Arbitrage Strategies: How Pros Make Money From Price Differences

Arbitrage is the practice of profiting from price differences for the same (or equivalent) asset across venues, instruments, or time. In efficient markets, arbitrage opportunities are rare and small. In crypto's fragmented market structure, they exist persistently, but most of the obvious ones are now captured by bots. Knowing which arbitrage forms remain accessible to retail (and which don't) clarifies what's worth pursuing.

The arbitrage landscape

Several distinct categories:

Cross-exchange spot arbitrage. Same asset trading at different prices on different exchanges. Simple in concept; hard in practice due to:

• Transfer time (moving BTC between exchanges takes minutes; price can change)
• Withdrawal/deposit fees (eat into the spread)
• Counterparty risk (each exchange leg is exchange risk)
• Bot competition (HFT bots capture most opportunities faster than humans)

Triangular arbitrage. Three currency pairs whose implied prices don't align. Trade through the cycle to capture the discrepancy. Mostly captured by bots within a single exchange.

DEX-CEX arbitrage. Same token at different prices on DEXs vs CEXs. Captured aggressively by bots; spreads typically very tight on liquid pairs.

DEX-DEX arbitrage. Same token at different prices across DEXs. Mostly captured by MEV bots in the same block your trade enters.

Funding-rate arbitrage / basis trading. Long spot, short perpetual. Captures the funding rate as yield. Not purely arbitrage but similar mechanic. Accessible to retail; covered separately in the basis-trading chapter.

Statistical arbitrage / pairs trading. Long one asset, short a correlated asset, profit from convergence. Requires sophisticated modeling.

Cross-chain arbitrage. Same asset at different prices across blockchains. Bridge or trade through to capture. Covered in the cross-chain chapter.

Some of these are accessible to disciplined retail; many are bot territory. Knowing the difference saves time chasing opportunities that don't exist for you.

Which arbitrage forms remain accessible to retail

Funding-rate / basis arbitrage. Yes, accessible. Doesn't require speed advantage; the funding pays periodically and persistent positive funding can be captured by holding the position. Covered in detail in the perpetual basis chapter.

Cross-exchange spot arbitrage on liquid pairs. Mostly not accessible. Bots capture the obvious opportunities in milliseconds. The arb might exist for 30 seconds; you won't beat the bots.

Cross-exchange spot arbitrage on illiquid pairs. Sometimes accessible. New listings, token issues, or small exchanges may have spreads that bots haven't captured yet (because the volume isn't worth their effort). Requires being early.

DEX arbitrage. Generally not accessible. MEV bots own this category.

Funding-rate arb on smaller exchanges. Sometimes accessible. The big arbitrageurs focus on big venues; smaller venues with persistent funding extremes can be captured by retail.

Stat arb / pairs trading. Theoretically accessible but requires real quantitative skills (modeling, risk management, execution infrastructure). For most retail, not realistic.

The honest assessment: most arbitrage is bot territory. Funding-rate / basis trading is the most retail- accessible. Niche opportunities exist on smaller venues and new listings.

How basis arbitrage works (the retail-accessible version)

The mechanic:

Setup:

• Buy 1 BTC of spot
• Short 1 BTC of perpetual at the same time, on a venue with high positive funding

Result:

• Net price exposure: zero (the spot gain equals the short loss when price rises)
• Income: you receive the funding rate every 8 hours from the longs paying funding

Returns:

• If funding averages 0.05% per 8h (~55% annualized), you collect that on the notional (minus fees)
• Realistic net returns: 5-25% annualized, depending on funding levels and execution efficiency

Risks:

• Counterparty risk on the exchange (you have spot and short positions there)
• Liquidation risk on the perp short if the price moves against it without spot offsetting fast enough
• Funding flips negative (you start paying instead of receiving)
• Transfer / management overhead

This is the boring real-edge arbitrage that pros use for steady returns. Not exciting; reliably profitable when funding is positive.

Cross-exchange spot arbitrage, when it's real

Even though bots dominate, some scenarios:

New listings. When a token first lists on a new exchange, the price often diverges from established venues for minutes-to-hours before bots catch up. Retail who's monitoring can capture the spread.

