Crypto Exchange Landscape 2026
Comprehensive research report — strategic intelligence for building a crypto exchange
Exchanges Analyzed
30+
CEX, DEX & Hybrid
Top 10 Perps Volume
$165B
Combined 24h derivatives
On-Chain Derivatives
$21.9B
24h total (80+ chains)
Open Source Projects
40+
Matching engines, DEX protocols, tools
📋 Executive Summary
Key Findings
- Binance dominates with ~$60.6B/day in derivatives — 3x larger than #2 OKX ($26.6B/day)
- On-chain derivatives hit $21.9B/day — Hyperliquid alone does ~$6.9B/day, rivaling mid-tier CEXs
- Hybrid architecture is the recommended model for new entrants: off-chain matching + on-chain settlement
- Estimated $100M–$500M+ needed over 3–5 years to break into the top 10
- US regulated perps remain the largest untapped market opportunity
- MEXC leads in pair count (879 perps) with zero maker fees — aggressive strategy
- Rust and C++ dominate matching engine development; Java (exchange-core) achieves 5M ops/sec
- All top 10 CEXs achieve 9–10/10 trust scores on CoinGecko
- Team of 80–120 people recommended for launch, with 40–50 engineers
- Post-FTX era favors transparency: Proof of Reserves, insurance funds, non-custodial options
🏆 Tier 1 — Top 10 Exchanges
Data from CoinGecko API, Feb 19, 2026. BTC price ~$96,000.
| # | Exchange | Type | Founded | HQ | 24h Spot (BTC) | 24h Perps (BTC) | Perp Pairs | Trust | Maker/Taker | Notable |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Binance | CEX | 2017 | Cayman Islands | 109,363 | 631,191 | 617 + 26 futures | 10/10 | 0.10% / 0.10% | Largest exchange; BNB ecosystem; Launchpad; copy trading |
| 2 | Bybit | CEX | 2018 | BVI | 22,021 | 191,018 | 654 + 60 futures | 10/10 | 0.02% / 0.055% | 60M+ users; copy trading; Web3 focus |
| 3 | OKX | CEX | 2017 | Seychelles | 18,277 | 277,006 | 281 + 28 futures | 10/10 | 0.02% / 0.05% | Unified account; DEX aggregator; Web3 wallet |
| 4 | Gate | CEX | 2013 | Panama | 23,553 | 181,351 | 673 | 10/10 | 0.02% / 0.05% | 4,300+ cryptos; $9.5B reserves; 48M users |
| 5 | Bitget | CEX | 2018 | Seychelles | 15,646 | 110,044 | 645 | 10/10 | 0.02% / 0.06% | 100M+ users; copy trading pioneer; LALIGA |
| 6 | MEXC | CEX | 2018 | Seychelles | 31,146 | 144,044 | 879 | 9/10 | 0.00% / 0.02% | Zero maker fees; most perp pairs (879) |
| 7 | KuCoin | CEX | 2017 | Seychelles | 27,275 | 36,878 | 561 | 9/10 | 0.02% / 0.06% | 41M users; 1000+ tokens |
| 8 | Crypto.com | CEX | 2019 | Malta | 26,433 | 34,651 | 234 + 10 futures | 9/10 | 0.015% / 0.05% | CRO token; Visa card; UFC, F1 sponsorships |
