Qobalt

Quant Finance
Agent Trading Systems

Sub for more Alpha

[Qobalt @Qobaltik]

two Nobel laureates at Chicago Booth can't agree on one question that decides how you should invest

Eugene Fama says markets are efficient. Richard Thaler says people are too irrational for that to always be true

Fama's argument is brutally simple: "prices reflect all available information"

and his advice is even more brutal

"I don't know any investors who shouldn't behave as if markets are efficient"

then Thaler brings up his two-year-old granddaughter

she runs around like crazy, and her behavior "isn't well captured by a model maximizing anything other than fun"

his point is that markets can behave the same way

Black Monday saw stocks fall 25% with no obvious news

Fama calls that an anecdote. Thaler sees a problem for models built around perfectly rational behavior

then they get to the Cuba Fund

it traded at a 70% premium, and Thaler calls it a bubble

"it's just a mistake"

Fama fires back: "it's a one-day bubble... it's an anecdote"

that's the entire debate in miniature

one Nobel winner sees temporary stupidity

the other sees insufficient evidence

and somehow both still agree on what most investors should do: stop assuming you're smart enough to consistently beat the market

[Livsun @L1vsun]

https://t.co/qOBb2XGkl3

[Qobalt @Qobaltik]

McDonald's charges seventy cents for more than double the soda, and an MIT professor says that pricing is the whole of human preference compressed into one cup.

The syrup and the extra cup cost the company about a penny. The other sixty-nine cents is a number they know about your brain that you don't.

"You're desperate for that soda, but you're not as desperate to have twice as much soda."

That's Jonathan Gruber, thirty-five minutes into a lecture that's been free on MIT's site for years.

The first sip on a hot day is worth real money to you. You'd pay almost anything for it. The fiftieth sip is worth almost nothing, and McDonald's has priced the cup accordingly, decades before you walked in.

He calls it diminishing marginal utility, and the whole idea sits on three assumptions so basic nobody writes them down.

You always have an opinion between two options. Your preferences don't loop back on themselves. More is always at least a little better.

Three sentences, and from them you can draw the entire shape of what a person wants, before they've spent a dollar.

Then the pricing follows the shape. Steep charge to go from zero units to some. Nearly free to go from some to a lot. Every all-you-can-eat buffet, every Costco size-up, every SaaS tier is the same curve wearing a different label.

Gruber has one test for whether you've actually understood any of it. Could you explain it to your mother, someone who's never opened an economics textbook, well enough that she gets it.

Most people who quote diminishing marginal utility at a dinner party would fail their own mother's test.

The lecture is free. MIT posted the whole course. The people paying seventy cents without knowing why have never pressed play.

[Livsun @L1vsun]

https://t.co/qOBb2XGkl3

[Qobalt @Qobaltik]

A ninety-year-old man defined the entire concept of risk in eleven words, on camera, and the clip has been sitting free since 2008.

"Essentially risk means we don't know what's going to happen."

That's it. That's the whole idea Wall Street charges billions to manage.

His name was Peter Bernstein. He spent decades running money before he spent his last ones writing the history of the idea instead, in a book called Against the Gods.

Risk means more things can happen than will happen.

Sit with that for a second. Every model, every scenario tree, every stress test is an attempt to enumerate a list that is provably larger than what actually occurs. You are always modeling more futures than the one you'll live through.

In 1952 a young economist said something Bernstein calls a thunderbolt. You have to think about risk as well as return.

Before that, investing was just picking winners. Markowitz made risk itself the second axis of the decision, and it took Wall Street years to stop treating it as a footnote.

Twenty years later three more men built the most sophisticated version yet. Fisher Black, Myron Scholes, Robert Merton, an option pricing model elegant enough to look like physics.

Then Merton and Scholes helped run Long-Term Capital Management, and it blew up anyway.

"The only thing they forgot was that all of those models were based on a world in which LTCM did not exist."

The fund's own size and leverage were never a variable in the equation pricing its trades. The model measured the risk of everyone else's market and never once measured itself.

A trader I know keeps that line pinned above his desk, not the equation, just that sentence.

The interview is thirteen minutes. Free on YouTube since 2008. The people still trusting a model that can't see itself have never pressed play.

[Livsun @L1vsun]

https://t.co/qOBb2XGkl3

[Qobalt @Qobaltik]

Every finance degree on earth is built on a model an MIT professor says is broken, and he isn't a critic on the outside throwing rocks. He wrote half the papers that model rests on.

Andrew Lo taught efficient markets for years before he started building the thing that replaces it.

"The traditional investment framework is flawed. It's not wrong, but it's incomplete."

Here's the piece it's missing. In stable environments the old relationships work fine. When the environment shifts, they temporarily break down, and nobody's model told them that was coming.

The reason traces back to a decision economists made decades ago, to build the whole discipline on a person who doesn't exist. Someone who calculates every option and always picks the best one.

Real humans don't do that. Herbert Simon named it before Lo was born.

"Individuals don't optimize. They satisfice. They use heuristics, they make mistakes, and they learn from those mistakes."

Lo took that and went looking for the actual mechanism, not in economics but in evolutionary biology. Markets aren't a physics equation that holds still. They're a population, competing for the same limited food, adapting generation over generation.

"Markets are highly competitive, innovative, and adaptive. Their degree of efficiency varies over time as the population of investors and the financial landscape change."

Which means efficiency isn't a permanent property of a market. It's a temporary condition, present when a system's strategies have had time to adapt, gone the moment something changes faster than traders can relearn.

"2008 was a watershed year. It showed us that the efficient markets hypothesis was on vacation."

Nobody's model failed in 2008. The environment moved and the models kept quoting an equilibrium that had already left the room.

A trader I know says this reframed every backtest he's run since. Not "does this strategy work," but "what environment was it adapted to, and is that environment still here."

The lecture is free. MIT posted the whole series. The people still treating 2008 as a black swan instead of a pattern have never pressed play.

[Livsun @L1vsun]

https://t.co/qOBb2XGkl3

[Qobalt @Qobaltik]

Ten million people have watched an MIT professor accidentally gut the executive coaching industry

he filmed it once in January 2018 and was dead eighteen months later

executive coaches bill fifteen thousand dollars a session to deliver about a third of what he gave away in a single hour

his name was Patrick Winston. He ran MIT Artificial Intelligence Laboratory from 1972 to 1997 and wrote the AI textbook every computer science major on earth read for thirty years

every January for four decades he stood up and gave the same lecture, called "How to Speak"

whole framework fits on a napkin

DO NOT READ. Be in the image. Keep images simple. Eliminate clutter. Open with an empathetic connection. Close with a punch line the audience can repeat over dinner. Never start with a joke. Never end with "thank you"

that final rule alone has probably cost the coaching industry a hundred million dollars.

"your success in life will be determined largely by your ability to speak, your ability to write, and the quality of your ideas. In that order"

that is how the lecture actually opens. Winston believed it enough to spend fifty years teaching computer scientists how to talk

founders drop $68,000 on an MBA, then pay a communications coach to walk them through material Winston filmed once and left online. Engineers write brilliant code and watch promotions go to teammates who played this lecture on the train

lecture sits free on MIT OpenCourseWare. textbook is free on his page

Winston died in 2019. almost none of those ten million viewers have ever used four rules on the napkin

napkin is free. willingness to actually pull it out in your next meeting is entire edge

[Livsun @L1vsun]

https://t.co/qOBb2XGkl3