There is a lot of investing advice written by people who have never lost money at scale.
These twelve have. Every one of them has stood in a drawdown big enough to end them and had to decide what to do next. That is the only filter I applied.
Some of what follows is famous. Several of these you probably have not heard, and those are usually the ones worth something.
1. Paul Tudor Jones
Tudor. 19.5% a year for 25 years, no losing year.
"The most important rule is to play great defense, not great offense. Every day I assume every position I have is wrong. I know where my stop risk points are going to be. I do that so I can define my maximum drawdown."
He wants 5:1 on every trade. Risk a dollar to make five.
That sounds like greed. It is the opposite. At 5:1 you can be wrong four times out of five and break even, which is the only way a strategy with a low hit rate stays alive.
His phrasing: "I can actually be a complete imbecile. I can be wrong 80% of the time, and I'm still not going to lose."
The part that never makes the quote roundups is when he loses. His worst drawdowns came right after his best runs.
The run makes you certain. The certainty makes you size up. The size is waiting for you at exactly the wrong moment.
"Don't be a hero. Don't have an ego. Always question yourself and your ability. Don't ever feel that you are very good. The second you do, you are dead."
So he uses the 200-day moving average. Below it, he cuts. No debate, no reassessment.
The indicator is not the point. The point is that he removed the moment where he would have had to argue with himself, because he knows he would lose that argument.
Market Wizards, 1989.
2. Paul Singer
Elliott. Two losing years since 1977.
"Our philosophy then is the same now: we try to never lose money no matter what. It's kind of brazen to use the phrase 'no matter what,' but in 47 years, we've had two losing years."
Between 1967 and 1974 he traded small amounts of tech and mining stocks with his father, a retail pharmacist. His summary of those years: "We found every possible way conceivable to lose money."
Then he started Elliott in 1977 with $1.3 million from friends and family. A dollar invested then became $165.
Here is the part that gets skipped. "never lose money" is not about avoiding risk. It is a rule about what you are allowed to hold.
You cannot own a position whose downside you have not measured and paid to cap. Elliott buys hedges every year, and in most years that spending is a drag.
In two years out of forty-seven it was the reason there was still a firm.
Which is uncomfortable, because it means the strategy costs you money most of the time and you only find out it was correct in the year you needed it.
Chicago Booth, May 2024.
3. Ken Griffin
Citadel. $71 billion. 19.2% a year since 1990, and that number includes losing 55%.
Sometime in the late 1990s Griffin sat across from Ace Greenberg, then chairman of Bear Stearns. Greenberg told him the biggest risk in capital markets is not losing your capital. It is losing your ability to operate.
In 2008 he found out what that meant. Citadel went into the crisis levered seven to one. The funds lost 55% against an industry that lost 19%.
"We were losing hundreds upon hundreds in capital a week - if not more."
When investors asked for $1.2 billion back that December he suspended redemptions and locked them in for ten months. He was attacked for it publicly for the better part of a year.
The positions that had done the damage produced 62% in 2009. He still owned them, because he had refused to sell them.
But the thing I keep coming back to is a number he gave a room of students at Georgetown:
"My colleagues that walk on water are 53/47. If they were brain surgeons they'd have very few patients."
And then: "How many people in this class have taken home a test in the last eight years where they got 53% right? We hire the best and brightest, they go from having 90s as their average test score to 53."
53%. These are people who got into Citadel.
If that is the ceiling on accuracy, then accuracy is not where the returns come from. The returns come from what happens in the other 47% and whether you are still solvent when the 53% shows up.
After 2008 the firm was rebuilt around that. Financing spread across counterparties so none of them controls the timing. Capital stress tested daily instead of quarterly.
"The key is when you're in hell, just keep walking forward each and every day."
Georgetown 2017. Miami, November 2025. Stanford, May 2026.
4. Ron Baron
Baron Capital. About $56 billion. Roughly $69 billion in cumulative profits for clients.
Everybody says hold for the long term. Baron did something stranger. He made it impossible for himself to sell.
"I told the board, 'if you let me invest a certain amount of money, then I will promise that I won't sell any of my stock. I will be the last person out of the stock.'"
"I will not sell a single share of my shares until my clients sold 100% of their shares. And I don't expect to sell in my lifetime Tesla or SpaceX."
