Equity Mates Investing Podcast The Data Signals That Predict Winning (and Losing) Stocks with Dr. David Allen
26 snips
Mar 23, 2026 Dr. David Allen, head of long/short strategies at Plato and quant investor who builds AI-driven red-flag systems. He unpacks a 150-flag framework, explains why eight flags often signal trouble, and describes using LLMs on earnings calls, geopolitical stress tests, and systematic quantamental approaches to spot winners and losers.
AI Snips
Chapters
Books
Transcript
Episode notes
Evasiveness In Earnings Calls Predicts Underperformance
- Use LLMs to analyse 25,000 earnings call transcripts a year and score Q&A evasiveness for predictive signals.
- Companies that evasively answer unscripted Q&A underperform markedly over the next 3–6 months.
Chart Recognition Models Find Nonlinear Price Patterns
- Convolutional neural nets applied to price/volume charts can detect complex non-linear patterns and produce strong predictive setups.
- Plato runs these models in staging to validate out‑of‑sample without deploying live capital yet.
AI Arms Race Forces Active Managers To Evolve
- Systematic and AI‑driven strategies are growing fast and will force funds to sprint to stand still as competition improves.
- Plato invests in hardware and in‑house GPUs because cloud costs outweighed owning infrastructure for heavy model workloads.
