the perpification of everything

published: December 26, 2025

updated: May 2, 2026

"because perpetuals do not require securing an underlying asset, markets can form around virtually anything." – coinbase ventures, 2026 investment outlook^[1]

if you've been following prediction markets and perpetual DEXs separately, you might have missed the most important trend of 2025: they're becoming the same thing.

prediction markets hit $44 billion+ in volume this year — nearly 3x year-over-year (vs ~$15.8B in 2024). perp DEX volumes peaked at $1.36 trillion in october 2025, on track for ~$7.9 trillion annual. but what's really interesting isn't the growth in isolation. it's the convergence.

the line between "will BTC hit $100k by december?" and a "BTC perpetual future" is blurring fast. both provide synthetic exposure. both use order books, margin systems, and settlement infrastructure. the distinction is increasingly artificial, and builders are starting to notice.

from binary bets to continuous exposure

the traditional model is breaking down

classic prediction markets work like this: you buy shares of a "yes" or "no" outcome. each share costs between $0 and $1, reflecting market-implied probability. if your outcome hits, you get $1 per share. if it doesn't, you get $0. simple, binary, zero-sum.

this model has serious limitations:

capital lock-up: your money sits idle until settlement. no yield, no composability.
limited exit options: you're betting on a specific outcome at a specific date. want out early? hope there's liquidity.
binary outcomes only: the world isn't yes/no. most traders want continuous exposure to price movements, not discrete events.
no leverage or margin: traditional prediction markets don't let you amplify positions or manage risk like derivatives.

now compare this to perp derivatives: continuous trading, early exits, leverage, yield on collateral, funding rates. the UX is fundamentally different. but it doesn't have to be.

what's actually happening

the technical evolution is already underway. polymarket abandoned AMMs in favor of a CLOB model in 2022 — the same infrastructure powering perp DEXs. and HIP-4 event futures^[2] (also called "event perpetuals" in secondary coverage — terms used interchangeably below), a community proposal targeting hyperliquid, spell out exactly what continuous trading with event-based settlement looks like.

here's what the convergence looks like in practice.

infrastructure overlap:

order books with maker/taker fees
margin systems and liquidation mechanisms
settlement infrastructure (just different trigger conditions)
risk engines for position sizing and portfolio management

new features emerging:

composable collateral: earn yield on deposited USDC while holding prediction positions
continuous liquidity: exit positions anytime, not just at settlement
funding rates: long-duration prediction markets testing funding mechanisms similar to perps
leverage: margin trading on prediction outcomes (still early, but technically feasible)

the RWA perpetuals connection

why coinbase ventures is all in

their thesis: perpetual futures can create synthetic exposure to literally anything without requiring custody of the underlying asset. private companies. macroeconomic data. athlete earnings. oil prices. inflation expectations.^[1]

sound familiar? that's exactly what prediction markets do.

the question "will SPY reach $600 by december 2025?" is functionally a synthetic SPY derivative with defined parameters. the only difference from a traditional perp is the settlement mechanism.

real numbers: injective's $6B experiment

injective isn't hypothesizing — they're already doing this at scale. their RWA perpetual markets have processed over $6 billion in volume^[3] across:

$2.4B in magnificent 7 equities (AAPL, MSFT, GOOGL, etc.)
pre-IPO exposure to OpenAI, SpaceX, Anthropic
$363M in treasury perpetuals (led by MicroStrategy's $313M)
24/7 trading on helix^[4] when traditional markets are closed

the key insight: these aren't "tokenized stocks." they're perpetual derivatives providing synthetic exposure. no custody required. no securities regulations (debatable, but that's the thesis). just price discovery through continuous trading.

prediction markets do the exact same thing, just with binary outcomes instead of continuous price feeds. that distinction is arbitrary, and it's disappearing.

the 24/7 problem prediction markets already solved

RWA perpetuals face a thorny question: what happens when traditional markets close? if you're offering 24/7 trading on SPY, where does the price come from at 3am EST when the NYSE is closed?

