sanraku!

money, odds & institutions / from the tape, not the summary

DM / offer, collab

[sanraku! @manwithoutfate]

Tractors blockaded the Federal Reserve in February 1982, and the man inside had just taken the policy rate to 20%.

Paul Volcker tells the story in this video, which runs about 4 minutes and has 119,000 views.

He'd become chairman in August 1979, inheriting a decade of rising prices and a government that had been fighting it with slogans and campaign buttons.

Important detail: inflation fell from 14.8% to roughly 3%, and unemployment peaked at 10.8% on the way there. Nobody thanked him at the time.

Save this before you say a central bank should just do what's necessary.

[sanraku! @manwithoutfate]

@ComputersRh lfg!

[sanraku! @manwithoutfate]

Brooksley Born was the only senior official asking about derivatives in 1998, and by 1999 she was gone.

This 6 minute video from 2009 is her own account, given while accepting the Profile in Courage Award.

Her agency wanted to know whether a market with hundreds of trillions in notional value should have any supervisor at all. Washington's answer was no, and it wasn't polite about it.

Important detail: the law Congress passed afterwards made the answer permanent for the next eight years.

Save this one, private credit has crossed $2T and the same questions are being brushed aside.

[sanraku! @manwithoutfate]

A model is a bridge from something that trades to something that doesn't, and Emanuel Derman builds his out of apartments.

In this video from the Institute of Physics, with about 25,000 views, he prices a 12 room penthouse using small flats in Battery Park City.

Those flats change hands every few days because Wall Street is constantly hiring and firing. The penthouse hasn't traded in years.

> find something that actually trades
> pick one intuitive unit, like price per square foot
> convert, and accept the first number is wrong
> then correct for everything the unit ignores

Bookmark this one, the four lines describe every valuation you'll ever be shown.

[sanraku! @manwithoutfate]

The expected value of a game paying 2, 4, 6 or 8 is exactly 5, and 5 is the one result you can never get.

Howard Marks uses that in this video to explain why averages mislead people who invest real money, and he's run Oaktree since 1995.

You don't experience the average, you experience a draw, and the draws are all that hits your account.

Important detail: the same logic applies to a career, a business and a portfolio, where one bad draw can end the sequence entirely.

Save this if you've ever planned around a number that can't actually occur.

[sanraku! @manwithoutfate]

@0xLeeem if that's rick guerin, buffett's line was that he was in a hurry and got margin called

[sanraku! @manwithoutfate]

@enzoxbt rebalancing two dead stocks. cover called it universal portfolios and nobody wanted it

[sanraku! @manwithoutfate]

@kursormaxx he says he'd do it again. that part never makes the screenshot

[sanraku! @manwithoutfate]

@0xLupenn the math was never hidden. it was just boring enough that nobody read it

[sanraku! @manwithoutfate]

@isssa0x math says take it, rent says no. both are right

[sanraku! @manwithoutfate]

"If you put $1,000 in a stock, all you can lose is $1,000."

Peter Lynch builds his whole approach on that sentence in a 1994 video, and it explains a record of about 29% a year over 13 years.

The downside is fixed at what you put in, while the upside on a company that works can be five or ten times your money. So a quarter of your ideas can carry the other three quarters.

That math is why he never tried to time the market. Magellan fell in all nine declines on his watch, including the $12B to $8B drop of October 1987.

Save this if you cut your winners early.

He lost in every decline of his career and beat nearly everyone anyway, which raises an awkward question about what you're paying anyone to protect you from.

[sanraku! @manwithoutfate]

Richard Thaler's most useful idea makes you save more without giving up any money today.

It's called Save More Tomorrow, built with Shlomo Benartzi, and he explains it in a 2010 video. You commit now to putting part of a future raise into your pension, so your take-home pay never falls.

People say yes to that because the sacrifice is always in the future, which is the same reason diets start next month.

The companion trick is automatic enrollment, which took participation in one plan from 49% to 86% and above 90% elsewhere.

Bookmark this if you've been meaning to increase your contributions.

The video has about 3,200 views, seven years before the man in it won a Nobel.

[sanraku! @manwithoutfate]

"My personal net worth was increasing ten billion dollars per week."

The AI trade has created paper fortunes again. This clip is about what paper fortunes do when the price turns.

He says it calmly, as a description of a time that was not normal. For three days in early 2000, Masayoshi Son was richer than Bill Gates.

Then the share price turned. Within six months, he tells the interviewer, it was down 99 percent, and SoftBank almost went bankrupt. His own net worth, which had reached roughly seventy eight billion dollars, fell to around one billion within about a year.

