The Bet With Buffett
What this covers
Patrick interviews Ted, who took the other side of Warren Buffett's ten-year wager that the S&P 500 would outperform a basket of hedge funds. The conversation covers how the bet was constructed, why Ted accepted it, and what followed. It then widens into how allocators should think about benchmarks, liquidity, and the cycle between passive and active management.
The middle stretch is technical. Ted notes that typical equity long/short funds carry about half the S&P's market exposure, and that "hedge fund" is a catchall covering debt, currencies, and other exposures, so the comparison only holds for one slice. Patrick presses on benchmarks. Choosing one is itself an allocation decision, and a custom benchmark would erase a manager's claim to alpha. They discuss crowding, with short interest rising from 2 to 3 percent to 15 to 20 percent, and the pattern of capturing upside and limiting downside that broke after Lehman. Ted's main reason for taking the bet was a high Shiller P/E, though Patrick stresses that valuation is a tilt on averages, not a timing tool.
The contested turns come later. Ted attributes Berkshire's long lag during the bet to passive inflows into cap-weighted indexes, and argues that once value and quality are packaged into ETFs, Buffett's record cannot be repeated. Patrick brings in AQR's Buffett's Alpha and the problem of retrofitting stories to past winners. They also cover permanent capital, David Swensen's diversification, David Salem's argument for only two buckets, and the gap between paper returns and what investors earn. Ted's advice for ordinary investors is a low-cost approach spread more widely than US large caps.
Seides argues the Buffett bet is not evidence that hedge funds are worthless or that the S&P 500 is the right default. The bet was a poorly specified comparison of a fee-laden basket against a market-cap index at a high valuation, and its real lessons are about allocator edge, permanent capital, and investor behavior.
- The bet compared fund-of-funds with the S&P 500, which has different market exposure, geography, and cap-size profile, so it was not an apples-to-apples test of fees.
- Fees were knowable in advance and were not the only driver; the S&P's starting valuation was the main reason Seides felt confident.
- Realized investor returns, shaped by drawdown tolerance and behavior, often diverge sharply from paper returns, and smaller drawdowns can matter more than headline returns.
Hedge fund value depends on manager skill and rotation, and should be judged before fees.
- Hedge fund fees are coming down, which shifts more of the gross return to investors and improves their net outcomes, especially in a lower-return environment where scrutiny of fees is rising.
“bringing the fees down, that means more of that return will go to the investors, and that accrues better for returns over the next period of time.”
- Most hedge fund allocation goes to multi-strategy funds rather than niche strategies, and the value of these funds comes from rotating capital among strategies through a combination of strategic asset allocation, tactical asset allocation, and bottom-up implementation.
“most of hedge fund investing, or most of the allocation to hedge funds, and most of the dollars, do not go to some niche strategy. They go to a multi-strategy fund”
- Even if hedge fund fees were zero or equal to Vanguard's, Seides would invest in some long/short funds but not others, meaning the value of hedge funds depends on manager selection rather than on fees alone.
“If hedge funds fees were zero, or effectively the same as Vanguard's fees, would you invest in long/short funds? And I would say, "I would invest in some and not others, even if there were no fees."”
- The best outcome combines the computer-like replicable strategy with a skilled practitioner, similar to Kasparov's loss to Deep Blue, which implies that manager judgment still matters even when the core strategy can be replicated cheaply.
“the optimal strategy isn't the computer, and it's not Kasparov. It's sort of the computer with a reasonably good practitioner sitting next to them.”
- Alternatives have wider return dispersion than public active management, so an investor needs an allocator's edge, meaning access to better managers, enough information, and the right type of capital, and without it they should not play the game.
“If you don't have that allocator's edge, and you don't think you should have it, you probably shouldn't play the game.”
- Niche alternatives such as music royalties, insurance settlements, and litigation claims have no index, so they cannot be replicated cheaply and require research and development to find the next opportunities, especially where banks have stopped participating.
“there won't be an index for music royalties or insurance settlements or litigation claims or those types. So there's all kinds of interesting niche things, and that's the R&D.”
- The fee burden of hedge funds was knowable ten years earlier, so asking whether hedge funds are terrible because of high fees is the wrong question, since the informative question is whether they add value before fees.