prediction markets have been dealing with this since day one. their solution:

pure price discovery: let the market set probabilities based on order flow, not external oracles
bounded movement: probabilities can't jump from 30% to 90% instantly without significant capital
liquidity-based pricing: market makers provide continuous quotes even when underlying reference data is stale

technical deep dive: what changes under the hood

order book architecture

traditional prediction markets used AMMs. you'd swap USDC for yes/no tokens through a constant product curve. simple, but terrible for capital efficiency.

modern prediction markets use CLOBs (central limit order books), just like perp DEXs:

polymarket: CLOB with on-chain settlement on Polygon
hyperliquid: the HIP-4 community proposal^[2] layers event perpetuals on top of their L1 order book

the data backs this up. kalshi's open-interest-to-volume ratio is 0.29 vs polymarket's 0.38 — a lower ratio means higher turnover relative to held positions. that's the perp DEX pattern (active trading), not the traditional prediction market pattern (bet and hold).

settlement mechanisms: the last remaining difference

right now, the main technical difference between prediction markets and perpetuals is settlement.

prediction markets:

binary settlement: YES shares → $1, NO shares → $0
oracle-based verification (UMA, Chainlink, etc.)
one-time settlement at event resolution

perpetuals:

continuous mark-to-market pricing
funding rates to keep synthetic price anchored to spot
never settle

but even this is converging. the HIP-4 proposal^[2] is continuous trading + one-time settlement at expiration. that's literally a prediction market with a perp UX.

you could implement the same thing in reverse: prediction markets with funding rates and continuous settlement windows. the infrastructure is the same. it's just a product design choice.

composability: the missing piece (until now)

coinbase ventures highlighted "composable perpetual markets" as a major 2026 opportunity. the idea: let traders earn yield on collateral while maintaining leveraged positions.^[1]

prediction markets are starting to do this:

collateral yield: deposit USDC into a lending protocol, use the receipt token as collateral for prediction positions
early exit mechanics: CLOB liquidity lets you sell positions before settlement (unlike traditional binary bets)
cross-margin: use the same collateral across multiple prediction markets (early)

the UX gap is closing. soon you won't think "should i trade perps or predictions?" — you'll think "what's the best synthetic exposure for this view?" and the platform will handle the rest.

what this means for builders

the opportunity: unified synthetic exposure platforms

if you're building a perp DEX, you're 80% of the way to a prediction market. if you're building a prediction market, you're 80% of the way to a perp DEX. the infrastructure overlaps almost completely.

here's what's missing from current platforms.

from perp DEXs:

event-based contracts (binary outcomes, specific dates)
oracle infrastructure for subjective settlement
natural-language market creation ("will X happen by Y?" instead of ticker symbols)

from prediction markets:

continuous price exposure (not just binary yes/no)
leverage and margin trading
yield on collateral
24/7 operation with robust price discovery

technical checklist for convergence

if you're building in this space, here's what you need:

unified order book: a CLOB that handles both continuous and binary markets
flexible settlement: support both perpetual funding rates and event-based oracle settlement
oracle infrastructure: price feeds for continuous markets, truth verification for binary events
margin system: cross-margin across both market types with proper risk calculations
composable collateral: let users earn yield while holding positions (via lending protocol integration)
market creation UX: support both ticker-based (BTC-PERP) and natural language ("will BTC hit $100k?") market creation

the market dynamics are already shifting

volume patterns tell the story

look at the numbers:

perpetual DEX volumes: $1.36T peak monthly (october 2025), ~$7.9T annual
prediction market volumes: $44B+ in 2025 (polymarket ~$21.5B + kalshi ~$17.1B + others) — nearly 3x year-over-year
RWA perpetuals (injective): $6B+ in 2025
kalshi (event contracts): $1.3B monthly as of september 2025

the gap is closing. RWA perpetuals are prediction markets in disguise. event contracts are perpetuals with binary outcomes. the categories are merging.

user behavior: trading vs holding

kalshi's 0.29 OI/volume ratio vs polymarket's 0.38 is telling. lower ratio = more trading activity. users are treating prediction markets like trading venues, not just outcome speculation.

