Hedge Fund Clips

All the best hedge fund content on the internet.

[Hedge Fund Clips @HedgeFundClips]

Dan Sundheim on his own industry: "I don't think hedge funds are a good business... it's horrible."

Dan Sundheim, founder & CIO of D1 Capital Partners.

Dan Sundheim explains:

"Money to me is a scorecard. I wanna have the best score."

"I don't think hedge funds are a good business. Our business, it's horrible."

"It's amazing cash flows, it cash flows really well — it has no terminal value."

"I told my companies I invest in: 'You have no cash flows and tons of terminal value. I have tons of cash flows and no terminal value.'"

"So we're good together. We can kinda arbitrage that."

"To have enterprise value you need hundreds of employees. I do not ever aspire to that."

[Hedge Fund Clips @HedgeFundClips]

"Why does a casino teach people how to play blackjack?" The framework isn't the edge.

Alex Sacerdote, founder & CIO of Whale Rock Capital, 20+ years in tech investing.

Alex Sacerdote explains:

"My mom said, 'Why do you tell everyone your secret?' It's like, why does a casino teach people how to play blackjack?"

"It's really hard to do. You have to be comfortable investing."

"No one's paid attention to hardware and chips at all, so you've got all these newbies coming into it."

"The charts are up, so it's scary. Can I buy? And you have to have the holistic view — if you don't have conviction..."

"Every time with NVIDIA over the last four years: 'they had a great year, it's gotta be a bubble.' Then another great year. 'It's gotta be a bubble.'"

"So many semi analysts missed it because they didn't see what was really happening at the foundational model layer."

"It helps to have decades and scores of S curves you're looking at."

[Hedge Fund Clips @HedgeFundClips]

"The machines are doing shitty this year. I'm kicking their ass."

David Tepper, founder of Appaloosa Management.

David Tepper explains:

"The machines are doing shitty this year. Really bad. I'm kicking their ass."

"When I went to Goldman, they had a trading model on the desk, and it was just wrong. I knew the option part was wrong."

"The machines are only as good as the people programming the machines."

"When times are changing — higher rates, out of the QE environment — people are continuously programming the same damn thing."

"They'll be less emotional than people, but when the times change, they don't change unless somebody reprograms them."

"When times change fast, that doesn't work."

"When you have a guy like Trump, you better know how to deal with people and how different emotions work."

[Hedge Fund Clips @HedgeFundClips]

Taiwan makes 90%+ of the world's leading chips. Break that supply chain and you get a Depression-scale economy.

Dan Sundheim, founder & CIO of D1 Capital Partners.

Dan Sundheim explains:

"We're on a collision course with China over semiconductors."

"Taiwan produces 90-something percent of the most advanced semiconductors. And everything we use is semiconductors."

"It's almost as if you went back 50 years and only one country produced oil. We went to war over oil even though you could get it all over the world."

"That supply chain is fragile. It's not easy to replicate. It's easy to destroy."

"If that supply chain got screwed up, we'd have a Depression-type economy."

"There is no scenario where China's happy, Taiwan's happy, and the US is happy. Somebody's going to be unhappy — either because the economy collapses or because their sovereignty is handed over."

"Replicating this supply chain takes 10 to 20 years. Usually when dictators say something religiously, you should believe them."

[Hedge Fund Clips @HedgeFundClips]

Google's real risk isn't losing search share. It's owning 99% of a bucket that no longer matters.

Philippe Laffont, founder of Coatue Management.

Philippe Laffont explains:

"I find that I am using search a lot less. But all of that goes to OpenAI and to X."

"I'm doing 5x as many queries, spread across different platforms. Instead of asking humans to do the work, I'm doing it myself."

"The total pie could be 5-10x bigger per person."

"I don't know if I care about having 99% of that, or if I'm better off having 80% of something 3x bigger."

"You lose 99% of one bucket, but you only get 10-20% of the new bucket. That's the bad scenario."

"That bucket is getting built right now, and they're in Nowheresville."

"How are you gonna show up in 18 months and say 'that new shiny bucket, pick me, pick me'? This is why it's a strategic error."

[Hedge Fund Clips @HedgeFundClips]

"If I ran Stripe, I wouldn't go public." A public-markets investor on why staying private beats an inflated ticker.

Dan Sundheim, founder & CIO of D1 Capital Partners.

Dan Sundheim explains:

"If I ran a private company like Stripe, I wouldn't go public. The public markets are kinda problematic at this point."

"Private tenders track value creation — the people building the value get paid for the value they created."

"Go public and, depending on what the retail crowd is doing that day, the stock trades at some insane value."

"Everyone's high-fiving. Our stock is 2X where it should be, we're all rich. The problem is you've pulled forward a ton of value."

"Employees are now overpaid — they didn't create that value, the stock gave it to them. And they're more likely to just cash out."

