Shock Propagation Through Cross-Learning in Opaque Networks

This paper studies information transmission in opaque networks with uncertain inter-firm linkages. Local traders can identify their firm's direct neighbors but are unsure about these firms' linkages to other firms. This uncertainty renders prices of neighboring firms difficult to interpret and leads to the propagation of shocks across firms. Payoff-relevant information diffuses slowly and there is excess comovement between fundamentally unrelated assets. Traders and firms in higher network layers suffer from less informative prices and invest less efficiently. Surprisingly, more volatile noise trading can render stock prices more efficient in opaque networks as it discourages cross-learning and reduces the propagation of unrelated shocks.