In today’s financial markets, investors often face a paradox: the more popular a stock becomes, the more capital it attracts, yet this collective behavior itself alters the asset’s risk–return profile. According to Morgan Stanley’s 2025 report, concentration in the U.S. equity market has reached historic highs, with the top ten stocks accounting for 32.5% of the S&P 500’s market capitalization—surpassing the 33% level seen during the dot-com bubble in 2000. While this capital concentration has boosted short-term returns, it has also embedded higher systemic risk, as collective withdrawals during a market reversal could magnify price volatility.

In the spring of 2025, BlackRock’s research team published a groundbreaking study in the Journal of Beta Investment Strategies, introducing “crowding” as a quantifiable risk factor in equities for the first time. The study not only revealed the systemic impact of crowded trades on stock performance but also provided investors with a practical framework to identify and manage this risk. Based on market data from 2014 through September 2023, the research covered 2,000 U.S. equities and analyzed institutional holding patterns disclosed in SEC 13F filings.

From Market Phenomenon to Risk Factor

Traditional equity risk models primarily focus on company fundamentals, such as value and quality, as well as market characteristics, including volatility and liquidity, while largely overlooking a critical dimension: investor behavior. The concept of crowding addresses this gap by uniquely capturing demand-side dynamics—that is, the collective impact of capital flows on asset prices. According to Morningstar analysis, this demand-side effect was particularly pronounced in 2025, as algorithmic trading accounted for 60% of U.S. equity volume, leading to a high degree of strategy synchronization.

To enable quantitative measurement, the BlackRock team analyzed U.S. institutional 13F filings and developed an innovative crowding framework. The framework classifies investors across three core dimensions: equity assets under management (AUM) to distinguish institution size, active share to gauge the level of active management, and turnover to assess trading activity. For example, institutions with a median equity AUM of around $1 billion tend to be more active, whereas large institutions—such as pension funds managing over $100 billion—typically track indices passively.

The study further segmented investors into two groups: “large, passive, low-turnover” and “small, active, high-turnover.” Findings show that the latter group—including small institutions, retail investors, and foreign investors without U.S. branches—most accurately represents the phenomenon of crowded trades. The holdings of these “crowding drivers” in stocks above the median tend to exhibit higher price sensitivity. In the first half of 2025, the holdings of such groups in tech stocks reached 45%, significantly above the overall market average.

The Systemic Impact of Crowdedness

After conducting empirical tests on market data from 2014 through September 2023, the research team reached a clear conclusion: crowding represents an independent systemic risk factor. It exhibits very low correlation with traditional risk factors, yet statistically has a significant impact on stock returns and demonstrates robustness across different market environments. Specifically, the factor’s correlation with the Fama-French five-factor model is below 0.05, highlighting its distinct independence.

Stocks with high crowding exposure generally exhibit abnormal return patterns, including negative alpha, higher volatility, deeper drawdowns, and positively skewed return distributions. The factor’s annualized return stands at -1.1%, with an information ratio of -0.91, indicating that crowding risk demands a corresponding risk premium.

In other words, investors chasing popular assets are effectively taking on additional fragility risk, as prices can reverse sharply when capital flows out. Data show that, when sorted by crowding quintiles, stocks in the highest exposure group have annualized volatility 25% higher than the benchmark and maximum drawdowns 15% deeper. This effect was particularly pronounced during the 2022 bear market, when high-crowding tech stocks experienced an average drawdown of 35%.

Why Crowding Leads to Risk

To understand why crowding has become a risk factor, one must start with supply–demand imbalances. When many investors simultaneously chase the same assets, prices are driven above fundamental values, creating what can be described as an “artificial premium.” This premium is inherently fragile, as any slowdown in inflows or sudden capital withdrawals can trigger a chain reaction of price declines.

Historical data show that such imbalances are particularly pronounced during bubble periods. During the 2000 dot-com bubble, for instance, the Nasdaq surged 500% before collapsing, wiping out nearly $5 trillion in market capitalization and pushing the unemployment rate close to 6%.

