Heterogeneous Returns and Wealth Tax Neutrality: A Fokker–Planck Framework

Battle Count: 71

The paper is primarily a public economics / tax policy paper with mathematical finance tools. Its direct relevance to quantitative trading is moderate: (1) The heterogeneous-ability framework affects equilibrium asset prices and risk premia, which is relevant for factor models and alpha generation. (2) The portfolio allocation results (Section 4.4) show that wealth taxes alter effective opportunity sets differentially across ability types, relevant for understanding how tax regimes affect market microstructure. (3) The Pareto tail analysis (Section 4.2) connects to tail-risk modeling. (4) The spectral portfolio theory extension (Section 6.6) could inform portfolio construction under heterogeneous agent populations. However, the paper does not propose trading strategies, does not use ML models, and does not provide actionable signals. The Fokker-Planck framework is more relevant to academic understanding of wealth dynamics than to practical trading.

The mathematical framework requires expertise in stochastic calculus (Itô's lemma, SDEs), Fokker-Planck equations, and PDE theory. The joint two-dimensional Fokker-Planck equation (Eq. 10) is non-trivial to solve analytically except in special cases. Numerical implementation would require finite-difference or spectral methods for 2D PDEs. The analytical propositions are elegant but the full joint dynamics (especially with mean-field feedback) would be computationally demanding. The illustrative example (Table 1, Figure 1) is simple to reproduce, but quantitative calibration to real data (Fagereng et al.) would require substantial econometric work. The theoretical framework is self-contained but builds on a series of 6 companion papers.

The paper is purely analytical with formal propositions and proofs. All mathematical derivations are self-contained. However, there is no code repository, no numerical simulations provided beyond one illustrative figure (Figure 1), and no empirical estimation performed. The framework builds on a series of companion papers (Frøseth 2026a-f) which would need to be consulted for full context. Parameters for the illustrative example are stated (σ=0.15, δ=0.08, τ_w=2.5%, τ_c=30%). Reproducibility of the analytical results is high; reproducibility of any quantitative application would require additional work.