Arbitrage isn't one strategy — it's a family of approaches. Each type works differently, requires different capital and technical knowledge, and carries different risks. Most people picture one thing when they hear "crypto arbitrage," but in practice there are at least five meaningfully distinct methods.
This article covers each type in detail: mechanics, real examples, pros and cons, and who each one is best suited for.
Overview: All Arbitrage Types at a Glance
| Type | Where | Transfer needed? | Speed | Complexity |
|---|---|---|---|---|
| Cross-exchange (spatial) | 2 different exchanges | Sometimes | Minutes | Low |
| Balance arbitrage | 2 different exchanges | No | Seconds | Medium |
| Triangular | 1 exchange | No | Seconds | Medium |
| P2P arbitrage | P2P platform + spot | No | Minutes | Low |
| CEX-DEX | Exchange + DeFi | Yes | Minutes | High |
| Statistical | Multiple exchanges | Sometimes | Hours | High |
Type 1: Cross-Exchange (Spatial) Arbitrage
How it works
The most intuitive type. The same coin trades at different prices on two different exchanges — you buy where it's cheaper and sell where it's more expensive.
Example:
- Binance: SOL/USDT = $142.50
- KuCoin: SOL/USDT = $143.80
- Spread: $1.30 (0.91%)
Buy SOL on Binance, transfer to KuCoin, sell. After trading fees (0.20%) and the Solana network fee (~$0.01): net profit approximately 0.70%.
Two execution approaches
With coin transfer — the classic approach. Buy on Exchange A, withdraw to Exchange B, sell. Main risk: the price on Exchange B may move while your transfer is in flight.
Without transfer (balance arbitrage) — the professional approach. Hold USDT on Exchange A and SOL on Exchange B simultaneously. When you spot a spread, instantly buy SOL on A and sell SOL on B. Both trades execute in seconds, no transfer needed. You rebalance the positions afterward at your convenience.
Pros and cons
✅ Simple logic — easy to understand and start
✅ Many opportunities — 600+ pairs across 17 exchanges in SpreadScan
✅ Fully market-direction neutral
❌ With coin transfer: price risk during the transfer window
❌ Requires accounts and capital on multiple exchanges
❌ Flat withdrawal fees can eat profits on small positions
Who it's best for
The best starting point for beginners. The logic is transparent, tools for finding opportunities are accessible (SpreadScan shows real-time spreads), and the entry barrier is low.
Type 2: Balance Arbitrage
Technically a variant of cross-exchange, but different enough in mechanics to cover separately.
How it works
You pre-position capital across multiple exchanges in two forms simultaneously: stablecoins (USDT) on some exchanges, target coins on others. When a spread appears, both sides of the trade execute instantly — no transfers anywhere.
Example capital distribution ($10,000 total):
- Binance: $2,500 USDT
- Bybit: $2,500 USDT
- Binance: 1 ETH (~$2,480)
- Bybit: 1 ETH (~$2,480)
You spot an ETH spread: Binance $2,480, Bybit $2,501.
- On Binance: buy ETH with USDT (USDT → ETH)
- On Bybit: sell ETH for USDT (ETH → USDT)
Both trades execute simultaneously in seconds. Profit is locked with no cross-exchange transfer.
Pros and cons
✅ No price risk from transfer time
✅ No withdrawal fees
✅ Very fast execution — seconds
✅ Profitable at smaller spreads (from 0.35%)
❌ Requires capital locked across multiple exchanges simultaneously
❌ Balances drift over time — periodic rebalancing needed
❌ Need to monitor multiple platforms at once
Who it's best for
Traders who understand basic cross-exchange arbitrage and want to eliminate execution risk. Requires more upfront capital ($5,000+), but gives much better trade control.
Type 3: Triangular Arbitrage
How it works
Triangular arbitrage is fundamentally different in nature: it exploits price misalignments between three trading pairs on a single exchange — not price differences across exchanges.
The loop: start with USDT, pass through two intermediate coins, return to USDT with more than you started.
Route: USDT → BTC → ETH → USDT
Why this works: In theory, if BTC/USDT = $83,000 and ETH/USDT = $2,490, then BTC/ETH should be exactly 33.33. If the actual BTC/ETH rate is slightly off — say, 33.45 — a triangular arbitrage opportunity exists.
