Understand trait floors, premiums, and fair value
Trait price is not just rarity. PunkPredictor connects cohort floors, same-trait sales, floor at sale time, current-floor equivalents, active listings, and fair-value estimates so each trait can be interpreted in market context.
Last updated Jul 5, 2026. This public page shows visible 3Y and 2W trait premium samples; the 6M premium, full trait analytics, and exact comp rows live in the app.
Public trait sample
Recent CryptoPunks trait premiums
These rows use PunkPredictor's same-trait sale premium signal: median sale premium over the collection floor. The 3-year and 2-week windows are visible; the 6-month premium stays locked as the bridge into the full trait analytics workflow.
| Trait | 3Y premium | 6M premium | 2W premium | 6M sample |
|---|---|---|---|---|
 Beanie | +268.3% | Locked | +268.3% | 1 sale |
 Hoodie | +260.2% | Locked | +334.8% | 13 sales |
 Pilot Helmet | +105.1% | Locked | +207.7% | 5 sales |
 Cowboy Hat | +84.5% | Locked | +210.4% | 8 sales |
 3D Glasses | +67.6% | Locked | +72.4% | 8 sales |
 Tiara | +52.6% | Locked | +102.7% | 3 sales |
 Top Hat | +50.1% | Locked | +88.4% | 5 sales |
 Choker | +27.7% | Locked | +67.7% | 3 sales |
 VR | +25.7% | Locked | +51.6% | 15 sales |
 Wild White Hair | +25.5% | Locked | +67.7% | 10 sales |
 Orange Side | +24.5% | Locked | +51.6% | 4 sales |
 Buck Teeth | +21.8% | Locked | +37.6% | 3 sales |
 Cap Forward | +18.2% | Locked | +27.3% | 16 sales |
 Pink With Hat | +17.3% | Locked | +53.8% | 7 sales |
 Luxurious Beard | +17.1% | Locked | +31.3% | 29 sales |
 Shadow Beard | +17.0% | Locked | +31.4% | 38 sales |
 Pigtails | +16.8% | Locked | +142.2% | 7 sales |
 Wild Blonde | +14.9% | Locked | +13.1% | 9 sales |
Unlock the exact 6-month premium, same-trait sale rows, current-floor equivalents, trait combinations, and Punk-specific fair value.
How trait prices should be read
Trait floor
The cheapest active listing in a trait cohort gives a market floor, not a complete valuation.
Premium over floor
The spread between collection floor and trait-adjusted value shows how much the market pays for a cohort.
Comps and recency
Recent sales matter more when the market is moving or when a trait has thin liquidity.
Combinations
A trait can behave differently when paired with another scarce or visually important attribute.
What PunkPredictor trait analytics tracks
The app's trait layer is built around actual same-trait sale rows, premium-over-floor math, current floor context, and trait-family filters. That matters because a trait's value is usually a distribution, not a single number.
Same-trait sales
Sale rows show Punk ID, sale date, sale price, collection floor at the sale date, premium or discount versus floor, and today's equivalent price.
Premium over floor
Premium is interpreted as a ratio versus the market floor, then translated forward against the current floor so older sales remain comparable.
Trait families
Identity, skin, primary head, eyewear, eyes, facial hair, lipstick, mouth, smoking props, face marks, and jewelry are grouped separately.
Market depth
Active listing counts, trait floors, sales counts, and floor history help distinguish a real premium from one stale or isolated comp.
Traits that need model context
Alien, Ape, Zombie, Hoodie, Beanie, 3D Glasses, VR, Cowboy Hat, 0-trait, and other cohorts can all move differently from the broad floor. A good trait-price page should make those differences explicit instead of flattening every Punk into the cheapest listing.
How the app builds a useful trait comp set
- Start with the Punk's meaningful traits Trait-count cohorts, identity or skin, primary headwear or hair, eyewear, expressions, marks, and props are ordered before noisy rollups.
- Prefer combinations when there is enough data A Hoodie Punk with 3D Glasses should not be judged only by the Hoodie floor or only by 3D Glasses sales if the combined cohort has usable comps.
- Back off when the cohort is too thin If a precise combination has no sales, the app relaxes the selected traits toward broader but still relevant cohorts instead of showing a false answer.
- Normalize comps to the market regime A sale from a different floor environment is shown with its historical floor and current-floor equivalent, not just the raw ETH price.
Trait price signals worth separating
Rarity
How few Punks share a trait or trait-count cohort. Rarity creates optionality, but it does not guarantee liquidity.
Demand
Repeated sales, recent bids, active lender interest, and collector preference decide whether rarity actually clears at a premium.
Liquidity
Thin cohorts need wider error bars. One listed Punk or one stale sale should not define a whole trait's market.
Composition
Traits interact. A clean Beanie, a cluttered Beanie, and a Beanie paired with another premium trait can clear very differently.
Related valuation paths
Value estimator
Estimate a specific Punk with trait and market context.
Floor price
Use collection floor as baseline, not final answer.
Underpriced listings
Find listed Punks where ask price is below model fair value.
Lending
Use trait-adjusted value when thinking about collateral and LTV.
CryptoPunks trait price FAQ
What is a CryptoPunks trait premium?
A trait premium is the value difference between a Punk with a specific trait or cohort and the broader collection floor, adjusted by sales, listings, market depth, and comparable Punks.
How does PunkPredictor calculate premium over floor?
PunkPredictor compares same-trait sale prices with the collection floor at the sale date, then translates that ratio into a current-floor equivalent so old comps can be read in today's market context.
Are rare traits always more valuable?
No. Scarcity matters, but actual value also depends on demand, comps, visual desirability, trait combinations, and current market liquidity.
What trait signals does PunkPredictor track?
PunkPredictor tracks trait family, trait-count cohort, same-trait sales, floor at sale time, premium over floor, current-floor equivalent price, active listing depth, trait floor history, and sales-count trends where available.