Mean squared error bundles three separate checks: bias, direction and volatility. See the MSE decomposition and why MSE reaches zero only when all three hold.

Review of Building AI Agents for Finance by Dupouy and El Mofatiche: production-grade multi-agent systems, framework comparison, and compliance-ready evaluation.

Research article on topological anomalies and return predictability in the S&P 500 accepted for publication in Bank i Kredyt. Preprint on arXiv and SSRN.

I will attend Cordell Tanny’s live Python workshop on building and stress-testing a 12-asset portfolio with CVaR, clustering and Johnson SU. Nov 28, 2026.

Mean, SD and range scaling turn one fitted regression coefficient into three interpretations, with no re-estimation.

Review of Options Pricing with Python by Mhamed Bettaieb (Packt): Black-Scholes, trees, Monte Carlo, Greeks and ML, with a teacher’s view of its limits.

Using 5-minute LOBSTER data for AAPL, this post visualises mean standardised bar returns over a 5×5 grid of jointly ranked trade count and volume. The resulting surface is markedly nonlinear: at 30-second bars volume dominates, while at 60-second bars a diagonal interaction emerges — a pattern inconsistent with purely additive specifications in trades and volume.

Review of Large Language Models in Finance by Miquel Noguer i Alonso (Packt, 2026) from a quant researcher and educator view: backtesting, agents, alpha.

A quick-reference guide to the R quant stack: 10 categories of AI/ML and algorithmic trading packages covering market data, modeling, risk, and execution.

My impressions of Packt’s hands-on GraphRAG workshop with Alessandro Negro: Neo4j, LangGraph agents, text-to-Cypher, evaluation, and lessons for quant finance.