| 9 | HTX (Huobi) | CEX | 2013 | Seychelles | 14,819 | 31,428 | 237 + 8 futures | 9/10 | 0.02% / 0.05% | Long history; deep liquidity |
| 10 | BingX | CEX | 2018 | BVI | 12,423 | 97,285 | 676 | 9/10 | 0.02% / 0.05% | Copy trading; Chelsea FC partner; 10M+ users |
Derivatives Volume Comparison (BTC/day)
📊 Tier 2 — Mid-Tier Exchanges
| Exchange | Type | Founded | HQ | 24h Perps (BTC) | Perp Pairs | Notable |
|---|---|---|---|---|---|---|
| BitMart | CEX | 2017 | Cayman Is. | 229,578 | 604 | Wide token selection |
| CoinW | CEX | 2017 | UAE | 215,750 | 155 | Dubai-based; growing fast |
| WEEX | CEX | 2018 | Singapore | 187,962 | 727 | High volume derivatives |
| XT.COM | CEX | 2018 | Seychelles | 95,551 | 629 + 13 futures | Fast-growing; aggressive listing |
| LBank | CEX | 2015 | BVI | 94,373 | 826 | NFA, MSB, AUSTRAC licenses; 9M+ users |
| WhiteBIT | CEX | 2018 | Lithuania | 93,545 | 290 | EU-focused; MiCA compliance |
| Phemex | CEX | 2019 | Samoa | 27,000 | 549 | 100% PoR; ex-Morgan Stanley team |
| Deribit | CEX | 2014 | Netherlands | 11,074 | 20 + 20 futures | Options leader; acquired by Coinbase |
| Bitfinex | CEX | 2012 | BVI | 2,729 | 83 | Tether affiliated; margin trading pioneer |
| Backpack | CEX | 2023 | BVI | N/A | N/A | Next-gen; Solana ecosystem; regulated |
🌱 Tier 3 — Emerging & DEX Exchanges
| Exchange | Type | Chain/Infra | 24h Perps (BTC) | Perp Pairs | OI (BTC) | Notable |
|---|---|---|---|---|---|---|
| Hyperliquid | DEX | Own L1 | 71,397 | 294 | 77,695 | Largest on-chain perps; CLOB; ~$6.9B/day |
| Aster | DEX | Multi-chain | 35,076 | 289 | 26,241 | New (2025); decentralized perps |
| edgeX | DEX | Own chain | 26,891 | 176 | 14,578 | Incubated by Amber Group; modular |
| Jupiter Perps | DEX | Solana | ~5,200 est. | 5+ majors | N/A | Solana DEX aggregator; LP vault model |
| GMX V2 | DEX | Arbitrum/Avalanche | ~2,100 est. | 40+ | N/A | GLP/GM pool model; oracle-based |
| Vertex | Hybrid | Arbitrum | ~1,050 est. | 40+ | N/A | Hybrid CLOB+AMM; cross-margin |
| Drift | DEX | Solana | ~845 est. | 30+ | N/A | Solana-native; DLOB hybrid |
| dYdX V4 | DEX | Own L1 (Cosmos) | N/A | 100+ | N/A | Pioneer; own appchain; governance token |
| Gains Network | DEX | Arb/Base/Polygon | ~220 est. | 150+ (multi-asset) | N/A | Synthetic leverage; crypto, forex, stocks |
| Paradex | DEX | Starknet | Active | 30+ | N/A | Starknet-based; CLOB |
⛓️ On-Chain Derivatives Volume
Source: DefiLlama, February 2026
24h Volume
$21.9B
30d Volume
$988.8B
1y Volume
$8.93T
All-Time
$14.2T
Active chains: 80+ (Hyperliquid L1, Arbitrum, Solana, Base, Ethereum, Starknet, Sui, etc.)