"You're robbing future employees to pay current employees."

"You're not being paid on value creation. You're being paid arbitrarily on multiples that have nothing to do with intrinsic value."

"If the stock overshoots, no one comes back and says 'I made too much money.'"

[Hedge Fund Clips @HedgeFundClips]

Why enterprise AI lags consumer AI: radio hit ~100% in 7yrs, the dishwasher crawled. It had to plug into the back end.

Alex Sacerdote, founder & CIO of Whale Rock Capital, commissioned Horace Dediu (ex-Clayton Christensen) to map 100yrs of adoption S-curves.

Alex Sacerdote explains:

"A lot of people think because we're in a modern world, everything's so fast, but there's a lot of factors that determine the pace of adoption."

"We commissioned Horace Dediu, who used to work with Clayton Christensen, to look in history. We have the big S-curves on our wall over the last 100 years."

"The radio S-curve was one of the fastest ever. It took seven years to reach 100% penetration."

"But the dishwasher S-curve is slow because it needs to be plugged into the back end."

"B2B stuff can take a long time because it needs to be plugged into the existing systems."

"Consumers generally tend to go a lot faster."

[Hedge Fund Clips @HedgeFundClips]

@cble416 @cble416 @ForgedUSAPod

https://t.co/1j0PDRuRDp

[Hedge Fund Clips @HedgeFundClips]

Single-manager funds turn over constantly. Diversified funds survive decades. Balyasny saw why from the allocator seat.

Dmitry Balyasny, co-founder & CIO of Balyasny Asset Management, started as a fund-of-funds analyst.

Dmitry Balyasny explains:

"There's a huge difference between two types of funds. One is a boutique structured around an individual risk-taker in a particular strategy."

"That's the vast majority of hedge funds. Then there's a business \a or a h
fund with a lot of different risk-takers. Completely different DNA."

"The single manager model had a tremendous amount of turnover. Some did great, then they didn't."

"The manager would want to do something else, lose motivation, or had business issues. Rare that we stayed with the same guys 10, 20 years."

"The diversified funds wound up being the same ones for decades."

"That type of firm, you can reinvent. Expand strategies, change the mix, allocate capital to better opportunities, refresh your teams."

"Just a much more stable and consistent business model. That's the way I wanted to build our business."

[Hedge Fund Clips @HedgeFundClips]

Hyperscalers will grow faster and become a worse business at the same time.

Dan Sundheim, founder & CIO of D1 Capital Partners.

Dan Sundheim explains:

"I'm more confident in the thesis that the hyperscalers are a worse business model going forward."

"Usually a worse business model implies growth slows and margins contract. I actually think you'll see the opposite."

"Their customer base was every corporation in the world — fragmentation, massive economies of scale no single company could get."

"LLMs are very likely to end up concentrated in four or five companies."

"At some point in the next 5-10 years, they'll be generating enormous free cash flow. When that happens, they'll insource the compute."

"Right now they look at hyperscalers as a financing mechanism, not because they're better at building data centers."

"Building CPU clusters is different than building GPU clusters. The LLMs are actually better at inference than the hyperscalers."

[Hedge Fund Clips @HedgeFundClips]

Retail thinks stock selection is 95% of the game. Balyasny: it's one-third.

Dmitry Balyasny, co-founder & CIO of Balyasny Asset Management.

Dmitry Balyasny explains:

"You get paid for stock selection. But selection, timing, risk management hose is one-third of that."

"People focus on selection as 95%. It's one-third."

"In a hedge fund, the reason you get paid a lot is you get the other two-thirds right consistently."

"At some point you're right on Amazon, at some point you're wrong. That's the retail world."

"In the institutional world, you're paid for making money consistently."

"You gotta get the timing and money management right, or the security selection doesn't really matter."

[Ethan Kho @ethanrkho]

He ran a desk competing head-on with Jane Street and Jump at Genesis. Asked what the secret sauce was, he says the tech wasn't the edge:

Lucas Schuermann built the electronic market-making system at Genesis. Now founder & CEO of @variational_io.

Lucas Schuermann (@variational_lvs) explains:

"Genesis, in many cases, was some of the best pricing you could get in the crypto market. But we all know the other players trading these assets at the time: the Jumps and the Jane Streets and many others of the world."

"People like to ask me: how were you guys competitive? What was the edge? What was the secret sauce?"

"While I like to think our infrastructure was great, and we built some very smart pricing systems, one of the biggest advantages a desk or a market maker of any kind can have is flow."

"Flow is just the size of the client base. It's the size and the economy of scale of the book of business you have."

"The bigger the economy of scale on external venues, the more you get. Maybe lower fee tiers, maybe better margin rates, maybe in some cases even preferential access to latency and matching."

"Most importantly, with natural incoming flow you can internalize: match risk on your own book without ever going out to hedge it externally, and manage portfolio-level risk much more intelligently."

