Insight Active diversification: Mitigating collective risks

Aggregate preferences and collective risks 

Among the most consequential innovations in finance has been the development of passive market-index strategies, which today provide investors with broad, low-cost diversification across hundreds, if not thousands, of securities. By weighting securities according to market capitalization, these strategies reflect the collective information and judgments embedded in market prices—what is often described as the “wisdom of crowds.” Yet the same dynamics that enable markets to aggregate dispersed information can also produce pronounced mispricing. Market dislocations in 1929, 2000, and 2008 illustrate that collective behavior is not invariably rational, a point documented thoroughly in Charles Mackay’s Extraordinary Popular Delusions and the Madness of Crowds (1841), Charles Kindleberger’s Manias, Panics, and Crashes: A History of Financial Crises (1978), and Robert Shiller’s Irrational Exuberance (2000). 

Relying exclusively on capitalization-weighted index investments, therefore, means embracing both the strengths and the limitations of the collective preferences and judgments of market participants. This might alternatively be viewed as bearing collective risk – risk that arises when a broad set of investors all do the same thing. However, collective risk today goes beyond simply accepting the risks of irrationality, something that is already a salient feature of financial markets. With the proliferation of passive products, it also means accepting the unintended risks that come from increases in concentration and from a growing share of asset flows into passive investments. 

There is an expanding body of research that highlights several structural concerns associated with the continued adoption of passive investing. For example, Wurgler (2010) explains how indexing inflates the prices of, and induces excess co-movement among, index members relative to non-members.¹ More recently, Jiang et al. (2025) document that incremental flows into passive funds exert a disproportionate influence on the prices and idiosyncratic risks of the market’s largest constituents. Complementing this evidence, Brightman and Harvey (2025) find that stocks with high passive ownership exhibit increasing sensitivity to market-wide risk, while securities more widely held by active managers display declining market betas. Sorensen (2025) also argues that capitalization-weighted index exposures can mask significant embedded concentration and diversification risks, underscoring the need for more deliberate portfolio implementations that include allocations to active strategies.

Lochstoer and Muir (2026) argue that passive investing has changed not only who owns stocks, but how markets trade. As assets have migrated into index funds, ETFs, and futures, investors increasingly buy and sell the market as a single instrument, concentrating activity in systematic exposures rather than individual securities. They find that index-based trading has coincided with higher market-wide volatility, greater stock comovement, and a larger share of returns driven by common market factors. They estimate that daily market volatility has increased by approximately 127% since the mid-1950s. Green, Krishnan, and Sturm (2026) also consider the impact of passive investing and model how rising passive ownership weakens the market’s self-correcting pricing mechanism: passive funds buy regardless of valuation, while active managers no longer fully offset their flows. The result is greater mispricing of fundamental characteristics, higher concentration, and increased price inelasticity. Their model suggests that if passive ownership rises to roughly 65% (currently at approximately 50%), volatility could increase sharply — a risk they describe as hiding in plain sight. 

Contrasting these views, Kritzman and Turkington (2026) caution against treating current concentration as inherently alarming. Based on a historical analysis using the effective number of constituents (ENC) measure, they argue that today’s concentration has historical precedent, little predictive power for future returns, and is less concerning because the largest companies are often more diversified entities and less volatile.² Clarke, de Silva, and Thorley (2026), however, show why the current episode may be different. ENC measures concentration in terms of security weights, but not whether those weights are attached to unusually volatile stocks. Their volatility-weighted ENC analogue finds that the largest constituents now combine high index weights with above-median realized volatility, concentrating risk in fewer names than the capital weights imply — the first sustained episode of such a reversal in a century. This directly challenges the reassurance implied by the historical ENC analysis: today’s concentration is not merely high by weight but is also concentrated in names with high volatility. They further show that this pattern is not limited to U.S. technology stocks but has appeared across major global markets on staggered timelines and in different industries. The implication is that cap-weighted index funds may now carry materially greater volatility than similar vehicles a decade ago. 

These findings extend the notion of collective risk by showing how passive investing has changed both the ownership and behavior of markets. Price-agnostic flows increasingly bind securities to common vehicles, common factors, and common sources of volatility, while concentration has become more consequential because the largest weights are now attached to more volatile stocks. In this environment, broad market ownership alone may no longer provide the diversification investors expect. Addressing this challenge requires a more intentional approach: active diversification, or the use of active strategies to diversify portfolios using exposures that are substantively different from the market. 

However, the premise of active diversification rests on the idea that active strategies can be sufficiently different from the market to provide diversification benefits.³ Markowitz (1952) explained that a portfolio of railway securities would not be as well diversified as a portfolio that included railroad, public utility, mining, and various manufacturing securities. The characteristics of the assets held – whether they behave the same or differently at different times – matter for diversification purposes. As such, diversifying collective risk requires being different from the market index. A key differentiator for active strategies is the information used for security selection and portfolio construction. Understanding the information sources used provides important insights into the return characteristics of active strategies and, more importantly, what types of risk they can diversify against. 

