Price Discovery - Executive Summary

For all the parallels being made between the dotcom bubble and today’s AI enthusiasm, one basic consideration is most often neglected: how market participation has transformed over the past two-plus decades. As well-known and often repeated as the marquee stats may be, they remain staggering. There’s passive: in 2000, passive investing accounted for roughly 10% of US equity market AUM; today, passive’s share exceeds 53%. There are systematic strategies: AUM managed by quant hedge funds has skyrocketed from the tens of billions around 2000 to over $1.2 trillion today. There’s high frequency trading: total trading volume by HFTs in the US has surged from roughly 20% in the early 2000s to roughly 50% today. There’s retail participation: the percentage of US households with capital market exposure has risen by nearly 10% over the past two decades and retail investors now account for roughly 20% of daily trading volume in the US, a share that has doubled just in the past decade. And there’s foreign participation: non-US investors now account for roughly 18% of US equity market ownership, up from roughly 7% at the turn of the century.

As we detailed in our December report on “GenAI & Productivity” (learn more here), we believe dotcom comparisons are overblown. Just to contextualize what a dotcom-like crash would mean today, former IMF Chief Economist Gita Gopinath published this calculation in The Economist in October: “A market correction of the same magnitude as the dotcom crash could wipe out over $20 trillion in wealth for American households, equivalent to roughly 70% of American GDP in 2024.” It’s hard to see how wealth-destruction on that level could transpire when the epicenter of any so-called “AI bubble” is the most profitable companies in the world and when there are no signs of a slowdown in AI infrastructure spending. However, as we also argue in the report, we do believe excessive enthusiasm is priced into the AI trade today, with market participants underestimating genAI’s weaknesses and overestimating its strengths and trajectory of improvement.

To quote the FT’s Editorial Board: “[2026] will be dominated by hard-headed evaluation as AI comes under intense scrutiny over its practical reliability and commercial viability.” We saw a glimpse of the potential potency of this threat with Microsoft’s stock selling off after its earnings release. How significant is the downside threat if more rigorous ROI accountability meaningfully depresses AI sentiment? While by many measures valuations do not yet suggest dotcom-levels of euphoria, they are elevated. As of writing this, the S&P 500's forward P/E of 22.1 topped both its five-year moving average of 21.15 and its 10-year moving average of 20.3. Meanwhile, equity market concentration is at an all-time high, with the top 10 companies in the S&P 500 by market cap accounting for roughly 40% of the index’s total market cap.

Of course, AI is not the only risk factor the US equity market faces. Geopolitical instability is rising in almost every corner of the globe. Skyrocketing US deficits and questions about the Fed’s future independence are threating to spike equity-market risk premia. And the Trump administration continues to take radical steps to challenge the US and global economic status quo. In recent months, evidence that “US exceptionalism” is cracking has continued to mount, from record inflows to non-US ETFs in 2025 to accelerating de-dollarization, with the share of USD-denominated assets held by foreign central banks dropping to the lowest level since 1994 in 3Q25.

To start 2026, US equity market bullishness was at an extreme—for one indication, average cash holdings in portfolios dropped to 3.3% in December, the lowest in records going back to 1999, according to Bank of America. Earnings season has indicated why that bullishness was merited—as of writing this, S&P 500 constituents had reported earnings roughly 9% above estimates, outpacing the 10-year average by two percentage points. Nonetheless, valuations combined with macroeconomic and geopolitical uncertainties merit caution. How do investors and allocators approach risk mitigation in this environment? To our mind, answering this question without careful consideration of the seismic shifts in market participation over the past decade is a path to misguided decision making.

For years, equity market participation shifts have inspired heated debate. The primary concern: Is price discovery eroding—the power of active managers to tether stock prices to fundamentals? From 2010’s Flash Crash to 2018’s Volmaggedon to 2021’s GameStop short squeeze, moments of market turmoil have been held up as warning signs of mounting vulnerabilities in the makeup of markets that, in the right bubble-like circumstances, could result in systemic disaster. The chart below suggests why these fears may not be unfounded. It shows the rising frequency of upside or downside moves of five standard deviations or more amongst the 100 largest stocks in the S&P 500. Last year alone, there were 47 instances of such severe single-stock selloffs, the most in Barclays records going back to 1998.

Yet, whether it’s analyst and academic research or pundit fearmongering, so often the evidence held up as indicative of eroding price discovery hinges on a fatal flaw: correlation as causation. Yes, markets have grown more concentrated, but how much of that is due to the momentum bias of passive and quant strategies versus fundamental factors like industry consolidation and the network effects of digital technologies? Yes, markets appear more prone to shirk off bad news, but how much of that is due to the price insensitive, buy-the-dip mentality of retail investors versus a more accommodating Fed and more sophisticated and therefore, resilient corporations? So much of the debate around the implications of market participation shifts has been muddied by attempts to quantify the incalculable. Cherry-picked data is leveraged to blame passive or retail or HFTs for a perceived increase in market fragility. To our mind, more holistic consideration is what is required to understand how market participation shifts are influencing price discovery and what that could mean for risk mitigation and alpha generation in this environment.

So, we dove deep into the last twenty years of market history. We dissected analyst and academic research across all market participation frontiers. It’s clear market participation shifts have compromised the market’s reaction function. Over the past decade-plus, market efficiency has declined, herd behavior has intensified, and short-termism has compromised the market’s ability to adequately “price in” longer-term risks. This is a particularly acute concern at a time of both elevated enthusiasm and geopolitical and macroeconomic uncertainty. What follows dissects how we see the market participation equation today and how that influences our investment convictions at this time of extreme uncertainty.