdown DISTRIBUTION, not just the typical
  path: reports `breach_probability` (share of simulated paths exceeding the stated tolerance)
  and flags material tail risk (`drawdown_tail_risk`) even when the median path passes.
- `factor_decomposition` — capital weight vs RISK contribution (where the risk actually sits).
- `portfolio_compare` — paired comparison of two portfolios (reference vs candidate revision)
  on IDENTICAL simulated paths, so every delta is attributable to the weights, not seed noise.
  Flags a candidate that deepens the worst-path drawdown or introduces a new diversification
  failure.
- `long_horizon_stress` — multi-year wealth paths for a savings plan (monthly contribution)
  or a withdrawal plan (inflation-indexed): terminal-wealth quantiles (nominal + real),
  ruin/shortfall probabilities, a sequence-of-returns diagnosis (same plan, bad vs good first
  two years), and a drift-sensitivity block. Long-run drift is a stated, overridable
  ASSUMPTION (re-anchored from the stress substrate — both disclosed); paths chain ~2y model
  blocks; costs on by default. Descriptive — no rate, allocation, or product is recommended.
- `regime_outlook` — model-conditional probabilities that an asset (SPY / QQQ / GLD / TLT) is
  in each regime (BULL / SIDEWAYS / BEAR / CRISIS) after a 5- or 21-trading-day horizon, with
  the persistence and unconditional baselines alongside. Preregistered + out-of-sample
  validated (annual seasonality was tested and falsified — deliberately excluded); only
  validated assets/horizons are accepted. Descriptive probabilities of operationally defined
  regime classes — not a market prediction.

**Trust layer — check the evidence before trusting a result:**

- `validation://{asset}` — the model's daily stylized
  facts measured vs empirical bands, with a per-fact verdict (in-band / caveat). Honest by
  construction: a missed fact is a caveat, not hidden (e.g. SPY/TLT match all 18 facts; GOLD/BTC
  carry one kurtosis caveat).
- `validation://value` — the numeric comparison vs a
  Gauss-copula baseline: where the cross-asset paths add value (regime-conditional correlation
  strengthening, joint-crash tail dependence) and where they do NOT (single-asset marginals).

Universe + schema: `portfolio://universe`. BaFin/WpHG: reports what the simulated portfolio DID
across stress regimes; no ranking, no buy/sell signal, no suitability claim.

## Architecture

- **Case catalog**: 63 stress cases — 31 empirical anchors (real historical OHLC: Lehman
  2008, Dotcom 2000, COVID 2020, Luna 2022, Volmageddon 2018, Taper Tantrum 2013, etc.)
  plus 32 synthetic stress probes (50 Monte-Carlo replicas each, across profile families:
  low-vol grind, controlled whipsaw, slow-stagflation, hyperinflation, demand-destruction,
  sharp-crash setup, vol-expansion setup, liquidity-stress setup, v-recovery setup,
  slow-decline-with-partial-recovery, slow-crash-no-recovery — the last two are intentionally
  distinct stress hardness levels).
- **Failure-mode taxonomy (10 axes; 9 score buckets since May 2026)**: TREND_UP, TREND_DOWN,
  SIDEWAYS, VOL_EXPANSION, VOL_COMPRESSION, SHARP_CRASH, SLOW_BEAR, V_RECOVERY, WHIPSAW,
  LIQUIDITY_STRESS. V_RECOVERY is treated as a diagnostic path pattern (decomposed into
  SHARP_CRASH down-leg + TREND_UP up-leg) for score aggregation, leaving 9 effective
  score buckets. The 10-axis label is retained for case-tagging and UI continuity.
  Every case is tagged with one or more failure modes; a single case (e.g. Lehman 2008)
  can contribute to multiple buckets simultaneously.
- **Conditional-Distribution-Engine (May 2026 — Phase 17/18a)**: profile names describe
  *imposed conditions* on the regime-switching simulator, not output guarantees. Per-replica
  ex-post FM-classification against operational gating definitions; replicas distribute
  across the FM spectrum, may be multi-tagged, may be sub-threshold. Per-FM-bucket
  scoring with shrinkage and dominance-weighted multi-FM attribution is implemented in
  V2 (`calculate_robustness_score_v2`). **V2 is live in production since
  2026-05-15** with a-priori-augmentation (Phase 18d) for assets with sparse
  ex-post coverage. V1 retained as `calculate_robustness_score` for
  diagnostics. See `/verifiability` for full architecture-status disclosure.
- **Aggregation**: Strategy performance is computed per case, then aggregated by
  failure mode to produce a robustness score (0-100) plus a per-mode breakdown.
- **Synthetic-probe simulator**: a regime-switching stochastic-volatility model; per-asset
  character informed by CFTC Commitment-of-Traders data (~3,000 weekly reports). Used only for the 32
  synthetic stress probes, not for replicating full instruments.
- **Backtest correctness**: Implements the Löw, Maier-Paape & Platen (2015) methodology
  for ambiguous OHLC candles, defaulting to worst-case execution.