Real example with numbers:
Starting capital: $10,000 USDT on Binance
| Step | Action | Rate | Result (before fee) |
|---|---|---|---|
| 1 | USDT → BTC | 1 BTC = $83,200 | 0.12019 BTC |
| 2 | BTC → ETH | 1 BTC = 33.45 ETH | 4.0184 ETH |
| 3 | ETH → USDT | 1 ETH = $2,500 | $10,046 USDT |
After three trading fees at 0.10% each (0.30% total ≈ $30):
Net profit: $16 (0.16%)
Small per cycle — but the cycle completes in seconds on a single exchange with zero transfer risk.
Why triangular opportunities exist
Exchange matching engines process orders independently for each pair. When a large trade hits one pair, the prices of three related pairs react with a brief lag. Triangular arbitrageurs exploit that lag.
Pros and cons
✅ Everything happens on one exchange — no transfer risk
✅ No withdrawal fees
✅ Independent of prices on other exchanges
✅ Can be automated as a single algorithmic unit
❌ Opportunities last seconds — manual execution is nearly impossible
❌ Three fees instead of two — higher minimum profitable spread
❌ Finding routes manually is difficult — requires a scanner
Who it's best for
Traders ready to move toward automation. Triangular arbitrage is an ideal first algorithmic trading task: clear logic, a closed loop, single exchange API. SpreadScan shows active triangular routes in real time.
Type 4: P2P Arbitrage
How it works
P2P arbitrage exploits the difference between exchange rates on P2P platforms (Binance P2P, Bybit P2P, HTX P2P) and the spot market price.
On P2P platforms, regular users post offers at rates they set themselves. These rates often diverge from the market price — especially in non-major fiat currencies (RUB, KZT, AED, BRL) or during periods of high regional demand.
Common scenarios:
Scenario 1 — Fiat to crypto and back:
- Buy USDT on Binance P2P at 87.2 RUB/$ (below market rate of 88.5 RUB/$)
- Sell USDT at market rate on spot or another P2P
- Difference: ~1.5%
Scenario 2 — Cross-platform P2P:
- Best seller on Binance P2P: USDT at 88.2 RUB/$
- Best buyer on Bybit P2P: USDT at 89.1 RUB/$
- Spread: ~1.0%
Scenario 3 — Cross-currency:
- Buy BTC for rubles on P2P below market rate
- Sell BTC for dollars on spot
- Net: obtained dollars more cheaply than official conversion rate
Pros and cons
✅ No millisecond execution needed — P2P trades take minutes
✅ Accessible with small capital
✅ Works especially well with RUB, KZT, UAH, AED
✅ No rapid price-movement risk
❌ P2P trades require counterparty trust and verification
❌ Limited volume — hard to move large amounts quickly
❌ Opportunity depends on P2P market activity
❌ May have tax implications depending on jurisdiction
Who it's best for
An excellent entry point, particularly for traders in CIS countries. No speed requirement, manageable risks, and small starting capital works fine. SpreadScan displays P2P offers across platforms for quick comparison.
Type 5: CEX-DEX Arbitrage
How it works
CEX-DEX arbitrage exploits price gaps between a centralized exchange (CEX) and a decentralized exchange (DEX) liquidity pool.
DEX pools price assets differently from CEXes: price is determined by a mathematical formula (x × y = k for AMM pools like Uniswap). When prices move significantly on a CEX, the DEX pool price hasn't updated yet — creating an arbitrage window.
Example:
- Binance: ETH = $2,480
- Uniswap (Arbitrum): ETH = $2,494
- Spread: 0.56%
Buy ETH on Binance → withdraw to wallet → sell on Uniswap. After trading fees (0.10% + 0.30% Uniswap pool) and gas: net profit ~0.15%.
Flash loans and MEV
Professional CEX-DEX arbitrageurs often use flash loans — a DeFi mechanism allowing you to borrow a large amount with no collateral, as long as it's repaid within the same transaction. This enables executing arbitrage without pre-positioned capital.
MEV (Maximal Extractable Value) is also relevant: bots operated by block validators or independent searchers can detect and front-run profitable transactions. This makes CEX-DEX arbitrage highly competitive at the professional level.
