Private Equity Returns in Public Markets
What this covers
Patrick interviews Dan Rasmussen about where leveraged buyout returns came from and whether listed stocks can reproduce them. Dan's evidence is a study of 390 buyouts that issued public debt. Revenue and EBITDA growth slowed afterward, margins barely moved, capital expenditures fell, and debt rose in 70% of cases from roughly 2x to 4-5x EBITDA. He reads this as mean reversion in firms picked for past growth, not as failure, and calls operational turnaround the first myth of the industry. He grants one exception: cost-cutting specialists such as 3G Capital and Vista have a real skill, because software and consumer goods businesses have fat to trim.
The conversation then widens into forecasting and economics. Dan draws on Phil Tetlock's research to argue that expertise raises confidence without raising accuracy, and that forecasts should start from base rates. Growth is hard to predict (about 50% persistence), while deleveraging is more knowable (about 60%). He also challenges Porter's five forces by tracing it to a theory the Chicago School tested and rejected. Other threads include why the high yield spread matters more than interest rate levels, how quarterly marks make private equity look less volatile than it is, and why Japan's institutional network makes cheap levered stocks there less risky.
The most contested claims are the forecast that deals at 11-12x EBITDA will lag public equities after fees and long lockups, and the argument that good ideas are the ones that sound bad.
Rasmussen argues private equity's 1980-2010 outperformance came from buying cheap, small, highly levered companies rather than operational skill, that this edge has eroded as capital flooded in, and that a systematic base-rate approach that screens for cheap, levered firms with deleveraging can capture the same returns in public markets.
- Private equity returns were driven by purchase price and leverage, which raise free cash flow yield, not by operational improvement, as shown by flat margins and slowing growth after buyouts
- Expertise does not improve forecast accuracy, so base rates and reference classes beat narrative judgment
- Rising competition has pushed entry multiples above 10x EBITDA, where historical returns no longer beat public markets
Private equity returns come from entry price and leverage, not operational value creation.
- 3G and similar firms (e.g., Vista Equity) show a real, underwritable skill in cost-cutting, because software and consumer packaged goods businesses have fat to cut and EBITDA margins often double after acquisition, whereas growth is very hard to improve systematically.
“EBITDA margins often double after their acquisition.”
- Private equity's reputation rests on three myths: that it creates value through operational change, that it is lower risk and less volatile than public markets because of quarterly subjective marks, and that past returns predict future returns because managers ignore the factors (price and leverage) that actually drive outcomes.
“I think the three persistent myths of private equity are first, that private equity makes money through operational improvements.”
- Private equity valuation matters more than public valuation because deals are roughly 65% debt-financed; raising the purchase price increases both debt and interest, which compresses free cash flow yield non-linearly and raises bankruptcy risk for expensive LBOs.
“I would say that private equity valuations are even more important than public equity valuations.”
- Post-acquisition operational claims made by private equity firms are largely window dressing; in aggregate the playbook is to buy fast-growing firms, let growth mean-revert, lever them up, and use cash to service and repay debt.
“Everything afterwards about operational improvements is to some extent, window dressing.”
- Firms whose average buyout multiple is above 10x EBITDA should not be expected to beat public equities in aggregate; deals at 11-12x EBITDA, where prices stood in 2018, will likely lag after large fees and 7-8 year lockups.
“any firm where you're seeing the aggregate or average buy out multiple for their deals north of 10 times EBITDA, you shouldn't expect those funds to return higher than the public equity markets.”
- Across 390 private equity deals that issued public debt, revenue and EBITDA growth slowed in the three years after acquisition, margins barely improved (a few basis points), debt rose in 70% of cases from about 2x to 4-5x debt/EBITDA, and capital expenditures fell.
“We looked back, we found out 390 private equity deals that issued public debt.”
- The number of businesses going up for auction has not changed; what changed is the number of competitors bidding for them, as institutional allocators rushed capital into private equity and firms raised huge funds to compete for the same limited deals, pushing prices to 11-12x EBITDA and leverage to 6-7x net debt to EBITDA.
“The number of businesses going up for auction has not changed. What has changed is the number of competitors in the auction.”
- Private equity firms select companies that grew fast in the past, and because growth rates mean-revert, post-acquisition growth slowing is expected rather than a sign of failure.
“every private equity firm in their underwriting is selecting for high growth firms”
- After acquisition, private equity portfolio companies showed almost no margin improvement, perhaps a few basis points, despite the industry's emphasis on cost-cutting and efficiency.
