High-Frequency Trading in Crypto: Risks and Rewards in HFT 

Last Updated: 17 October, 2023
5 min read

high frequency trading in crypto

High frequency trading (HFT) has become an integral part of modern financial markets, with HFT crypto trading firms accounting for over 50% of equity trading volume in the US. As cryptocurrency markets have grown, HFT strategies have started entering this new domain as well.

How does High-Frequency Trading work?

High frequency trading refers to sophisticated algorithmic trading systems that execute a large number of orders in fractions of a second and try to profit from small price movements. The key characteristics of HFT strategies are:

  • Speed: HFT systems need to have ultra-low latency connectivity to execute orders within milliseconds. They covet speed advantages over other market participants.
  • Short holding period: HFT firms typically do not accumulate large directional positions. They close out positions within seconds or minutes after opening them.
  • Volume: High-frequency traders engage in extremely high daily volumes, accounting for a significant chunk of overall market volume.
  • Math and technology: HFT strategies rely heavily on mathematical models and programming. Trading algorithms are constantly tweaked and updated to enhance profitability.

Speed Matters in High-frequency Trading

Speed is integral to many HFT strategies because it allows firms to detect and capitalize on fleeting arbitrage opportunities in the market before others can. Some examples include:

  • Latency arbitrage: Buying and selling between two markets based on microsecond time delays in price updates across exchanges/ECNs.
  • Liquidity rebates: Earning exchange fee rebates by rapidly submitting and cancelling limit orders.
  • Order anticipation: Deducing the existence of large pending orders and incorporating that in cryptocurrency trading strategies.

Advanced networking technology like co-location and direct data feeds provide HFT firms with speed advantages to capitalize on these short-lived opportunities.

Types of HFT Strategies

Some common categories of HFT strategies include:

  • Market making: Providing simultaneous buy and sell quotes for a security to capture the the bid-ask spread.
  • Arbitrage: Exploiting price differences between related securities/assets/markets.
  • Structural: Exploiting structural inefficiencies in market microstructure.
  • Directional: Using statistical models to make short term directional bets.
  • Agent: Using HFT algorithms as “agents” to execute large parent orders to reduce market impact and costs.

Within these broad categories are hundreds of specific strategies for automated trading.

The Advent of HFT strategies in Crypto

Cryptocurrency markets have many properties that make them conducive to HFT type strategies, including:

  • 24×7 availability: Crypto markets operate 24 hrs/day, 7 days/week, unlike traditional markets. More trading time allows HFT strategies to capitalize on more profit opportunities.
  • Volatility: Crypto assets are notoriously volatile, with prices often moving +5-10% intraday. This creates more price aberrations to exploit.
  • Fragmentation: There are hundreds of cryptocurrency exchanges with varying liquidity and pricing. This creates arbitrage opportunities.
  • Young markets: Crypto markets are relatively new with unsophisticated participants. This allows professionals to capitalize on structural edges.

Market makers were the first entrants, looking to capture spreads on major crypto pairs. But crypto markets now see activities similar to traditional HFT strategies:

Latency Arbitrage

Exploiting millisecond pricing discrepancies across different trading platforms is a pure latency arbitrage strategy. Differences in exchange matching engines, geographic latency in sending orders, etc. create short-term distortions to capture.

Direct market data feeds like FIX API are used to get real-time price updates. CME even offers a HFT-friendly simulated trading platform for crypto derivatives.

Liquidity Rebates

Many crypto exchanges, like Coinbase Pro, offer maker-taker fee models where market makers get paid for providing liquidity through rebates.

HFT firms exploit these liquidity rebates through strategic order placement and cancellation. By rapidly submitting and canceling limit orders, rebates can become a key source of profitability.

Order Anticipation

Detecting large pending buy/sell orders allows HFTs to incorporate that directional information into their trading strategies.

For example, if a large impending buy order on Coinbase Pro is detected, an HFT firm can simultaneously buy BTC on another exchange beforehand, anticipating the price impact.

Large orders are usually broken up into smaller chunks. But statistical detection techniques can be used to <reverse engineer> information about the full order.

Structural Strategies

Cryptocurrency markets actually aid certain structural strategies that are dying out in equity markets:

  • Spoofing: Placing non-bonafide orders to artificially influence the order book and deceive other participants is illegal in most developed markets. But it still exists in crypto markets.
  • Wash trading: Buying and selling securities with yourself to manipulate trading volume and prices is banned in many jurisdictions. But still prevalent in crypto.
  • Frontrunning: Front running client orders as a broker by trading ahead is prohibited in traditional markets. But unregulated crypto exchanges allow certain forms of frontrunning.
  • Quote stuffing: Flooding the market with orders/cancellations to slow down one exchange is harder to execute. But feasible in some crypto exchanges.

