Public vs. private equity: The widest equity return gap in two decades

• Public equities have repriced the future while private equity is still monetizing the past. Since the October 2022 trough, the S&P 500 has compounded at more than 20% a year for three straight years, while Private Equity buyout portfolios assembled at the 2020-2021 peak, and now carrying more expensive debt into a thin exit market, have lagged on every benchmark and horizon. At this point in time, the gap between liquid and illiquid equity returns seems to be the widest in two decades.
• The rally that opened the gap is narrow and earnings-led, not a liquidity melt up. The Magnificent Seven alone delivered more than half the S&P 500's three-year total return, strip them out and the index is an ordinary performer, and the equal weighted version has compounded at about half the pace, a gap unseen since the late 1990s. But the move rests on delivered earnings and a real AI capex super cycle rather than multiple expansion, so the premium looks more durable than the dot-com peak.
• Yet private equity's long-run premium is real, this cycle has just interrupted it. On a like for like basis, matched to public indices for geography, size, sector and leverage and net of fees and carry, buyout has beaten public markets over three decades, with MSCI estimating pooled direct alpha of roughly 400bps a year since 1994. The exception is the 2021-2023 vintages, which currently show a negative direct alpha of about ~800bps against the MSCI ACWI, the first consecutive run to trail public markets in the series. Consequently, the data points to rough PE performance ahead even allowing for the fact that these vintages are still young and J-curve distorted.
• Private equity can no longer count on rising valuations and must now drive returns by growing the underlying business. For over a decade, buyout returns came mostly from rising valuations, selling companies for a higher multiple than was paid. That is mostly gone as borrowing costs have roughly doubled and exit valuations have stopped climbing, so returns now come from growing the business. The contrast on the financing side is also large, with the largest listed companies still holding more cash than debt, so benefiting from higher rates, while buyout owned companies sit on floating rate debt due to be refinanced in 2026-2028, where higher rates only add to their interest bill.
• As capital drains from private equity, private debt is catching the outflows. The same jump in interest rates that made buyout deals more expensive made lending the more attractive trade. Big long term investors continue to like senior private credit because it pays a steady, contractual cash yield, its value barely moves and losses have been low, none of which buyout has delivered lately. So new money is rotating out of private equity and into private debt.
• 2026 will decide whether the public-private gap was a passing extreme or a lasting regime change. Our central case assumes an economic mix that keeps returns positive but well below the post-Covid peaks with public equity settling in the low teens, while buyout follows in the mid-teens through 2027-2028. All in all, the way forward for private equity seems to be less about how much to own and more about how to own it. This means favoring managers and vintages with a demonstrable edge over generic exposure, treating liquidity and valuation discipline as features to be tested rather than assumed and resisting the temptation to buy the label without the underlying skill.

Since the October 2022 trough, public equities have staged one of the sharpest re-ratings in modern memory.  The S&P 500 has compounded at more than 20% a year for three straight years, with other large-cap benchmarks close on its heels. But the headline flatters the market. The rally has been extraordinarily narrow: The Magnificent Seven (Mag 7 - Apple, Microsoft, Alphabet (Google), Amazon, Nvidia, Meta (Facebook) and Tesla) alone delivered more than half of the S&P 500's three-year total return. Stripped of those seven names, the index is a thoroughly ordinary performer. The dispersion shows up in valuation, too. The top ten names now trade at a roughly 50% forward price/earnings premium to the rest of the index, wide by any historical yardstick. But that premium rests on an earnings path well above the dot-com peak of the 2000s, a sign that the fundamentals are more resilient now than they were then (Figure 1). Earnings breadth is mending as well (Figure 2), if grudgingly: Barely a third of S&P 500 members beat the index in 2025, up from a record low the year before. At the same time, small caps have scarcely shown up at all. The equal weighted cut of the same indices tells a more sober story still, with the S&P 500 equal weight compounding at about half the pace of its capitalization weighted parent, a gap unseen since the late 1990s. All in all, what reads at the index level as a broad bull market is, mechanically, the re-rating of a very small number of balance sheets.

