The survival secret in volatile markets! An article to help you understand the "quantitative stock selection" strategy!

What advantages does quantitative stock selection have over index enhancement in volatile markets?

With the increasing abundance of domestic financial data and advancements in AI technology, combined with the significant fluctuations in China’s capital markets in recent years, efficient, objective, and disciplined quantitative stock selection strategies have become prevalent in the private equity fund sector to tackle complex markets with systematic advantages.

So, what are the specific advantages of quantitative stock selection strategies? How do they differ from index enhancement? What has been their historical performance over the past few years? How can ordinary investors select outstanding quantitative stock selection strategies? In the following text, the author will provide detailed answers.

1. The Essence of Quantitative Stock Selection: From “Experience-Driven” to “Data-Driven”

Quantitative stock selection is an investment method based on mathematical models and computer algorithms, systematically analyzing vast amounts of data to select stocks. Unlike traditional subjective stock selection, it does not rely on analysts’ subjective judgments but instead objectively screens stocks across the entire market by establishing systematic rules. In simple terms, quantitative stock selection is “predicting the future using historical patterns and overcoming human weaknesses with algorithms.”

The logic of traditional subjective stock selection often involves fund managers forming “bullish” or “bearish” judgments on individual stocks based on macroeconomic conditions, industry trends, or company fundamentals (such as financial reports and research), and then constructing a portfolio. However, this model is limited by human cognitive biases and information processing capabilities—human brains struggle to analyze vast amounts of data simultaneously and maintain absolute discipline amidst market fluctuations.

Quantitative stock selection completely overturns this process: it transforms investment strategies into computable “factor models,” identifying key variables that influence stock prices (such as valuation levels, profitability, K-line volume-price relationships, etc.) through historical data mining, training models to predict future stock returns, and ultimately generating trading instructions automatically through algorithms.

2. How Does Quantitative Stock Selection Work? 3 Mainstream Methods

Quantitative stock selection strategies vary widely, but the core can be summarized into three main methods, with different private equity firms choosing to focus on different directions based on their research capabilities.

Multi-Factor Models: The Classic “Stock Selection Formula”

The multi-factor model is the “cornerstone” of quantitative stock selection, and its core idea is that the future returns of a stock can be explained by multiple “factors” (i.e., key indicators affecting returns). “Factors” are the basic units of quantitative stock selection. The predictive power of a single factor is limited and prone to failure. Strategy researchers conduct extensive historical backtesting to identify combinations of factors that have proven effective in the A-share market over the long term.

Common factors include: value factors (valuation indicators such as PE and PB), growth factors (performance growth rates), momentum factors (price and volume trends), etc.

Statistical Arbitrage: Capturing Price Discrepancies from “Mispricing”

The logic of statistical arbitrage is based on “mean reversion”—the price relationships of certain correlated assets (such as stocks in the same industry, companies in upstream and downstream supply chains, ETFs and component stocks) tend to remain stable over the long term, but may deviate from the norm in the short term due to emotional fluctuations, and quantitative models can capture this deviation and profit from it.

For example, if A and B are leading companies in the same industry, and their historical price ratio stabilizes at 1.5:1, if one day due to market speculation, the A/B price ratio rises to 1.8:1 (deviating from the historical mean), the model would short A and long B, waiting to close the position for profit when the price ratio returns to 1.5:1. This strategy relies on strict statistical testing to ensure that the “correlation” genuinely exists and is not coincidental.

Event-Driven: Uncovering Immediate Opportunities from “News”

Events that affect stock prices frequently occur for publicly listed companies (such as earnings releases, mergers and acquisitions, executive share buybacks, and favorable policies). Event-driven strategies use quantitative models to monitor these events in real time and quickly assess their impact on stock prices in terms of direction and magnitude, generating trading signals. The key to this type of strategy is having “clearly defined events + quantifiable impacts” to avoid subjective interpretation.

3. Advantages and Potential Challenges of Quantitative Stock Selection

Compared to subjective stock selection, the core advantages of quantitative stock selection are:

1. Discipline: It avoids interference from human emotions (such as chasing highs and selling lows or premature profit-taking) and strictly implements model signals;

2. Efficiency: Computers can process multidimensional data for thousands of stocks in seconds, covering a breadth unattainable by human effort;

3. Diversification: Quantitative stock selection strategies typically hold hundreds of stocks, helping to reduce the risk of individual securities.

In the past five years, 2021 and 2025 were structural bull markets, 2022-2023 were bear markets, and 2024 was a year of significant volatility transitioning from bear to bull. Based on these advantages, in the past five years (2021-2025), private equity funds employing quantitative stock selection displayed lower drawdowns, higher returns, and a higher Sharpe ratio compared to subjective stock selection.