## Supported Assets

SPY (S&P 500), QQQ (Nasdaq 100), BTC (Bitcoin), ETH (Ethereum), GOLD, WTI (Crude Oil),
VIX (Volatility Index).

Each asset has its own case set and CoT-derived agent calibration where applicable.

## Supported Indicators

SMA, EMA, RSI, MACD, Bollinger Bands, ATR, Stochastic, CCI, Williams %R, Supertrend,
Parabolic SAR, Aroon, Donchian, and others.

Not yet supported: volume-based indicators (OBV, VWAP, MFI, CMF). Strategies
referencing volume parse correctly, but volume-dependent signals do not fire.
Synthetic case OHLCV data includes volume columns; the engine does not generate
signals from volume.

## Pre-Computed Strategy Reports

Thirteen curated strategies have pre-computed reports available at /s/{slug}, designed to
span contrasting failure-mode profiles rather than indicator redundancy. Nine are calibrated
against the SPY case set; four are asset-diversified variants on BTC and GOLD.

The full catalog is browsable as a hub at /strategies (grouped by instrument), which carries
ItemList JSON-LD enumerating every /s/{slug} report. Faceted landing pages narrow the catalog
to a single instrument or strategy style — each is prerendered with its own canonical URL and
ItemList JSON-LD, so they can be cited or indexed individually:

- /strategies/asset/spy  — strategies for the S&P 500
- /strategies/asset/btc  — strategies for Bitcoin
- /strategies/asset/gold — strategies for Gold
- /strategies/style/trend-following — moving-average crossovers, Supertrend, trend-filters
- /strategies/style/mean-reversion  — RSI, Bollinger bounces, oscillator-based entries
- /strategies/style/momentum        — MACD crossovers, momentum filters
- /strategies/style/passive         — buy-and-hold baselines

- /s/buy-hold-spy — Buy and Hold S&P 500 (passive equity baseline)
- /s/buy-hold-btc — Buy and Hold Bitcoin (passive crypto baseline)
- /s/sma-200-crossover — SMA 200 Crossover (long-term trend-following)
- /s/sma-200-trend-gold — SMA 200 Trend Following (GOLD; secular-trend asset)
- /s/ema-12-26-crossover — EMA 12/26 Crossover (short-term trend)
- /s/ema-12-26-crossover-btc — EMA 12/26 Crossover (BTC; crypto momentum)
- /s/macd-crossover — MACD Signal Line Crossover (composite trend + momentum)
- /s/macd-rsi-confirmation-btc — MACD + RSI Confirmation (BTC; momentum + filter)
- /s/rsi-30-70 — RSI 30/70 Mean Reversion (oscillator-based mean-reversion)
- /s/rsi-30-70-btc — RSI 30/70 Mean Reversion (BTC; crypto mean-reversion)
- /s/bollinger-bounce — Bollinger Band Bounce (volatility-band mean-reversion)
- /s/rsi-sma-trend-filter — RSI with SMA-200 Trend Filter (hybrid)
- /s/supertrend — Supertrend Strategy (volatility-adaptive trend)

Each report contains: robustness score (0-100), failure-mode breakdown per regime,
worst-case simulated drawdown (WCDD-95), distribution statistics, score decomposition
(drawdown resilience, failure rate, regime consistency, return stability, path
sensitivity), best/median/worst path equity curves, and example trades per regime.