The scale of the recent gap is easier to appreciate against private equity's long-run record. On a like-for-like basis, comparing each buyout fund with a public index matched for geography, size, sector and leverage, and measuring excess return net of fees and carry, private equity has, in fact, beaten public markets over the past three decades. MSCI estimates pooled buyout direct alpha of roughly 4% a year since 1994. That long-run premium is precisely what the current cycle has interrupted. Broken out by vintage, buyout funds delivered positive alpha against global equities in almost every cohort since the mid-1990s. The exception is the 2021-2023 vintages, which currently show direct alpha of roughly -5% to -12% against the MSCI ACWI, the first run of consecutive vintages to trail public markets in the series. Of course, those readings come with an important caveat since the funds are young, and early fee drag (the J-curve) mechanically depresses returns that typically recover as portfolios mature, so those returns are provisional rather than final. But even allowing for that, the message is consistent: The vintages bought into the cheap-money peak are, so far, the worst relative performers in a generation and may still lead to some struggling years ahead (Figure 3).

The character of the move matters as much as the magnitude. This has not been a liquidity-driven melt up of the 2020-2021 variety. It has been an earnings- and margin-led re-rating built on top of a genuine capex super cycle (Figure 4). Along these lines, forward earnings for the S&P 500 are up by roughly a third since the 2022 trough, and the bulk of the index's price return is explained by delivered earnings rather than multiple expansion. That distinction matters for how we read the gap to private markets performance (Figure 5). Public equity has not simply re-priced risk, it has also re-priced a specific, identifiable growth thesis around AI infrastructure, hyperscale compute and the operating leverage of asset-light platforms. Private equity, holding a portfolio assembled for a different regime, has had no comparable thesis to ride, and software, one of its biggest post-Covid allocations, has suffered from major potential disruptions. The second feature worth flagging is what the rally has done to the investable opportunity set itself. Free float in the largest names has shrunk as buybacks have accelerated and insider holdings have stayed concentrated, even allowing for the supply that the latest mega-IPOs have brought back. The effective supply of mega-cap equity available to incremental buyers is therefore tighter than headline market capitalization suggests (Figure 6). At the same time, the US now has roughly half as many listed companies as it did at its late 1990s peak, and the firms that do go public are older and larger than they used to be. So when a private equity sponsor comes to sell a company today, the public market it can exit into is smaller, more concentrated in a few large names and pickier about what it will absorb than the market sponsors were counting on when they bought in over 2018–2021. Consequently, the gap between public and private valuations in that light does not look like a temporary mispricing that will simply close, but a lasting result of a deeper shift in where growth capital is raised and how it is priced.

The narrowness of the public equity rally deserves closer inspection, because it is both the source of the divergence with private equity and a risk factor in its own right. The Mag 7 now represent roughly a third of S&P 500 market capitalization, up from around an eighth a decade ago, and their combined market value exceeds the GDP of every country except the US and China. Of course, the concentration is also self-reinforcing, because as these companies grow, passive and benchmark tracking flows mechanically direct more capital toward them, deepening their index weight independently of any fresh view on fundamentals.

Earnings concentration is also at extremes, and it now runs in parallel with the concentration in price. The same handful of mega-caps that dominate index weight also generate a disproportionate and rising share of S&P 500 earnings, so the benchmark's profits, not just its market value, increasingly depend on a single cohort. On valuation, the index trades well above its long-run average forward multiple, but the premium is almost entirely a mega-cap phenomenon; strip out the largest names and the rest of the market sits much closer to its historical norms, leaving a two-tier index in which the most expensive part is also the largest part (Figure 8). Because so much of the index now sits in long duration growth names, the multiple is also unusually sensitive to moves in rates and growth expectations; a small shift in either can swing the level materially. As an anchor for that sensitivity, a reversion of the cluster's valuation premium to its long-run average alone, before any change to underlying earnings, would cost the S&P 500 close to 10-20% of its value.

