Algorithmic Trading

Introduction

Market Microstructure: What Is Actually Different

No central exchange. In equities, you have consolidated tape and RegNMS routing. In crypto, liquidity is fragmented across dozens of venues — Binance, Coinbase, OKX, Bybit, Kraken, plus a growing universe of decentralised exchanges. Your execution quality depends heavily on how you aggregate and route across venues.

Perpetual futures dominate. The most liquid instrument in crypto is not spot — it is the perpetual futures contract. Perpetuals have no expiry date and use a funding rate mechanism (periodic cash settlements between longs and shorts) to keep the contract price anchored to spot. For anyone who has traded equity swaps or FX forwards, the funding rate is a familiar concept wearing different clothes.

Latency characteristics are different. Centralised crypto exchanges operate with matching engine latencies in the low milliseconds — similar to equities. But the co-location model is less mature, and the latency arms race is less intense than on CME or Nasdaq. For strategies that do not require microsecond execution, the playing field is more level.

On-chain settlement adds a new dimension. Decentralised exchanges (DEXs) like Uniswap, dYdX and GMX settle on blockchain, introducing block-time latency (typically 0.5–12 seconds depending on the chain) but also creating unique arbitrage dynamics between centralised and decentralised venues.

---

Strategies That Transfer Directly from TradFi

Statistical Arbitrage — Funding Rate Basis Trading

This is one of the cleanest institutional strategies in crypto and maps directly to basis trading in futures markets.

The setup: buy Bitcoin spot (or a spot ETF), sell a Bitcoin perpetual futures contract. You are delta-neutral. Your income is the funding rate — the periodic payment from longs to shorts (or vice versa) that keeps the perpetual anchored to spot.

When the market is in a sustained uptrend, longs dominate and the funding rate becomes significantly positive. Shorts receive this payment. In extended bull markets, annualised funding rates have exceeded 30–50% — creating a basis trade with returns that dwarf anything available in traditional futures markets at equivalent risk.

The risk: funding can flip negative in bear markets, at which point the carry is negative and the position must be unwound. Active management of the carry is essential.

# Simplified funding rate arbitrage monitor
# Connects to exchange API and calculates net carry in annualised terms

import requests

def get_funding_rate(symbol: str) -> float:
    """Fetch current funding rate from exchange API."""
    url = f"https://api.exchange.com/v1/fundingRate?symbol={symbol}"
    response = requests.get(url).json()
    return float(response["fundingRate"])

def annualise_funding(rate_per_8h: float) -> float:
    """Convert 8-hour funding rate to annualised percentage."""
    periods_per_year = 365 * 3  # Three 8-hour periods per day
    return ((1 + rate_per_8h) ** periods_per_year - 1) * 100

symbol = "BTCUSDT"
rate = get_funding_rate(symbol)
annualised = annualise_funding(rate)
print(f"Current funding rate: {rate:.6f} ({annualised:.2f}% annualised)")

---

Market Making

The mechanics of market making in crypto are identical to equities and FX: post bids and offers, capture the spread, manage inventory risk. The differences lie in execution environment:

• Spread width: On major pairs (BTC-USDT, ETH-USDT), spreads are 1–3 bps on tier-one venues — comparable to large-cap equities. On smaller tokens, spreads widen significantly.
• Adverse selection: Crypto markets have high levels of informed order flow during news events (protocol upgrades, macro data releases, whale transactions visible on-chain). Adverse selection models need to account for on-chain signal.
• Inventory management: Without a delta-one hedge instrument as liquid as equity futures, inventory risk in spot market making requires careful position sizing. Perpetuals serve as the primary hedge.

For firms with existing market making infrastructure, porting to crypto APIs (which are REST and WebSocket-based, not FIX) is an engineering project rather than a research project.

---

Momentum and Trend Following

Crypto markets exhibit strong momentum characteristics — more persistent than equities, comparable in duration to commodity trend. The reasons are structural: retail-driven sentiment cycles, information asymmetry between on-chain data and market prices, and the absence of short-term mean-reversion forces like dividend yields.

Classic CTA-style trend following (moving average crossovers, breakout systems, Donchian channels) has performed well in crypto — particularly when applied to Bitcoin and Ethereum with monthly or weekly rebalancing horizons.