For long-warm-up indicators (e.g. SMA-200, requiring 300 days), synthetic cases are
automatically skipped if the case duration cannot accommodate warm-up plus a tradeable
window. The skip is reported transparently in the response.

JSON access: `GET /api/v1/strategies/{slug}`.

## Per-Asset Case Coverage

- **SPY**: Bull Run 2017, China Sideways 2015, Fed Bear 2022 H2, Lehman GFC 2008,
  COVID 2020, Vol Shock Feb 2018, plus 4 synthetic probes
- **QQQ**: Dotcom Crash 2000, Tech Rally 2017, COVID + Tech V-Recovery, Tech Bear
  2022, Vol Shock Feb 2018, plus 4 synthetic probes
- **VIX**: Low Vol 2017, Volmageddon 2018, COVID Spike 2020
- **BTC**: 2017 Parabolic ATH, Crypto Winter 2018, Luna Collapse 2022, FTX Collapse
  2022, Sideways 2023, plus 1 synthetic probe
- **ETH**: Crypto Winter 2018, Vol Shock 2018, COVID + DeFi Recovery 2020, plus 1
  synthetic probe
- **WTI**: Pre-GFC Oil Boom, Saudi Oil War 2014-2015, Sideways 2018, Negative Oil
  2020, plus 1 synthetic probe
- **GOLD**: Post-GFC Inflation Hedge 2010, Sideways 2014, Taper Tantrum 2013, COVID
  + ATH 2020, plus 2 synthetic probes

API endpoints:
- `GET /api/v1/strategies/{slug}/cases?asset={asset}` — all cases for an asset
  (Strategy vs Buy-and-Hold reference)
- `GET /api/v1/strategies/{slug}/cases/{asset}/{case_id}` — single case with
  equity curves
- `GET /api/v1/strategies/case-assets/list` — assets with case studies

## Example Queries This Site Answers

- "How does SMA 200 crossover perform in the 2008 financial crisis?"
- "Is RSI 30/70 robust during the Luna collapse?"
- "Which trading strategies survive sharp crashes historically?"
- "How fragile are mean-reversion strategies under volatility expansion?"
- "Strategy behavior during whipsaw markets — what works?"
- "Worst-case drawdown for momentum strategies in COVID 2020?"
- "Buy-and-hold vs trend-following during slow bear regimes?"
- "Cross-asset robustness: does an SPY-tuned strategy work on Bitcoin?"

Each case-study response provides: strategy total return, maximum drawdown,
Sharpe ratio, equity curve (single curve for empirical anchors; median plus
p25/p75 band for synthetic cases), and buy-and-hold reference for direct
comparison.

## Failure-Mode Definitions

The 10-axis failure-mode taxonomy is defined at `/methodology` with operational
thresholds, historical examples, common strategy failure patterns, and
distinctions between similar regimes. Each definition is published as a
Schema.org DefinedTerm to support machine extraction.

Direct anchors: `/methodology#failure-mode-{slug}` (e.g. sharp-crash, vol-expansion,
whipsaw, liquidity-stress).

## Verifiability — Falsifiable Model Claims

Synthetic stress profiles are accompanied by measurable statistical claims.
For each profile-asset combination, the site publishes claimed operational
properties (drawdown ranges, realized-volatility bounds, sign-change
frequency, etc.) alongside the measured aggregate over 50 Monte Carlo
replicas — including transparent disclosure of where measured values
deviate from the claimed range.