The same concentration invites two sharply opposing readings, and the contest between them is itself a market risk. The bullish case argues that the Mag 7 are nothing like the 2000 dot-com leaders as they generate hundreds of billions of dollars of free cash flow a year between them, sit on net cash balance sheets and dominate the decade's most powerful secular trend: AI compute, cloud infrastructure and advertising-data networks. Consensus forward earnings growth comfortably exceeds the broader market's, and on a PEG (Price/Earnings-to-Growth ratio) basis the cluster trades at parity or a discount. Nonetheless, the bearish case argues that the AI thesis rests on hyperscale capex sustaining at current levels, that enterprise AI revenue has yet to scale into the returns that capex implies and that the concentration itself is a source of fragility. An earnings disappointment from one or two names could de-rate the whole index in days. Both views can be right at once but which one prevails comes down to whichever is dominating the narrative at the time.

Why the concentration matters for the public-private comparison is subtler than it first appears. The Mag 7 are themselves “long duration” growth assets, capex heavy and R&D intensive, with revenue streams that compound over a decade rather than monetize in a five-year LBO window. In effect, US public markets have absorbed the kind of growth equity exposure that used to migrate to private markets. Combined hyperscale capex now approaches the entire global venture capital (VC) deployment. Private investors who, in earlier cycles, would have funded the next generation of compute or AI infrastructure through late-stage venture rounds can now access that exposure through publicly listed mega caps, at any size and at low friction.

The corollary is uncomfortable for private markets. The universe of high-growth, scaled, profitable technology companies available exclusively to private investors has shrunk. At the time of writing, some of the private mega caps remain private holdings (though not for long, as SpaceX, Anthropic and OpenAI have IPO'd or filed to do so), but the depth of comparable opportunity in private markets has narrowed precisely as fundraising and deployment has slowed.

Private equity, by contrast, is still digesting the previous cycle. The asset class entered the 2022 rate shock with peak fundraising, peak entry multiples, peak leverage and aggressive growth assumptions. The 2020-2021 vintages, in particular, were priced for a world of cheap money and continued multiple expansion, a world that ended within 18 months. Even top-quartile buyout funds returned single-digit IRRs in 2025, well behind both the S&P 500 and MSCI World, and the average vintage of the past decade barely cleared mid-single-digits.

The lag has many causes, but the most mechanical is a reset in the cost of capital. A typical deal a decade ago cleared at around 10 times EBITDA, with half its capital in single-digit yield debt; today's clears closer to 13 times, with materially less leverage at far higher financing costs. Sponsors paid those record prices expecting to exit into further multiple expansion. That expansion has not arrived, and because thinner leverage cushions the entry, the drag on equity returns runs deeper than the change in multiples alone would suggest.

The mismatch is as much about sector mix as about timing. Public benchmarks are now dominated by mega-cap technology, semiconductors and communication services, the very cohort the AI capex cycle has re rated, while financials, healthcare and industrials make up the balance. Private equity's deployment skews the other way, with software and technology being the single largest buyout sector for most of the past decade, peaking at roughly a third of deal value in the 2020-2021 vintages, with healthcare, business services and industrials filling out most of the remainder.

The catch is that the software the public market rewards and the software private equity owns are not the same exposure. Listed technology leadership is concentrated in hyperscale compute, AI infrastructure and platform businesses with net-cash balance sheets; the typical Private Equity-owned software asset is a mature, leverage-financed vertical or horizontal SaaS business bought on a high revenue multiple, precisely the cohort most exposed to slowing seat growth, AI-driven displacement risk and a higher discount rate. So even where the two universes appear to overlap on the word "software," they sit on opposite sides of the AI trade: public markets long the enabler, private portfolios long the potentially disrupted incumbent.