🏗️ Architecture Comparison
| Dimension | CEX | DEX (AMM) | DEX (CLOB) | Hybrid |
|---|---|---|---|---|
| Order Matching | CLOB on proprietary servers | AMM pools | On-chain order book | Off-chain CLOB, on-chain settle |
| Settlement | Internal ledger (custodial) | On-chain atomic | On-chain | Periodic on-chain |
| Latency | <1ms – 10ms | 200ms – 12s | 200ms – 2s | 10–100ms match; block settle |
| Throughput | 100K – 1M+ ops/sec | Chain TPS limited (15–65K) | Chain TPS limited | Millions/sec match; chain settle |
| Custody | Exchange holds funds | User retains (smart contracts) | User custody | Smart contract deposits |
| MEV Risk | None (private book) | HIGH — sandwich attacks | Moderate | Low |
| KYC/AML | Required | Usually none | Usually none | Varies — gateway level |
| Liquidation | Centralized; instant | Keeper-dependent | Validator-dependent | Hybrid; fast |
| Funding Rates | Exchange-calculated; 8h | Smart contract; variable | On-chain calculation | Off-chain calc; on-chain enforce |
| Margin System | Cross/Isolated; real-time | Pool-based / account-based | On-chain margin accounts | Off-chain risk; on-chain collateral |
| Insurance | Centralized fund ($1B+) | Protocol treasury / LP pool | On-chain insurance fund | Hybrid |
⚖️ Pros & Cons by Model
CEX Centralized Exchange
- Ultra-low latency (<1ms)
- Millions of orders/sec
- Familiar UX; fiat ramps
- No gas fees
- Advanced order types
- Regulatory clarity
- Custodial risk (FTX, Mt. Gox)
- Single point of failure
- Opaque operations
- Requires KYC
- Expensive compliance
DEX AMM Model (GMX, Jupiter)
- Non-custodial
- Transparent; all on-chain
- Permissionless
- DeFi composable
- Higher latency
- MEV / front-running
- Oracle dependency
- Gas costs
- LP impermanent loss
- Smart contract risk
DEX CLOB (dYdX, Hyperliquid)
- Non-custodial
- Real order book UX
- Better price discovery
- Lower slippage
- Near-CEX performance
- Requires high-perf chain
- Spam/DoS risk
- Throughput limited
- Complex to build
- Validator centralization
Hybrid Off-chain Match + On-chain Settle
- CEX-like speed + on-chain security
- Non-custodial
- Low latency matching
- Reduced MEV
- Complex order types
- Centralized sequencer
- Liveness risk
- Regulatory ambiguity
- Complex architecture
- Settlement delays
🛡️ MEV & Liquidation Mechanisms
MEV Exposure by Model
| Model | MEV Risk | Details |
|---|---|---|
| CEX | None | Private matching engine. Internal front-running by exchange possible but not MEV. |
| DEX (AMM) | HIGH | Sandwich attacks cost users $100M+/year on Ethereum. |
| DEX (CLOB) | Moderate | Reduced if chain uses fair ordering (FCFS). Hyperliquid's L1 controls ordering. |
| Hybrid | Low | Off-chain matcher eliminates MEV at matching layer. |
Liquidation Mechanisms
| Model | Detection | Execution | Speed |
|---|---|---|---|
| CEX | Centralized risk engine, real-time | Internal matching, partial liquidation | Near-instant |
| AMM DEX (GMX) | Keeper monitoring | Oracle price trigger, keeper bounty | Block-dependent |
| CLOB DEX (dYdX) | Validators | Liquidation orders on book | Block-dependent |
| Hybrid | Off-chain risk engine | Submit liquidation to chain | Fast detect, chain settle |
💻 Open Source Projects
Matching Engines
Matching Engines (14 projects)
| Project | Language | Stars | Maturity | Description | GitHub |
|---|---|---|---|---|---|
| exchange-core | Java | ★ 3.8K | Production-ready | Ultra-fast matching, LMAX Disruptor. 5M ops/sec, <1ms | ↗ GitHub |