“In fact, the fee burden was knowable 10 years ago. So the existential question of. "Are hedge funds terrible because they charge high fees?" to me, it's not the right question.”
- Long/short equity funds continue to generate returns mainly in Asia and in less-trafficked sectors like consumer names, while crowding has made large US technology names harder to win in.
“where I've seen long/short equity funds continue to generate the types of returns their investors expect are outside the US, particularly in Asia.”
- High hedge fund fees finance extensive research and development, so some of the smartest investors sit in hedge funds, and winners are likely to be found there, though the question is whether their edge justifies the fees.
“Because of the high fees, the research and development effectively in hedge funds is much more extensive than it is more broadly in active management.”
- Around 2013 to 2016 hedge funds appeared to outperform net of fees under a careful benchmark, but then they stopped, which shows that there is always something in the markets the observer cannot fully explain.
“then one day, they stopped. And so there's something always going on in the markets that we can look back and think we understand and not.”
- Fund of funds imposed two layers of hurdles, where hedge fund managers must justify their fees by picking securities that beat the market, and fund of funds managers must justify their fees by picking hedge funds that add enough value.
“You had these two layers. You said, okay, are the hedge fund managers picking securities that'll beat the market? And they have to justify their fees.”
The bet's design and benchmark choice compare unlike risk exposures, making it an unfair test of hedge funds.
- The hedge fund basket differs from the S&P 500 by being more globally diversified and weighted toward mid and small caps, so its market risk is spread across different exposures rather than concentrated in US large caps.
“large-cap US stocks versus other. And in this case, the other is a lot more globally diversified, a lot more weighted to mid and smaller caps.”
- Typical equity long/short hedge funds carry about half the market exposure of the S&P 500, so a 50 percent S&P gain would be expected to produce about a 25 percent hedge fund gain, which makes the bet a comparison of different risk levels.
“let's think about equity long/short hedge funds. The two key differences are the level of market exposure, where the S&P 500, every dollar you put to work is a dollar exposed to the market.”
- The bet is not a clean test of whether hedge funds are worth their fees, because hedge funds and the S&P 500 differ in market exposure, geography, and capitalization, so the right question is whether hedge funds are valuable even if they charge no fees.
“the existential question of. "Are hedge funds terrible because they charge high fees?" to me, it's not the right question. Because you can ask the question, "Are hedge funds valuable if they charge no fees?"”
- Comparing a basket of hedge funds to the S&P 500 is like asking which Chicago team is better, the Bulls or the Bears: the two are different games, so the comparison measures category choice more than quality.
“I think the right analogy is to ask, "Which is better sports team? The Chicago Bulls or the Chicago Bears?"”
- When the fund-of-funds are measured against a fully invested global index, the result is almost a tie over the nine-year period because global stocks outside the US were also down, so the apparent gap shrinks under a closer comparison.
“If you measured these fund of funds against the Morgan Stanley World Index, which is not the right index, because it's 100% invested, it's almost a tie.”
- The bet used five fund-of-funds mainly for logistical reasons: they were stable enough to remain in business and keep track records over 9.5 years, whereas individual hedge funds would require rotation and introduce selection problems.
“Part of the reason for using fund of funds in the bet was just logistics. That the five fund of funds, and Warren put the numbers in, haven't been all that stellar but 9.5 years later, they're all still in business”
- Because Berkshire stock is more volatile than a hedge fund basket, Seides argued Buffett should post more collateral than Protégé, which started a negotiation about risk-adjusted collateral in the bet.
“if we're collateralizing hedge funds and you're collateralizing Berkshire's stock, one's more volatile than the other, so you need to post more collateral than we do.”
- The charitable wager was structured through Long Bets, a nonprofit for long-term wagers, because betting is legally difficult even for charity, and the handshake agreement ended up as a 25-page legal contract, far more complex than Buffett's typical single-page acquisition deals.
“he has bought companies on a contract on a single sheet of paper, and yet this kind of handshake charitable wager ultimately was a 25-page legal contract.”
