Exposure scores
tell you the risk.
OM Atlas turns it into a decision.
OM Atlas combines filing-anchored exposure intelligence with a rules-based decision engine. Regime state, portfolio exposure and user-defined constraints map to transparent, sized model actions - each supported by historical evidence from the same portfolio.
Regime Status
Active
Signal
Rate-Cycle Shift
Model Action
Rotate - Sized
Illustrative model output
An exposure score identifies the risk.
It doesn't define the response.
The Scoring Trap
Company-level geopolitical and macro risk scores are increasingly available. But a score without a decision framework simply moves the judgment call downstream - the PM still has to determine what to adjust, by how much, and when.
The OM Atlas Answer
OM Atlas closes that gap with a rules-based decision layer. Regime state, portfolio exposure and user-defined constraints map to a sized, auditable model action - together with the historical evidence used to evaluate it.
Four Engines, One Closed Loop
Each engine produces a defined, auditable output - signals identify regimes, exposures measure portfolio sensitivity, decision rules interpret the combination, and the action engine sizes the response.
01
Forecast
Observed macro signals map to a defined set of tracked regimes, each progressing through calibrated status thresholds.
02
Exposure
Company exposure is grounded in reported company disclosures rather than sentiment alone, then aggregated to sector and portfolio levels.
03
Decision
Regime state, portfolio exposure and user-defined constraints map to a rules-based action state and sizing level - a reproducible decision rather than a qualitative risk flag.
04
Action
Eligible portfolio actions are ranked under transparent model rules. Historical evaluation then compares the same portfolio with and without the model action, isolating the decision framework's contribution.
We grade the evidence behind every model action - including where it's weak.
Every model action carries an evidence tier based on predefined criteria such as sample size, directional consistency and statistical confidence - not a headline return alone. Limited or inconclusive evidence is shown as such, rather than being absorbed into a single composite number.
Evidence Classification
Strong Evidence
Larger sample; predefined evidence threshold met
Directionally Consistent
Outcomes directionally aligned; not yet at the highest tier
Limited Sample
Too few observations for a reliable conclusion
No Observed Differentiation
No meaningful historical separation observed
Classification derived from sample sufficiency · directional consistency · statistical confidence. See Methodology for definitions and limitations.
OM Atlas provides research, analytics and model-generated portfolio decision support. Outputs are based on defined methodology rules and should not be construed as a guarantee of future investment outcomes. Historical analyses may include hypothetical or backtested results and are subject to assumptions, limitations and transaction-cost estimates described in the Methodology. Historical or simulated results are not indicative of future results.