The number that matters is not the peak. It is the speed. Ten billion a week on the way up means the same machinery can remove ten billion a week on the way down, and there is no mechanism in between that converts paper into money in time.

Four things worth knowing about wealth that grows that fast:

> It is a quote on a share, not a sum you can spend.
> The faster it rises, the larger the share of it that is other people's enthusiasm.
> Selling enough to matter would itself move the price you are trying to capture.
> The only protection is deciding in advance what fraction is real to you.

If someone you know is watching one position multiply week after week, that first line is the conversation to have before it turns.

It is a mainstream television interview, which is exactly why the ten billion a week line slides past almost everyone who hears it.

Save the post now, because this situation repeats and you will need it again.

Ten billion a week up, ninety nine percent down in six months. At which point would you have called any of it yours?

[sanraku! @manwithoutfate]

Two axes on a sheet of paper, return on one and risk on the other, and a student who did not know the formula he needed.

One Wall Street firm called the AI boom near its end this month, and the framework everyone uses to answer that came out of this afternoon. The formula arrives about six minutes in.

Harry Markowitz was at the University of Chicago, hunting for a dissertation. The prevailing theory said value equals the present value of expected dividends, which implies a portfolio of one holding. He knew that was wrong and could not yet prove it.

He knew the expected value of a weighted sum but not its variance. A textbook on the library shelf gave him the answer, and the answer contained covariances.

That is the whole origin of the idea that a portfolio's risk is not the average of its parts. Two years later it was a paper. Thirty eight years later it was a Nobel prize.

What the graph did to finance:

> It made risk and return one picture instead of two arguments.
> It defined a set of portfolios that cannot be improved.
> It moved the question from which asset to what mix.
> It gave the industry a defensible process for the first time.

If you have ever sat in a meeting where people argued about a single holding for an hour, that third line is what was missing.

The interview has about 74,000 views, eight years after a university put it online for free.

Bookmark this. Next time someone asks you about it, you will want the numbers.

He drew the axes before he had the maths. Most people wait for the maths and never draw the axes.

[sanraku! @manwithoutfate]

Bomber crews in the Second World War had to choose between a flak jacket and a parachute. They could not wear both.

Andrew Lo tells this story carefully, and the first thing he says about it is that he cannot verify it. He heard it from Steve Ross, who said he heard it from Amos Tversky.

The reason to hear it this month, while markets argue about concentration in a handful of stocks, is what the crews reportedly did.

Shrapnel was the cause of death about seventy five percent of the time, missiles about twenty five. The optimal choice is obvious: flak jacket, always. According to the story, the crews took the jacket about seventy five percent of the time and the parachute about twenty five.

Lo uses the story as an illustration and immediately grounds the phenomenon in things that are documented: the same pattern in goldfish, ants, pigeons and primates.

What to take from an unverified story told honestly:

> He flags the sourcing before he uses it, which is rare.
> The illustration is separated from the evidence.
> The evidence is the animal experiments, not the anecdote.
> A good lecturer tells you which is which.

If you repeat business anecdotes for a living, watch how he handles this one before you tell your next.

The lecture has around 29,000 views and this passage is about two minutes of it.

Bookmark it now, because tomorrow you will not remember the numbers.

He could not verify the story and told the audience so. How many of the stories you repeat would survive that test?

[sanraku! @manwithoutfate]

Tomorrow's flight from Austin to Dallas has three empty seats, and the airline will not sell you one of them at any price.

That is not a glitch. It is the example Robert Crandall, who ran American Airlines from 1985 to 1998, gives in a 2013 conversation at DePaul University's law school. "We won't sell you those seats to go to Dallas at any price," he says.

The reason is three other people. The airline's data says that somewhere in Austin, three travellers will want to fly to Dallas and connect to Europe. If there is no seat for them on the short hop, a seat to Frankfurt, London or Paris flies empty. The last seat out of Austin is worth many times more to the system than a passenger who only wants Dallas.

His definition of the whole discipline fits in one sentence: establishing "the probable value of every seat on every airplane for the next 365 days."

The logic behind the empty seats:

A seat has no fixed value, only a probable one.
Its value depends on who else will want it, not on what it cost.
Selling cheap now can kill a more expensive sale later.
Sometimes the right move is refusing money on the table.

For anyone who has been told a flight was sold out while the seat map showed gaps, this is the explanation nobody at the counter gives.

The recording has been online since 2013. About 2,800 people have watched it.

By the early 1990s American estimated the system had added 1.4 billion dollars over three years. Would you have turned away a paying passenger to find out?