- Buffett's bet framing treated the S&P 500 as an easy hurdle for hedge funds because it was the wrong index for a manager with market exposure, making the choice of benchmark a major part of why the bet looked lopsided.
“I thought Warren was patsy at the poker table, because he threw out the S&P as the index. I thought that was the wrong index to be picking, and it was something that was going to be an easy hurdle to overcome.”
- For the bet to be favorable, hedge funds only needed to return a mid-to-high single-digit return, so as long as the S&P underperformed that level the odds favored Seides's side.
“if you thought that hedge funds were going to do something independent and churn along, maybe generate a mid-to-high single-digit return, as long as you thought that the S&P would do worse than that, that was a good bet.”
- The bet was structured to compare Berkshire's pick of the S&P 500 against a basket of five fund of funds, with Protégé taking the hedge fund side and the results tracked annually, collateralized, and funded by a zero-coupon bond.
“I'll make it fund of funds instead of hedge funds. And there're a whole bunch of reasons why that was the case. And I picked five.”
- Hedge funds are a catchall category covering debt, commodities, currencies, rates, and esoteric financing exposures, so the S&P 500 comparison only makes sense if restricted to equity long/short strategies.
“Hedge funds, it's an interesting starting point, because hedge funds can be anything. It's a catchall.”
Investor behavior, especially drawdown behavior, determines realized returns more than vehicle choice.
- CalPERS's abandonment of hedge funds was about poor fit rather than a verdict on hedge funds, because its governance made it hard to deploy meaningful capital, it had subpar results, and the required sophistication consumed board time.
“CalPERS does things in a certain way. And it's a difficult governance entity.”
- Most investors earned much less than the S&P 500's 9.5-year return, because a 50 percent drawdown causes many to sell near the bottom and miss the recovery, so dollar-weighted returns fall well below time-weighted returns.
“almost no one has earned the returns of the S&P 500 over the last 9.5 years.”
- Ted Weschler left private equity because exits forced him to sell businesses he increasingly understood to be great, so he created a permanent public equity vehicle and adopted a style similar to Buffett's.
“he hated the fact that once you really believed that this was an incredible business, you had to sell it.”
- Rebalancing adds value when asset classes are more volatile relative to each other than they are directional, and the ability to communicate targets to a committee is as important as the allocation itself because it keeps people from abandoning the plan.
“I think that was as big of a component as just that particular asset allocation strategy.”
- Carol Loomis's 1971 article on hedge funds described the same challenges hedge funds face today, which shows the industry's problems are recurring rather than new.
“The things that she talks about in that article, about the challenges that hedge funds were facing in 1971, are the exact same things we're talking about today.”
- Salem's formulation that alternatives should be classified into just two buckets, total return and hedging, is a useful framework, but it risks the behavioral problem of making decisions in the middle of a cycle without a structure that keeps people disciplined.
“So total return and hedging is one. When I had Brett Barth on, he talked about high-risk strategies and low-risk strategies is another.”
- Because hedge funds had roughly half or less of the S&P's drawdown, investors were more likely to stay invested, so the realized dollar-weighted outcome could have been better for hedge fund investors even though their paper returns were lower.
“the drawdown over that period was half or less than half, which just means it was a little bit easier because the pain was far less severe for the investors to stay in.”
- Hedge funds protected capital poorly in 2008 but much better than other alternatives, and Buffett remained silent about the bet during years when the market rally made hedge funds look bad, suggesting he would have stayed quiet otherwise.
“Hedge funds protected capital, poorly in '08, but much, much better than other alternatives.”
- The current trend toward passive investing, especially in the S&P 500, is setting investors up for disappointment in the same way as ten years earlier, because they are not prepared for drawdowns.
“I just think that we have a trend that's happened now with passive investing, and particularly the S&P 500 that is setting those people up for disappointment in the same way that it did 10 years ago.”
- Hedge fund allocation consumes significant board time, and less sophisticated decision-makers tend to react to recent performance, which is a recipe for disaster in alternatives.
“if the less sophisticated the decision-making body, the more they're just going to react to performance. And as we know, that's always a recipe for disaster.”
Ordinary investors should choose a globally diversified, low-cost approach rather than default to US large caps.
