Investing is Einstein’s Cat, or, The Thales Fallacy

Two or three thousand years ago, Indo-European speakers who had turned back at the Khyber Pass decided that some of the gods of their fathers, the daevas, were malevolent spirits. Their relatives in the Indus Valley disagreed and so the languages and religions diverted. Twenty or thirty years ago, I was growing up a science-fiction fan without ever having met one in person. From yellowing paperbacks and early websites I became a twentieth-century humanist well inoculated against some of the follies which clever young men are prone to. I did not know that in California the same movement was festering into something proudly un-democratic, full of longings for yesterday’s tomorrows and resentment that not everyone recognized them as the rightful rulers. I cannot stop them, that work is for people in the United States and UK. I cannot bridge that gap. But this month I can explain what is wrong with two of their core beliefs: that intelligence is the ability to predict the future, and that predicting the future leads to success. The story starts with Thales of Miletus and an olive-press, and ends with Albert Einstein and a cat carrier.

After this, seeing that Alyattes would not give up the Scythians to Cyaxares at his demand, there was war between the Lydians and the Medes for five years; each won many victories over the other, and once they fought a battle by night. They were still warring with equal success, when it chanced, at an encounter which happened during the sixth year, that during the battle the day was suddenly turned to night. Thales of Miletus had foretold this loss of daylight to the Ionians, fixing it within the year in which the change did indeed happen.⁠ So when the Lydians and Medes saw the day turned to night they ceased from fighting, and both were the more zealous to make peace. Herodotus, Histories, 1.74 tr. A.d. Godley

Thales of Miletus was one of the Seven Sages of Ionia, and later Greeks decided that he must have been good at predictions too. Herodotus told a story that he predicted a solar eclipse, and Aristotle taught his students that Thales used his wisdom to predict that the olive-harvest would be bountiful, and rented all the olive-presses in Miletus and Chios in advance (Politics 1.1259a). When the olives needed pressing, he could charge what he liked. Fun-ruining historians point out that not even Babylonians could predict a solar eclipse in the sixth century BCE, and the economy in Aristotle’s story looks much more like the world he lived in than the world Thales lived in. The Aegean around 600 BCE was not a place where everything had a price, but a place where landlords were proud that their neighbours had to come to them and beg to use their olive-presses or borrow some barley before the harvest. The important part of the story is that later generations believed that if Thales was wise he must have been able to use his wisdom to predict the future and get rich. If you can predict something in the natural world, like the movements of the planets, you can predict things in the world of markets and get rich. Anyone who flips through a news magazine or opens YouTube can find worshipful interviews with someone who made a successful prediction in the past. But when you ask the wisest and most mathematically qualified people in investing, they have a very different attitude towards predictions.

Forecasting is for losers

In the spring of 1971, I was about to become a newly minted Ph.D. in abstract, or ‘pure,’ mathematics. … Jobs that would have challenged and fascinated me were, for me, tainted because they only contributed to a war (in Vietnam) I didn’t believe in. … I … was offered a job (at a brokerage firm in Chicago). I thought, I don’t know anything about the stock market— I don’t even know what it is— but I may as well learn about it. Besides, I should easily be able to get rich using my knowledge of mathematics, and why not? … Little was I to know how many people I would meet over the years with the same idea, all of whom would be wrong. Mathematics PhD, long-time finance-industry consultant, and sustainability expert Michael Edesess, The Big Investment Lie (2007) pp. 1-2

One day a friend of my father-of the rich and confident variety- called me during his New York visit (to set the elements of pecking order straight, he hinted right away during the call that he came by Concorde, with some derogatory comments on the comfort of such methods of transportation). He wanted to pick my brain on the state of a collection of financial markets. I truly had no opinion, nor had made any effort to formulate any, nor was I remotely interested in markets. … I … did not make predictions, period. Hedge-fund quant Nicholas Taleb, Fooled by Randomness (2004) pp. 102, 103

The important turning points in markets are never identified with precision in advance by ‘experts’ and policymakers. This lack of foresight is not surprising, because markets and the course of the economy are not model-able scientific phenomena but rather are examples of mass human behavior, which are never predictable with anything like precision. Hedge-fund manager Paul Singer, interview with Forbes, 2016

But people don’t always want the best risk adjusted returns, sometimes people want to speculate … They want to express an informational belief, and who am I to say that’s wrong? Certified Financial Planner Ben Felix, Rational Reminder podcast episode 201 (2022, automated transcript has not been verified)