“we found almost no improvement in margins, at least nothing particularly noticeable, maybe a few basis points.”
- For 3G's buyouts, EBITDA margins often doubled after acquisition, which Dan argues shows a real cost-cutting skill investors could underwrite, though 3G uses its own capital so outside investors cannot access those returns.
“EBITDA margins often double after their acquisition.”
- The first myth of private equity is that it creates value through operational change, the idea that brilliant Harvard-trained managers can turn businesses around, and the rebranding from leveraged buyout industry to private equity to growth equity obscures this.
“the first myth of private equity is that private equity creates value through operational change”
- After leveraged buyouts, capital expenditures decline because cash flow shifts to interest payments and debt repayment, so the buyout's real effect is a change in capital structure rather than investment in the business.
“we also observe is a decrease in capital expenditures, which makes sense. If you just took out a lot of debt, you're probably not going to want to invest in capital expenditures”
- Institutions happily pay two and twenty fees to lock up capital in private equity for ten years, yet refuse lockups in smart factor-based public strategies, an inconsistency that suggests their preferences are about structure rather than economics.
“Oh, no, no, no we would never do that. Why would we want to lock up our money? We hate lockups.”
Cheap, small, levered value investing explains private equity returns and can be replicated publicly.
- PE deals bought at less than seven times EBITDA, levered around 65%, and with small market caps near $200 million produced strong returns, and the vintages at those valuations dramatically beat public equities, while deals above 10x EBITDA did not.
“the deals done at less than seven times EBITDA do really, really well.”
- Backtests of levered small value in Japan showed about 17% standard deviation, roughly half the volatility of the Nikkei, which seemed impossible until the explanation was found in Japan's bankruptcy backstop dynamics.
“the volatility on this thing is like half that of the Nikkei”
- For levered equities, what matters most is the direction of leverage: a company actively paying down debt, especially toward its industry median capital structure, captures returns from debt repayment and multiple expansion as lenders and investors grow comfortable.
“It's the direction of leverage... How much debt did they pay off in the last year? Let's divide that by a market cap.”
- Size does not appear strongly predictive of returns in Rasmussen's research, especially in Japan and recently in the US, so the fund does not weight it heavily, and 'small' is kept in the strategy name only because most levered value opportunities are small.
“size, at least in our research, especially in Japan and outside the US, and more recently in the US does not seem to be overly predictive.”
- The cheap 1980s and 1990s LBOs done at about seven times EBITDA worked because the increase in leverage far outweighed the increase in interest payments, while the same mechanism reverses at high valuations.
“the 1980s, 1990s LBOs, done at seven times EBITDA. In those cases, you had the increase in leverage far outweighing the increase in interest payments.”
- Buying a company at six to seven times EBITDA with 50 to 60 percent leverage implies a free cash flow yield near 20 percent to equity, and returns should approximate that yield, which is the core logic of the levered small value strategy.
“What leverage does, is it just Jacks up free cash flow yields.”
- Believing that industry structure and competitive advantage predict returns leads investors to buy large-cap glamour stocks with elaborate justifications, a strategy that Dan argues does not work in practice.
“It's going to lead you to come up with elaborate intellectual justifications for buying large-cap glamor stocks.”
- High-yield market panics are liquidity-driven, with people deciding no one will ever lend again, yet lending typically resumes within about twelve months, so investors must hold through them rather than sell at the bottom.
“These are liquidity-driven panics where people say, 'Oh, my gosh, nobody's going to lend to a company ever again,' and of course, 12 months later, they're lending to companies again.”
- The statistically significant predictors of buyout performance are size, value, and leverage, which Dan would select as the three key variables for forecasting private equity outcomes.
“I think probably the first three would be the ones that I would pick size, value and leverage.”
- Dan argues that his firm's public equity portfolio can replicate the economic attributes of private equity buyouts, namely small size, cheapness, and leverage, without needing operational change.
“buying small, cheap, highly levered companies should be a good investment strategy based on the historical evidence from private equity.”
- Japan has about 3500 listed companies, nearly as many as the US, and many cheap stocks, making it a large and low-valuation opportunity set for levered small value investing.
“There are 3500 publicly listed companies in Japan, so you have almost as many companies to choose from in Japan as you do in the US”
Forecasts should rest on base rates and knowable inputs, not expert narratives or manager confidence.
- Good investment ideas are those that sound bad but are actually good, so they scare away enough other investors; anything with consensus is unlikely to offer a good opportunity, which is why investment opportunities are almost always found in non-consensus contrarian places.