According to data from Private Equity Ranking, except for 2024, which was a transitional year, the median returns of quantitative stock selection slightly lagged behind those of subjective stock selection, while in all other years, quantitative stock selection significantly outperformed subjective stock selection. If we look at the average returns, quantitative stock selection has led over the past five years.

In terms of drawdown control, whether from the median or average perspective, quantitative stock selection only had a greater overall drawdown than subjective stock selection in 2024, while performing better in the remaining four years.

The Sharpe ratio, a classic metric for measuring “how much return can be earned for each unit of risk taken,” serves as a “cost-performance benchmark” for fund products, comprehensively considering the volatility of fund returns and the final returns. Since, over the past five years (excluding 2024), quantitative stock selection has achieved higher overall returns and better drawdown control, it also has a higher Sharpe ratio.

Of course, the above is merely a comparison of overall performance data and does not imply that subjective stock selection strategies are not viable; many subjective stock selection strategies from private equity funds also perform exceptionally well. At the same time, quantitative stock selection strategies face some challenges:****

1. Factor Failure Risk: Changes in market conditions (such as regulatory policies or adjustments to trading rules) may cause historically effective factors to fail. For instance, during previous market hotspots, funds may have favored companies with market capitalizations below 5 billion, but at some point, funds might prefer stocks with market capitalizations above 20 billion, indicating the risk of factor failure related to market capitalization.

2. Model Homogenization: If multiple private equity firms use similar factors (such as placing emphasis on “low PE + high ROE”), it may lead to crowded strategies, making it difficult to profit from market discrepancies and diluting excess returns.

3. Black Box Risk: The decision logic of certain complex models may be difficult to explain, and real-world performance may differ significantly from backtesting results. It’s like hiring a “stock guru” to trade for you, but he never explains why he buys or sells. Because you cannot understand his operational logic, you don’t know if he has true skill or just luck, nor do you know when he might suddenly “malfunction.”

For example: a quantitative model discovers that stocks with “numbers in their names” have an 80% probability of rising every month over the past five years. If the model buys stocks with “numbers in their names” this month according to this strategy, but ends up incurring substantial losses, this illustrates the “black box” risk; you wouldn’t know it merely identified a coincidental pattern of “having numbers.”

4. Overfitting Trap: A stock selection model may perform perfectly on historical data but fails to adapt in real-world trading. This happens because the quantitative model appears perfect during historical data testing not because it has learned “profitable patterns,” but because it mistook previous noise (coincidental occurrences) for patterns. Once the market environment changes slightly, such a model will quickly become “out of sync” and start to incur losses.

4. What Are the Differences Between Quantitative Stock Selection and Index Enhancement?

As a strategy within quantitative long positions, how does quantitative stock selection differ from the index enhancement strategy, which also uses quantitative models?

Quantitative stock selection and index enhancement share the same origin and underlying genetics: both employ quantitative models, machine learning, and other quantitative methods to screen stocks and construct portfolios. The core difference lies in whether they invest “with an anchor.”

Index enhancement strategies have a clear benchmark index (e.g., CSI 300), akin to dancing while wearing “shackles,” seeking enhancement while closely tracking the index and strictly controlling tracking errors. In contrast, quantitative stock selection strategies are entirely unconstrained by specific index components and styles, offering greater freedom, with the goal of maximizing absolute returns.

The unique advantage of quantitative stock selection arises from this “unrestrained” nature. By removing constraints, it can more fully leverage the advantages of quantitative models in breadth coverage, flexibly capturing rotation opportunities across different styles and industries, aiming for greater return elasticity, and providing a purer tool for investors seeking a higher risk-return ratio.

Therefore, quantitative stock selection seeks absolute superiority through “sifting through the sand,” excelling in flexibility, while index enhancement seeks relative superiority through “adding finishing touches.” Investors who prefer clear indices may choose index enhancement, while those seeking higher absolute returns and who can tolerate greater volatility may consider quantitative stock selection.

5. How Should Ordinary Investors Screen Quantitative Stock Selection Strategies?

For ordinary investors, there is no need to delve into the mathematical details of models, but they can observe the effectiveness of quantitative stock selection strategies through the following dimensions:

1. Stability of Excess Returns: Focus on whether the strategy can outperform the market across bull and bear cycles (e.g., whether the annualized excess return has remained positive over the past three years);

2. Risk Control Capability: Pay attention to the maximum drawdown and Sharpe ratio of the strategy; (Refer to:)

3. Team Research and Development Strength: The core competitive advantage of quantitative stock selection lies in its ability to discover factors and iterate models, so it is important to consider whether the team has a background in financial engineering and model iteration capabilities.

In summary, quantitative stock selection is not a “sure-win” magic trick; it is a combination of science (data, models, statistics) and art (factor selection, parameter tuning, responding to market changes).