In this article, we make the case that active strategies are an increasingly important source of diversification for investors that can provide portfolio benefits beyond benchmark-relative returns. We first define collective risk, provide an overview of its relationship with endogenous risk, and discuss why it is a key concern today. Empirical evidence of collective risk in passive investments is also provided to further motivate the need for active diversification. We then explore two forms of diversification offered by active strategies: uninformed and informed. This includes a review of foundational finance theory, which points investors toward active investing more readily than is often assumed, as well as empirical evidence on the persistence of active returns in equity markets around the globe. Finally, we provide guidance on implementing active diversification in practice. 

Collective risk, uncertainty, and diversification

We define collective risk broadly as risk that arises when many investors collectively do the same things. It is closely tied to the concept of endogenous risk in that it is a fragility that emanates from within the financial system as a function of collective investor behavior. Two examples illustrate the implications of endogenous risk. The first is Black Monday in 1987, when the failure of portfolio insurance strategies held by a large share of institutional investors induced a feedback loop of selling that sent the market tumbling 22.6% in one day. The second is the quantitative crisis of 2007, when many quantitative equity market-neutral funds were all running near-identical factor strategies that aggregated into a common risk factor. When one fund began deleveraging, it triggered margin calls and forced selling across the other funds. In neither case was the danger visible in any single portfolio; it emerged from how widely the same positions were shared. 

Collective risk is distinct from the traditional notion of endogenous risk in both scope and expression. The latter is mechanism-specific, as in the examples above: a particular strategy or risk-management rule compels many investors to trade in the same direction, producing a feedback loop that amplifies price movements. It is episodic and requires a trigger. Collective risk requires neither. Rather than being tied to any specific strategy, it is a function of collective ownership and asset flows, and it accumulates continuously, eroding effective diversification whether or not a dislocation occurs. Endogenous risk is what happens when collective positioning unwinds; collective risk is the condition that makes such an unwind possible. 

Collective risk builds as more investors, and more money, do the same thing. When price-agnostic asset flows, driven by investor demand rather than any assessment of value, dominate price discovery, the securities within a strategy become increasingly tied to the flows into and out of the strategy itself. Risk becomes less about the characteristics of the individual holdings and more about the risk of collective behavior. While capitalization-weighted market index portfolios theoretically have the greatest capacity, capacity is not the same as price neutrality, and new passive flows are not necessarily price neutral.⁴ Recent research, cited previously, indicates that collective risks are now extending into the realm of even large capitalization passive strategies. 

Compounding the issue, market-capitalization weighted exposures of all varieties, along the size spectrum and across regions, have become ubiquitous. They are now widely considered the default investment option for implementing asset-allocated portfolios and see increasingly regular flows from mechanical, buy-only retirement funding – purchases made regardless of valuation. 

A simple illustration provides insight on why this matters. Consider a single small capitalization fund with assets under management (AUM) approaching an estimated capacity limit. Liquidity and price impact from trades can be managed based on a fund’s holdings and trade impact estimates. Now consider there are 1,000 funds with identical holdings, weights, and AUM. The risk from the aggregate holdings of all the funds is not considered by any of the individual fund managers. No manager is imprudent, but all are exposed to the collective risk. This is a practical example of what is occurring across the passive investing landscape. 

Collective risk is therefore not a property of any individual holding, but of the investor base that owns it. When market participants share the same price-agnostic exposures, the same algorithms, the same vehicles, and the same buy-and-hold approach, their behavior ceases to be independent, and diversification depends on independence. 

If collective risk is a function of shared positioning, its consequences depend on the environment in which those positions are held – and that environment has grown increasingly volatile and markedly less certain. The past three decades have seen the proliferation of a broad range of new technologies across societies and economies. There have been massive increases in both computing power and the data available to extract answers to increasingly complex questions. A global data network infrastructure now allows ideas – both good and bad – to be transmitted virtually instantly to almost any point across the globe. And we are currently in the midst of an artificial intelligence revolution that is expected to have a significant impact on how we live, work, and continue to innovate. In short, this means that new ideas and innovations can be developed quicker, scaled faster, and can have a greater and broader impact in a much shorter time than at any other point in history. All of this produces change at a rate we have never before experienced, and change produces uncertainty. 

One expression of the resulting uncertainty faced by investors is the set of economic policy uncertainty (EPU) indexes, based on the work of Baker, Bloom, and Davis (2016), that reflect the relative frequency of country-specific newspaper articles containing terms pertaining to the economy, uncertainty, and policy-related matters. Exhibit 1 presents standardized EPU indexes for the world, the United States, Europe, the United Kingdom, and China over the period from 1998 through 2026. The trend is difficult to ignore: across every region shown, uncertainty has moved steadily higher and reached levels well above its long-run average in recent years, with spikes that are both larger and more frequent than those observed previously. Rising uncertainty is precisely the environment in which collective risk is most dangerous, because it widens the range of outcomes for which a portfolio must be prepared at the very moment that shared positioning has narrowed the range of risks that portfolio actually diversifies against. 

Taken together, these dynamics make the case for diversification – and, more specifically, for diversification that is substantively different from the market. If the dominant exposures investors hold are increasingly the same, and if the environment in which they hold them is increasingly uncertain, then the value of holding something different increases accordingly. The remainder of this article argues that active strategies are an increasingly important source of that difference. We begin, in the next section, by examining the clearest empirical evidence of collective risk in practice: the dramatic increase in concentration within passive products themselves.