Exchange-specific events. During exchange outages, deposit/withdrawal halts, or other disruptions, prices can diverge meaningfully. Retail with capital pre- positioned at the affected venue can sometimes capture the spread when normalcy returns.

Geographic restrictions. Tokens listed in different regions can have persistent regional spreads. Capturing requires accounts in multiple jurisdictions.

Stablecoin depeg events. When a stablecoin briefly depegs (USDC during SVB, etc.), various spreads open up. Retail with the right setup can capture portions.

These are rare events, not steady opportunity. Plan for them; don't rely on them.

A common mistake: chasing visible spreads

A trader sees BTC trading at $67,000 on one exchange and $67,150 on another. They calculate: "$150/BTC arbitrage!" They start the process, buy on cheap exchange, transfer to expensive, sell.

By the time the BTC has transferred (minutes to hours on the chain), the spread has closed. The trader ends up flat or negative after fees. The "arbitrage" they saw was a snapshot that didn't persist long enough to capture.

The fix: visible spreads on liquid pairs are usually not capturable by retail. Bots get them in the same block. Don't waste time on these. Focus on basis trading and other forms with longer-duration opportunity.

A common mistake: ignoring transfer and execution costs

A trader calculates the spread but doesn't include:

• Withdrawal fees (sometimes $5-50 per transfer)
• Network fees (gas)
• Trading fees on both legs
• Slippage on both legs
• Price movement during transfer

The actual capture is much smaller than the visible spread. Often negative after all costs.

The fix: build a complete cost model before any arbitrage trade. The realized spread after all costs needs to be meaningfully positive. Many "arbitrage opportunities" are net losers after costs.

A common mistake: counterparty concentration

A trader runs a basis trade with $50k spot and $50k short on the same exchange. Total notional exposed: $100k on a single counterparty. If the exchange fails, both sides are at risk.

The fix: split exposure. Spot on one venue (or self-custody), short on another. The funding capture is slightly more complex but the counterparty risk is distributed. For larger sizes, this becomes critical.

A common mistake: forgetting margin maintenance

A basis trade has spot offset against short perp. As price moves, the short perp's margin shifts but the spot side just sits there. Without active management, you can be liquidated on the short side even though your overall position is hedged.

The fix: monitor margin actively. Add collateral to the short side as needed. Some exchanges have unified margin that uses spot as collateral for the short, those make basis trades easier.

A common mistake: treating funding as guaranteed

Funding rates change. They can flip negative. They can go to zero. A basis trade returning +0.05% per 8h this week might return -0.02% per 8h next week.

The fix: monitor funding continuously. Be willing to exit the basis trade when funding is no longer attractive. Don't assume the rate persists.

Mental model, arbitrage as a closing window

Every arbitrage opportunity has a closing window. The opportunity exists because the market hasn't yet arbitraged it away. As soon as enough capital flows to capture it, the spread closes.

Some windows close in milliseconds (bot territory). Some close in hours (sometimes capturable by fast retail). Some close in days or weeks (more accessible). Some persist for months at varying intensities (basis trading territory).

Match your strategy to the window length. Don't try to capture millisecond windows manually. Do focus on the longer-window opportunities where retail can compete.

Why this matters for trading

Most "arbitrage opportunities" advertised online are fake or already-captured. The real arbitrage available to retail is mostly basis trading on perps with positive funding. Hex37's perpetual support and multi-asset infrastructure makes basis trades straightforward to set up; the discipline is in managing the position over time and recognizing when the funding stops being attractive.

Takeaway

Arbitrage is profiting from price differences for equivalent assets. Most categories (cross-exchange spot, DEX-DEX, triangular) are dominated by bots and not accessible to retail. The most retail-accessible form is funding-rate / basis trading: long spot, short perp on positive-funding venues. Niche opportunities exist around new listings, exchange events, depeg events. Always model complete costs (transfer, fees, slippage). Manage counterparty risk. Monitor funding continuously. Match strategy window to opportunity window. Most retail's "arbitrage hunting" is wasted effort on opportunities that bots already captured.