"Having flow also gives you information about where the price might be going on a very, very short time horizon."

"This is the biggest misconception: that HFT firms or market makers have to win purely by nature of being fast. It's definitely a component. But having flow, just having an economy of scale, is one of the biggest reasons Genesis was winning at the time in that space."

[Hedge Fund Clips @HedgeFundClips]

$100/day in tokens per Anthropic engineer × 20M coders = a half-trillion-dollar market. On 7-9 month old tech.

Alex Sacerdote, founder & CIO of Whale Rock Capital Management.

Alex Sacerdote explains:

"Anthropic came out mid-year and it could do so much more. It started to run agentically, and the coding market just exploded."

"People who could use it unfettered were spending a hundred dollars a day on tokens."

"Do the math, that comes out to twenty or thirty thousand dollars a year."

"Think about how many coders there are in the world, twenty million. You've got a half a trillion dollar market just from coding alone."

"And mind you, that was on seven, eight, nine-month-old technology."

[Ethan Kho @ethanrkho]

The most important industries of the last century were among the worst investments ever made. Why that shouldn't change where you build your career:

Lucas Schuermann (@variational_lvs) built the electronic market-making system at Genesis, one of crypto's largest desks. Now founder & CEO of @variational_io.

Lucas explains:

"Asset prices are almost entirely uncorrelated over long horizons with expansion economics in specific industries."

"During the invention of the automobile there were dozens if not 100 companies in the United States. Many of them very, very large. Only two made it through."

"Airlines are the counterintuitive one. An incredible area of the economy, an incredibly important area of technology. But holding airline stocks, even the ones that exist now, has been a horrible investment over the lifetime. You're losing money left and right."

"Does that mean it's not worthwhile to start businesses in aviation, or take a career in aviation as a pilot, as an employee, as a designer of jet engines? No. There are incredible businesses being built in all parts of aviation."

"The price of memory, the price of Nvidia, the price of GPUs: none of it has anything to do with the durability of these as industries, or as sectors that are interesting to build businesses in."

"As an investor in memory stocks, no idea. I'm a horrible beta predictor. But looking at price is absolutely the wrong way to go about it."

[Ethan Kho @ethanrkho]

"The biggest misconception is that HFT firms or market makers have to win purely by nature of being fast."

Lucas Schuermann (@variational_lvs) started a quant fund out of a Columbia dorm room, dropped out when a New York hedge fund sent a term sheet, and sold it to DCG in 2019. Built the electronic market-making system at Genesis, DCG's trading arm, when the desk was still hedging by phone.

That desk moved hundreds of billions in volume.

He now runs @variational_io, which raised $50M led by Dragonfly in May and has cleared $300 billion in volume since launching .

"It looked like Wolf of Wall Street. Traders on the phone, traders in Bloomberg chats and Telegram chats, hedging on screen onto Coinbase, relatively manually."

We cover:
- ⁠Everyone thinks HFT wins on speed. He says it wins on flow. Client order flow buys you internalization, information, and fee tiers that no amount of latency does
- The three real moats in institutional finance: flow, capital, and trust. Why trust is the one nobody models
- ⁠What it takes to electronify a live desk: rewiring the engine while the plane is flying, with no option to land
- ⁠Why a risk system is just very high-frequency accounting wearing a quant costume
- ⁠Moats vs. durability. "Will A beat B" and "should this exist in 20 years" are different questions, and most people conflate them
- ⁠Funding rates are the last unsolved problem in perps. His answer is an instrument TradFi has used for decades, brought on-chain
- ⁠The honest case against his own product, from someone who's been the market maker, the broker-dealer, and now the platform
- ⁠Why the price of a trend tells you nothing about the trend. A hundred car companies, two survivors, and airlines as the counterexample
- His contrarian take: the cypherpunks won. We counted the 98% that died and forgot the 2% that went mainstream
- "Ed and I are the worst macro traders, the worst kind of traders in human history"
- His one-word answer on how to differentiate yourself: hubris

Highlights:
00:00 Intro
00:50 "The biggest mistake was starting the fund at all"
03:03 Crypto in 2017 was a nightmare of a market, which is where the alpha was
05:00 Good alpha makes a little every day, not a lot at once
06:36 The term sheet that made him drop out of Columbia
09:04 Taking a phone-and-Telegram desk electronic without turning it off
13:09 Why a risk book is high-frequency accounting
14:29 The most underappreciated edge at a large firm: flow
15:52 The biggest misconception about HFT firms
17:12 Flow, capital, trust. Where moats actually come from
20:14 "Genesis stopped being fun the moment it went over 150 people"
23:20 You don't make money on a 3-year view. You make it on a 10-year view nobody can see
29:44 Perps ate the world, and everyone built the same order book
38:40 Total return swaps, and how TradFi actually gets its leverage
41:14 Internal vs. external market makers, and the case against him
47:19 Contrarian take: the cypherpunks won
53:56 How to tell a real trend: dig deeper, find more exponentials
58:14 Price is not economics: the car manufacturer and airline trap
01:01:10 Three tools to get expert in any domain fast
01:09:20 He turns the final question back on me

[Hedge Fund Clips @HedgeFundClips]

LLMs are Netflix + Spotify: huge fixed cost to train, near-zero incremental cost to sell. But models aren't differentiated content — the moat is personalization.