| CppTrader | C++ | ★ 1.5K | Library | High-perf trading platform components | ↗ GitHub |
| dingir-exchange | Rust | ★ 600 | Prototype | High-performance crypto trading engine | ↗ GitHub |
| gitbitex-spot | Go | ★ 500 | Prototype | Open source crypto exchange (Coinbase-like) | ↗ GitHub |
| gitbitex-new | Java | ★ 400 | Active Dev | Rewrite of gitbitex, active development | ↗ GitHub |
| trading_engine | Go | ★ 400 | Active Dev | Financial securities trading system | ↗ GitHub |
| order-matcher | C# | ★ 300 | Library | Feature-rich: limit, market, stop-loss, iceberg, IOC, FOK | ↗ GitHub |
| orderbook-rs | Rust | ★ 300 | Reference | Basic order matching engine | ↗ GitHub |
| limit-order-book | C++/Python | ★ 300 | Library | C++ LOB with Python bindings | ↗ GitHub |
| LightMatchingEngine | Python | ★ 250 | Educational | Lightweight matching engine | ↗ GitHub |
| matching-core | Rust | ★ 200 | Active Dev | High-perf core library, multiple order types | ↗ GitHub |
| nodejs-order-book | TypeScript | ★ 150 | Active Dev | Ultra-fast LOB for Node.js/HFT | ↗ GitHub |
| open-outcry | Go | ★ 100 | Active Dev | Multi-asset matching engine | ↗ GitHub |
| DistributedATS | C++ | ★ 100 | Niche | FIX Protocol multi-matching engine over DDS | ↗ GitHub |
Full Exchange Platforms
| Project | Language | Stars | License | Description | GitHub |
|---|---|---|---|---|---|
| OpenDAX | Ruby/Docker | ★ 1.2K | Apache 2.0 | Cloud-native full exchange platform (spot) | ↗ GitHub |
| Peatio | Ruby | ★ 3K+ | MIT | Core exchange engine behind OpenDAX (archived) | ↗ GitHub |
DEX Protocols (On-Chain Derivatives)
| Project | Language | Stars | License | Description | Volume | GitHub |
|---|---|---|---|---|---|---|
| dYdX V4 | Go/Cosmos | ★ 500 | AGPL-3.0 | Full L1 appchain for derivatives CLOB | Was top 3 DEX | ↗ GitHub |
| GMX | Solidity | ★ 400 | MIT | Perps on Arbitrum/Avalanche; GLP model | ~$200M/day | ↗ GitHub |
| Drift Protocol | Rust (Anchor) | ★ 400 | Apache 2.0 | Solana perps/spot with DLOB | ~$81M/day | ↗ GitHub |
| Hyperliquid | Rust | — | Proprietary | Own L1 with on-chain CLOB; highest volume | ~$6.9B/day | Closed source |
| Vertex | Solidity | ★ 100 | BSL 1.1 | Hybrid CLOB+AMM on Arbitrum | ~$100M/day | ↗ GitHub |
| Gains Network | Solidity | ★ 200 | MIT | Synthetic perps multi-asset | ~$21M/day | ↗ GitHub |
| Synthetix V3 | Solidity | ★ 300 | MIT | Liquidity layer for derivatives | Infrastructure | ↗ GitHub |
| Perpetual Protocol | Solidity | ★ 200 | GPL-3.0 | vAMM-based perps on Optimism | Declining | ↗ GitHub |
| OpenBook | Rust | ★ 200 | Apache 2.0 | Solana on-chain CLOB (Serum successor) | Active | ↗ GitHub |
| Injective | Go | ★ 200 | Apache 2.0 | Cosmos-based L1 for derivatives | Active | ↗ GitHub |
Trading Infrastructure & APIs
| Project | Language | Stars | License | Description | GitHub |
|---|---|---|---|---|---|
| CCXT | JS/Python/PHP/C#/Go | ★ 35K | MIT | Unified API for 100+ exchanges. Essential tool. | ↗ GitHub |
| Freqtrade | Python | ★ 30K | GPL-3.0 | Crypto trading bot; backtesting; ML strategies | ↗ GitHub |
| Hummingbot | Python | ★ 8K+ | Apache 2.0 | Market-making bot; 140+ venues; $34B+ vol generated | ↗ GitHub |
| Jesse | Python | ★ 5K | MIT | Algo trading framework with backtesting | ↗ GitHub |
| Catalyst | Python | ★ 2.5K | Apache 2.0 | Algorithmic trading library for crypto | ↗ GitHub |
| Tribeca | Rust | ★ 150 | MIT | Solana market making | ↗ GitHub |
Bridge / Cross-Chain
⚙️ Technology Stacks
What Top Exchanges Use