Most successful investors do not try to predict the future, except in the way you predict that it is more likely to be sunny in summer than winter and you are more likely to get the job if you apply for it. Markets are so unpredictable that it can take fifteen or twenty years to tell the difference between skill and luck. When people sit down to determine “Can Stock Market Forecasters Forecast?” they get the same answer as Alfred Cowles in 1933: no. The last century or so of investment history is dotted with unprecedented events, from the Russian Revolution to the decline of dividend yields in the 1950s to the Zero Interest-Rate Policy. You can learn to expect a sequence of booms and busts by studying the past, but not foresee these black swans which are only in the future. Moreover, you have to be able to survive the times that things do not happen the way you expect, as one 24-year-old investor found out. He was so confident that AI would let him buy entire galaxies like an E.E. “Doc” Smith character (and in the excellent grades he had received on standardized tests) that he did not worry that his stocks might fall in value in the short term. Rather than trying to guess the one thing which will happen, executives focus their thinking about the many things which might happen and how to survive each. So while being able to predict the future would be nice, there seems to be no reliable way to do it, except in the way an insurance company predicts about how many clients will make claims in the next fiscal year. Trying to do what almost no one can do is not not a good strategy.

Instead, the way to get and stay rich is to follow a few simple rules: save every month, invest it in a diversified mix of revenue-generating assets without paying too much to middlemen, protect it from the taxman and friends and relations with open hands, and repeat for decades while those assets grow. Exactly what those assets are matters surprisingly little, as long as most of them are stocks from profitable companies and bonds from stable governments, and as long as you spread your investments widely and keep your costs low. Investors since Jane Austen’s day have been used to returns of about 4% a year plus inflation, whether they held farmland, railroad bonds, or electronics stocks. Over decades, wars, market crashes, and periods of high inflation become bumps in a general upward trend. Done properly, investing is as exciting as watching grass grow, and requires about as much attention as replacing the batteries in your fire alarm once a year.

Anyone interested in personal finance can list people who followed this strategy who were not intellectual or educated. Dave Chilton gave the example of his father, who forgot about a mutual fund for years and had trouble deciphering the difference between units and dollars. Eventually they figured out that it had quadrupled in value since he opened it. Dividend investors tell the story of Ronald Read, a gas-station mechanic in Vermont who started to buy dividend stocks in the 1950s and gave away about $8 million when he died in 2014. Read still had his paper stock certificates sitting in a safe-deposit box, and his way of vetting potential investments was to read the Wall Street Journal every day. That was enough to get rich on a mechanic’s wage, because he did not get discouraged when his stocks went down, did not sell what was lagging to buy the latest cool thing just before it dived, and did not splurge when his investments were doing well then take risks to maintain his lifestyle in a bear market. Financial planner Allan S. Roth taught his son to be a better investor than most of his clients in second grade, because his son trusted arithmetic and did not watch business news and buy and sell based on guesses about the future.

Is there anything better?

In 44 years of Wall Street experience and study I have never seen dependable calculations made about common-stock values, or related investment policies, that went beyond simple arithmetic or the most elementary algebra. Whenever excalculus (sic) is brought in, or higher algebra, you could take it as a warning signal that the operator was trying to substitute theory for experience, and usually also to give speculation the deceptive guise of investment. Benjamin Graham, “The New Speculation in Common Stocks,” The Analysts’ Journal, 1958 p. 20

In hundreds of years, nobody has found anything which will grow money faster or with fewer ups and downs than this simple method. The closest I know are five:

• You can up your exposure to (uncertain, high-expected-yield) stocks when you are young, and hope that this time is good for the stock market and you don’t have to draw on your investments in a trough in the market or get discouraged as your savings shrink and shrink. This lifecycle model works well in practice especially when it is automated.
• You can adjust your asset allocation based on macroeconomic statistics, like buying fewer bonds when interest rates are very low and fewer Japanese stocks when the Japanese stock market is exploding like kduzu (dynamic asset allocation or contracyclical rebalancing). This is easier said than done! Done sensibly, it will drag down your returns in good times in exchange for reducing your losses in the bad. Done carelessly, it will tend to trigger taxation as you sell one asset and buy another.
• You can pick stocks (or hire experts to pick stocks) which have specific mathematical properties or factors. The most famous models were created by Eugene Fama and Kenneth French in the United States. Their five-factor model or smart beta works great on historical data in the USA, but their attempts to apply the model have not been very successful, and others have even more trouble. If you rely on agents, they will be tempted to bend those simple mathematical rules and become just another monkey throwing darts at the finance pages or chasing the latest hot thing, and if you calculate yourself, you have to face the same temptation.
• You can construct a portfolio with a comfortable balance of risk and return and borrow to invest (use leverage). Finance academics love this idea, but most people have trouble borrowing money at low interest, and investing borrowed money is stressful for many people.
• You can get involved in running the businesses you partially own. Most people can’t afford to own a large part of even one business (and tying up all that money in one investment is risky), and would have to give up their old job to start working as an executive. It is not a very useful insight that if you were brilliant at your job you would make more money, and if you were paid more you could save more!

The remarkable thing is that these all take some time and math skills, and they do not put you very far ahead of a waitress putting $500 a paycheque towards blue-chip dividend stocks. The waitress might have more ups and downs, but the economist might be ruined when managed futures don’t move as independently from stocks and bonds as the computer said they would. Most honest advocates for these methods estimate that they can gain a fraction of a percent in annual return, or reduce annual volatility by a few percent while keeping expected returns the same. That is something but not nearly as much as you gain by realizing that there is no point in paying a fund manager 2% a year to pick stocks for you, or moving from one fund to another because of something you heard on the radio.

Man is the rationalizing animal

A little learning is a dangerous thing;
Drink deep, or taste not the Pierian spring:
There shallow draughts intoxicate the brain,
And drinking largely sobers us again. Alexander Pope

The investor’s chief problem – and even his worst enemy – is likely to be himself. Benjamin Graham

There are also risks in looking for anything better. Humans are very good at deceiving themselves and finding patterns in noise (pareidolia). Very many intelligent people have convinced themselves that they know how to pick stocks or buy high-yield bonds and sell them just before they crash, and very many have ruined themselves because of this. The financial services industry is full of people whose job is to convince you to give them control of your money, from columnists telling you what to sell to the nice man who promises he has a way to make securitized mortgages a safe investment. They will keep a percentage regardless of whether their advice is good. Everyone is lying to you for money, and many are lying to themselves. Another twentieth-century humanist, William J. Bernstein, suggests treating anyone in the financial services industry like a hardened criminal. Intermediaries such as the news industry are engaged in entertaining you with future babble and selling sales pitches disguised as popular science. Trading sites and apps are often designed to look sophisticated and stimulating not help you focus on the few relevant facts a few times per year (after all, the people who think they can pick stocks or guess which way interest rates will go are willing to pay for tools, whereas the people who admit they are no good at either just want a few publicly-available numbers in a simple display, so the money is in serving the stock-pickers). The risk of looking for something better is not just that it takes time you could spend reading a book or riding your bike. It is that you will convince yourself or be convinced to do something which is riskier than you know.

Most successful investors do not try to predict the future. Instead, they decide on a pattern of behavior which will have good outcomes in most possible futures, and follow it for decades until randomness has time to average out. If the stock market crashes and does not recover for a decade, your bonds will save you. If the government defaults on its debt, your stocks will save you. If your country is taken over by thieves and ignoramuses, your international assets will protect you (Robeco in the Netherlands saw which way the wind was blowing and spent the 1930s investing in the United States, and were able to report that their surviving clients had done very well by 1946). You don’t need to know which of these will happen when to create an investment policy which can defend against them, any more than a farmer moving her house off a floodplain has to know when the river will rise. She just has to know how high the water has been in the past fifty or one hundred years, and whether there is any reason to think it could go even higher soon. Leopold Aschenbrenner, the 24-year-old who turned a viral essay into a hedge fund, had to watch one of his institutional investors bidding for his assets when he sold at fire-sale prices. They were ready to make money if he was right, and money if he was wrong, and they did not care which because there were a hundred other people with more money than sense in their portfolios and they expected that most of them would blow up.

