Cross-sectional identification of private information

We propose a new approach to identifying private information based on a model of
strategic trade optimization in the cross-section of securities. Investors are exposed to
liquidity and private information shocks and strategically optimize their trading across
securities, accounting for price impact (Kyle’s λ). The model yields a simple private
information measure: λ×OIB (order imbalance). Intuitively, observed order imbalance
is more likely to be information-driven when trading is expensive. Consistent with our
measure capturing private information, we show empirically that it is greater for smaller
firms with higher analyst dispersion, peaks with insider trades, helps explain return
reversals, predicts return volatility, and increases before M&A announcements and after
analyst coverage terminations.