| Exchange | Core Languages | Database | Message Queue | Performance |
|---|---|---|---|---|
| Binance | C++ Java Go | MySQL, Redis, custom | Kafka, custom | 1.4M orders/sec (2019) |
| Coinbase | Go Ruby Python | PostgreSQL, DynamoDB | Kafka, SQS | ~100K orders/sec |
| Bybit | Java Go | MySQL, Redis | Kafka | Handles $18B+/day derivs |
| OKX | Java C++ | MySQL, Redis, RocksDB | Kafka | Unified account system |
| Kraken | Rust C++ Python | PostgreSQL, Redis | Kafka | <10ms matching |
| Hyperliquid | Rust | Custom (in-memory) | L1 consensus | ~20K TPS; sub-second finality |
| dYdX V4 | Go (Cosmos) | In-memory order book | Tendermint consensus | ~2K TPS; sub-second finality |
Common Patterns
🔧 Languages
- Matching: C++, Rust, Java
- Backend: Go, Java, Python, Node.js
- Frontend: React, Next.js, TypeScript
- Contracts: Solidity, Rust, Move
🗄️ Databases
- Hot: Redis, Memcached
- Warm: MySQL/PostgreSQL
- Cold: S3, Parquet, TimescaleDB
- Time-series: InfluxDB, TimescaleDB
📨 Message Queues
- Dominant: Apache Kafka
- CN exchanges: RocketMQ
- In-process: LMAX Disruptor
- Lightweight: NATS, ZeroMQ
🏗️ Infrastructure
- Matching: Bare metal, CPU pinning
- API: Cloud (AWS/GCP)
- Network: DPDK, io_uring
- Pattern: Active-passive + event sourcing
Performance Benchmarks
| System | Throughput | Median Latency | P99 Latency | Notes |
|---|---|---|---|---|
| exchange-core (Java) | 5M ops/sec | 0.5µs | 4µs @ 125K/s | Single order book |
| LMAX Disruptor | 6M events/sec | ~1µs | ~10µs | Ring buffer pattern |
| Binance (claimed) | 1.4M orders/sec | <1ms | N/A | Production (2019) |
| Typical CEX | 100K–500K/sec | 1–5ms | 10–50ms | Industry standard |
| Hyperliquid L1 | ~20K TPS | ~200ms | ~500ms | Blockchain consensus |
| Solana | ~65K TPS (theory) | ~400ms | ~2s | Network-wide |
| dYdX V4 | ~2K TPS | ~1s | ~3s | Cosmos chain |
🎯 Strategic Analysis — Building a Top 10 Exchange
Capital Requirements
| Phase | Capital | Purpose |
|---|---|---|
| Seed / MVP | $2M – $5M | Core team (10–15 engineers); MVP matching engine; basic compliance |
| Launch | $10M – $30M | Licenses; security audits; market making seed; marketing |
| Growth | $50M – $200M | Liquidity provision; institutional onboarding; global expansion |
| Top 10 Scale | $200M – $1B+ | Brand building; sports sponsorships; custody infra; global compliance |
Total estimated: $100M–$500M+ over 3–5 years to reach top 10.
Regulatory Requirements by Jurisdiction
Jurisdiction Licensing Details (10 jurisdictions)
| Jurisdiction | License Type | Timeline | Capital Req. | Key Regulator | Notes |
|---|---|---|---|---|---|
| 🇺🇸 United States | MSB + MTLs + SEC/CFTC | 12–36 months | $5M–$50M+ | FinCEN, SEC, CFTC | Most complex; BitLicense (NY); state-by-state |
| 🇪🇺 EU (MiCA) | CASP License | 6–18 months | €125K–€350K | MFSA, BaFin, etc. | Unified framework from 2024 |
| 🇸🇬 Singapore | MPI License | 6–12 months | S$250K+ | MAS | Well-regarded; strict AML/CFT |
| 🇦🇪 Dubai | VARA License | 3–9 months | $1M–$5M+ | VARA / ADGM | Fast-growing hub; favorable tax |
| 🇭🇰 Hong Kong | VATP License | 6–18 months | HK$5M+ | SFC | Retail allowed; insurance mandatory |
| 🇯🇵 Japan | JFSA Registration | 12–24 months | ¥10M+ | JFSA | Strict; max 2x leverage |
| 🇧🇸 Bahamas | DARE License | 3–6 months | $100K+ | SCB | FTX was here; reputation damaged |
| 🇰🇾 Cayman Islands | VASP Registration | 3–6 months | Varies | CIMA | Popular for offshore structures |
| 🇻🇬 BVI | Crypto License | 2–4 months | ~$50K | FSC | Many exchanges domiciled here |
