Valuation engine

Prism valuation overview

Prism is MANTL's internal valuation framework used to estimate fund-level fair value and calculate NAV updates. We share the structure publicly to show rigor and consistency, while keeping implementation details and calibrations proprietary.

Coming soonPrism cert lookup

Get a Prism fair value for your slab

Add your email so we can notify you when cert valuations go live. Optionally include a PSA cert number you care about — same Prism engine as fund NAV, no valuation generated yet.

Built for reliability

Multiple market signals reduce single-source bias and improve valuation stability across both liquid and thinly traded cards.

Designed for transparency

Each valuation carries traceable confidence and freshness metadata so investors can judge precision vs directionality in context.

Connected to NAV

Per-card fair values roll up to gross portfolio value, then risk controls are applied to produce fund NAV used throughout the product.

Three-lens framework

Exact comps, grade extrapolation, and player market signal blended dynamically by confidence and structural priority.

Lens 1 — Exact grade comps

Anchors valuation to verified market transactions for the exact card and grade, with recency emphasis and data-quality filtering.

Primary anchor

Lens 2 — Grade extrapolation

Bridges valuation between exact sales by referencing nearby grades of the same card and applying controlled pass-through logic.

Continuity layer

Lens 3 — Player market signal

Adds directional context from the broader player market to keep illiquid assets from going stale between direct comps.

Directional context

Blending approach (high level):

The engine assigns dynamic weights to each lens using confidence and structural importance. As exact data improves, the model naturally leans harder on Lens 1; when markets are less liquid, supplemental lenses carry more of the signal.

Update cadence

How Prism stays current over time.

- Pulls fresh market signals on a recurring schedule.

- Applies data quality checks before values enter the model.

- Tracks staleness and confidence decay when data becomes sparse.

- Refreshes portfolio-level NAV from updated per-card estimates.

Risk controls and guardrails

Framework behavior under uncertainty.

- Outlier handling to avoid one-off prints dominating valuation.

- Confidence caps for indirect signals relative to direct comps.

- Fallback behavior for low-liquidity or sparse-data assets.

- Explicit uncertainty signaling instead of fabricated precision.

What we disclose vs keep proprietary

Public model view

Framework structure, signal categories, confidence philosophy, and how values roll into NAV.

Proprietary layer

Exact feature weights, threshold parameters, platform normalization curves, and internal calibration logic.