Trading bots influence crypto market liquidity through a set of interacting mechanisms that can both improve everyday trading conditions and amplify instability during stress. Empirical work shows that algorithmic traders often narrow spreads and facilitate price discovery, yet their behavior can also lead to sudden liquidity withdrawals when markets move rapidly.
Mechanisms that increase and concentrate liquidity
Research by Igor Makarov and Antoinette Schoar at MIT Sloan demonstrates that arbitrage bots connect fragmented order books across exchanges, which tends to compress price differences and improve price efficiency across venues. In practice, these bots post opposite orders on multiple platforms, effectively supplying liquidity by buying where prices are low and selling where they are high. More broadly, studies of algorithmic trading in traditional markets by Terrence Hendershott at the University of California Berkeley, Charles M. Jones at New York University, and Albert J. Menkveld at VU Amsterdam find that algorithmic liquidity provision typically reduces bid-ask spreads and speeds information incorporation into prices. Those findings help explain why crypto markets with active bot activity often show tighter quoted spreads and higher apparent depth between manually posted orders.
Fragility, withdrawal effects, and fragmentation
At the same time, research focused on electronic markets highlights a recurring pattern: bots supply liquidity in normal conditions but withdraw it during stress. Andrei Kirilenko at the U.S. Commodity Futures Trading Commission and coauthors studied high-frequency trading in extreme events and found that liquidity dry-up can follow rapid price moves as automated strategies pull back to avoid losses, exacerbating volatility. In crypto, Makarov and Schoar document that cross-exchange frictions—deposit and withdrawal delays, differing regulatory regimes, and counterparty risk—create arbitrage costs that limit continuous liquidity. The result is conditional liquidity: plentiful when markets are calm, fragile when they are not.
Human and territorial nuances shape these dynamics. Exchanges domiciled in different jurisdictions have disparate rules and banking connections, so a global arbitrage bot that operates from a low-latency data center in a financial hub will behave differently than a retail trader in an emerging market. Research by Nicholas Easley and Maureen O'Hara at Cornell University points to information asymmetries in crypto markets; when informed trading rises, liquidity providers demand wider compensation, which can increase spreads despite automation.
Consequences for traders, markets, and policy
For retail traders, the presence of bots often means lower execution costs and quicker fills in ordinary conditions, but unexpected slippage during volatile episodes. For institutional participants, bots create both opportunities for cheap execution and risks from sudden liquidity migration across exchanges. Environmental and cultural factors also matter: the geographic concentration of low-latency infrastructure alters who can effectively run market-making bots, reinforcing advantages for technically sophisticated actors in regions with better connectivity. Regulators and exchange operators must weigh these trade-offs when designing rules on order types, latency, and capital requirements; empirical evidence from academic and regulatory studies suggests that encouraging fair access and reducing cross-border frictions can preserve the beneficial side of bot-driven liquidity while mitigating systemic fragility.