Macro regime classification API for institutional risk monitoring
Macro regime classification across every major crisis since 1998
Validated against every major crisis since 1998. Outperformed the benchmark in 15 of 16 backtested episodes.
6.0:1
net preservation ratio
15.9%
average capital preserved per episode
15/16
episodes outperforming 60/40
Backtested signal validation
Classification accuracy measured as net capital preservation per episode. 16 stress episodes across 5 validated regions, 28 years of historical data (1998–2024). JP classified but unvalidated. Net of transaction costs and opportunity cost.
| Episode | Period | Benchmark | RegimeR | Net preserved | Ratio |
|---|---|---|---|---|---|
| US GFC | 2007–2009 | -22.6% | +16.1% | +35.8pp | 13.5:1 |
| AU GFC | 2008–2009 | -25.2% | +12.1% | +31.4pp | 6.4:1 |
| DE GFC | 2008–2009 | -31.8% | +7.0% | +35.5pp | 11.6:1 |
| GB GFC | 2007–2009 | -29.1% | +10.4% | +36.6pp | 13.3:1 |
| US COVID | 2020 | +2.8% | +8.0% | +1.2pp | 1.3:1 |
| AU COVID | 2020 | -5.5% | +6.4% | +8.9pp | 4.0:1 |
| DE COVID | 2020 | -2.5% | +6.7% | +6.1pp | 3.0:1 |
| GB COVID | 2020 | -4.5% | +8.1% | +11.0pp | 7.8:1 |
| US 2022 | 2022 | -24.2% | -15.6% | +6.0pp | 3.3:1 |
| AU 2022 | 2022 | -20.3% | -13.7% | +3.3pp | 2.0:1 |
| DE 2022 | 2022 | -31.6% | -18.4% | +10.0pp | 4.1:1 |
| GB Gilt Crisis | 2022 | -21.0% | -5.3% | +13.1pp | 6.3:1 |
| DE Eurozone | 2011–2012 | +3.4% | +23.0% | +7.3pp | 1.6:1 |
| CA GFC | 2008 | -22.5% | +1.4% | +23.9pp | 8.2:1 |
| CA COVID | 2020 | -5.4% | +3.0% | +8.4pp | 4.1:1 |
| CA Oil Rout | 2014–2016 | -2.6% | -4.0% | -1.4pp | — |
Detect
Multiple macroeconomic indicators classify 4 regimes — Contraction, Stagflation, Transition, and Expansion — using fixed thresholds calibrated on historical validation. No learned weights in classification, no black boxes.
Validate
Independent regime classification that runs against your internal macro view. Divergence between RegimeR and your team is an early signal that is a data point for internal analysis.
Preserve
Net capital preservation measured per backtested episode, after all transaction costs and opportunity cost deductions. Every episode outperformed the 60/40 benchmark across 28 years of historical data.
What you can’t get elsewhere
Fragility scoring
Not just what regime you’re in — which signal would flip you out, how far away each boundary is, and what’s binding. The system computes distances to every adjacent regime boundary in real time. A portfolio manager with their own macro view can see exactly where the vulnerabilities are.
Divergence monitoring
Early warning when economic signals show stress that credit markets haven’t priced in. When divergence fires, a regime change follows within 12 months 27.5% of the time — 5.3x the base rate. Fires roughly once per decade per region. An attention signal, not a trading trigger.
Private credit composite
Yield curves and VIX are available anywhere. Our proprietary ACCT indicator captures stress in private credit markets through multiple channels before it surfaces in public market indicators.
Three validation layers
Not one statistical test — three independent methodologies using different data, different timescales, and different frameworks, all finding signal. Permutation testing, daily streaming confirmation, and sub-type discrimination on observed outcomes. Convergence across methods is harder to dismiss than any single p-value.
Known limitations
2022 is the weakest out-of-sample window. Japan's zero-rate environment breaks the example baskets. Some sub-type classifications have zero confirmed episodes. We publish what doesn’t work alongside what does.
The result
$306M
$100M under RegimeR after 16 stress episodes
$150M
Same $100M under 60/40
+104%
Cumulative advantage, averaged across 4 regions
Each preserved crisis raises the capital base for the next. In backtested analysis over 28 years and 16 stress episodes across the US, Australia, Germany, and the UK, the compounding advantage doubled the terminal value of the portfolio relative to a standard 60/40 allocation.
These figures include the 2022 rate shock — our weakest episode — where the system still lost money in two regions. The 104% advantage is calculated across all episodes, not a selection.
Developed 2026. All performance figures are backtested against historical data (1998–2024) using walk-forward methodology with no look-ahead bias. Prospective validation live since March 2026. Past performance does not guarantee future results.
Every classification is cryptographically logged before the outcome is known. The hash chain is independently verifiable — due diligence teams can audit the prospective record without trusting our reporting.