The comfortable belief, and it is held by serious people, is that we have outgrown the bubble. We have real-time data, deep liquidity, an alphabet of regulators, and eight centuries of documented disasters to learn from. Surely the crude manias of the past, tulip bulbs and railway shares, cannot happen to investors this well-equipped. John Kenneth Galbraith, the economist who spent a career studying these episodes, spent a slim book taking that belief apart. In A Short History of Financial Euphoriahe sets out the uncomfortable counter-fact: the protection we count on, memory, has a short shelf life. “For practical purposes,” he wrote, “the financial memory should be assumed to last, at a maximum, no more than 20 years.” That is roughly the time it takes for the last catastrophe to fade and for some new version of the old folly to capture the financial mind again. Twenty years is not an accident of arithmetic. It is about the length of one working career.
Read that way, the question is not whether better tools have ended the mania. It is who is in the room when the next one starts. The people placing the largest bets at the top of a cycle are usually the ones who were too junior, or not yet born, the last time the same bet blew up. The data is better. The memory is not. And memory, Galbraith argued, is the only defense that ever mattered.
The two features that never change
Galbraith claimed that every episode of euphoria, however different on the surface, shares two features. The first is the brevity of the financial memory we just met: each generation arrives convinced it is seeing something genuinely new. The second is what he called the specious association of money and intelligence (the conviction that the people who happen to be getting rich as prices rise must possess a special insight, rather than simply standing in the path of a rising market). During the boom, being right and being lucky are impossible to tell apart, so the crowd credits the winners with genius. The genius, Galbraith noted dryly, is always discovered before the fall, never after.
Underneath both features sits a single mechanism, and naming it is the whole point. The “innovation” that each cycle celebrates as unprecedented is, on inspection, almost always leverage in a fresh wrapper, debt arranged so that a small move in the underlying asset produces a large move in the speculator’s equity. Galbraith put it with a shrug: the world of finance “hails the invention of the wheel over and over again, often in a slightly more unstable version.” All financial innovation, he added, comes down to the creation of debt secured, well or badly, against real assets. Strip the wrapper off and you find the same engine every time. What changes is the packaging, which is exactly what hides the leverage from regulators, the press, and often the speculators themselves until the unwind exposes it.
A toy version makes the mechanic legible. Put in $10 of your own money, borrow $90, and buy an asset for $100. A 10% rise hands you a 100% gain on your stake; you look like a genius. A 10% fall wipes you out. Nothing about that math is new or clever. Dress the same structure up as a trust, a tranche, a token, or a special-purpose vehicle, and a fresh cohort of investors will treat it as a financial discovery rather than as the oldest trick in the book.
One career, then the clock resets
Lay the documented episodes end to end and the cadence is hard to unsee, even allowing for loose edges. The investment trusts of the late 1920s were the headline innovation of their day, layered companies that held shares in other layered companies, each tier borrowing against the one below. By 1929 a new trust was appearing at the pace of roughly one per business day, and the leverage that magnified the gains magnified the collapse just as faithfully. Irving Fisher, the most respected economist in America, told an audience in the middle of October 1929 that stock prices had reached “what looks like a permanently high plateau.” The plateau gave way within days, and the market fell by the great majority of its value over the next three years. Galbraith treated that line not as a personal failure but as a recurring symptom.
Then, roughly a working generation apart each time, the wrapper changes and the structure returns: the conglomerate-building of the 1960s, the junk-bond and buyout wave of the 1980s, the late-1980s asset bubble in Japan, the dot-com mania at the turn of the century, the mortgage and structured-credit boom of the mid-2000s, and the SPAC-and-crypto enthusiasm of the early 2020s. I am keeping the dates deliberately broad, because the exact start and end of any bubble is arguable and that argument is a trap. What is not arguable is the family resemblance. Each had a genuine kernel of innovation, real railroads, a real internet, real cryptography, which made the “this time is different” story feel like analysis rather than excuse.
That phrase has its own pedigree. Carmen Reinhart and Kenneth Rogoff built an 800-year study of financial crises around it and called their book This Time Is Different. Their epigraph comes from a trader with, in their words, an uncharacteristically long memory, who told a room of Federal Reserve governors during the 1998 crisis: more money has been lost because of four words than at the point of a gun, and those words are “this time is different.” The recurrence of the claim is itself the tell. When the case for a price rests on the argument that the old rules of valuation no longer apply, the burden of proof should rise, not fall.
Where the crowd leaves a fingerprint
Here is the problem with all of this as practical advice: “it looks like a bubble” is not a timing signal. Manias inflate for years, sometimes most of a decade, and the investor who calls the top early can be right about the destination and ruined by the route. Calling tops is also a reliable way to miss large and entirely legitimate gains; plenty of expensive-looking companies kept compounding for a generation. So the lesson is not market-timing bravado. It is humility plus position discipline, and that is where a measurement helps more than a forecast.
The crowd Galbraith describes leaves a measurable fingerprint, and at Obermatt the Sentiment rankis the instrument we use to read it. We score a company on how analysts, owners, and the recent flow of revisions are leaning, then express it the same way we express every rank: from 1 to 100, relative only to the company’s true peers, the handful of businesses competing for the same capital. A name or a whole sector priced on euphoria rather than on fundamentals shows up as a sentiment reading running far ahead of what value and safety can justify. That gap, sentiment outpacing the fundamentals, is the closest thing to a dial for the “specious association of money and intelligence” that Galbraith could only describe in prose. It does not tell you the top is tomorrow. It tells you that you are being asked to pay for a story, which is precisely when the historical record says to slow down.
A sentiment reading on its own is only half the defense. The other half is the balance sheet, because leverage is what turns a popping bubble into a personal catastrophe; the same euphoria that lifts the price tends to ride on debt that the cheerful narrative conveniently ignores. A peer-relative safety reading is the antidote, and reading the safety rank shows how to use it as mania defense: prefer the businesses that would survive the unwind over the ones that only thrive in the boom. The behavioral sibling of all this is the disposition effect, the pull that makes a crowd hold on long after the story has curdled.
The durable principle is the one Galbraith left us: financial memory is short by default, so you have to supply your own. You cannot will yourself to remember a crash you never lived through, but you can build the memory into your process. The concrete action is small and repeatable. The next time an investment is sold to you as a genuinely new kind of thing, ask the two questions the historical record always rewards. Where is the leverage hiding, and am I being asked to pay for a fact or for a story? If the honest answer to the second is “a story,” let the Sentiment rank confirm what your unease already suspects, and size the position as though this time is exactly like all the others. Because, so far, it always has been.