Methodology & Data Sources
Transparent documentation of our empirical approach to regime classification and performance analysis. All regime performance claims are grounded in historical backtesting using publicly available data.
Empirical Grounding Approach
How we validate regime performance claims
VantMacro's regime performance data is empirically derived from factor-based historical regime classification spanning 2003-12-17 to 2026-01-20 (8,039 daily classifications). We classify each day across four dimensions using publicly available economic indicators, then calculate actual asset performance during each regime period.
Data Sources
All data from authoritative, publicly available sources
| Factor | Source | Series ID | Purpose |
|---|---|---|---|
| Growth (CFNAI) | FRED | CFNAI | Chicago Fed National Activity Index - measures economic activity above/below trend |
| Inflation (CPI) | FRED | CPIAUCSL | Consumer Price Index - calculated year-over-year % change |
| Liquidity (Fed BS) | FRED | WALCL | Federal Reserve balance sheet total assets - QE/QT proxy |
| Market Risk (VIX) | Yahoo Finance | ^VIX | CBOE Volatility Index - equity market implied volatility |
| Credit Risk (HY) | FRED | BAMLH0A0HYM2 | ICE BofA High Yield Option-Adjusted Spread - credit market stress |
| Asset Prices | Yahoo Finance | S&P 500, Nasdaq 100, Small Caps, Long-Term Treasuries, Investment Grade Corporate, High Yield Bonds, Gold, Oil, Copper, Europe, UK, Japan, China, India, Emerging Markets, BTC-USD, ETH-USD | Daily closing prices for 15 asset classes spanning US equities (S&P 500, Nasdaq 100, Small Caps), bonds (Long-Term Treasuries, Investment Grade Corporate, High Yield Bonds), commodities (Gold, Oil, Copper), global equities (Europe, UK, Japan, China, India, Emerging Markets), and crypto (Bitcoin, Ethereum) |
All data sources are free to access and widely used by financial professionals for research and analysis.
Four-Dimensional Classification Framework
How we identify historical regimes
π1. Real Cycle (Growth)
π₯2. Inflation
π°3. Liquidity/Policy
β‘4. Market Risk
Composite Regime Mapping
The four dimensions combine using priority-based logic:
- 1.Crisis conditions override all - If VIX > 25 or HY spreads > 500bps β CrisisLiquidation
- 2.Post-crisis recovery - If risk elevated + policy easing β PostShockRecovery
- 3.Stagflation check - If growth slowdown + inflation high β StagflationarySqueeze
- 4.Normal regimes - Map based on growth + inflation + policy combination
- 5.Mixed signals - Unclear patterns β Transitional
Historical Validation
Verifying classifications against known episodes
We validate our regime classifications against major historical episodes to ensure accuracy:
2008 Global Financial Crisis
2020 COVID Pandemic
Post-GFC Recovery
Post-COVID Recovery
Pass criterion: β₯50% of days in expected regime during validation period
Performance Calculation Methodology
How we compute empirical returns
For each continuous regime period (same regime, no gaps > 35 days), we calculate:
CAGR
Sharpe Ratio
Max Drawdown
Regime Distribution (2003-2026)
Coverage across composite regimes
| Regime | Days | % of Total | Occurrences | Avg Duration |
|---|---|---|---|---|
| ReflationaryExpansion | 1,942 | 24.2% | 21 | 92 days |
| StagflationarySqueeze | 1,617 | 20.1% | 16 | 101 days |
| PostShockRecovery | 1,461 | 18.2% | 20 | 73 days |
| DisinflationarySlowdown | 1,295 | 16.1% | 24 | 54 days |
| LateCycleInflationaryBoom | 804 | 10.0% | 19 | 42 days |
| CrisisLiquidation | 638 | 8.0% | 21 | 30 days |
| Transitional | 268 | 3.3% | 17 | 16 days |
| Total | 8,039 | 100% | 138 | 58 days |
Balanced distribution across regimes provides robust coverage for statistical analysis. All 7 composite regimes represented.
