Jasper Capital Fundamental Thoughts (Part One) | Quantitative Fundamentals: A Niche Path Less Travelled

In 2013, the American hedge fund AQR published an academic paper titled Buffett's Alpha, in which the authors conducted a quantitative analysis of Warren Buffett’s investment decisions over the years and developed a mathematical model to characterize his stock-picking preferences.

Interestingly, the authors attempted to deconstruct the "Oracle of Omaha's" moneymaking logic using six categories of factors: quality, beta, market, size, value, and momentum. An investment portfolio modeled on these factors, following a “Buffett-style” approach, significantly outperformed market indices over the long term, with performance comparable to Berkshire Hathaway itself. This paper demonstrates the effective combination of quantitative thinking and subjective fundamental stock selection. It raises an intriguing question: if we have a sufficient sample size, can we replicate the stock-picking logic of outstanding fundamental investors using quantitative models and consistently beat the market? This leads us to the topic at hand: quantitative fundamental analysis.

Chart 1: Performance of the stock market, Berkshire Hathaway, and the simulated Buffett-style portfolio.

Quantitative Fundamentals: A Road Less Traveled

Quantitative fundamental strategies experience a slower decay in alpha and can handle larger capacities. However, they remain relatively niche within China's quantitative hedge fund space. Due to the high difficulty of research and modeling, managers often use these strategies as a supplement to price-volume strategies. Quantitative fundamentals are like an unattainable ideal—everyone acknowledges their potential, yet few can truly master them.

Effectively integrating quantitative and fundamental thinking requires a complementary balance of internal and external skills. The universe breadth of a active fundamental equity manager is typically narrow, while their research dives deep into the operating models of the companies that they invest in. In contrast, quantitative investors use their data advantage to develop more generalized models across a broader range of stocks with more diversified alpha sources. However, without a deep understanding and forward-looking judgment of both macro and microenvironments, the breadth of information may fail to compensate for the lack of depth in individual stock research and could even lead to pitfalls.

Take financial indicators as an example: financial data fields are abundant, but only investors who have deeply studied the industry and understand accounting details can discern which fields accurately describe a company’s true fundamentals. Moreover, the methods of financial fraud are constantly evolving, with many financial statements hiding intricacies. For instance, some companies may bury inflated revenues within different balance sheet items, such as inventories in agriculture or goodwill and intangible assets in TMT sectors. Capturing these opportunities amidst layers of obscurity requires a deep understanding of the roles played by different accounting categories.

Jasper’s Quantitative Fundamentals: The "Iron Triangle" Team Balancing Breadth and Depth

As a quantitative manager that has consistently balanced fundamental and price-volume factors since 2017, Jasper has accumulated substantial experience on this niche path. The company's "Iron Triangle" fundamental team has developed a strong rapport over the years, with members boasting interdisciplinary backgrounds in neural networks, artificial intelligence, computing, and accounting, ensuring both research breadth and depth.

Chart 2: Jasper’s quantitative fundamentals research team.

In terms of breadth, the company’s proprietary IT systems and intelligent research platform track various databases in real-time, including industry and corporate data, extracting core information from vast datasets. For industries that are hard to predict, we also build separate models using alternative data, complementing the overall market model to ensure strong stock-picking capabilities across different sectors.

The advantage in depth is reflected in the precision and differentiation of research. To maintain the long-term effectiveness of our strategy, we must explore more "uncharted territories." First is the deep analysis of corporate characteristics. Our team's diverse academic background ensures a high sensitivity to information about listed companies. For example, in the case of the earnings quality factor, we have built a series of financial indicators based on a deep understanding of accounting categories such as cash flow, receivables/payables, related party transactions, and inventory costs, resulting in an "earnings shock detector." This detector can automatically spot "cosmetic traps" in financial data, identifying companies that have manipulated their reports, helping the model avoid stocks with high potential for earnings shocks.

Furthermore, attention to the details of factor research is crucial. Below, we use the analyst factor as an example to illustrate the uniqueness of Jasper’s fundamental research. Analyst report indicators reflect the marginal changes in market expectations for company performance. There is a prevailing view that the analyst factor has been in decline since 2021—for instance, the performance of analyst coverage factors has seen a significant pullback since 2021, indicating a reduced market response to analyst expectations.

Chart 3: Analyst coverage factors have underperformed overall since 2021. Data source: Dongfang Securities.

Chart 4: The live performance of Jasper’s analyst-related factor, showing consistent predictive value. Data source: Jasper Capital.

Analyst-related Factor_Pure Alpha

We further dissect and restructure the analyst factor. For instance, as market participants, analysts themselves may make irrational decisions. Some analysts, upon seeing a significant rise or fall in a stock before publishing a report, may subjectively adjust their forecasts—this can be understood as analysts being influenced by short-term price-volume changes in the stock.

Therefore, analyst reports are further broken down into two parts: forecasts based on in-depth company research and predictions influenced by price-volume movements. The latter, often irrational, is filtered out. By extracting the essence of analyst reports and combining it with our own price-volume research, our model becomes more "objective" than the analysts themselves, allowing us to select higher-quality companies.

Conclusion

Quantitative investing is like an autopilot ship. Most of the time, we can trust the intelligent system to steer the course. But when the seas are rough beyond the system's recognition limits, an experienced navigator is needed to keep the ship on track. Fundamental information is akin to this navigator, preventing the ship from veering off course.

Today, globally renowned asset management institutions rarely rely solely on price-volume factors for investment decisions. As China's market matures and efficiency improves, the difficulty of extracting excess returns from price-volume strategies will gradually increase. We also see regulators repeatedly emphasizing the trend of lowering the frequency of quantitative strategies. As a manager who has deeply cultivated this field for nearly a decade, we believe that quantitative fundamentals will fill the missing pieces in price-volume strategies, ushering in the next golden era of quantitative investing in China.