AI & Investing

Quantitative & AI Investing Research

We deconstruct the marketing hype surrounding "AI in finance" to analyze the actual mathematical and statistical methods used by quantitative funds.

Beyond Sentiment: NLP in Earnings Calls

Natural Language Processing

Why simply counting positive vs. negative words is obsolete. How quantitative researchers use transformer models (like FinBERT) to detect executive evasion, measure management uncertainty, and identify thematic shifts before they manifest in fundamental data.

The Overfitting Epidemic in Algorithmic Trading

Machine Learning / Statistics

If you backtest enough parameters, you will inevitably find a strategy that looks like a money printer. How to apply Deflated Sharpe Ratios, walk-forward optimization, and out-of-sample testing to prevent curve-fitting in neural networks.

Alternative Data Exhaust and Alpha Decay

Data Engineering

Satellite imagery of parking lots and scraped credit card receipts used to provide massive edge. Now they are commoditized. We analyze the half-life of alternative data signals in modern quantitative finance.

Analyzing AI-Managed ETFs: Marketing vs Reality

Fund Analysis

A deep dive into funds that claim to use proprietary AI to pick stocks. Why do their holdings often look identical to a standard momentum or quality factor portfolio? We deconstruct the black box.

Market Regime Detection with Hidden Markov Models

Statistical Modeling

Markets don't behave linearly; they shift abruptly between risk-on and risk-off states. How quantitative researchers use Hidden Markov Models (HMMs) to classify latent market states and adjust risk exposure dynamically.