- The audience of ordinary investors, including Seides's parents, should use a low-cost approach, but that approach should be globally diversified across more securities than the S&P 500 rather than defaulting to US large caps.
“people who don't know that they're making that bet should probably invest in a more diversified portfolio globally, and probably across a wider selection of securities than what the S&P 500 really represents.”
- Choosing the S&P 500 is an active bet that the US will outperform the rest of the world, so the index is not a neutral passive default, and investors who do not realize this are taking an unexamined position.
“That is a bet that, as you said, Jack makes and Warren likes to make that the US is the best country in the world.”
- If Seides designed the bet today he would not use fund-of-funds and would instead use a more cost-effective vehicle, possibly lower-cost hedge funds, because the fund-of-funds layers add fees without necessarily adding value.
“Would you use fund of funds? No. You would use something that's far more cost-effective. You might use hedge funds. You might use lower-cost hedge funds.”
Crowding has changed shorting economics, eroding equity long/short performance.
- Equity long/short hedge funds have become less effective because crowding has increased short interest from 2 to 3 percent to 15 to 20 percent, higher rates raise short-rebate costs, and more capital chasing the same stocks changes the economics of shorting.
“a crowded short had 2% or 3% short interest outstanding. And today, those numbers are 15%, 20%. It's just a completely different game.”
- As strategies such as value, quality, and leverage are packaged into ETFs, the advantage they once provided disappears, so replicating a past strategy like Buffett's will not reproduce the same record over the next 50 years.
“by the time you can replicate most of those strategies, the advantage that they have has probably gone away.”
- Money flowing into market-cap-weighted index funds forces the index to buy the same large names regardless of value, so during periods of heavy passive inflows it is nearly impossible for active managers to outperform, and active management is cyclical.
“As long as money is flowing there into exactly that strategy and buying those names more than it's buying other names, it is next to impossible to outperform that index.”
- Equity long/short funds used to capture about 50 to 60 percent of market upside and 20 to 30 percent of downside, but that pattern broke down after Lehman, possibly because of competition and asset flows, and hedge fund performance has been erratic since.
“if the S&P went up 10%, you might expect them to make 5% or 6%. And if it went down, you'd expect them to only lose 2% or 3%. And that kind of upside-downside capture didn't really work.”
Allocation frameworks matter through the discipline they impose, not through a single correct structure.
- Once an endowment moves beyond a 60/40 stock-bond portfolio, every other asset is by definition less liquid, so the liquidity premium was a secondary consequence of Swensen's diversification rather than its driver.
“if you were going to diversify away from 60/40, by definition, everything else had to be less liquid. And that was one thing, I think, people misinterpreted in what he did.”
- Asset allocation frameworks are valuable mainly when they impose structure and discipline on the practitioners, rather than the particular bucket schema, because clear objectives help people stick to the plan.
“the discipline that having any asset allocation, or in this case asset allocation, but a structure imposed on the practitioners, the investments office, and the committee, such that they knew what they were trying to accomplish and they could stick to it.”
- There is no single correct way to structure alternatives; the right allocation depends on the liability needs of the institution or person, though Seides is biased toward the structure that worked well at Yale.
“So I don't feel that there is one way to do it. I just have seen a few ways that have worked so well”
The bet's starting point before the 2008 drawdown shaped its early outcome.
- Seides's main reason for taking the bet was that the S&P 500 was expensive near its historical high on a Shiller P/E basis, so the market was likely to underperform and almost any alternative had good odds of beating it.
“the S&P 500 was trading near its historical high on a Shiller P/E basis”
- From the start of the bet until about February 2009 the S&P 500 dropped roughly 50 percent, meaning the bet's starting point came just before a severe drawdown and the 2008 losses set the early pattern.
“from the beginning of the bet till, I'll call it, February 2009, the S&P dropped 50%.”
- In 2008 the S&P 500 fell 37 percent while the fund-of-funds basket fell about 24 percent, so hedge funds protected capital relative to the index during the crisis even though they still lost money.
“in 2008, the S&P was down 37%. The hedge funds, after Lehman, tt was a big difference before and after, but after Lehman had a tough time but were down, I don't remember the number, maybe it was 24%.”