“I think anything, where there's consensus, is unlikely to provide a good investment opportunity.”
- The strategy accepts only locked-up capital, typically three years, because it forces investors to commit to the full cycle; extending investors' time horizon is the easiest way to keep them from losing money, since bootstrap simulations show longer holding periods dramatically improve outcome distributions.
“The easiest way to help people not lose money in equity markets is to extend their timeframe.”
- Forecasts should start from base rates and statistically significant predictors rather than individual narratives; a person's fit to a stereotype matters less than how often that category actually occurs (nurses far outnumber librarians), and PE deals should be forecast by comparable size, value, and leverage outcomes.
“What Tetlock and Kahneman found is you have to use base rates.”
- Tetlock's study of thousands of experts over about 10 years and roughly 30,000 forecasts found that field expertise does not improve forecast accuracy; it only raises confidence, creating a dangerous calibration error for investors.
“expertise in a given field does not improve forecast accuracy. All it does is increase your confidence in the accuracy of your forecasts.”
- For deep value, positions should be averaged in over one to three quarters, because large price drops often signal weaker next-quarter fundamentals, so buying a full position at once risks buying on a bad signal.
“If you see a stock that's dropped 50%, most likely the next quarter's earnings are not going to be too hot.”
- Relying on a manager's 30-year track record or confidence in a specific deal, such as 'Seth thinks this is a good investment', is a flawed way to forecast compared with reference-class base rates.
“You should use that reference class judgment rather than saying, "Well, Seth has been doing private equity for 30 years."”
- Investors should weight forecasting inputs by the product of their importance and their knowability, because many important variables such as five-year growth are essentially unknowable, while others like debt paydown are both important and somewhat predictable.
“I think that investors have to multiply importance by knowability.”
- Tetlock's super forecasters do not stop at base rates; after completing a base-rate assessment they make individual qualitative adjustments, and this has been found to improve their forecasts.
“his super forecasters don't tend to just stop at the base rates, they actually do make individual more qualitative adjustments at the end”
- A company that has grown above the median for three years has only about a 50 percent chance of growing above the median next year, so growth is largely unpredictable even though it is very important to outcomes.
“A company that has grown above the median for the past three years has about a 50% chance of growing above the median next year. It's just pure chance.”
Private equity's reported edge reflects cheap leverage and judgment-based marks, not skill.
- Mitt Romney's early investing success likely came from operating in cheap, levered, small private businesses rather than from superior stock picking, as shown by his letters noting that he bought at 4-6x EBIT in an era when competitors were bidding up auctions.
“We used to buy things at four to six times EBIT, but we're having trouble finding things at those valuations”
- Private equity looks less volatile only because firms mark illiquid portfolios quarterly by judgment; if the same leveraged small-company portfolios were held publicly they would be considerably more volatile than public markets.
“The private markets don't experience that because it's just based on the judgment of an accountant or the private equity firm itself.”
- Private equity was probably the best place to invest from 1980 to 2010, beating the public market by about 6% a year net of fees, but it has underperformed public equities since 2010 according to Cambridge Associates benchmarks.
“From 1980 till 2010, it was probably the best place you could possibly invest, returns, net of fees for about 6% in excess of the public market.”
- Institutional investors hold a firmly entrenched belief that private equity will beat the market by about 4% a year; roughly half hold that view, while under 10% expect parity.
“I think it's something like 50% of institutional investors believe private equity will outperform the market by 4% per year.”
Disciplined reading and out-of-sample testing beat expert opinion.
- Ray Dalio's Bridgewater method is to start with a simple logical principle, test it across many market cycles and geographies because most good ideas fail, then trade it live to see whether it works out of sample.
“Ray's approach to investing is, so first come up with a logical principle. ... Then third, if both of those two things work, try it out in live trading”
- Reading about 50 pages a week compounds over a year and a decade into an exponential learning advantage over people who do not study, and Google Scholar and meta-analyses make the research base accessible.
“if you read even 50 pages a week, over a year and then over 10 years, the exponential amount you'll have learned”
- Most good ideas in investing and business do not work, so the discipline of testing them across cycles, geographies, and out-of-sample periods is essential, and experts who purvey failed ideas are no better than laymen with logic and common sense.
“since most good ideas don't work, test that logical, simple principle across as many market cycles”
- The kindest thing anyone has done for Dan was his mother spending an hour or two each day reading to him as a child, which instilled a lifelong love of learning.
