Anonymity in a Limit-Order Market: An Experimental Analysis

In this paper, we adopt an experimental approach to evaluate the impact of pre-trade anonymity in order-driven markets. More specifically, we build an experimental design of an electronic limit-order market, and compare two settings: one in which traders observe the identities of agents placing orders in the order book, and the other one where this information is not available on market screens. We find that ID code disclosure does not alter liquidity, efficiency or traders' profits. This is in contrast to the finding of Perotti & Rindi (2006) that anonymity improves market liquidity in a design where information acquisition is endogenous. When we exogenously increase the proportion of informed traders in our experimental markets, anonymity remains non-significant but we observe a clear improvement in liquidity and efficiency. Our experiments thus cast doubt on the real role played by anonymity in stock markets