Pros and cons
✅ Larger opportunities during significant market moves
✅ Operates across 7 blockchains (SpreadScan tracks major ones)
✅ Interesting for those wanting DeFi exposure
❌ High technical barrier — requires understanding DeFi
❌ Competes with professional MEV bots
❌ Gas costs on Ethereum mainnet can be high
❌ Requires non-custodial wallet management
Who it's best for
Advanced users with DeFi experience. Not recommended as a starting point.
Type 6: Statistical Arbitrage
How it works
Statistical arbitrage is more complex: it's based not on immediate price gaps but on statistical relationships between assets over time.
Core idea: some coins are historically correlated (they tend to move together). If their price ratio temporarily deviates from the historical norm, it's likely to revert. You buy the "cheap" coin in the pair and sell the "expensive" one, waiting for convergence.
Pairs trading example:
- BTC and ETH historically maintain roughly a 33:1 ratio
- The ratio shifts to 35:1 — ETH is relatively underpriced versus BTC
- Buy ETH, short BTC
- Wait for reversion to 33:1
Pros and cons
✅ Less dependent on instant execution speed
✅ Can work over longer timeframes (hours, days)
✅ Diversified over time
❌ Requires statistical knowledge and historical data
❌ Correlations can break down — especially in crypto
❌ Closer to directional trading with real loss potential
Who it's best for
Traders with a quantitative or analytical background. Not a good starting strategy for beginners.
Full Comparison Table
| Criteria | Cross-exchange | Balance | Triangular | P2P | CEX-DEX |
|---|---|---|---|---|---|
| Complexity | Low | Medium | Medium | Low | High |
| Min. capital | $500 | $3,000 | $1,000 | $200 | $500 |
| Execution speed | Minutes | Seconds | Seconds | Minutes | Minutes |
| Transfer needed? | Sometimes | No | No | No | Yes |
| Competition level | Medium | Medium | High | Low | Very high |
| Transfer risk | Medium | None | None | None | Medium |
| Automation | Helpful | Helpful | Essential | Not needed | Essential |
| Best for | Beginners | Experienced | Algo traders | Beginners | Pros |
How to Choose Your Starting Strategy
If you're new to arbitrage, the recommendation is clear: start with cross-exchange arbitrage.
Reasons:
- The logic is transparent — easy to understand what's happening and why
- Tools are accessible — SpreadScan shows opportunities in real time
- No automation required to start
- Mistakes are understandable and recoverable
After 20–30 successful trades, move to balance arbitrage — it removes transfer risk. In parallel, study triangular for the transition to automation.
P2P arbitrage works well as a complement to your main strategy, particularly for CIS-region traders.
CEX-DEX and statistical arbitrage are separate specializations — worth pursuing consciously after building a foundation in simpler types.
Conclusion
Each type of arbitrage is a distinct tool with its own mechanics, risks, and requirements. There's no universally "best" type — there's the one that fits your current level, capital, and available time.
The key principle: start simple, scale to complexity as you build experience. The market will still be there, and mistakes in complex strategies are more expensive than in simple ones.
SpreadScan supports four arbitrage types — CEX-CEX spreads, triangular, P2P, and CEX-DEX — all in one place.
This article is for educational purposes only and does not constitute financial advice.
Frequently Asked Questions
Which type of arbitrage is most profitable? It depends on capital and experience. CEX-DEX and statistical can produce larger absolute profits but require higher skill. For most traders, cross-exchange and balance arbitrage offer the best return-to-effort ratio.
Can you combine multiple arbitrage types? Yes — experienced arbitrageurs typically do. A common combination: cross-exchange as the primary strategy + P2P for fiat operations + triangular running automatically.
Is automation required for triangular arbitrage? Practically, yes. Opportunities last seconds, and manually executing three consecutive orders is too slow. Triangular arbitrage without automation only works if you're monitoring extremely closely and can act very fast.
What's the difference between cross-exchange and balance arbitrage? Mechanically, only the absence of transfers. But this changes everything: it removes transfer risk, lowers the minimum profitable spread, and requires more capital locked across exchanges simultaneously.
Does arbitrage work in a bear market? Yes. Cross-exchange, balance, and triangular arbitrage are all market-direction neutral — they exploit price gaps between venues, not price movement itself.
Which type requires the least starting capital? P2P arbitrage — you can start with $200 and no special technical setup. Cross-exchange is next at around $500+.