[sanraku! @manwithoutfate]

A woman he knows lost her husband, remarried, and ten years later divorced the second one. For years afterwards she described the marriage as a bad decision.

The man telling the story disagrees with her, and not out of politeness.

"Good decisions never become bad and bad decisions never become good," says Ronald Howard, the Stanford professor who named the field of decision analysis in 1964. In his reading, she made a good decision to marry him, with what she knew then, and a good decision to divorce him later, with what she knew by then. Two good decisions. No contradiction.

He gives the same structure in money: buying a stock at one point can be a good decision, and selling it later can be a good decision too, and the second one says nothing about the quality of the first.

Worth writing down:

Judge a decision on the information and the alternatives available at the time.
New information produces a new decision, not a retroactive verdict on the old one.
If you would not have known it then, it cannot count against you now.
Keep a record of what you knew, or memory will quietly rewrite it.

If you know someone still re-litigating a choice they made five years ago, this is the four minutes to send them.

It is a conference recording that fewer people have opened than were sitting in the room.

She kept calling it a bad decision. He kept telling her she had made two good ones. Which version would you apply to your own last big choice?

[sanraku! @manwithoutfate]

Fifty miles decided whether the storm cost fifteen billion dollars or sixty.

Hurricane Andrew came ashore in August 1992 south of Miami, near Homestead. The insured loss was about fifteen billion, which was already three times what the American insurance industry believed its worst case to be, and enough to bankrupt eleven companies.

Karen Clark, who had built the first commercial hurricane model, makes the point without drama: had Andrew hit about fifty miles further north, into the built up part of Miami, the loss would have been around fifty or sixty billion. Anyone could do that arithmetic afterwards on a calculator.

That is the entire argument for modelling rather than remembering. A catastrophe is not a fixed quantity attached to a storm. It is the storm multiplied by whatever happens to be underneath it.

What follows from that:

Your loss depends on geography and on timing, not on the severity rating alone.
Two identical events produce completely different bills.
History under samples the expensive versions, because most storms miss the expensive places.
Model where the value sits now, then move the same event across it.

If your business has one site that holds most of the value, that last line is a half day exercise with a large payoff.

It is a video call recorded for an insurance company's channel, which is exactly as promoted as it sounds.

Fifty miles of coastline stood between an expensive disaster and one that would have taken the industry apart. Nobody chose that margin. It was weather.

[sanraku! @manwithoutfate]

"The price of land always goes up, especially in places like London, where land is scarce."

Robert Shiller quotes that sentence back to a room in London, and then takes it apart in one move: just because land is scarce does not mean the price will always go up.

He calls this kind of belief a naive theory, and says it behaves like a thought virus. It is not an economist's model. It is a sentence that people repeat to each other because it explains what prices are already doing, and it spreads fastest exactly when prices are rising.

His own record here is unusual. The first edition of his book on bubbles appeared in March 2000, days after the Nasdaq top. The second, in 2005, added housing, and American house prices peaked the year after.

Four questions to run on any sentence of that shape:

Does the claim contain a mechanism, or only a fact about supply.
Has the same sentence been true in every period, including the falls.
Who benefits from you believing it right now.
What would the price have to do for the believers to abandon it.

If your city has a version of the London line, that second question is a twenty minute check in a price archive.

The reason he calls it a virus rather than an error is that it spreads by being repeated, not by being tested. Nobody checks the claim before passing it on, because it arrives attached to a price that is currently proving it.

Free, on a university channel, viewed by a few tens of thousands since it was posted.

Scarce land, rising prices, and a sentence that has survived every crash it was supposed to explain. Which one do people repeat where you live?

[sanraku! @manwithoutfate]

He is on every list of the century's great mathematicians and on almost none of the lists of great investors, which is strange, because from the late 1950s to 1986 his portfolio returned about twenty-eight percent a year.

Claude Shannon worked at Bell Labs and then at MIT. In 1948 he published the paper that created information theory, invented the bit, and made digital communication possible.

He ran his own money at the kitchen table with his wife Betty. In 1986 Barron's ranked more than a thousand mutual funds and only one beat him.

He knew the mathematics of betting better than anyone alive. He gave Ed Thorp the Kelly paper. He described, in an MIT lecture, the rebalancing puzzle that now carries his name and that people still write papers about.

And he made his fortune on three stocks. Teledyne, Motorola, Hewlett-Packard, held for decades. His stated reason for the largest position was that he thought well of the man running the company.

A half-hour film about his life sits on YouTube with four hundred thousand views and barely a sentence about any of this.

He spent his career proving how much can be extracted from a noisy signal. Then he bought three things and stopped listening.