The Effect of Disclosure Opacity on Trading Opacity : New Evidence from ADR Trades in Dark Pools

As volume increasingly migrates away from exchanges to dark pools, we revisit the link between disclosure opacity and volume, conditional on the choice to trade in a dark venue. We exploit the exogenous variation in home-country reporting opacity to examine how disclosure opacity affects the fraction of ADR volume traded in dark pools. Using multiple metrics for reporting opacity and controlling for firms' information environment and country-specific governance issues, we find that dark pool volume is positively correlated with home-country reporting opacity. This relation holds after controlling for observable differences between ADRs and other securities that trade in dark pools (matched sample analysis), volume migration to dark pools during earnings announcements, informed trading in lit versus dark venues, ADR levels, IFRS reporting requirements, and the possible endogenous determination of home-country reporting opacity and dark pool volume. The positive relation is stronger for ADRs held by institutions with low turnover and diversified holdings (quasi-indexers), and weaker for ADRs favored by institutions that trade frequently (transient) and institutions that hold concentrated portfolios (dedicated). Bid-ask spreads are greater for higher opacity ADRs that trade in dark pools, indicating that disclosure opacity magnifies the negative effect of dark pool trading on market quality