Such strategies rely on immature market microstructure and regulatory arbitrage between crypto and traditional asset markets.

Key Enablers of High-frequency Crypto Trading

High frequency trading strategies in crypto markets are enabled by many of the same key infrastructure components used in traditional HFT:

Co-Location: Establishing trading servers in the same data center as the exchange matching engines reduces latency in sending/receiving orders. This enables arbitrage strategies dependent on microseconds and nanoseconds.

Direct Market Access: Accessing raw market data feeds directly using FIX APIs/WebSockets provides the real-time price updates needed for low latency trading.

Smart Order Routing: Sophisticated order routing algorithms break up orders across multiple liquidity pools to disguise trading intentions and minimize execution costs.

Low Latency Networks: Dedicated communication networks using microwave, fiber optics, and laser technology provide the low latency connectivity between trading servers and exchanges required for latency arbitrage strategies.

FPGAs and GPUs: Custom hardware using GPUs and FPGAs provide the data parallelism needed for very rapid statistical computations on ticker data.

Issues Around Crypto HFT Practices

While HFT provides liquidity and tight spreads in markets, some of its practices in the crypto space remain controversial:

  • Sub-penny front running: Frontrunning clients by jumping ahead of pending orders from related accounts and profiting on price changes of less than a penny.
  • Wash trading: Buying and selling assets with themselves or counterparties to manipulate reported trading volume and skew prices. Still occurs in crypto markets.
  • Unsupported leverage: Providing hidden leverage by misreporting trading positions and collateral risk considerations, in ways that would not be allowed in regulated markets.
  • Unchecked market power: Just a few dominant HFT players account for a controlling share of crypto market volumes, order flow, liquidity provision, etc. There is insufficient competition.
  • Unfair latencies: Normal traders are handicapped by speed advantages of colocation and direct feeds available to only HFT firms. Asymmetries exist.
  • Regulatory arbitrage: HFT firms apply strategies like spoofing or aggressive order pinging that would not be legal in many equity and futures markets. Cryptocurrencies allow regulatory loopholes.

Many industry experts have called for reforms to create fairer market microstructure and trading practices in cryptocurrencies.

Use HFT in Bitcoin and Ethereum Markets

Bitcoin and Ethereum are the top two cryptocurrencies where HFT-style trading method dominates. Here’s an overview of HFT activity in each market.

HFT Trading in Bitcoin

As the first and largest crypto blockchain, Bitcoin led the way for HFT market making and arbitrage trading opportunities between exchanges. Here are some stats on the current state:

  • Up to 70% of Bitcoin spot trading volume attributed to quant funds and HFT strategies.
  • Half of BTC futures trading activity on CME from “High Frequency & Algorithmic Trading” category of traders.
  • Primary exchanges connected by HFT include Bybit, Bitget, Binance, gate.io.
  • Primary arbitrage spread is between Coinbase Pro as the retail exchange and Binance as the institutional exchange.
  • HFT scalping exploits inefficiencies from having multiple BTC order books across different exchanges.
  • Large OTC trades between institutions and funds often broken up via HFT smart order routers.

HFT Trading in Ethereum

Ethereum has also proven to be a fertile ground for HFT type trading, with ample arbitrage and structural opportunities. Some key aspects:

  • Up to 60% of ETH spot volume estimated to come from HFT-style firms.
  • Primary exchanges include: Coinbase Pro, Binance, Bybit, Bitget, gate.io.
  • On-chain Ethereum transactions used for inter-exchange arbitrage and settlement.
  • ETH/BTC is the most liquid trading pair and efficiently priced across exchanges.
  • But stablecoins like USDT, USDC, DAI, etc. lead to fragmented liquidity and pricing discrepancies.
  • Both spot and futures ETH markets targeted for different latency arbitrage and order anticipation strategies.

So in summary, Ethereum lends itself well to the same HFT strategies and infrastructure as used by Bitcoin HFT traders.

References

[1] Greg N. Gregoriou, 2015, Handbook of High Frequency Trading.

[2] High-Frequency Trading: New Realities for Traders, Markets and Regulators. Risk Books, 2013.

[3] Wah, Elaine, et al. Latency Arbitrage, Market Fragmentation, and Efficiency: A Two-Market Model. Management Science.

Before you go…

This is just one of the many crypto trading strategies we examined.

Wanna unlock more tested strategies? Check out:

Gianluca Lombardi

Gianluca is the editor-in-chief of this site. A finance graduate, he is an active trader who has tested all trading platforms and knows all their secrets. Technology is his passion; he spends much of his free time in the metaverse. Gianluca loves learning new things, researching, discussing and writing about technology, especially when it comes to cryptocurrency and blockchain technology.