CASA-RL: Exact Set Policies for Sparse Portfolio Control

Sparse portfolio reinforcement learning couples unordered asset selection with continuous allocation, while financial priors complicate attribution of learned gains. We introduce CASA-RL, a framework for $\underline{C}$ardinality-$\underline{A}$ware $\underline{S}$parse $\underline{A}$llocation built around three requirements. First, an exactly normalized fixed-cardinality distribution assigns probability directly to the selected asset set. Second, a support-conditioned logistic-normal allocation yields feasible stock/cash weights and an explicit joint likelihood for PPO. Third, matched initialization, support-prior ablations, and checkpoint comparisons separate prior-based performance from learned changes. In a 45-run study on reconstructed historical S&P 500 constituents, the validation-selected configurations of CASA-RL and two ordered-policy implementations achieve approximately $\text{315.92}$% mean cumulative net return on a retrospective test interval. The selected trained policies match their initialized counterparts' stock supports on every test day. Removing the explicit support prior exposes substantial policy changes and checkpoint-dependent method rankings. The resulting framework aligns feasible portfolio actions, likelihood-based learning, and empirical attribution.