- Berkshire Hathaway underperformed the S&P 500 for a prolonged period during the bet, which Seides says is a symptom of passive-flow dynamics rather than evidence that Buffett became a worse manager.
“Berkshire Hathaway, for the most part, is underperforming the S&P 500 for a prolonged period of time”
- The bet's collateral, pre-funded as a zero-coupon bond, grew from about $640,000 to $950,000 in several years and was then moved into Berkshire stock, which made roughly 300 percent, so the charities' best outcome came from the collateral rather than the S&P.
“the real winner was the charities and cash, not even the S&P 500.”
Permanent capital's patience lets owners buy businesses that others must sell.
- A business or investor with no need to sell can tolerate near-term pain to pursue longer-term profits, which Tom Russo calls the capacity to suffer, something public companies find hard because they are scrutinized quarterly.
“The ability of a business to suffer in the near term, to basically encourage A curve to reap longer-term profits, in this world, and certainly a public company world, it's just hard for companies to do that.”
- Permanent-equity investors such as Brent Beshore and Chenmarks buy small family-owned businesses at around four times earnings, but the index for this strategy would return roughly negative 100 percent because most purchases fail, so success requires exceptional selection and network building.
“buying small family-owned businesses at four times earnings. And yeah, you have to find the right ones.”
- Permanent capital is advantaged because it does not need to sell a business on a fixed timeline, making the owner a more favorable buyer for founders who want their companies preserved, which is the core of Buffett's model at Berkshire.
“If you can make the case credibly that you don't have to sell someone's business, you become a more favorable buyer.”
- Seides argues that investing as a long-term equity owner with permanent capital is the first-principles idea behind alternatives, because a long-horizon investor should own equity-like streams that look different from the market.
“I can circle back to first principles of what I learned working for David Swensen, which is he has long-term capital. And therefore, you should have equity ownership in things.”
The bet's origin, publicity, and terms were deliberately negotiated by both sides.
- Seides was prompted to write Buffett after a transcript showed Buffett claiming no one took him up on his claim that five hedge funds couldn't beat the market, which Seides thought was cheeky enough to challenge in the summer of 2007.
“And in this transcript, he said, "Well, no one took me up on it, so I must've been right."”
- André Perold's case teaching on Buffett emphasized studying how Buffett manages his brand as well as how he invests, which Seides took seriously when he later approached Buffett.
“André had said something about you should pay attention to how he manages his brand as well as how brilliant is his investing.”
- Protégé Partners chose to make the bet public under its own name because partners believed association with Buffett would be good for the firm, despite Seides initially wanting to remain anonymous.
“This is only going to be good for us to be associated with Warren.”
- Seides wanted to avoid a horse-race presentation of the bet and proposed a trigger based on a 10 percent S&P decline, while Buffett insisted on yearly annual-meeting updates, showing the PR dimension was negotiated by both sides.
“the horse race component to his is kind of exactly what is wrong with asset management.”
- Buffett did not publicly discuss the bet for several years after 2009, but he resumed reporting results at the annual meeting and in the letter once the S&P crossed above hedge funds, and Seides theorizes Buffett used the moment to tell a clean fee story when the data supported it.
“I have a theory of why he did that now... the story that he wanted to make, this simple story of fees, at this moment in time, the data supports it in every way it possibly could.”
Seides would repeat the bet, valuing the people and annual meeting most.
- The most enjoyable part of the bet was getting to know the people involved, such as Buffett, Ted Weschler, Todd Combs, and Carol Loomis, rather than the outcome, while the least enjoyable part was constant questions from friends about losing.
“Easily the most enjoyable aspect has been getting to know the people involved.”
- Knowing what he knows now, Seides would make the bet again and again because the odds were favorable, the S&P's starting valuation made it one of the best 10-year starts for the bet, and the losing outcome was partly unexpected.
“I would make the bet again over and over and over again, enough times so that the odds played out.”
- The Berkshire annual meeting has evolved from a small, substantive investment-lessons event into a mix of entertainment and company substance, and the bet's annual scoreboard was especially theatrical.
“And the first time I went after the bet, it was much more of a cartoon show.”