That promise got tested. When clients and the press went after the firm over concentration, Baron Funds sold 30% of its Tesla position.
Baron sold none of his. "We sold 30% for clients. I did not sell personally a single share."
Roughly 40% of his net worth is Tesla, 25% SpaceX, 35% his own funds. He has made about $8 billion on Tesla and thinks he makes five times that over the next decade.
The mechanism here has nothing to do with patience. Conviction is a structure, not a feeling.
He made a public promise to a board decades ago and that promise now does the work that willpower cannot do on day 400 of a drawdown, when the position is down 40% and everyone is telling you that you have lost it.
If your thesis runs in years, build something that takes the daily decision away from you. A rule, a lockup, a commitment somebody else witnessed.
CNBC, 14 November 2025.
5. Carl Icahn
On how he thinks about where to put money at all:
"When I started out, I learned the hard way, playing the market. It's a competitive area, it's gambling. Too many variables, too many people competing with you. You can't believe that the gambling will carry you."
"So you look for something I call a 'no-brainer.' Very little risk, but a lot of reward."
That is the whole method. Not better analysis of the same crowded thing. A different thing, where the asymmetry is so lopsided you do not need to be clever.
Where he finds them:
"My investment philosophy, generally, with exceptions, is to buy something when no one wants it."
And what he looks at once he is there. This is Icahn on TWA after he took control:
"One of my executives was out giving speeches all the time. I said, what do you talk about? How to lose $400 million?"
The mispricing he hunts is not in the financials. It is in the incentive structure of the people producing them, and that structure is published every year in the proxy.
In 1989, before the LBO wave came apart:
"It's patently absurd that a management that can't manage a company decides to pay a very high price with borrowed money because they're afraid of losing their jobs. It's like throwing water on a drowning man."
"Our economy has major time bombs built into it - you're going to see bankruptcies, you're going to see problems ahead."
Drexel collapsed that February. The recession showed up in July.
He also has the least sentimental line about losing I have come across:
"You're gonna lose and you gotta be willing to say, hey, if I lose, so what?"
MacNeil/Lehrer NewsHour, PBS, 1989, and later interviews.
6. Bill Ackman
Pershing Square. Peaked at $18 billion, fell to $4 billion, best eight years in firm history since.
At the bottom he did the thing you are not supposed to do.
"I borrowed $300 million. It was the biggest personal unsecured loan ever granted to an individual by J.P. Morgan."
He used it to buy his own collapsing fund. Then he called the staff together:
"I called all the employees together and said, we are better positioned today than at any time in our history, because we have the holy grail - we have permanent capital."
Pershing Square Holdings is closed-end. Investors cannot redeem. That is why a 78% decline stayed a decline instead of becoming a liquidation.
It is Griffin's gates arrived at from the other side. Griffin had to impose the lockup in the middle of the crisis. Ackman had built it in advance.
And the bit nobody mentions when they talk about him. He raised his first fund by cold calling the Forbes 400.
"I figured if I want to raise $10 million, why not go to the richest people in the world and ask them for a really small amount of money. 90 plus percent turned us down."
Four of his first six investors came off that list.
Concentration and conviction are not a strategy on their own. They are a large bet. What makes them a strategy is the third thing, the structure that lets you be wrong for three years without being removed from the table.
7. Gavin Baker
Atreides.
"Open source models fundamentally just shift margin from the model layer to the infrastructure layer. If you have cheaper tokens we're going to consume more of them - we are structurally short compute."
And the harder version of the same idea:
"If you're a foundation model company and you do not have unique data and internet scale distribution, you are the fastest depreciating asset in human history. Most of these companies are zeros, and there's like 10 of them."
Every technology cycle does this. The thing everyone watches is almost never where the money lands.
When an input gets cheaper, people consume more of it than the price fell, and the profit moves to whatever became scarce instead. Right now that is power, land, memory, and the physical ability to plug a GPU in and switch it on.
The other thing he said is stranger, and I have not seen anyone else make the point:
"Every piece of news gets fed into Claude. Claude is kind of Walter Cronkite for the stock market, and everybody just believes whatever it says. A huge chunk of people trade on Claude's view - and it's not always right."