In today’s markets, this risk is even more acute. ETFs and index funds concentrate capital in a handful of large-cap stocks, social media accelerates the spread of collective behavior, and algorithmic trading synchronizes similar strategies across the market. Together, these factors amplify the market impact of crowding.

In 2025, net inflows into equity ETFs reached approximately $750 billion, fueling market concentration and driving some stocks to extremely high levels of crowding. According to JPMorgan’s Global Strategy and Quantitative Research team, the current crowding of high-beta stocks has reached the 100th percentile, and the jump from the 25th to the 100th percentile occurred in just three months—the fastest rate in 30 years.

Adding further complexity, crowding interacts dynamically with other risk factors. It can sometimes amplify traditional factors—for example, boosting momentum returns by 20%—or it can offset them, such as weakening value factor recoveries by 30%. This helps explain why traditional risk models often show biases in empirical tests. According to BlackRock’s multi-factor model, incorporating crowding improves overall explanatory power by more than 20%.

The Value of Crowding in Investment Management

Crowding is not merely an academic concept; it carries broad practical applications in investment management.

First, in risk management and portfolio construction, incorporating a crowding factor into risk models allows investors to better identify potential vulnerabilities, thereby reducing losses from concentrated exposures during market adjustments. For example, in 2025 stress tests, portfolios that included crowding experienced 15% lower drawdowns and a 0.2 improvement in the Sharpe ratio.

Second, in strategy optimization, quantitative investment strategies often degrade as they become widely adopted. Monitoring crowding can help managers detect trends of strategy overcrowding in a timely manner, enabling them to adjust model parameters or explore alternative approaches. BlackRock recommends embedding crowding as a “meta-factor” within models, which can prevent up to a 30% decay in alpha.

Finally, in asset allocation, investors can enhance portfolio robustness and achieve more sustainable long-term risk-adjusted returns by diversifying across assets with different crowding levels. Specific approaches include tilting toward low-crowding small-cap value stocks, which can deliver annualized returns 2–3% above the benchmark, or using equal-weight ETFs to mitigate large-cap concentration risk—flows into such ETFs doubled in 2025.

Rethinking the Risks of Herd Behavior

BlackRock’s research serves as a wake-up call for today’s markets. The excessive concentration of large-cap tech stocks, the continued expansion of passive investing, and social media-driven trading decisions all contribute to increased market crowding. For investors, this means that simply chasing popular stocks or strategies not only exposes them to traditional market risks but also to additional crowding risk.

This risk typically manifests most forcefully during periods of market stress. In 2025, market concentration and the AI bubble ranked among the top five risks, and according to BBH Capital Partners, these factors could amplify volatility through crowding channels.

Although crowding has now been established as a new systemic risk factor, related research remains in its early stages. Future areas of exploration include the behavior of crowding across different market environments, cross-border transmission mechanisms, and interactions between crowding and other market anomalies.

As data availability improves and computational methods advance, the measurement and application of crowding will become more sophisticated, with related investment tools and hedging products gradually emerging. BlackRock’s investment outlook indicates that in the second half of 2025, liquidity buffers and crowding-monitoring tools are expected to become standard components of portfolio management.

In Conclusion

Overall, BlackRock’s research represents a significant advancement in market understanding: investor behavior itself has become a source of systemic risk. In today’s highly interconnected, algorithm-driven markets, understanding and managing crowding risk is no longer optional—it is essential.

For investors, this means shifting from passively following market trends to proactively analyzing capital flows and collective behavior. Beyond identifying potential opportunities, it is critical to assess the degree of crowding and the associated risk exposure. Goldman Sachs projects that if earnings growth slows to 6.5% in 2025—below the expected 8%—highly crowded positions could face additional pressure.

Ultimately, successful investing is not just about choosing the right markets, but also about recognizing when those markets have become overly crowded. By integrating crowding metrics, investors may achieve a transition from defensive to more proactive strategies, capturing more resilient returns amid uncertainty.