The Twenties,
again
A pandemic landed almost exactly one hundred years after the last one. Then the speculation came back. Then the tariffs. The pattern is real — and the most honest thing to say about it is also the most unsettling.
“History doesn’t repeat itself, but it often rhymes.” — a line everyone gives to Mark Twain, who almost certainly never wrote it. Which is, itself, the whole problem with patterns: we hand them to an authority so they’ll feel like proof.
It started as a joke about being “due”
Before any of us had heard the word coronavirus, the centennial of the 1918 influenza was already a minor genre — anniversary pieces, a documentary or two, the occasional uneasy aside that we were, statistically, about due for another one. Then 2020 happened, and the joke stopped being funny.
The 1918 pandemic killed roughly 675,000 Americans. COVID-19 killed more than a million. The onsets sit 102 years apart — close enough that the coincidence does something to your attention. It makes you start looking. And once you look past the pandemic, the part that doesn’t go away isn’t the virus. It’s everything that came after the last one.
That is usually where a story like this turns to hindsight. The unsettling part is that it wasn’t hindsight at all — the murmurs were loud, and they were on the record.
The tracks are matched not by calendar year but by years elapsed since each pandemic. Read vertically: the events rhyme. Read the position of “now” — 2026 sits only six years in. In the first lane, the reckoning didn’t arrive until year eleven.
Borrowed money, buying a market priced for perfection
The 1920s ran on a belief that stocks only went up, financed by money people didn’t have. Investors bought on margin — ten or twenty cents on the dollar, the rest borrowed — and for a while it felt like free money. It is the most familiar rhyme in the whole sequence.
By June 2026, the Shiller CAPE ratio sits near 41 — above where it stood on the eve of the 1929 Crash, second only to the dot-com peak of 44. Margin debt has hit a record $1.3 trillion, an all-time high as a share of the economy. None of that is a timer. But it describes a market with very little give.
The Shiller CAPE ratio measures price against ten years of inflation-adjusted earnings. Today’s reading sits well above the 1929 pre-Crash level and trails only the dot-com summit. A high CAPE has never guaranteed a crash — but every prior reading this high preceded a lost decade for returns.
A high valuation never set a date for the fall. It only sets the stakes for one.
Margin debt is money investors borrow against their holdings — the same mechanism that turned 1929 from a sell-off into a spiral. As a share of GDP it now stands at an all-time high. Records aren’t timing signals, but they describe a market with little slack if confidence cracks.
Here is the first place the rhyme breaks — and it matters. The machinery that turned 1929 into 1933 has guardrails now: circuit breakers that halt panic selling, deposit insurance, a Federal Reserve that floods the system with liquidity within hours. A rhyme is not a repeat. The same setup can meet a very different ending.
America First, the first time and now
After the last pandemic, the country turned inward. The Senate rejected the League of Nations. Immigration quotas arrived in 1921 and hardened in 1924. The Fordney–McCumber tariff went up in 1922, and in 1930 Smoot–Hawley raised the wall higher — drawing retaliation that helped turn a downturn into a decade.
A century on, the wall is going back up at a speed the 1930s never saw. The U.S. average effective tariff rate reached roughly 11% in 2026 — the highest since the early 1940s — after touching nearly 17% in late 2025. The slogan is the same one Harding’s era would recognize, and so is the question underneath it: who counts as us.
The U.S. average effective tariff rate on all imports. The 1930 Smoot–Hawley Act lifted duties on dutiable goods toward 59% and drew retaliation that deepened the Depression. After 2025, the rate climbed faster than at any point in modern history — to roughly 11% in 2026, the highest since the early 1940s.
The chief business of the American people is business.
— Calvin Coolidge, 1925
That line captured a whole decade’s mood — low taxes, light regulation, business in the driver’s seat — and it could be printed today without anyone blinking. The difference is the plumbing. In 1930, a country could wall itself off and mostly function. In 2026, the supply chains run through everyone’s living room, so the same impulse lands harder and faster.
Now the honest part
It is tempting to draw a clean line — pandemic, boom, backlash, crash — and stamp a hundred-year interval on it. Two real bodies of theory invite exactly that: Kondratiev’s long economic waves, and Strauss and Howe’s “Fourth Turning,” which times an eighty-to-hundred-year cycle of crisis and renewal. Both are seductive. Neither is accepted as something you can set a watch by.
And the cleanest test of a hundred-year clock — the 1820s — doesn’t cooperate. There was a cholera pandemic, but it ran from 1817 to 1824. There was a financial collapse, the Panic of 1819, the first great American crash. There was an identity rupture, the Missouri Compromise of 1820. The echoes are there. But the dates smear across a whole decade, and if you allow that much slack, almost any decade can be made to rhyme with almost any other.
Why the hundred-year cycle is probably a story we tell ourselves
So the cycle, as a clock, almost certainly isn’t real. That is the disciplined conclusion, and it is worth saying plainly before drawing the one that remains.
What awareness actually buys you
Strip away the mystical calendar and something stubborn is still standing. Stretched valuations. Record leverage. A trade regime in open retreat. Politics fractured along the same fault lines as a hundred years ago. Each of those is independently, soberly documented. You don’t need a cycle to be worried about the pile.
What the pattern buys you isn’t a prediction. It’s a posture. It is the difference between being surprised and being positioned. The conditions of 2026 are the kind that have, in the past, ended in a correction — maybe sharp, maybe shallow, probably not 1929, because the guardrails are real. The move was never to panic. It’s to diversify, to distrust any story that begins with “this time it’s free,” and to make sure you’re not the last one holding the margin call.
History doesn’t owe us a rhyme. But the conditions don’t need a calendar to do damage.
That is the whole case, kept honest: the century-clock is a story, and the risks it points at are not. You can let go of the first and still take the second seriously. In fact, that’s the only responsible way to hold both.