01
Observe
Every analysis begins with an exact market timestamp, underlying level, option expiry and bounded strike range. Missing or stale observations remain visible.
Methodology
The evidence stays beside the result.
A smooth chart is not enough. VolForge retains the observations, conventions, quality state and calculation version required to understand what a result means—and where it should not be trusted.
02
Fit within the market
Observed midpoints, bid–ask ranges and excluded observations remain distinct from interpolated values. Fit quality is measured against the quoted market.
03
Check the shape
The fitted view is tested for economically inconsistent shapes over its declared domain. A certificate records checks and any violations.
04
Derive carefully
Local volatility is produced only from a certified implied-volatility path and carries convergence, repricing error, bounds and warnings.
05
Reprice the book
Portfolio explanations and scenarios use full option repricing. Contributions reconcile back to the total change and retain a residual.
06
Keep provenance
Outputs retain market and position timestamps, inputs, calculation version, creator and tenant. Historical comparisons never silently use future observations.
Limitations
Models describe assumptions, not certainty.
Interpolation and calibration do not make an illiquid market precise. Portfolio results depend on data quality, instrument coverage, conventions and the selected scenario. VolForge exposes these dependencies so a practitioner can apply judgment.