Features
Every layer of the model, built for scrutiny.
Orèvance combines structured data ingestion, rules-based analysis, and disciplined output formatting into a single pipeline. Below is what actually happens between the moment you define a mandate and the moment you receive a portfolio construct.
Core Modules
Four functions, one continuous process.
Each module below handles a distinct stage of portfolio construction. They are designed to run in sequence, with each output feeding the next stage's input requirements.
01
Mandate Intake
Define objectives, constraints, time horizon, and risk tolerance through a structured intake form. This becomes the fixed parameter set that governs every downstream calculation — nothing is inferred beyond what you specify.
02
Data Aggregation
Relevant market and asset data is pulled from configured sources and normalized into a common structure, so comparisons across asset classes and time periods are handled consistently rather than manually reconciled.
03
Model Analysis
A rules-based engine applies the logic tied to your mandate — screening, weighting, and constraint checks — and produces a candidate allocation along with the reasoning trail behind each inclusion or exclusion.
04
Structured Output
Results are returned as a readable construct: allocation breakdown, rationale notes, and flagged constraints. Nothing is auto-executed — the output is a document for your review, not an instruction to a broker.
Why It's Structured This Way
Analysis without opinion injection.
Most modelling tools blend data retrieval, judgment, and presentation into a single opaque step. That makes it difficult to know which part of an output is fact-derived and which part is assumption.
Orèvance keeps these stages separate and visible. You can see what data fed the model, what rule produced a given weighting, and where the model's fixed logic reached the edge of your specified constraints.
The result is a portfolio construct you can interrogate line by line — not a black-box recommendation you're asked to trust outright.
Methodology
What the engine actually checks.
The analysis engine runs a fixed sequence of checks against your mandate: exposure limits, correlation between selected positions, concentration thresholds, and alignment with the time horizon you defined.
None of these checks are adjustable mid-run — they're set at intake, which means the same mandate run twice against the same data will produce the same construct. Consistency is treated as a feature, not an afterthought.
Where a constraint can't be satisfied within your inputs, the model flags the conflict explicitly rather than silently loosening the rule.
Checks applied per run
- Exposure limits — per-asset and per-sector caps honored from intake
- Correlation review — flags overlapping risk across selections
- Concentration threshold — caps single-position weighting
- Horizon alignment — matches asset liquidity to stated timeframe
Output logic
Mandate → parameters locked
Data → aggregated and normalized
Rules → applied without override
Conflict → flagged, not resolved silently
Output → returned for your review
In Practice
What the output construct looks like.
Below is a simplified view of the kind of structured breakdown Orèvance returns once a mandate has been processed.
Each line item ties back to a mandate constraint, visible on request.
Weightings reflect the rules run, not a discretionary override.
You adjust and re-run, or approve the construct as delivered.
See the full construct for your own mandate.
Initialize PortfolioIllustrative process only. Orèvance does not execute trades or guarantee outcomes.