# Simple momentum signal: MA crossover
# Returns +1 (long), -1 (short), 0 (flat)

import pandas as pd

def momentum_signal(prices: pd.Series, fast: int = 20, slow: int = 200) -> int:
    """Generate directional signal based on moving average crossover."""
    if len(prices) < slow:
        return 0
    ma_fast = prices.tail(fast).mean()
    ma_slow = prices.tail(slow).mean()
    if ma_fast > ma_slow:
        return 1   # Long signal
    elif ma_fast < ma_slow:
        return -1  # Short signal
    return 0

This is basic — production systems add regime filters, volatility scaling and transaction cost models. But the underlying signal logic is immediately recognisable to anyone who has run a systematic equity or commodity strategy.

---

Cross-Exchange Arbitrage

Price discrepancies between exchanges are a persistent feature of crypto markets. Unlike equities (where RegNMS enforces best execution), crypto has no equivalent rule. The same Bitcoin futures contract can trade at different prices on Binance, OKX and Bybit simultaneously.

Pure arbitrage — buy on the cheaper exchange, sell on the more expensive — is dominated by HFT firms with low-latency infrastructure. The more accessible opportunity for institutional players is triangular arbitrage (BTC/ETH/USDT price relationships across venues) and spatial arbitrage (geographic price differentials in periods of high regional demand).

---

Where Crypto Algo Trading Is Structurally More Challenging

Data quality: Historical data in crypto is noisier than equity market data. Wash trading, exchange outages and data gaps require more rigorous cleaning than a Bloomberg feed.

Exchange risk: Centralised exchanges carry counterparty risk. The FTX collapse in 2022 was a defining event — many systematic funds lost significant capital held as margin on-exchange. Capital management across venues is now a first-class risk category in crypto algo operations.

Regulatory uncertainty for automated strategies: The regulatory treatment of algorithmic trading in crypto is jurisdiction-dependent and evolving. Strategies that are routine in equities may require careful legal review in crypto contexts.

API reliability: Exchange APIs are generally less robust than institutional equity infrastructure. Rate limits, maintenance windows and occasional outages are part of the operational reality and require defensive engineering.

---

Dimension	Equities (EU/US)	Crypto (Centralised)	Crypto (On-Chain / DEX)
Market hours	08:00–16:30 local	24/7/365	24/7/365
Execution protocol	FIX, OUCH	REST + WebSocket	On-chain transactions
Typical spread (large cap)	0.5–2 bps	1–3 bps	3–20 bps (+ gas)
Settlement	T+2	Near-instant	Block time (0.5–12s)
Counterparty risk	CCP-cleared	Exchange risk	Smart contract risk
Historical data quality	High (regulated tape)	Medium (wash trading risk)	High (on-chain, immutable)
Latency sensitivity	Very high (HFT dominant)	High	Low (block time dominates)

Source: Industry estimates, own analysis.

---

Getting Started: The Practical Path

For a TradFi quant or systematic trader exploring crypto, the path of least resistance is:

1. Start with CME Bitcoin and Ethereum futures — familiar infrastructure, regulated venue, standard FIX connectivity available through prime brokers
2. Add centralised exchange exposure via institutional APIs (Coinbase Advanced, Kraken Institutional) for spot and perpetuals
3. Explore funding rate strategies as a first foray into crypto-specific dynamics
4. Build toward DEX strategies once on-chain infrastructure and wallet management is operationally comfortable

The learning curve is real, but the foundational knowledge transfers. The professionals who have moved from systematic TradFi to crypto algo have consistently found that the core skills — signal research, execution optimisation, risk management — are fully applicable. The new learning is in market microstructure, exchange infrastructure and the on-chain data layer.

---

Key Takeaways

• Crypto algo trading uses the same mathematical foundations as systematic TradFi strategies
• Perpetual futures and funding rate mechanics are the central crypto-specific instruments to understand
• Funding rate arbitrage, momentum and market making all have direct TradFi analogues
• Key differences: fragmented liquidity, 24/7 operation, exchange counterparty risk, noisier data
• CME futures are the lowest-friction starting point for TradFi systematic traders

---

Next: What the largest banks and asset managers are already earning from crypto in 2026 — custody revenues, structured products and lending — and what that means for the competitive landscape.