“my mother from the age, probably when I was little, barely able to talk until I could read on my own, would probably spend an hour or two a day reading to me.”
- The top-ranked name on Dan's global screen was a small-cap Canadian mining company trading at 1.2 times EBITDA, which deep reading showed to be unbuyable for about half a dozen reasons, illustrating why human review is needed to sanity-check the algorithm.
“the number one thing would be a small-cap Canadian mining company that trades at 1.2 times EBITDA.”
Simple fundamental checks beat elaborate forecasting and strategic planning.
- The useful fundamental checks are simple: covenant breaches, maturities due within a year, or loss of a major customer that would cut EBITDA by about 20 percent, not elaborate DCF models or management meetings.
“'Oh, I didn't realize they'd already breached their covenants,' or 'Oh, Holy smokes, all their maturities are due in a year'”
- Building a DCF model that forecasts five years of revenue, income, and cash for a complex business like Microsoft is the height of idiocy and hubris, since people cannot even forecast their own finances two years out.
“It's just the height of idiocy and hubris.”
- Because the future is hard to predict, elaborate strategic plans and models are a poor basis for decisions, and it is better to focus on craftsmanship and serving customers well than on consultant-driven planning.
“predicting the future is really, really difficult. I think most efforts to plan things are bad because those are not going to come true”
- Tetlock argues it is nearly impossible to predict outside roughly a one-year horizon, which implies investors should limit their thinking to shorter horizons rather than planning two or three years out.
“if Tetlock says it's nearly impossible to predict outside of a one-year horizon, certainly we should be only thinking”
- Reading 10-K filings can reveal covenant breaches, debt maturities due within a year, and loss of a major customer representing about a 20% hit to EBITDA, which change a company's prospective valuation relative to its trailing valuation.
“Oh, I didn't realize they'd already breached their covenants... Or I didn't see that they actually lost their biggest customer and that's a 20% hit to EBITDA”
Private equity is overpriced, flow-driven, and shielded from accountability.
- Private equity's 10-year lockup combined with subjective quarterly valuations is a perfectly designed wrapper that prevents institutions from making the buy-high, sell-low mistakes they make in public markets, though it hides bad outcomes for five to ten years.
“Private equity in some sense has the perfect wrapper for an equity product.”
- PE firms assert operational value creation but never offer systematic evidence such as tracked results or forecast accuracy, so their claims rest on anecdotes, which is equivalent to claiming an Olympic sprint time without ever timing oneself.
“The fact that they never offer any evidence for their operational improvements other than anecdotes”
- Small private businesses with 15 to 100 million dollars of EBITDA are selling for 11 to 12 times EBITDA and about 6 to 7 times net debt to EBITDA, a competitive market Dan compares to a gold rush.
“these tiny, little businesses, 15, 20, 30, 100 million of EBITDA and they're selling for absolutely insane prices. 11, 12 times EBITDA, six to seven times net debt to EBITDA.”
- Large institutional investors such as Gulf sovereign and pension funds are increasing private equity allocations, and the resulting flows will keep driving purchase prices higher for years.
“you're going to see massively increased flows into the asset class for years to come. That's going to continue to drive purchase prices higher.”
Experts' added confidence exceeds their accuracy, so most good ideas fail.
- Financial advisors who make poor recommendations for clients also make the same poor decisions with their own portfolios, which suggests the problem is adherence to ideas that don't work rather than bad intent.
“they actually made the same bad decisions that they made for their clients.”
- Most good ideas do not work, and experts who champion them are often no better or worse than laymen armed with common sense and logic, because they are more confident rather than more accurate.
“One is that, most good ideas don't work. I think the second would be that experts who are often purveyors of good ideas that don't work are often no better or worse”
- Over three years of quarterly forced rankings, Rasmussen's team could barely distinguish its best picks from its 30th-ranked picks, but the least-liked names slightly underperformed the rest of the portfolio.
“We found actually very little ability to distinguish between say, number one and number 30.”
Opportunity depends on regional credit spreads, not interest rate levels or home bias.
- In Japan, the bank, government, and company network effectively prevents public companies from going bankrupt, with perhaps one of 3,500 companies failing per year, so almost any cheap levered Japanese stock is a lower-risk way to run a levered small-value strategy.
“they will not let public companies go bankrupt... we can basically buy anything in Japan, because nothing is going to go bankrupt, except for maybe one out of 3500 companies a year”
- High-yield spreads in mainland Europe are so tight that levered value opportunities are hard to find there, and the fund has almost no exposure to France or Germany, though some UK opportunities exist.
