Entropic Value-at-Risk parity for tempered stable returns

Battle Count: 84

Highly relevant to quantitative trading and portfolio management. The paper directly addresses portfolio construction using tail-risk measures (EVaR) under realistic non-Gaussian return distributions (tempered stable). The risk parity framework is widely used in institutional portfolio management. The empirical evaluation across cross-asset, momentum, and sector universes provides practical guidance. However, the higher turnover of ICA-based EVaR portfolios and the lack of a uniform advantage over CVaR temper the practical impact. The methodology is more suited for strategic asset allocation than high-frequency trading.

High implementation complexity due to: (1) fitting MNTS and ICA-based tempered stable models requires specialized numerical optimization; (2) computing EVaR involves minimizing over the moment-generating-function domain with weight-dependent admissible domains; (3) Euler risk contributions require either analytic derivatives (Propositions 2-3) or finite-difference approximations with domain boundary handling; (4) solving the ERC system of equations for portfolio weights; (5) the ICA decomposition and component fitting add further complexity. The mathematical framework is well-specified but requires careful numerical implementation.

The paper provides detailed mathematical derivations, parameter estimation procedures, and experimental design (Table 1). Data sources are publicly available (Yahoo Finance, Kenneth R. French Data Library). However, no code repository is provided, and the finite-difference procedure and parameter fitting details would require careful implementation. The backtesting methodology (monthly rebalancing, 12-month estimation window, tail probability eta=0.05) is clearly specified.