Adnan KhanAdnanProduct & AI
San Francisco Bay Area
Get in touch

Source note3 min read

Quant Convergence: Bridging Classical Value Investing and Modern Factor Models for Systematic Equity Selection

Source

arXiv

arxiv.org/abs/2606.24575v1

Read the source

Source

Plain-Language Brief

This research explores the effectiveness of Benjamin Graham's classic value investing rules in comparison to complex machine learning models in the context of equity selection. The study found that the pure Graham Random Forest achieved the highest overall return with less risk, while the Combined Random Forest successfully mixed momentum with Graham's rules, balancing returns and risk better than other models tested.

What the Document Says

  • The research is about comparing Benjamin Graham's classic value investing rules to complex machine learning models for equity selection.
  • Three sets of features were built: pure Graham rules, modern market factors, and a mix of both.
  • These features were tested against XGBoost and AutoGluon using 20 years of S&P 500 data.
  • A four-year test period (March 2022 to March 2026) was applied with a buy-and-hold strategy.
  • The AutoGluon model captured high returns but suffered a substantial drop due to buying volatile tech stocks before the market crashed.
  • The pure Graham Random Forest achieved the highest overall return with less risk.
  • The Combined Random Forest successfully mixed momentum with Graham's rules, making a higher return while keeping the lowest maximum drop.
  • The study concludes that Benjamin Graham's 'margin of safety' is effective in preventing modern AI from taking on too much risk.

Key Concepts Mentioned

  • Benjamin Graham's value investing rules
  • Modern machine learning models
  • S&P 500 data
  • XGBoost
  • AutoGluon
  • Buy-and-hold strategy
  • Calmar Ratio

Claims Made by the Document

  • Benjamin Graham's classic value investing rules can act as a mathematical 'low-pass filter' for modern machine learning models.
  • The AutoGluon model captured high returns but suffered a substantial drop due to buying volatile tech stocks before the market crashed.
  • The pure Graham Random Forest achieved the highest overall return with less risk.
  • The Combined Random Forest successfully mixed momentum with Graham's rules, balancing returns and risk better than other models tested.

Evidence Used

  • Results from testing sets of features against complex machine learning models using S&P 500 data.
  • Four-year test period results with a buy-and-hold strategy.
  • Overall return and risk ratios for each model

Methods / Approach

  • Building sets of features, testing them against complex machine learning models using S&P 500 data, applying a buy-and-hold strategy over a four-year test period

Limitations Stated by the Authors

The research is limited by the specific data and time frame used (20 years of S&P 500 data from March 2022 to March 2026).

Limitations Noted by the Agent

Agent interpretation — not from the source.

The agent may have limitations in understanding the specific market conditions during the test period.

Open Questions

  • How would the results vary with different data sets or longer time frames?

Notes for Future Q&A

May include interpretation. Labeled where it does.

It is important to note that while this research provides evidence of the effectiveness of Benjamin Graham's rules, further testing and validation with additional data sets and longer time frames may be necessary to fully understand their applicability in various market conditions.

Raw Abstract / Excerpt

Modern finance relies heavily on complex machine learning models to find patterns in the stock market. However, as these AI models get more complicated, they often memorize short-term market noise instead of finding companies with real, lasting value. We designed this research to test if Benjamin Graham's classic value investing rules could act as a mathematical 'low-pass filter' to keep these modern models in check. We built three different sets of features - pure Graham rules, modern market factors, and a mix of both - and tested them against highly complex models (XGBoost and AutoGluon) using 20 years of S&P 500 data. By applying a strict buy-and-hold strategy over a four-year test period (March 2022 to March 2026), the results showed that more complex algorithms do not always win. While the AutoGluon model captured high returns (222.68%), it suffered a substantial 39.78% drop because it bought volatile tech stocks right before the market crashed. On the other hand, the pure Graham Random Forest achieved the highest overall return (232.13%) with much less risk (1.38 Calmar Ratio). Furthermore, the Combined Random Forest successfully mixed momentum with Graham's rules, making a 202.91% return while keeping the lowest maximum drop (34.53%) of any model tested. Ultimately, this research proves that Graham's 'margin of safety' isn't outdated; it is actually a highly effective way to prevent modern AI from taking on too much risk.