Why recessions surprise almost everyone

Every recession is obvious in hindsight and invisible beforehand. The warning signs, once you know the ending, look glaring, and yet the professionals whose job is to see them coming, armed with more data than anyone in history, miss the turn again and again. It is tempting to put this down to incompetence. The truer, more uncomfortable explanation is that economies do not slow down smoothly, and the tools we lean on are built for a world that changes gradually.

§ 1Economies turn, they do not glide

The intuitive model of a downturn is a gentle deceleration, growth easing from fast to slow to negative like a car coasting to a stop. Real economies rarely behave that way. They hold up, and hold up, and then tip, because so much of economic life runs on confidence and momentum that reinforce each other until they suddenly don’t. Hiring supports spending, which supports hiring; when that loop reverses, it reverses fast. The turn is nonlinear, and linear expectations are structurally unprepared for it.

§ 2The data arrives late, and revised

Compounding the problem, we steer using a rear-view mirror. Most headline economic data describes the recent past, is published with a lag, and is then revised, often heavily, months later.¹ Some of the sharpest downturns were officially recognised only well after they had begun, once revisions revealed the economy had been weakening while the real-time numbers still looked fine. Reacting to the latest print feels prudent; it can also mean reacting to a world that no longer exists.

§ 3Feedback loops the models understate

Standard forecasts tend to assume relationships hold in their normal proportions. Recessions are precisely the moments when they don’t. A modest fall in demand leads a firm to cut jobs; the newly unemployed cut their spending; that deepens the demand shortfall; more jobs go. Each step is individually reasonable, but together they amplify rather than settle. Models built around gentle, self-correcting adjustment consistently underweight these spirals, which is why the forecast error tends to be largest exactly when it matters most.

§ 4Everyone is positioned for continuation

There is a behavioural layer on top of the economic one. After a long expansion, nearly everyone, businesses, investors, forecasters, is arranged for the expansion to continue, because for years that was the correct bet. That shared positioning is itself stabilising until the moment it isn’t, and when the regime finally shifts, the crowd all tries to reposition at once. The surprise is not only that the economy turned; it is that so many were leaning the same way when it did.

§ 5What to do with a signal you cannot time

None of this means the warning signs are useless. An inverted yield curve, deteriorating credit, softening in the most cyclical corners of the economy, these genuinely tilt the odds. What they cannot do is tell you when. The mature response is not to predict the date but to notice when risk is rising and to hold your assumptions more loosely, so that if the turn comes you are surprised by its timing rather than its direction.

This is educational content, not investment advice. Recessions will keep surprising people, and no framework changes that. But understanding why they surprise, nonlinearity, lagged data, feedback, crowded positioning, turns the surprise from a shock into something you were at least braced for.

Note. Educational content only. This working note is general information and does not constitute investment, financial, tax, or legal advice, and is not a recommendation to buy or sell any security. Cases and figures are constructed for exposition.