| 🇱🇹 Lithuania | VASP + MiCA | 3–6 months | €125K | Bank of Lithuania | EU passporting |
Competitive Moats
| Moat | Description | How to Compete |
|---|---|---|
| Liquidity | Self-reinforcing: more liquidity → better prices → more traders | Aggressive MM; incentive programs; maker rebates |
| Brand Trust | Years of operating without hacks | Security-first; transparent PoR; insurance fund |
| Licenses | Multi-year head start on compliance | Start licensing early; hire ex-regulator advisors |
| Token Ecosystem | BNB, OKB create lock-in | Launch token with real utility |
| User Base | 40–100M+ users | Target underserved markets; superior UX |
| Infrastructure | Years of engineering investment | Build on open-source; focus on specific edge |
| Product Breadth | Spot, perps, options, earn, NFT, etc. | Start focused (perps); expand |
Market Gaps & Opportunities (2026)
- Regulated US Derivatives: No major exchange offers full perps to US retail. Massive TAM.
- On-Chain + Off-Chain Hybrid: Best of both worlds not yet delivered at scale.
- RWA Trading: Tokenized stocks, bonds, commodities alongside crypto.
- Institutional-Grade DeFi: Compliant DeFi access with permissioned KYC pools.
- Multi-Chain Unified: Trade across 50+ chains from one account.
- AI-Native Trading: Integrated AI for strategy, risk management, portfolio optimization.
- Emerging Markets: Africa, Southeast Asia, LATAM still underserved.
Recommended Model: Hybrid Architecture
🏆 Strong Recommendation: Hybrid
- Off-chain matching engine (Rust/C++, LMAX Disruptor pattern)
- On-chain settlement on fast L2 or L1 (own rollup, Arbitrum/Base)
- Non-custodial — users deposit to smart contracts, trade off-chain, settle on-chain
- Regulatory compliance built in — KYC at onboarding
- Multi-chain support — accept deposits from any chain via bridges
Why it wins: Avoids CEX trust issues (post-FTX), achieves CEX-like performance, "trustless" marketing appeal, composable with DeFi.
Liquidity Bootstrapping
- Market Maker Partnerships: Wintermute, GSR, Jump — fee rebates, dedicated API
- Maker Fee Rebates: Pay makers (negative fees) initially
- Liquidity Mining: Distribute exchange tokens to LPs
- Cross-exchange Aggregation: Route orders to other exchanges initially
- Institutional API: Low-latency co-location, FIX protocol, prime brokerage
- Copy Trading: Attract retail (Bitget/BingX model)
Team Composition (80–120 at launch)
| Function | Headcount | Key Roles |
|---|---|---|
| Engineering | 40–50 | Matching engine (Rust/C++), Backend (Go/Java), Frontend (React), Blockchain, DevOps, Security |
| Product | 5–8 | PMs, UX designers, data analysts |
| Compliance/Legal | 8–12 | CCO, AML analysts, legal counsel, licensing specialists |
| Operations | 10–15 | Support, KYC ops, treasury, market ops |
| Business Dev | 5–8 | MM relations, institutional sales, token listing |
| Marketing | 8–12 | Growth, community, social, content, paid |
| Finance | 3–5 | CFO, accountants, treasury |
| Leadership | 3–5 | CEO, CTO, COO, Chief Risk Officer |
📈 Perpetual Futures Deep Dive
How Perps Work
💰 Funding Rate
Keeps perp price anchored to spot. Rate = (Mark - Index) / Index + Interest. Longs pay shorts when positive; shorts pay longs when negative. Typically every 8h (CEX) or 1h (Hyperliquid).
📊 Mark vs Index Price
Index: VWAP from multiple exchanges. Mark: Fair price for liquidations (index + decaying basis). Prevents manipulation — liquidations use mark price.