Dan Sundheim, founder & CIO of D1 Capital Partners.

Dan Sundheim explains:

"Your business is some kind of combination between Netflix and Spotify. Netflix in that, unlike other tech companies, you spend a ton of money upfront to train these models."

"Once trained, you sell at extremely high incremental margins. You don't know the revenues from that fixed asset, but you want to sell as much as possible to fund the next model."

"You invest heavily, get the revenues, spread it over more people, invest more — a flywheel: more revenue, more content, more revenue."

"Eventually it's almost impossible to compete. The first mover advantage is just too great."

"The difference: Netflix's content was differentiated. The models are more similar than different — OpenAI may have a better model, Anthropic may have a better model, but the innovation disseminates quickly."

"The differentiating factor won't be that Google gives you a better answer. It's personalization — the more these models know about how you live, your health, all the things important to you."

"Spotify's music is no different than Apple or Amazon — theoretically a pure commodity. Personalization is what lets you charge a premium for a commodity."

[Hedge Fund Clips @HedgeFundClips]

Whale Rock went from 50% software to net short entering the year. The thesis: even if AI never disrupts software, the incumbents still lose.

Alex Sacerdote, founder & CIO of Whale Rock Capital Management.

Alex Sacerdote explains:

"5 years ago we might've had 40-50% of our portfolio in software."

"Early on we thought: huge sales forces, AI APIs, the data — this is gonna be amazing for software."

"Pretty quickly we realized their AI products weren't very good. Nobody could charge for them."

"We sold almost all our application software. Entering this year we were net short — it really helped us in Q1."

"Even if AI isn't disruptive, software's fallen way down every CIO's priority list — they're spending on Anthropic tokens, faster ROI there."

"Software companies raised price every year. Now they're nervous about doing that."

"The old way of software is pen and paper. The new way is a jet engine — it feels like it has to be disruptive."

[Hedge Fund Clips @HedgeFundClips]

Index investors have handed the future to "bureaucrats in the basement of MSCI and S&P."

Philippe Laffont, founder of Coatue Management.

Philippe Laffont explains:

"There are times where we've held 80% cash in our fund, not fully invested at all times."

"Who had a lot of cash, a lot of publics, and a lot of privates? Berkshire Hathaway. Nobody asked Warren why."

"Public markets have completely surrendered their intellectual creativity to the indices."

"The indices are decided by bureaucrats that live in the basement of MSCI and S&P."

"How is it possible we're allowing these people to tell us what to invest in the future?"

"It's all disguised under the fact that fees are low, but you're basically dying by a thousand cuts."

[Hedge Fund Clips @HedgeFundClips]

Fed not cutting = economy strong, not weak. Coatue's news/sentiment ratio at an all-time extreme: news great, sentiment terrible.

Philippe Laffont, founder of Coatue Management.

Philippe Laffont explains:

"What if the Fed is not cutting because the economy is really strong? The Fed not cutting is actually not that bad of a message."

"I'm surprised at how bad sentiment is, but how good the hard data is."

"We have this ratio at Coatue where we divide hard news as the numerator and sentiment as the denominator."

"It's the first time where the news is so good and the sentiment is so bad."

"The Fed said: we're not gonna cut just to bail out the equity market."

"But if the market's liquidity is no longer functioning, then we'll step in to restore liquidity."

"A tariff correction, a tariff tantrum, but not a tariff crisis."

[Hedge Fund Clips @HedgeFundClips]

Booking: 40% growth at 9x earnings. Now high-teens growth at ~3x that multiple. The money was never in the earnings.

Dan Sundheim, founder & CIO of D1 Capital Partners.

Dan Sundheim explains:

"It was growing 40% a year and traded at 9 times earnings. The narrative back then was they're just arbing Google, it's a flimsy business model."

"Now it's growing a lot slower, and it trades at 3x the multiple."

"Sometimes the market is so skeptical about the business model that even as the business slows, the multiple goes up."

"You're always gonna make the most money from multiple expansion."

"It's hard to have a differentiated view on a company's growth rate. But the multiple over the long term is a function of the perception of the stability of the cash flows."

"People are skeptical of a business model, then over time it gets proven out, and the multiple expands."

"When I sell companies too early, it's almost always because we used too low of a multiple at exit."