Brains are overrated

You don’t need a rocket scientist. Investing is not a game where the guy with the 160 IQ beats the guy with 130 IQ. Warren Buffet, Fortune 1990 Investor’s Guide, as quoted by Chris Leithner

Some kinds of Americans (but probably not my gentle readers) like to talk about IQ scores. A contrarian in Australia has gathered the evidence that whatever IQ tests measure is not a help for investors at all. The traits which make a good investor are old-fashioned moral virtues like the ability to live within your means, ignore groupthink and social pressure, acknowledge your own limits, and act rationally when you are terrified. Very clever people without those virtues end up blowing their money studying candle charts or investing in the South Sea Bubble near the peak. If you want to know whether someone will be a good investor, focus on those virtues and behaviors not on things which are loosely correlated with them. If you want to be a better investor, meditating on the Stoics is worth a dozen textbooks on Modern Portfolio Theory. Randall Munro warns of the Engineer Syllogism: some of the worst investors are people with quantitative training, self-confidence, and a love of analyzing systems to find and exploit the hidden rules.

Cat!

I’m not saying that professional money-managers are not smart or knowledgeable people. They are usually well educated. They try to keep themselves well informed about the economy, market sectors, technologies, and current public affairs. This doesn’t mean they they can predict the market or the price of stocks. Mathematics PhD, long-time finance-industry consultant, and sustainability expert Michael Edesess, The Big Investment Lie (2007) p. 91

When Maciej Ceglowski discovered people dreaming of superintelligence, someone introduced him to Einstein’s cat. Albert Einstein was much more intelligent than a cat, at least as people who dream of superintelligence understand smarts, but that did not let him just tell his cat to get in the carrier. He could beg and offer treats using the best feline psychology the 1940s had to offer, but he could not make the cat do what he wanted without pitting sweater and skin against maws and claws. There are fundamental limits to what you can do just by being smart or intelligent. Investment is another example. The very basics of investment can be grasped and carried out by a child. Once you know those things, success comes down to luck and temperament. And this raises a big question.

Many people try very hard to make as much money as possible from their investments. Serious academic work on the field begins with Louis Bachelier in 1900. And yet being more intelligent than average does almost nothing for an investor. Once you understand basic arithmetic and can imagine different possible futures at the same time, what matters is saving regularly, investing according to a clear strategy, and ignoring anyone who encourages you to change course after less than ten years. You cannot use your intelligence to predict the future of markets any better than anyone else, and you don’t need to make predictions to make money. Hard work is not a virtue in investing either: as Jack Bogle used to say, “don’t do something just stand there!” How many other areas of life are like this? Intelligence and hard work are useful in many areas of life, but not in this one.

Herodotus and Aristotle were wrong that being wise lets you predict the future and get rich. The science-fiction fans in California who think they can have success without effort if they just have a big IQ are wrong too. But Iranians and Indians never agreed again about the daevas, even after wave after wave of Iranian immigration into India. I don’t think this post will convince these people either. What I do think is that skepticism and humbleness about your own limits can save you from many follies. It can be profitable to go along with the latest nonsense. Sometimes you even manage to get out before it all falls apart with your life and your reputation intact. But the jug that is brought to the well too often breaks, and these people keep returning to dangerous follies from a hundred years ago. Some have already ended up in prison, or burned out from the drugs and the obsessive talk about the depravity of mankind and the immanence of the Machine God, and the rest are closer to the precipice than they want to believe.

I beseech you, in the bowels of Christ, think it possible that you may be mistaken. Oliver Cromwell before the Battle of Dunbar

In lieu of donations, I will ask that you listen to one more thing. If you ever meet people who are lively conversationalists and very confident in their brains but say horrible things with a polite tone, the rest of your life may pivot on whether you distance yourself from them or stay in the room. I stopped following their websites and podcasts and found a different martial-arts club, and stayed a poor but honest man. These people were too scared of conflict to say “no Nazi shit,” or too greedy for a chance to pitch their own crank idea to object to the other cranks in the room, and it gained them a few pieces of Bitcoin but cost them their souls. You can always earn more money, but souls are issued one per person.