🔐 Margin System
Initial: Required to open (e.g., 10% for 10x). Maintenance: Min to keep open (0.5–5%). Cross: Entire balance as collateral. Isolated: Only allocated margin at risk. Portfolio: Hedged offsets.
⚡ ADL (Auto-Deleveraging)
When insurance fund depleted, most profitable opposing positions auto-reduced. Ranked by profit × leverage. Unpopular but necessary. Binance: 1–5 light indicator.
Implementation Comparison
| Feature | Binance | Bybit | OKX | Bitget | Hyperliquid |
|---|---|---|---|---|---|
| Max Leverage | 125x | 200x | 125x | 125x | 50x |
| Funding Interval | 8h | 8h | 8h | 8h | 1h |
| Margin Modes | Cross/Iso/Portfolio | Cross/Iso/Portfolio | Cross/Iso/Portfolio (unified) | Cross/Isolated | Cross/Isolated |
| Perp Pairs | 617 | 654 | 281 | 645 | 294 |
| Collateral | USDT, USDC, coin | USDT, USDC, coin | USDT, USDC, coin (unified) | USDT, USDC | USDC |
| Insurance Fund | $1B+ (SAFU) | ~$300M | ~$500M | ~$300M | Protocol treasury |
| Open Interest | 289,693 BTC (~$27.8B) | 146,160 BTC (~$14B) | 93,866 BTC (~$9B) | 85,277 BTC (~$8.2B) | 77,695 BTC (~$7.5B) |
On-Chain Perps Protocols
Detailed Protocol Breakdowns (7 protocols)
1. Hyperliquid — DEX
Custom L1 (HyperBFT, Rust). Full on-chain CLOB. ~$6.9B/day, 294 pairs, 77,695 BTC OI. Sub-second finality. No gas fees. HYPE token for governance + staking + fee sharing. Hourly funding. Own L1 eliminates MEV.
2. dYdX V4 — DEX
Cosmos SDK appchain. Validators maintain off-chain order book. V3 total: $1.2T all-time. Fully decentralized governance. MegaVault for protocol-owned liquidity. Declining volume post-migration; Hyperliquid competition.
3. GMX V2 — DEX
Arbitrum/Avalanche. Pool-based — GM pools + Chainlink oracles. ~$200M/day. No order book; no slippage for small trades. V2 introduced per-market pools (isolated risk). LP risk: oracle dependency, adverse selection.
4. Drift Protocol — DEX
Solana (Anchor). Hybrid DLOB + AMM backstop. ~$81M/day perps, $145.6B all-time. JIT liquidity — makers provide at execution. Cross-margin with spot + perps + borrow/lend.
5. Vertex Protocol — Hybrid
Arbitrum. Off-chain sequencer (5–15ms matching) + on-chain settlement. Cross-margin unified account. Vertex Edge: cross-chain orderbook. Direct competitor to dYdX and Hyperliquid.
6. Jupiter Perps — DEX
Solana. JLP pool model (similar to GMX). ~$500M/day estimated. Built by largest Solana DEX aggregator. Limited pairs (BTC, ETH, SOL). Pyth oracle pricing.
7. Gains Network (gTrade) — DEX
Solidity on Arbitrum/Base/Polygon/MegaETH. Synthetic trades against gDAI vault. ~$21M/day, $28.3B/year. Multi-asset: crypto, forex, stocks (150+ pairs). Up to 1000x leverage on forex.
Insurance, Socialized Loss & ADL
| Mechanism | CEX | DEX | Details |
|---|---|---|---|
| Insurance Fund | Binance SAFU >$1B; exchanges contribute fees | Protocol treasury or smart contract (transparent) | Absorbs losses from bankrupt positions |
| Socialized Loss | Historically used (BitMEX, OKX); rare now | Some protocols distribute to LPs | Losses spread across profitable traders |
| ADL | Tiered indicator; rarely triggered at top CEXs | On-chain mechanism (Hyperliquid); GMX: none (LPs absorb) | Most profitable opposing positions auto-reduced |