“high yield spreads in Europe are so tight that, it's just hard to find great value opportunities in mainland Europe.”
- The high-yield spread, rather than the level of interest rates, is the volatile signal that matters for levered equities: when the spread jumps from a historical 3-4% to around 9% in a year, levered equities face painful drawdowns, and high-yield panics are liquidity-driven and short-lived.
“the much more volatile time series other than interest rates is actually high yield rates and the spread between high yield and the prevailing interest rates.”
- Limiting investing to one's home country, when about 15,000 equities are accessible, has no strong justification, since headquarters location should not predict returns; Rasmussen's global fund is about 40% US, 35% Japan, with the rest in emerging markets and the UK.
“why would you limit yourself to the 3000 that happened to be in your country”
- Going back to 1964, there is essentially no correlation between interest rate levels or changes and equity returns, and small value and levered small value have done well in rising-rate environments, because rates tend to rise with good growth and inflation.
“What we saw is that there's no correlation really, between interest rates and equity returns.”
Judgment, overlooked corners, and tracked results matter more than screens alone.
- Rasmussen's team reviews financial statements and earnings transcripts for red flags on top of its quantitative top-40 screen and phases in positions over one to three quarters after large price drops; over three years of tracking, about 70% of their judgment calls added value relative to the equal-weighted screen.
“we found that about 70% of the time, we're right. We can add value relative to our simple screen.”
- Screens for likely fraudulent small-cap biotech and tech stocks, with huge market caps and no revenue, generate alpha against a negative market index but still lose money outright, so shorting them is hard to profit from.
“they all lose money, which I think is interesting because I like making money.”
- Rasmussen wrote his undergraduate senior thesis, later a book, about the largest slave revolt in American history in New Orleans, reconstructing participants' origins and movements from property, legal, and court records mapped in Excel and Google Maps.
“I wrote my senior thesis about a slave revolt in New Orleans that was actually the largest slave revolt in American history.”
- Reading empirical research widely and thinking while walking, instead of producing strategy documents, is a more effective way to generate investment insight, because compounding reading over years produces an outsized learning advantage.
“if you read even 50 pages a week, over a year and then over 10 years, the exponential amount you'll have learned”
- Great investors such as Buffett, Romney, Kravis, and Roberts likely succeeded less by being superior stock pickers than by locating corners of the market that were undervalued and that others dismissed as too risky.
“they found a corner of the market that was deeply undervalued?”
Leverage lifts equity yield, and deleveraging persists more predictably than growth.
- Levered small value needs a red team that asks why a cheap, levered firm might be cheap for a reason, and a blue team that applies quantitative quality and bankruptcy filters, such as excluding C-rated credits, high short interest, and low Piotroski F-scores, to avoid bankruptcy.
“to almost think like Blue Team, Red team.”
- Leverage mechanically raises free cash flow yield to equity: a firm bought at 10x cash flow is a 10% unlevered yield, but at about 50% debt it becomes roughly 20% minus interest, so levered cheap firms can match high-yield unlevered firms and widen the opportunity set.
“What leverage does, is it just Jacks up free cash flow yields.”
- Firms that paid down debt last year pay down debt again next year about 60% of the time, versus about 50% for growth persistence, so deleveraging is both important and more knowable than growth; a Bayesian boosted tree model using US data back to 1964 lifts accuracy to about 70%.
“If you just said, "Let's look at all firms that paid down debt last year, what percent of them will pay down debt next year?" The answer is 60%.”
Porter's five forces and industry-structure theory lack empirical support.
- The most dangerous thing in investing is increasing confidence without increasing accuracy, and frameworks like Porter's five forces produce elaborate justifications for buying large-cap glamour stocks, which is not a strategy that works.
“anything that increases your confidence without increasing your accuracy is the most dangerous thing in investing.”
- Porter's five forces framework is a reframing of the structure-conduct-performance theory, which assumes market share drives market power and profit margins; that theory was empirically tested and found to fail in the 1970s by Chicago School economists, so the framework has been empirically wrong for decades.
“It's been empirically wrong since the '70s.”
- Chicago School industry-concentration and profit-margin studies found no reliable relationship between a firm's market share or industry structure and its profitability, which undermined the antitrust rationale that high share yields market power.
“they found that it's not true, it's not true at all. There's no relationship and industry as a factor does not predict almost anything about a company.”