Investors trade in various types of venues. When demanding immediacy they face a basic tradeoff between price impact and execution uncertainty. Venues can be sorted accordingly along a "pecking order," with mid-point dark pools and lit markets at the top and bottom, and non-midpoint pools in the middle. A simple model formalizes this pecking order hypothesis. We test it using a unique dataset that disaggregates U.S. dark trading into various categories. A higher VIX or larger earnings surprise tilts trading volumes from the top of the pecking order to the bottom, confirming the hypothesis.

A breakdown of cross-market arbitrage activity could make markets more fragile and result in price crashes. We provide suggestive evidence for this novel channel based on a high-frequency analysis of the most salient crash in recent history: The Flash Crash. We further show that such an event can be extremely costly for a large seller trading in a particular venue as the seller effectively relies on local liquidity supply only. These findings highlight the vulnerability of today’s highly fragmented markets. This paper was accepted by Gustavo Manso, finance.

No abstract

Investors trade in various types of venues. When demanding immediacy they face a basic tradeoff between price impact and execution uncertainty. Venues can be sorted accordingly along a "pecking order," with mid-point dark pools and lit markets at the top and bottom, and non-midpoint pools in the middle. A simple model formalizes this pecking order hypothesis. We test it using a unique dataset that disaggregates U.S. dark trading into various categories. A higher VIX or larger earnings surprise tilts trading volumes from the top of the pecking order to the bottom, confirming the hypothesis.

No abstract