Trove × Pudgy Penguins
Trove is partnering with Pudgy Penguins to bring Pudgy Floor Perpetuals on-chain.A cash-settled perp that tracks the Pudgy Penguins floor via Trove’s oracle framework.
The product is designed around a single principle:
thin underlyings demand strong reference pricing + conservative risk logic.
What You’re Trading
Pudgy perps are cash-settled and track a Pudgy Floor Index.
You are not trading a specific Pudgy token ID
No NFTs move, no NFT custody, no delivery mechanics
PnL is realized in the settlement asset based on the index move
You’re trading the collection floor index, not a single NFT.
Why NFT Perps Failed Before
Two years ago, “NFT perps” largely broke for predictable reasons:
Self-referential pricing: last trade / internal prints became “truth” in thin markets
Wick-driven liquidations: low liquidity created extreme prints → liquidation cascades
Bad debt events: weak reference pricing + aggressive parameters amplified risk
Trove’s Pudgy perp is built specifically to avoid those failure modes by using a tri-layer oracle plus a smoothed risk mark (ORACLE + MARK + SMA).
Oracle Design

Oracle = (Mark + Validator + EMA) / 3
The oracle price is formed by averaging three independent signals, each starting at ⅓ weight:
Mark (internal market signal)
Validator (external floor consensus)
EMA (trend + smoothing model)
This tri-layer design anchors pricing in external reality while remaining responsive to live market conditions.
Design Principles
External data grounds the oracle in real-world pricing. Internal signals ensure responsiveness to live market sentiment.

Oracle Components
1) Mark (last traded price on Trove)
Mark reflects real-time trader sentiment and executed price discovery on Trove.
Source: on-chain executions for the Pudgy perp market
Strength: responsiveness to live demand
Risk it avoids: ignoring actual flow and order-book reality
2) Validator (cross-market floor consensus)
Validator is the floor price aggregated across multiple NFT marketplaces.
This is the “reality anchor.” The goal is not to chase every listing, it’s to create a robust floor reference that isn’t dominated by any single venue.
Validator logic (high level):
pull floor listings across marketplaces
filter suspicious / stale / abnormal floors
aggregate into a cross-market floor value
Validator is not “random external prices.” It’s a cross-market floor.
3) EMA (HyperPS-style smoothing model)
EMA is an exponential moving average model (HyperPS-inspired) that:
smooths short-term noise
tracks trend
reduces impact of one-off spikes
Dynamic Weighting (Beta)
As more volume and deeper price discovery forms on Trove, the oracle can evolve:
if internal liquidity becomes strong, Mark/EMA weighting may increase (>33%)
if external markets become noisier, Validator weighting may decrease
weights are tested and may vary during the Trove beta to optimize robustness
Weights can change over time as liquidity and reliability change.
ORACLE MARK SMA (Risk Mark)
Pricing for trading is one thing. Liquidations and risk are another.
Trove uses ORACLE MARK SMA as the risk mark, a smoothed mark built from the oracle stream.
Why: thin markets produce wicks. If you liquidate off last trade, you invite manipulation and cascades.
So the system references a mark that is:
composite (3-signal oracle)
smoothed (SMA)
harder to game via single prints
Liquidations are based on ORACLE MARK SMA, not last-price wicks.
Funding
Funding is the mechanism that keeps the perp aligned with the oracle reference over time.
If perp trades above the reference, longs pay shorts
If perp trades below the reference, shorts pay longs
Funding prevents permanent drift while letting the market trade freely.
Market Integrity and Safety
Pudgy perps are designed with thin-underlying rules in mind:
robust oracle (Mark + Validator + EMA)
smoothed risk mark (ORACLE MARK SMA)
conservative market parameters during early liquidity
Typical safeguards include (parameterized per market):
max leverage caps
open interest caps
funding clamps
circuit breakers / oracle staleness rules
Last updated