Free Money: The Effects of Foreign Exchange Market Microstructure on Returns to the Bellman-Ford Arbitrage Strategy

Abstract

The foreign exchange market is one of the most liquid in the world with many unique properties. One such property is the network property, which allows a very large number of possible arbitrage through a disparity in rates, which is commonly referred to as triangular arbitrage. The Bellman-Ford Algorithm is a network path-finding algorithm, which will always yield an arbitrage opportunity on complex networks if one exists. This algorithm allows us to study the efficiency of the market and what conditions will lead to the existence of such arbitrage. This paper uses regression analysis to test whether market microstructure given by the volatility, spreads and volume present in the network affect the existence and profitability of arbitrage. Such analysis finds that market microstructure is a poor explanation for variation in the duration over which a given arbitrage trade is continuously profitable, but is a fair predictor of the returns to that trade with and without accounting for trading costs. The regression analysis also suggest that these unique arbitrage profits exist as compensation to traders who adjust the network to rate parity if liquidity or volatility are relatively higher in other parts of the network.