AI quants posting huge returns
Quant shops are publicly touting ML + traditional quant stacks that they say uncover patterns and improve risk management — one thread cited AI quant trading returns up to 198.55% annualized x.com on ML+models in quant trading x.com on touted returns. Firms are also rolling AI portfolio managers for buy/sell alerts and rapid risk-off shifts — names like QuantAI Alpha are being flagged as examples x.com on AI PMs.
I Know First’s “up to 198.55%” stat refers to one stock, SIFY, inside its Computer Industry package where SIFY returned 198.55% over the forecast window March 10, 2025–March 10, 2026 and the package averaged 79.86% versus the S&P 500’s 17.53% for the same period. iknowfirst.com QuantAI Alpha’s product literature says the engine curates roughly 30–50 AI-selected stocks with an ~one‑month horizon and recommends a minimum investable size of about $10,000. intellectia.ai Its published playbook also specifies a 15% fixed stop‑loss and automated “risk‑off” cash shifts plus buy/sell alerts and daily portfolio updates. intellectia.ai Leading quant research has explicitly pushed larger, ML‑heavy models: AQR’s Journal of Finance paper “The Virtue of Complexity in Return Prediction” (2024) documents complex models (thousands+ parameters) outperforming simple benchmarks in their tests. ideas.repec.org That finding has drawn sharp pushback from academics who argue the results can reflect momentum or overfitting rather than durable predictive power. fa-mag.com Regulatory and governance examples are already visible: Two Sigma settled SEC allegations in January 2025 after model‑change weaknesses coincided with roughly US$165 million in client losses, and reporting in September 2025 says a former Two Sigma researcher was later indicted in a separate case tied to alleged model manipulation and multimillion‑dollar impacts. investmentexecutive.com
