Multidimensional stochastic liquidity in Kyle's model of informed trading

Battle Count: 67

Highly relevant to quantitative trading in several ways: (1) provides a rigorous multi-asset framework for understanding how informed traders optimally time their trades against stochastic liquidity; (2) characterizes equilibrium price impact matrices and market depth processes; (3) the economic interpretation of informed trading as optimal liquidation of information is directly applicable to execution algorithm design; (4) the stochastic liquidity framework generalizes Collin-Dufresne-Fos to multiple assets. However, the paper is primarily theoretical and the general matrix case remains open, limiting immediate practical implementation.

Extremely high complexity. Requires advanced knowledge of: matrix-valued stochastic analysis, causal optimal transport, Kalman-Bucy filtering, forward-backward SDEs, stochastic maximum principle, Doob-Meyer decomposition for matrix-valued processes, convex analysis in infinite dimensions, and Gaussian bridge theory. The general case involves a coupled n²-dimensional FBSDE with quadratic and singular drivers. Even the scalar verification requires solving BSDEs. No code or numerical implementation is provided.

The paper is entirely theoretical with complete proofs provided in appendices (B, C, D, E). All derivations are self-contained with explicit formulas. However, the general matrix-valued case remains open (reduces to a coupled FBSDE). No numerical experiments or code are provided. Mathematical reproducibility is high given the detailed proofs, but the open problem limits full verification.