Quantitative Investing: Myths, Realities, and How AI Fits In

01 Quantitative Investment Basic Concepts

Quantitative investment lacks a precise definition; broadly, any investment method that relies on mathematical models and computers can be called quantitative investment.

Common domestic methods include stock multi‑factor (alpha) strategies, futures CTA strategies, arbitrage strategies, and high‑frequency trading.

Before 2010 it was a niche field, but the emergence of CSI‑300 index futures sparked a boom in quant funds. From 2010 to 2014 the sector enjoyed a dividend period with substantial profits and rapid growth in fund size.

Misunderstandings arose, such as the 2015 “stock crash” where high‑frequency trading was blamed as a cause, leading to regulatory restrictions that sharply reduced industry profits and ushered in a downturn.

The key takeaway is that quant is a tool to improve investment performance; one does not need to be a professional quant to use quantitative methods.

02 Features of Quantitative Investing

Quantitative investing is not a black‑and‑white opposite of subjective investing. Investors often blend both approaches, which can be described along two axes: perception (intuitive vs. quantitative) and decision (intuitive vs. quantitative), forming four dimensions:

Intuitive reception, intuitive decision – e.g., reading news and gauging market sentiment.

Intuitive reception, quantitative decision – e.g., scraping web text and building models for decisions.

Quantitative reception, intuitive decision – e.g., analyzing financial statements but deciding based on intuition.

Quantitative reception, quantitative decision – e.g., statistical analysis and multi‑factor models for decisions.

The diagram below illustrates these four dimensions.

03 Advantages of Quantitative Investing

Quantitative investing offers three main advantages: objectivity, big data, and fast response.

Objectivity : Strategies are often validated through backtesting, and many are executed via programmatic trading, reducing emotional bias.

Big Data : A single codebase can analyze thousands of stocks simultaneously, a scale unattainable by traditional research.

Fast Response : Automated analysis enables second‑level reactions, with high‑frequency trading operating at microsecond speeds. Even non‑HFT applications benefit from rapid news parsing and announcement analysis.

04 Quant, AI, and Their Limits

Despite hype around AI after AlphaGo’s success, AI alone cannot surpass top fund managers in investing because investment is a prediction problem, not merely an optimization problem.

AI suffers from limited usable sample data and high‑dimensional feature spaces, making financial prediction challenging. Many so‑called AI funds merely repurpose mature quant models.

Therefore, one should not over‑rely on quant or AI; instead, leverage their strengths to gain incremental advantages in the competitive investment landscape.