What ultimately separates public and private equity today is not how much they have returned, but what is driving those returns. Strip a buyout down to its parts and the value created comes from three places: the business growing its earnings, the multiple on those earnings rising between purchase and sale and the debt used to finance the deal. For most of the past decade, one of those did the heavy lifting. It was not, as is often assumed, leverage; across successive cohorts of exited deals, debt has been close to neutral and frequently a quiet drag once its rising cost is counted. The real engine was the multiple. Sponsors bought well, watched valuations climb and sold into a market that reliably paid more than they had, with growth in the underlying business playing only a supporting role.

That engine has stalled. Purchase multiples sit at or near record highs, so there is little room left for further re-rating, and exit multiples are no longer drifting up to reward patient owners. With cheap leverage gone and the multiple tapped out, only one source of value is left standing: the business itself. Revenue growth, once the junior partner, is now the dominant contributor to private equity returns, and the deals that work are the ones where the company genuinely grows rather than the ones where the multiple obligingly expands.

This is exactly where the two universes part ways. Public market returns are still, overwhelmingly, a re-rating story, but a narrow one, concentrated in a handful of net-cash mega caps that the AI trade has repriced upward. Private markets have been denied that same rerating, and the businesses sponsors own carry debt rather than cash, so their returns now have to be earned the hard way, one point of revenue growth at a time. Put simply, public equity is still being lifted by the market; private equity increasingly has to build its returns from the ground up. The bar has risen accordingly: only operationally well-versed managers in genuinely growing sectors can clear it.

The exit market is where the Private Equity cycle's hangover shows most clearly. Buyout funds hold tens of thousands of unsold portfolio companies worth trillions of dollars, and median holding periods have stretched markedly. The starkest symptom is liquidity, the cash investors actually get back. Unlike a public equity, which can be sold on any trading day, a private holding returns capital only when the fund exits a deal, and that channel has narrowed sharply. Distributions to LPs (limited partners, the investors who fund the buyout) stayed below the 20% mark for a fourth straight year, the longest such stretch on record, and a depth last seen during the 2008-09 crisis. After three years of investors paying in more than they took out, net cash flow edged only modestly above breakeven in 2025, a change in direction rather than a true recovery.

The other exit routes tell the same uneven story. Deal and exit values rose sharply from a low base in 2025, but the rebound was narrow, driven by a handful of mega-deals while the mid-market stayed constrained. The IPO window has cracked open, but aftermarket pricing suggests public investors are still cautious on sponsor owned businesses, with Private Equity-backed listings tending to trade less buoyantly than venture backed ones. The secondary market has absorbed part of the gap, transacting a record volume in 2025 at a blended 87% of NAV (net asset value, the carrying value of the fund's holdings), with continuation vehicles taking a meaningful share, but it clears only a small slice of the unsold inventory each year. The plumbing is working again; the backlog behind it will probably take years to drain.

Private credit deserves its own section because it has been the structural beneficiary of the very forces that have pressured private equity. The same rate reset that raised LBO yields and made buyout math harder turned lending into the more attractive side of the trade, and as banks retreated from leveraged lending under regulatory pressure and sponsors still needed acquisition finance, direct lending stepped into the gap at unprecedented scale. Global direct-lending AUM has more than quadrupled since 2018 and is on track to keep growing through 2028. All-in yields on senior unitranche loans to mid-market, sponsor-backed borrowers remain in the low double digits, equity-like returns earned from a senior, secured position in the capital structure rather than the bottom of it. That combination is exactly what private equity has failed to deliver over the past three years: For long-duration insurers and pension funds, private credit has provided contractual cash yield, low mark-to-market volatility and a credit-loss experience that has stayed below long-run averages despite higher rates.