His evidence is Japanese capacitor stocks running a three year sector cycle in six weeks.
If everyone runs the same news through the same model, the disagreement that normally slows a repricing just is not there anymore. The move happens all at once.
Which raises an awkward question about your own process. If your read on a story came out of the same tool as everyone else's, that is not an edge. It is consensus arriving faster.
8. Jamie Dimon
Most of what he says publicly is about markets. The useful material is from a talk he gave at Harvard, where he laid out how he actually runs decisions.
"The job of a leader is not to make the decision. It's to make sure the best decision is made, and therefore you need other people."
That sounds soft until you see what he does to enforce it.
"I tend to tell people, if you can't say it in the room, you don't deserve the job. Say it politely, say it reasonably, but say it."
The tell he watches for is people queuing outside his office after a meeting to say the thing they would not say inside it. He treats that as a failure of the meeting, not a courtesy.
"Bureaucracy is the Petri dish of politics."
And on how meetings should run:
"At management meetings, emphasize the negatives. You can always have a party about the positives. What are we not doing well? How come the competition's doing better?"
There is one habit in there I have not seen anywhere else, and it is trivially copyable.
"I always make a list on Sunday of all the issues I'm trying to avoid, and I write them down, and I try to deal with them on Monday. I avoid them half the time."
The admission at the end is the part that makes it real. He built a system for the things he is dodging, and it still only works half the time.
On being fired from Citigroup in 1998, after fifteen years working for the man who fired him:
"It was my net worth, not my self-worth."
He called the guy a year later to have lunch, told him he had made the wrong call for Citigroup, then listed his own mistakes.
"Whether it was 40/60 or 60/40 wasn't important. I wanted to learn from the experience and move on."
And a test for promotions that takes two seconds and is brutal to apply honestly:
"Ask yourself a question if you promote people - would you have your child work for that person? Because half the time we promote people, we would not let our kid work for them."
Harvard Business School.
9. Chamath Palihapitiya
He asked an AI model how much AI had lifted S&P 500 earnings growth since 2024. It told him 50%.
So he pushed. But you are counting the money Nvidia gets from selling chips to Amazon. What did the S&P 493 do?
9%. He pushed again, and most of that turned out to be pricing power sitting on inflation, with another 3% from buybacks.
"The actual ROI was somewhere between zero and 2%."
Then he went and checked his own company:
"I asked my CTO how we're doing on token spend. He said our token costs are doubling every 45 days. I asked what the downstream productivity is. He said maybe 5% max."
The explanation his CTO gave him: "We've effectively already asymptoted - you need a lot more tokens to get to this next iteration of improvement."
Circular revenue. A sells to B, both book growth, and at the index level it reads as productivity. Strip out the internal transfers and most of it is gone.
This is not an AI thing. It is what every capex boom looks like from inside.
The move worth copying is not the index analysis. It is that he asked his own CTO for a number and accepted the answer when it contradicted what he wanted to believe.
10. David Sacks
"Trying to re-engineer the unit economics or culture of a business that is already operating at massive scale is brutally hard. Searching for the scalable model when you're already at scale is a contradiction in terms."
Growth locks things in. The pricing you launched with, the cost structure you accepted, the culture you tolerated when there were thirty people.
By the time you have thousands of customers all of it is load bearing, and every month makes the repair more expensive and less likely to happen.
Same shape as the leverage lesson, different room. What decides whether you survive gets decided before the thing you are afraid of shows up.
He also has a hiring filter I like more than most:
"You have to be comfortable with ambiguity. If you're the type who likes to very carefully weigh 99% of the data before you make a decision, you're not cut out to run a start-up."
Prove the model small, then scale it. The other order, scale first and fix the economics with volume, is the most reliable way to destroy a large amount of capital.
It is always defended with the same sentence about operating leverage.
11. Jason Calacanis
$25,000 into Uber at a $5 million valuation came back as roughly $100 million.
He has the smallest fortune on this list and says so out loud. Which is why his rule is the most useful one here, because it costs nothing to implement.
Never say yes in a meeting.
Tell them you have to finish diligence and compare against your other options. Then walk out.
The pitch is built to get a decision out of you under social pressure, in a room, with the best version of the story fresh and no alternative in front of you.