Limitations & Disclaimers
Understanding the boundaries of this analysis
β οΈ Important Limitations
- Data begins 2003-12-17 (limited by VIX and HY spread availability)
- Monthly/weekly indicators forward-filled to daily (assumes state persistence)
- Four-factor model captures most regime nuances but not all edge cases
- Fixed thresholds (e.g., CFNAI -0.7, CPI 2.0-4.0%) may not apply equally across all eras
- Asset class coverage: 15 total (US equities, bonds, commodities + 6 global equity indices: Europe, UK, Japan, China, India, Emerging Markets + crypto: Bitcoin/Ethereum added Dec 2025; crypto data from 2014/2017 onwards)
π Educational Use Only
This methodology is provided for transparency and educational purposes. All historical performance is descriptive, not predictive.
- Past performance does not guarantee future results
- Not a recommendation to buy or sell securities
- Market conditions evolve and historical patterns may not repeat
- Consult a qualified financial advisor before making investment decisions
Reproducibility & Code
Open methodology for verification
All regime classification and backtest code is available in the project repository under web/backtests/empirical_grounding/
Classification Script
historical_regime_classifier.pyFetches FRED/Yahoo data, classifies daily regimes, validates against historical episodes
Backtest Script
asset_regime_performance.pyLoads regime history, fetches asset prices, calculates CAGR/Sharpe/drawdown by regime
Detailed methodology documentation available in HISTORICAL_REGIME_METHODOLOGY.md
Statistical Validation Results
Out-of-sample testing and asset-specific significance
Out-of-Sample Testing (2015-2026)
We validated our regime framework by training on 2003-2015 data and testing on 2015-2026 (held-out data the model never saw).
Use regime signals for directional guidance (should I be risk-on or risk-off?), not for precise return forecasts. Out-of-sample results are strongest for equities and weakest for long-duration Treasuries.
Asset-Specific Regime Effects
| Asset | Statistical Significance | Effect Size (Ξ·Β²) | Recommendation |
|---|---|---|---|
| S&P 500 (SPX) | p < 0.001 | 1.08% | β Use for regime positioning |
| Nasdaq-100 (QQQ) | p < 0.001 | 0.96% | β Use for regime positioning |
| Russell 2000 (IWM) | p < 0.001 | 0.94% | β Use for regime positioning |
| High Yield (HYG) | p < 0.001 | 0.57% | ~ Moderate regime effect |
| Commodities (DBC) | p < 0.001 | 0.57% | ~ Moderate regime effect |
| Long Treasuries (TLT) | p = 0.12 | 0.18% | β Weak regime effect |
| IG Bonds (LQD) | p = 0.28 | 0.14% | β Weak regime effect |
| Gold (GLD) | p = 0.45 | 0.11% | β Weak regime effect |
Statistical significance tested via one-way ANOVA (F-test) with Tukey HSD post-hoc analysis. Ξ·Β² (eta-squared) represents variance explained by regime classification. p < 0.05 indicates statistically significant regime effect.
What does RΒ² = 0.68 mean here?
Understanding the Net LiquidityβNASDAQ relationship
We regressed log(NASDAQ) on log(US Net Liquidity) using weekly FRED data from 2003β2025 (650 observations). The RΒ² of 0.68 indicates a strong fit between the two series in log-levels over this period.
This is a descriptive association, not a prediction or causal claimβyear-over-year changes show a much weaker relationship (RΒ² β 0.16). Because both series trend over decades, this specification can inflate RΒ²; we present it as context, not a mechanical pricing model.
US Net Liquidity Definition
Where WALCL = Fed Balance Sheet, TGA = Treasury General Account, RRP = Reverse Repo Facility
Regime Transition Patterns
Empirical transition probabilities from 427 historical regime changes
Based on 427 empirical regime transitions from 2003-2026. Shows the probability of transitioning from one regime (rows) to another (columns).
Regime Transition Matrix
Empirical probabilities of transitioning from one regime (rows) to another (columns). Based on 427 historical transitions (2003-2026).