- `/verifiability` — browsable per-dataset claim validation with aggregated
  metrics across all 32 synthetic stress datasets, plus the full off-band
  calibration disclosure (BTC empirical-identifiability verification, Wilson
  95% CI for marginal cases, 9-point methodology-limitations list including
  score definition-dependence)
- `/verifiability_snapshot.json` — raw machine-readable snapshot
  (CC-BY 4.0, refreshed on data regeneration)
- Schema.org Dataset markup published on the verifiability page

This is the falsifiability layer: model behavior is not asserted but
measured against published claims.

## Pages

- `/` — Investment-thesis catalog (home): curated theses, each a framework → portfolio →
  stress-evidence case study
- `/catalog/{slug}` — a single thesis in full (framework, derived portfolio, stress evidence);
  live: `/catalog/regime-balanced-debasement`, `/catalog/free-rider-portfolio`
- `/contact` — commission a bespoke analysis or portfolio framework ("work with me")
- `/s/{slug}` — Pre-computed strategy reports
- `/methodology` — How it works, conceptually: the regime-switching simulator,
  stylized facts, the clustering gate, the failure-mode taxonomy, scope of claims
- `/reference` — Deep technical reference: architecture, per-asset bands,
  the validated regime catalog, documented limits
- `/legal/impressum` — Impressum (German)
- `/legal/datenschutz` — DSGVO/GDPR privacy policy
- `/legal/disclaimer` — Legal disclaimer

## API Endpoints (read-only, public)

**Investment-thesis catalog (home page):**
[https://crashtestyourstrategy.com/](https://crashtestyourstrategy.com/)
— curated investment theses, each a framework → portfolio → stress-evidence case study.
Descriptive, not advisory. The machine-readable agent capability declaration lives at /interop (below).

**Full machine-readable capability declaration:**
[https://crashtestyourstrategy.com/interop](https://crashtestyourstrategy.com/interop)
— canonical input/output schemas, controlled vocabularies (behavior, severity,
impact_type, regime classes), example agent workflows, methodological boundaries.

**Query pages (natural-language retrieval, FAQPage JSON-LD):**
[https://crashtestyourstrategy.com/queries](https://crashtestyourstrategy.com/queries)
— question-shaped pages answering common queries with live catalog data:

- `/queries/does-rsi-work-on-bitcoin`
- `/queries/how-do-strategies-behave-in-liquidity-stress`
- `/queries/how-do-strategies-behave-in-sharp-crashes`
- `/queries/most-robust-spy-strategies`
- `/queries/most-robust-btc-strategies`
- `/queries/buy-and-hold-vs-active-strategies`
- `/queries/worst-case-drawdowns-catalog`
- `/queries/how-trading-strategies-are-stress-tested`

**Failure-mode ontology (stable URLs, citation-ready):**
[https://crashtestyourstrategy.com/ontology](https://crashtestyourstrategy.com/ontology)
— DefinedTermSet root listing all 10 failure-modes. Each term has its own
stable page with full operational definition + DefinedTerm JSON-LD:

- `/ontology/trend-up`         · TREND_UP
- `/ontology/trend-down`       · TREND_DOWN
- `/ontology/sideways`         · SIDEWAYS
- `/ontology/vol-expansion`    · VOL_EXPANSION
- `/ontology/vol-compression`  · VOL_COMPRESSION
- `/ontology/sharp-crash`      · SHARP_CRASH
- `/ontology/slow-bear`        · SLOW_BEAR
- `/ontology/v-recovery`       · V_RECOVERY (diagnostic only)
- `/ontology/whipsaw`          · WHIPSAW
- `/ontology/liquidity-stress` · LIQUIDITY_STRESS

### Agent-optimised (recommended for LLM/agent consumers)

- `POST /api/v1/agent/analyze` — one-shot strategy analysis with structured
  robustness semantics (behavior labels per regime, top failure modes with
  impact-type ontology, evaluation confidence, methodological limitations).
  Accepts either natural-language description or canonical strategy JSON.
  Output is descriptive, not advisory — the agent decides what "suitable" means.
  Schema family: `ctys-agent`, version: `ctys-agent-v1`.

