AIQF Daily digest — 3 new papers

Quant Paper Radar · 3 new

1. 1

   Deep Learning of Robust Market Making under Regime-Switching Order Flow

   Felipe Moret, Fabrizio Lillo · ARXIV

   quant relevance 0.98

   The paper develops a regime-aware deep-reinforcement-learning quoting policy that balances spread capture and inventory risk as order-flow conditions change. Its profitability and risk-return comparisons come from a calibrated zero-intelligence limit-order-book simulator, not a real-data trading backtest. The mechanism cannot be backtested with bar-only equity data.

   original paper

2. 2

   Market-Informed Networks for Modeling and Forecast Evaluation of Financial Extremes

   Ayla Jungbluth, Johannes Lederer, Simon Trimborn · ARXIV

   quant relevance 0.82

   The paper presents a market-informed network Hüsler-Reiss model for forecasting multivariate financial extremes. Its JEAM method weights observations by global extremeness and similarity to historical joint-tail episodes, producing improved out-of-sample lower- and upper-tail log scores on one-minute US equity returns rather than measured trading returns.

   original paper

3. 3

   Short-maturity skew stickiness ratio under local volatility

   Masaaki Fukasawa · ARXIV

   quant relevance 0.78

   The paper proves that the short-maturity skew stickiness ratio converges to two in general time-dependent local-volatility models. The result addresses implied-volatility smile dynamics and cross-gamma risk relevant to short-dated options hedging and relative-value volatility trades, but includes no trading strategy or empirical performance.

   original paper