Frequent Batch Auctions investigated by Agent-Based Model

The operating speed of order matching systems on financial exchanges has increased due to competition between markets and strong demands from investors. While this increase is beneficial for liquidity thanks to the increasing liquidity provided by market maker (MM) strategies, an undesirable effect is the socially wasteful arms race for speed, the costs of which are passed down to other investors as execution costs. A frequent batch auction (FBA) that reduces the value of speed advantages to mitigate the arms race has been proposed, but the disadvantage with FBA is that the MM strategies providing liquidity are exposed to more risks; moreover, as they continue to provide liquidity, many MMs retire and thus the liquidity will ultimately be reduced. In this study, we implement a price mechanism that can be continuously changed between a comparable continuance double auction (CDA) and FBA, analyze the profits/losses and risks of MM, and investigate whether MM can continue to provide liquidity even on FBA by utilizing an artificial market model, which is an agent-based model for a financial market. Our simulation results demonstrate that the execution rates of MM on FBA become smaller and this causes the MM to reduce the liquidity supply. Moreover, MM on FBA cannot avoid both an overnight risk and a price variation risk intraday, and it is very difficult for MM to be rewarded for risks and continue to provide liquidity. In contrast, MM on CDA is rewarded for risks and continues to provide liquidity. This means that while MM can provide liquidity on CDA, it cannot continue to provide liquidity on FBA, which leads to many MMs being retired and ultimately to liquidity being reduced.