Two cautions are worth registering. First, private credit's default experience has been benign, with realized losses well below long-run averages, but the asset class has not yet been tested through a full credit cycle at current scale. The borrower base is also more levered than at any prior measurement point, and payment-in-kind (PIK) structures, in which coupons accrue rather than pay in cash, have grown sharply, suggesting some borrower-level stress that is being deferred rather than recognized. Second, private credit is increasingly intertwined with the insurance balance sheets that own much of the asset class. The IMF and the ECB have flagged this as a systemic concern; when liquidity pressures emerge, insurance regulators may impose forced de-risking, creating sudden price-discovery events in a market built on stable, marked-to-model valuations. Neither concern is a near-term catalyst. Both argue for portfolio construction that favors senior secured exposure, granular underwriting and clear pricing discipline over chasing the highest-yielding mezzanine and second-lien strategies.

The broader implication for private credit is that the asset class should be treated as a core, structural holding rather than a tactical reach for yield. The case does not depend on rates staying high. Even as financing costs normalize, the bank retreat from leveraged lending looks structural, the borrower base is largely captive and the asset's contractual cash yield and low mark to market volatility answer precisely the liquidity problem that has made private equity so hard to hold. For insurers and pension funds matching long-dated liabilities, that makes private credit a natural anchor allocation rather than a trade. But “core” tends to mean senior, not adventurous. The allocation belongs at the top of the capital structure, in senior secured exposure run by managers who can underwrite through a full cycle. Mezzanine, second-lien and opportunistic strategies still earn their place as return enhancers, but they are best treated more tactically, sized around that senior core rather than forming it. The suggested posture, in short, is a larger and more permanent allocation held more conservatively.

The next 12 months will decide whether the 2022-2025 divergence between public and private market returns was a cyclical extreme or the start of a longer regime change. The central case assumes a prolonged but contained Middle East conflict, the Federal Reserve maintaining high rates and the AI capex cycle sustaining without a major guide down, a configuration that keeps returns positive across asset classes but well below the peaks of the post-Covid recovery. Under those conditions, public equity returns would be moderate into the low teens. Consumption and construction sentiment, the two biggest supports for the S&P 500 since 2010, hold up, while higher for longer rates shift from neutral to a modest drag. Private equity follows with a lag, with buyout returns settling in the mid-teens through 2027-2028 as financing conditions stabilize and the income flows that drive most of the variation in buyout returns recover from their 2023-2024 trough .

Three drivers will decide which scenario plays out, each operating on a different horizon and through a distinct channel. In the near term, the resilience of the Middle East deal is the dominant variable. If oil and gas trade continue to normalize through the negotiated settlement, the path for both asset classes would look very different than if disruption to the Strait of Hormuz reignites beyond September. In that second case, a forced Fed tightening would turn rates from a modest drag into an active headwind for public equity. In this regard, consumption which is itself sensitive to rates through mortgage costs and credit availability, and the US consumer would enter any oil driven tightening from an already fragile position. The shock would also hit the real economy directly, with construction activity stalling as input cost visibility would collapse and commodity pass through would erode real incomes. Under that path, public equity would turn negative in 2026, with buyouts following one to two quarters later as valuations reset with their characteristic lag.

Over the medium term, the path of rates and the sustainability of AI capex are likely to become the binding constraints, and their interaction will determine whether the baseline recovery holds or extends. A higher for longer rate environment would delay the recovery of buyout profits, holding the downside path well below the cost of illiquidity in 2026. If AI capex were to guide down, whether through enterprise revenue disappointment, regulatory friction or balance sheet discipline, large cap tech margins would be hit first, and operating leverage means the damage would reach returns faster than the analyst revision cycle would suggest. From there it would propagate into buyout exit valuations. At that point a 10 to 15% public market de-rating is the single factor most capable of turning a baseline buyout year into the 4% downside outcome.

