Trend-Following Trading Strategies and Financial Market Stability
Abstract: We study the interaction of trading strategy and financial market stability, through a combination of agent-based modeling and game-theoretic reasoning. Using a high-fidelity simulator of financial market environments employed in prior studies of algorithmic trading, we generate data about candidate strategy profiles, and identify equilibria over heuristic trading strategies from a game model induced from that data. Our model features incomplete information about common components of value, an essential element to incentivize agents to learn from market information. This provides a transmission path for market shocks, which we demonstrate through a scenario with trend followers: agents that continue price trends rather than oppose them. The presence in equilibrium of trend following is an economic consequence of delayed market access for background traders. Our main result shows that equilibrium trend following alters the market’s response to an external shock. Absent trend followers, market shocks cause a mild drop in price followed by a long recovery time, whereas in the presence of trend followers, there is a significant drop in price that also recovers quickly.