Liquidity depth and information efficiency: High-frequency evidence from leading U.S. technology stocks

Using modern high-frequency data, we study how intraday liquidity depth and informational efficiency co-move in leading U.S. technology stocks. We estimate daily Kyle’s price-impact coefficient, λ, from 1-minute transactions and measure informational efficiency with a normalized Shannon-entropy index. Consistent with classic microstructure predictions, λ exhibits sharp spikes around earnings announcements, major macro releases, and option-expiration (OPEX) dates, indicating temporary contractions in market depth when information asymmetry and inventory risk intensify. In both firm-level and pooled fixed-effects regressions, higher λ is associated with lower entropy-based efficiency after controlling for market-wide shocks (market return), trading conditions (realized volatility and dollar volume), and firm size (market capitalization). Rolling and subsample evidence shows that this depth–efficiency relation is state-dependent: the negative slope strengthens in turbulent episodes and attenuates in calmer periods. Allowing firm-specific slopes reveals economically meaningful heterogeneity, large negative effects for some firms (e.g., Apple and Google) but an indistinguishable-from-zero relation for Tesla. Causality tests based on the Toda–Yamamoto approach suggest predominantly contemporaneous co-movement, with feedback dynamics present only for a subset of firms. Stock splits generate little systematic change in either λ or entropy.