Existing empirical studies provide little support for the theoretical prediction that market makers rebalance their inventory through revisions of quoted prices. This study provides evidence that the NYSE's specialist does engage in significant inventory rebalancing, but only when not constrained by the affirmative obligation to provide liquidity imposed by the Price Continuity rule. The evidence also suggests that such obligations are associated with better market quality, but impose significant costs on the specialist. The specialist mitigates these costs through discretionary trading when the rule is not binding. These findings shed light on how exchange rules affect market makers’ behavior and market quality.

Though trading halts are a common feature in securities markets, the issues associated with the coordination of these halts across markets are not well understood. In fact, regulations often allow traders to circumvent trading halts through the use of alternative venues. Using a sample of order imbalance delayed openings on the NYSE, we examine the costs and benefits of continued trading on alternative venues when the main market calls a halt. We find that trades routed to off-NYSE venues during NYSE halts are associated with significant price discovery and lead to an improved post-halt trading environment. In addition, limit orders routed through ECNs reflect price-relevant information even prior to the halt, with limit book imbalances decreasing and depth filling in during the halt around the eventual reopening NYSE price. However, these informational benefits come at a substantial cost, as both execution costs and volatility are extremely high on off-NYSE venues during NYSE halts.

Using European stock data from two different venues and time periods for which we can identify each trade's aggressor, we test the performance of the bulk volume classification (Easley et al. (2016); BVC) algorithm. BVC is data efficient, but may identify trade aggressors less accurately than “bulk” versions of traditional trade-level algorithms. BVC-estimated trade flow is the only algorithm related to proxies of informed trading, however. This is because traditional algorithms are designed to find individual trade aggressors, but we find that trade aggressor no longer captures information. Finally, we find that after calibrating BVC to trading characteristics in out-of-sample data, it is better able to detect information and to identify trade aggressors. In the new era of fast trading, sophisticated investors, and smart order execution, BVC appears to be the most versatile algorithm.