Financial Epiplexity: A Theory of Learnable Market Structure under Bounded Computation

Battle Count: 50

Highly relevant to quantitative trading. The paper directly addresses core quant concerns: what alpha is theoretically extractable from data (monetization bounds), how to set Sharpe ratio ceilings as backtest consistency checks, how alpha decays over time (computational depth), how crowding degrades returns (mutual compressibility), how capacity limits strategies (bit-leak budgets), and how competitive dynamics shape alpha extraction (game theory). The framework provides a principled way to value data and representations before running backtests. However, it is purely theoretical with no empirical validation, and practical implementation requires market-specific modeling of constants and standardized compression schemes that do not yet exist.

The theoretical framework is mathematically sophisticated, combining information theory, MDL, algorithmic complexity, Kelly criterion, game theory, and stochastic processes. Practical implementation faces significant challenges: (1) computing description lengths for neural model classes lacks standardized compression schemes; (2) estimating the epiplexity profile requires specifying representation, budget, target, and horizon; (3) the monetization constants (kappa_W, kappa_SR) require market-specific modeling; (4) measuring mutual compressibility of forecast streams requires compression algorithms; (5) the game-theoretic equilibria involve complex strategic interactions. The toy model in Section 11 is analytically tractable but highly stylized. No code or implementation guidance is provided.

This is a purely theoretical paper with no empirical validation, no code, and no datasets. It contains mathematical definitions, propositions, theorems, and proofs. A closed-form toy model (Section 11) provides numerical interpretability but does not constitute empirical testing. Reproducibility of the theoretical results is high (proofs are self-contained), but there is no empirical pipeline to reproduce. The framework requires market-specific modeling of constants (kappa_W, kappa_SR) and representation choices that are not standardized.