this matches the perp DEX pattern:

frequent position changes
active management, not buy-and-hold
liquidity-seeking behavior (concentrated in high-volume markets)
early exits when probabilities move

the liquidity paradox and how perps solve it

prediction markets have a brutal liquidity problem. unlike crypto liquidity pools where both sides retain value, prediction contracts expire worthless for losers. no rebalancing, no salvage value.

polymarket spent ~$10M on market-making incentives (at one point $50k/day). kalshi spent $9M+. do you think that's sustainable?

perp DEXs solved this with:

funding rates: market makers get paid for providing liquidity
no expiry: positions don't go to zero, they just mark-to-market continuously
yield on collateral: LPs earn returns even in sideways markets

prediction markets adopting these mechanisms could fix their liquidity problem. and they're starting to: composable collateral, continuous trading, early exit options. the perpification is happening because it has to.

why this matters: the bigger picture

synthetic exposure is eating finance

we're seeing a fundamental shift: from tokenizing assets (hard, regulatory nightmare) to creating synthetic exposure (easier, more flexible).

tokenized stocks: need custody, securities licenses, KYC/AML, limited to accredited investors
synthetic perpetuals: no custody, arguably derivatives not securities, global access, 24/7 trading

perpetuals and prediction markets are both implementing this same vision. that's why they're converging — they're solving the same problem from different angles.

the total addressable market

think about what becomes tradable:

private companies: trade OpenAI, SpaceX, Stripe exposure before IPO (injective already doing this)
macro data: GDP prints, CPI, unemployment — all tradable before official release
political events: policy decisions, elections, regulatory outcomes
weather: temperature, rainfall, hurricanes (already active on kalshi)
sports: not just game outcomes, but player stats, season performance, MVP odds
cultural moments: awards, releases, announcements, trending topics

coinbase ventures put it simply: "markets can form around virtually anything."^[1]

the platform that makes this easiest — unified infrastructure for continuous and binary exposure — wins massive market share. you're not competing in "prediction markets" or "perp DEXs." you're competing in synthetic exposure infrastructure.

the first-mover window is closing

right now, the categories are still separate in people's minds. polymarket is "prediction markets." dYdX is "perp DEX." but builders are starting to see the convergence:

hyperliquid's community-proposed event perpetuals^[2]
injective launched RWA perps that are basically continuous prediction markets
limitless focused on pre-TGE hedging (prediction market meets perp DEX)

in a year, most major platforms will offer both. the question is who gets there first with the best UX, deepest liquidity, and most robust infrastructure.

conclusion: everything is becoming a perpetual

the perpification of prediction markets isn't a trend — it's an inevitability. the technical infrastructure is converging. the use cases overlap. user behavior is shifting from "bet and hold" to "trade actively."

for builders:

if you're building a perp DEX, you're 80% done with a prediction market
if you're building a prediction market, you're 80% done with a perp DEX
the winner will be whoever realizes this first and builds unified synthetic exposure infrastructure

for traders: stop thinking "should i use a prediction market or a perp DEX?". start thinking "what synthetic exposure do i want?". the platform that makes that easiest — regardless of whether the underlying mechanism is binary or continuous — will capture the most value.

for the crypto ecosystem: this convergence unlocks the perpification of everything. any asset, any event, any outcome will be tradable 24/7 with leverage, yield, and composability.

the future isn't prediction markets OR perpetual futures. it's synthetic exposure infrastructure that handles both — and everything in between.

the perpification isn't coming. it's already here.

references

Coinbase Ventures. 2026 Crypto Outlook. Cited at ↩a, ↩b, ↩c, ↩d.
Bedlam Research. HIP-4 Event Futures. bedlamresear.ch/posts/hip-4-event-futures. Cited at ↩a, ↩b, ↩c, ↩d.
Messari. Injective: Building the Infrastructure Layer for Onchain RWA Derivatives. messari.io/report/injective-building-the-infrastructure-layer-for-onchain-rwa-derivatives. Cited at ↩.
Helix futures (TSLA-USDT-PERP). helixapp.com/futures/tsla-usdt-perp. Cited at ↩.