(scheduled August 2026)

1. The philological and religious debate about the chronology is exciting but not for this blog! Encyclopedia Iranica has no entry on pre-Islamic traditions about the div but the Wikipedia entry for daeva is not bad.
2. This can be seen in their obsession with prediction markets and futurism and IQ scores, and their boasting that they were only a week behind the US government in deciding that COVID would be a big deal (they don’t boast that they gave up infection control in 2022 or so, although they are very keen to debate whether China was to blame).
3. Robert Rollinger, “The Western Expansion of the ‘Median Empire’: A Re-Examination,” in Giovanni Lanfranchi et al., eds., Continuity of Empire (?) (S.A.R.G.O.N. Edetrice: Padova, 2003) pp. 307-310
4. Nathan Vardi, “David Tepper, Andrew Beal and Bill Conway: The Top People In Finance Speak About The Markets,” Forbes, 12 May 2016
5. Alfred Cowles 3rd., “Can Stock Market Forecasters Forecast?” Econometrica 1, no. 3 (1933): 309–24. https://doi.org/10.2307/1907042 For subsequent research see William J. Bernstein, The Four Pillars of Investing, Second Edition (McGraw Hill: New York, 2023) pp. 68, 69.
6. Berber Jin, Ben Cohen, and Anissa Gardizy, The Connections That Turned a Precocious Teen Into the Fallen ‘Nostradamus of AI’, The Wall Street Journal, 25 August 2026. This article does not discuss Aschebnrenner’s shorts on stocks which he thought would be hurt by chatbots, giving him a portfolio of borrowed money in holdings which would all go up or down together. Even a chatbot could have told him this was very risky. The authors did not even try to explain how Aschenbrenner thought really smart computers would lead to faster-than-light travel because Silicon Valley is a silly place (I can’t explain it either although American science fiction fans have hated relativity for a long time).
7. Mebane T. Faber’s Global Asset Allocation (The Idea Farm LP, 2015) is one good place to start. He found that from 1973 to 2013, all sensible portfolios returned investors in the USA somewhere between 4% a year and 6% a year after inflation if you picked one and stuck with it. If you switched from whatever strategy was having a bad decade to whatever was doing well this year you would perform poorly, because all strategies have good decades and bad decades, and changing course to whatever has done well recently just locks in your losses while giving up some of the good times.
8. Dave Chilton, The Wealthy Barber Returns (Financial Awareness Corp, 2011) pp. 170-171
9. Efficient Market Theorists have a Mirror Universe version of this policy, where if stocks in Syldavia are booming, that is not a sign to cash in some of your winnings but the rational market deciding that Syldavia is the best place to invest. I don’t take investment advice from people with goatees under a sword-and-planet logo but you do you
10. I can’t resist putting a European company called AI Alpha Lab in to the historical record. Its marketing pitch is breathtaking: “Initializing AI Alpha Lab… Connected. Thank you for visiting our site. I am an AI model specifically developed for investing, nothing else. I use data and probabilities to make investment decisions. I have no biases or preferences. And no humans interfere with my investment decisions. Welcome to the future of investing. Today.” Anyone with Math 12, a Raspberry Pi, and a brokerage account can apply factor investing, but the problems are always getting humans to stick with it and the possibility that the future is not like the past. Many investing strategies work brilliantly until they don’t, and many people say they will stay the course then withdraw their money the first time their account drops 20%. More experts, more data, and more number-crunching cannot avoid these basic problems.
11. One site pushing “return stacking” proposes for the sake of argument that you can borrow money at 0.5% greater than the yield of a 90-day U.S. government treasury bill. Indeed, if I could borrow money as cheaply as the Irvings can, all kinds of wonderful opportunities would open up for me!
12. Eg. Dimensional Fund Advisors’ ten-year annualized return for Canadian equities is 13.03% which is a quarter of a percent higher than a plain old TSX index like XIC (12.74%) or VCN (12.63%). When you read the prospectus you realized that to invest in Dimensional’s products you need a paid advisor who will generally charge at least 0.75% a year. Cullen Roche spent ten years developing a rule to switch between stocks and bonds based on macroeconomic statistics which has historically returned 0.12% more a year than a simple 60/40 portfolio (Your Perfect Portfolio p. 212). This strategy requires frequently selling stocks to buy bonds or bonds to buy stocks, which is a taxable event in a non-registered account. Meb Faber has his own version of this idea with the same impressive backtest and the same problems of tax drag and how to know if the strategy will continue to work: Meb Faber, “A Quantitative Approach to Tactical Asset Allocation,” The Journal of Wealth Management, Spring 2007 https://ssrn.com/abstract=962461.
13. William J. Bernstein, If You Can: How Millennials Can Get Rich Slowly (self-published, 2014)
14. Bernstein calls this the engineer’s mistake (Four Pillars of Investing, second edition p. 194) and uses phrases like “less math, more Shakespeare.”