Execution risk and price improvement under dark pools

Abstract

We develop a theoretical dynamic trading model with heterogeneous agents to examine how introducing a dark pool—running in parallel to a traditional lit exchange organized as a limit order market—primarily affects execution risk and price settings, and how these changes, in turn, influence the welfare of different trader types. Our model’s computational simulations show that the addition of a dark pool reduces liquidity on the traditional lit exchange. This liquidity reduction is evidenced by longer order execution times and a wider effective spread in the traditional lit exchange, driven by the migration of trading activity to the dark pool. We also identify two opposing channels that influence traders’ performance: the execution delay channel and the price improvement channel. Regarding the execution delay channel, dark orders lead to longer execution times for impatient traders (who seek to trade quickly) and shorter execution times for speculators (who wait for favorable execution prices relative to the asset’s fundamental value). This is because dark orders generally have longer (shorter) execution times than market (limit) orders. Regarding the price improvement channel, dark orders offer more favorable prices for impatient traders than market orders, while dark orders can result in less advantageous pricing for speculators. This is because dark orders are typically executed at the midquote of the bid and ask prices from the limit order market. Ultimately, the effect on execution risk, price improvement, and welfare for both impatient traders and speculators depends on which of these two opposing channels prevails.