AIQF Daily digest — 17 new papers

Quant Paper Radar · 9 of 17

1. 1

   Dynamic portfolio selection under range value-at-risk with logarithmic returns

   Hui Mi, Yuzhe Pan, Zuo Quan Xu · ARXIV · Portfolio Optimization / Quantile Risk Management

   we backtested it — Sharpe 0.18 · quant relevance 0.97

   The paper derives dynamic allocations that balance expected logarithmic growth against Range Value-at-Risk, producing state-dependent terminal payoffs and replicating strategies. It is directly relevant to portfolio risk management, but its model-based numerical illustrations are not evidence of realized trading performance. A daily implementation is backtestable, although it cannot reproduce exact continuous-time replication.

   original paper

2. 2

   Lineage-Aware Search Control for LLM-Based Evolutionary Factor Mining

   Xintong Liu, Chenkai Zhang, Yanzhao Ma et al. · ARXIV · LLM-Based Evolutionary Alpha Factor Mining

   quant relevance 1.00

   LineageEvo uses lineage-conditioned LLM mutation and crossover, together with lineage-balanced search allocation, to discover predictive symbolic equity factors. It reports predictive and portfolio-performance improvements on CSI 500 and S&P 500 stocks. The US application is buildable with daily market data, but the Chinese-market experiment is not reproducible with our coverage.

   original paper

3. 3

   Distributions or Quantiles? A Convex Variational Theory of Worst-Case Risk

   Jianjun Gao, Jonathan Li, Zhen Wang et al. · ARXIV · Distributionally Robust Portfolio Optimization

   we backtested it — Sharpe 0.67 · quant relevance 0.94

   The paper develops worst-case spectral risk methods and an exact conic formulation for robust expected-shortfall portfolio selection under mean-anchored Gaussian-reference KL ambiguity. This is directly applicable to portfolio allocation using historical returns; the supplied material establishes theoretical results rather than measured trading performance.

   original paper

The rest of the batch

• Market Regime-Based Multi-Asset Portfolio Optimization Using Nested Clustered Optimization for Tactical Asset Allocation

  Supriadi, Bambang Siswoyo · ARXIV · relevance 1.00

  The framework combines causal HMM regime detection with regime-conditioned covariance estimation, clustered minimum-variance optimization, and adaptive rebalancing.

• A game-theoretic analysis of the market microstructure in decentralized exchanges with concentrated liquidity

  Mikhail Belokon, Rostislav Berezovskiy · ARXIV · relevance 0.98

  The study derives optimal concentrated-liquidity range widths, boundary-triggered rebalancing policies, and a volatility threshold for profitable DEX liquidity provision.

• Arithmetic Models for Commodity Futures, Vanilla, and Calendar Spread Options

  Malthe Rauh Johansen, Nina Lange, David Skovmand · ARXIV · relevance 0.98

  The study estimates an arithmetic stochastic-volatility model for WTI futures and options and evaluates daily option hedging.

• Is sophistication a panacea? The learning capacity of random feature models for financial forecasting

  Martial Laguerre · ARXIV · relevance 0.98

  The study uses random-feature ridge regression to forecast the equity premium and construct mean-variance portfolios.

• LOB-Impact Benchmark: Emergent Market Impact in Generative Limit Order Book Models

  George Nigmatulin, Valentin Mohl, Sascha Frey et al. · ARXIV · relevance 0.97

  The benchmark evaluates generative limit-order-book simulators for execution-agent training by injecting metaorders and separating learned market impact from mechanical book depletion.

• Deep catastrophic-risk-aware reinforcement learning

  Carlos Pérez Octavio Mendoza, Frédéric Godin · ARXIV · relevance 0.96

  The method trains neural hedging policies using a differentiable extreme-value approximation to CVaR, directly targeting catastrophic hedging losses.

and 8 more · see the full radar