The upside case requires both the near- and medium-term constraints to ease together, and when they do, the dynamics are mutually reinforcing in a way that produces outsized returns. In this scenario, with the Middle East conflict de-escalating after the June 2026 agreement, the Fed gains leeway and AI investment accelerates toward the 1990s capex cycle, with digital technology spending and business investment rising sharply. In that configuration the outcomes are assertive. Public equity peaks above 20% in 2027 as the consumption, margin and Fed rate drivers align at once, historically the setup that produces the sharpest single year public equity returns. Buyout surpasses 20% shortly after as multiple expansion feeds through into exit valuations and money supply conditions ease, with the equity market PE ratio and the interest and profit rate drivers all contributing at the same time. Europe participates, but structurally less so. With the ECB's more gradual easing delivering less impulse than the Fed, Eurozone growth stays subdued and the AI preparedness gap, lower adoption capacity and smaller hyperscaler exposure mean the productivity gains accrue disproportionately to the US, compounding the transatlantic divergence of the post-2020 period.

The central message of the downside is not the magnitude of the return shock but its asymmetric transmission across asset classes. In the downside scenrio, public equity would reprice immediately and sharply as margins and consumption deteriorate together under the stagflationary mix of input cost pressure and slowing growth, then would recover gradually as base effects and eventual policy normalization rebuild the factor stack. Private equity NAVs would adjust with a two to three quarter lag, technically positive in 2026 but close to the cost of illiquidity. The losses would be more contained in return space yet materially worse in liquidity terms, and nowhere is that gap more visible than in distributions. Historically, borrowing spreads and the CAPE ratio drive distribution activity, with rates the most persistent drag since 2022 and equity issuance, a proxy for IPO activity, the cyclical amplifier that has been conspicuously absent through the drought. The mega-IPO pipeline is where that factor becomes concrete. Successful mega-IPOs at premium multiples would lift the equity issuance driver, validate private AI marks across the buyout book and reopen the exit channel for sponsors holding inventory, converting years of paper gains into LP cash on a scale that dwarfs any prior exit event. Nonetheless, a collapse of the June agreement and renewed Iran escalation would break the sequence entirely, with spreads widening, the CAPE ratio deteriorating, rates staying higher and the IPO window closing before the queue clears.

All in all, the story is, at heart, a story about the fading power of beta. For most of the last two decades, simply being in the room was enough, public markets rode a long bull run and private equity compounded on cheap leverage and steadily expanding multiples. That era has closed on both sides of the divide. Public benchmarks now carry their returns on a narrow cohort of AI-exposed mega caps, while private equity is still working through vintages bought at the top of the cycle. What used to be a rising-tide question, how much to allocate to private markets, has become a question of which managers, which vintages, and which exposures. The center of gravity has shifted from beta to selection.

That shift is not a matter of conviction but of arithmetic. The dispersion between the best and worst private equity managers has widened to the point where the median fund tells you very little, and the returns that justify the asset class increasingly sit in its upper reaches. Skilled selection, on the evidence, is now worth more than the allocation decision itself, but it is a genuine skill, not a default setting, and an allocator without a durable edge is as likely to give that premium back as to capture it. Beta has become a commodity, the scarce input is judgement.

For investors looking forward, that reframes the central question. The debate worth having is no longer "public or private," but "differentiated or undifferentiated" within each, because passive index beta and generic, broad-exposure buyout have quietly become the same crowded trade, priced for a tailwind that is no longer blowing. The democratization of private markets, through evergreen and semi-liquid wrappers, only sharpens the point as it widens access to an asset class precisely as access stops being the thing that pays, and the funds that actually generate the premium remain the hardest to reach. The way forward is less about how much to own and more about how to own it. Investors should favor managers and vintages with a demonstrable edge over generic exposure, treating liquidity and valuation discipline as features to be tested rather than assumed, and resisting the temptation to buy the label without the underlying skill. The next market correction will be the first real test of which private market structures were built to endure and which were built to sell, and the distinction between robust and fragile will become visible quickly. The alpha, on both sides of the liquidity divide, now belongs to those with the conviction and the capability to select it.