Everything the other side has going for them is time bound. Leave and it evaporates.
Underneath that is the same arithmetic as Jones' 5:1, arrived at from venture. Most investments return nothing, a few return everything, and no amount of diligence tells you which is which because the outcome depends on execution that has not happened yet.
So you make enough bets to be holding the one that works, and you size them so that being wrong repeatedly does not take you off the board.
Angel, 2017.
12. Jim Simons
Renaissance Technologies. 66% a year before fees from 1988 to 2018. Over $100 billion in trading profits. Around 300 employees. He died in May 2024.
He never took a finance class. He was a topologist, then a code breaker for the NSA during Vietnam, then chair of the maths department at Stony Brook. He did not trade seriously until he was forty.
Then he built the most successful fund in the history of the industry, and did it by removing himself from it.
"The only rule is that we never override the computer. No one ever comes in any day and says the computer wants to do this and that's crazy and we shouldn't do it."
The reason he gives is not about ego or discipline. It is about what you can and cannot test.
"You don't do it because you can't simulate that. You can't study the past and wonder whether the boss was gonna come in and change his mind about something. So you just stick with it, and it's worked."
That is the strictest version of what Jones and Baron both arrived at.
Jones uses the 200-day so he does not have to argue with himself. Baron made a promise to a board so he cannot sell.
Simons removed the human from the loop entirely, because a human who might intervene makes the whole backtest meaningless.
Once you allow the override, you no longer know what your system does. You only know what it did on the days you happened to leave it alone.
The other rule of his is the one people skip, and it is the first of the five principles he wrote down in 2010:
"Do something new; don't run with the pack. I am not such a fast runner. If I am one of N people all working on the same problem, there is very little chance I will win. If I can think of a new problem in a new area, that will give me a chance."
He is describing the same thing Icahn calls a no-brainer, arrived at from pure mathematics rather than forty years of hostile takeovers.
Do not compete where everyone competes. Find a problem where the number of people working on it is small.
He applied it to hiring as well.
"We hire physicists, mathematicians, astronomers and computer scientists and they typically know nothing about finance. We haven't hired out of Wall Street at all."
Three hundred people, none of them from the industry, producing more trading profit than most of the industry combined.
The last of his five principles is the one that sounds like a joke and is not:
"Hope for good luck."
Coming from the man who built the most systematic operation in finance, it reads as the honest ending to the other four.
Do something nobody else is doing, hire better people than yourself, build it well, refuse to quit. Then accept that the rest is not up to you.
San Francisco State University, 2014. Guiding principles, 2010.
The one thing they disagree about
Read them together and there is a conflict sitting right in the middle.
Ron Baron will not sell. He built a structure specifically so that he cannot, and it produced $69 billion.
Paul Tudor Jones exits below the 200-day, mechanically. He also built a structure to take the decision away from himself, pointed the opposite direction.
Both work. The reason both work is the most useful thing here.
Baron owns businesses. Jones owns positions.
If your return comes from a company compounding over twenty years, the drawdown is noise and selling is the only way to lose. So you engineer selling out of reach.
If your return comes from being on the right side of a price move, the drawdown is information and holding is the only way to lose. So you engineer the exit to be automatic.
The failure is running one of these with the other's structure. Holding a trade because you like the founder. Exiting a twenty year thesis because the chart broke.
Work out which one you are before you enter, not while you are down.
And the thing they agree about
Eight of them are saying a version of the same sentence from different chairs.
Jones sets his maximum loss before entry. Singer pays for hedges in the years it costs him. Griffin says his best people are right 53% of the time. Ackman borrowed $300 million rather than sell.
Baron made selling unavailable. Dimon writes down the problems he is avoiding. Calacanis leaves the room. Sacks fixes the model before scaling it.
None of that is about being right.
All of it is about what you decided before the thing you are worried about arrived. By the time the drawdown is happening your options are already set. The leverage is on, the terms are signed, the size is what it is.
The other four are working the opposite side. Icahn hunts for the trade where he does not need to be clever. Baker looks a layer below the excitement. Chamath strips out the transfers inside the reported growth. Friedberg followed who actually paid instead of who he expected to.
Find something others are not looking at, then build a structure that lets you be wrong about it for three years without getting taken off the board.
That is most of it.