  Request example (natural language):
  ```json
  {
    "input_type": "natural_language",
    "strategy_text": "Buy BTC when RSI(14) falls below 30, sell when above 70",
    "asset": "BTC",
    "num_runs": 30
  }
  ```

  Request example (canonical JSON):
  ```json
  {
    "input_type": "strategy_json",
    "strategy": { /* same shape as /backtest/robustness/quick */ },
    "asset": "SPY"
  }
  ```

  Response (excerpt):
  ```json
  {
    "robustness_score": 53.2,
    "robustness_class": "moderate",
    "evaluation_confidence": { "level": "moderate", "classified_ratio": 0.59,
                                "sparse_bucket_count": 5, "augmented_replicas_total": 305 },
    "summary": "Robustness 53/100 (moderate). Strong in TREND_UP, SLOW_BEAR. Weak in LIQUIDITY_STRESS.",
    "per_regime": [{ "regime": "TREND_UP", "behavior": "stable", "score": 81.6, ... }, ...],
    "top_failure_modes": [{ "mode": "LIQUIDITY_STRESS", "severity": "high",
                             "impact_type": "drawdown_expansion",
                             "expected_behavior": "drawdowns deeper than the pool baseline",
                             "bucket_score": 37.0, "delta_from_pool": -14.5 }],
    "regime_class_summary": { "trend_regimes": { "avg_score": 73.7, "behavior": "stable",
                                                  "buckets": ["TREND_UP", "TREND_DOWN", "SLOW_BEAR"] }, ... },
    "caveats": [...], "methodological_limitations": [...],
    "report_url": "https://crashtestyourstrategy.com/s/buy-hold-btc",
    "schema_version": "ctys-agent-v1"
  }
  ```

- `GET /api/v1/catalog/query` — semantic discovery of pre-computed reports
  by indicator, asset, failure-mode coverage, regime class, or robustness class.
  Query params: `indicator`, `asset`, `failure_mode`, `regime_class`,
  `robustness_class`, `min_score`, `max_score`.

  Examples:
  - `?indicator=RSI&asset=BTC` — RSI strategies on BTC
  - `?failure_mode=SHARP_CRASH` — strategies scored against SHARP_CRASH bucket
  - `?robustness_class=robust&min_score=70` — top-tier strategies

### Read-only canonical reports

- `GET /api/v1/strategies/` — list of pre-computed strategies
- `GET /api/v1/strategies/{slug}` — full strategy report (JSON, UI-shape)
- `GET /api/v1/strategies/{slug}/cases?asset={asset}` — case-study results per asset
- `GET /api/v1/strategies/{slug}/cases/{asset}/{case_id}` — single case detail

### Lower-level execution (for callers that want to compose steps)

- `POST /api/v1/nl/parse` — parse a natural-language description (UI-shape; agents
  should prefer `/agent/analyze`)
- `POST /api/v1/backtest/robustness/quick` — run a strategy and return UI-shape JSON

## Technical Details

- Backend: FastAPI (Python), async SQLAlchemy, deployed on Hetzner
- Frontend: React 19 + TypeScript + Tailwind CSS + Vite
- Strategy parsing: LLM-based with structured-output JSON-Schema enforcement
  (GPT-4o-mini via OpenRouter, Llama via Groq, Claude as fallback)
- Synthetic-probe simulator: Numba JIT-compiled, per-asset CoT calibration
- Backtest engine: Löw, Maier-Paape & Platen (2015) ambiguous-candle methodology

## Important Legal Notice

This tool is for informational and educational purposes only. It does not
constitute financial advice, investment recommendations, or forecasts. Results
are based on model-driven simulations under simplified assumptions. Real market
outcomes may differ significantly. Individual financial circumstances are not
considered.

Operated under German jurisdiction (BaFin/WpHG framework). All published
analyses use neutral framing — no rankings, no directive language, no buy/sell
signals.

## Contact

Website: https://crashtestyourstrategy.com
Legal: /legal/impressum