Who decides, and how long they take#
There’s a rule of thumb — two consecutive quarters of falling GDP — and then there’s the actual process. In the US, the National Bureau of Economic Research’s Business Cycle Dating Committee makes the call, weighing employment, income, industrial production and spending rather than GDP alone.
The committee waits for enough data to be confident, which means announcements come long after the fact. From 1979 to 2021 the average lag was about 6.3 months from onset, and across the turning points since 1980 it has run closer to 11.7 months, ranging from 5 to 21.
The December 2007 recession wasn’t declared until November 2008 — eleven months later. The February 2020 recession was called in June 2020, and that was considered unusually fast.
So “we’re in a recession” is a historical statement. It’s never news you can act on.
And the market isn’t the economy anyway#
Even if the announcement were instant, it would be the wrong signal, because the two things being confused run on different clocks.
The market is forward-looking. It prices what people expect. Economic data is backward-looking — it measures what already happened.
Which means bad economic news and a bad time to invest are not the same thing, and frequently the opposite. The market bottoms while unemployment is still rising, because unemployment is a lagging measurement of a decline that’s already priced.
March 2020 is the cleanest demonstration on record. The S&P 500 fell roughly 30% in weeks, bottomed around 23 March, then began one of the fastest recoveries in modern history — while headlines still read “worst crash since 2008”, lockdowns were spreading, and unemployment was still climbing.
Anyone waiting for the news to improve waited well past the recovery.
The best and worst days share a postcode#
The reason waiting is so costly: volatility runs in both directions, because uncertainty does.
In March 2020, two of the S&P 500’s ten worst trading days and two of its eleven best days occurred in the same month. That isn’t a coincidence — extreme uncertainty produces extreme moves either way, and they cluster in the same weeks.
This holds more broadly. Nine of the ten largest single-day gains landed during recessions, six of them during bear markets. Selling to avoid the worst days reliably removes the best ones too, because they’re the same fortnight.
Recessions are shorter than they feel#
Some context that helps when you’re inside one. The US has had 34 NBER-dated recessions since 1854 — roughly one every five years.
The long-run average length is around 17 months, skewed by pre-modern crises. Since World War II the average is nearer 10–11 months. The extremes bracket it usefully: 2007–2009 ran 18 months; the 2020 pandemic recession was over in two.
Recovery takes longer than the recession. The S&P 500 bottomed in March 2009 and didn’t fully recover until early 2013 — about four years. That’s the number worth internalising, because it’s the period during which a panicked seller waits for things to “feel safe” and misses most of the rebound.
The three mistakes, and they’re all the same mistake#
Recency bias. Every recession feels uniquely fatal while you’re in it. 2020: a global pandemic, the economy physically shut. 2008: the banks, the financial system broken. 2000: the internet bubble exposed as fraud. Each story was genuinely different and genuinely true. The underlying pattern repeated almost identically anyway.
Waiting for good news. The confirming data arrives, by construction, after the best days have passed. This isn’t bad luck; it’s what “confirmation” means.
Confusing the economy with the market. Covered above, and the root of the other two.
All three are attempts to make a decision while frightened, using information that arrives late.
The actual solution: decide now#
The people who come through recessions well didn’t call the bottom. Nobody calls the bottom — not professionals with research teams, not anyone. They decided their rules in advance and followed them mechanically.
Three decisions, settled today while you’re calm:
What happens to your contributions if the market drops hard? Continue unchanged is a valid answer. Increase by a set amount is a valid answer. “I’ll see how I feel” is not an answer — it’s a decision you’re deferring to your most frightened self.
What’s your emergency fund threshold, in months of expenses? This is what makes the first decision possible. Without it, a downturn plus a job loss forces a sale regardless of your intentions.
What’s the concrete trigger for rebalancing? A percentage drift or a calendar date. Something mechanical that doesn’t require a market view.
None of these require predicting anything. They require deciding once, in advance, instead of repeatedly under pressure.
Summary — and what to do about it#
The drop is temporary. Selling is what makes it permanent.
- Ignore the recession declaration. It arrives 5–21 months late and describes the past.
- Stop reading economic data as a market signal. One looks forward, the other looks back.
- Write down your contribution rule now — unchanged, or increased by a set amount, if the market falls hard.
- Set your emergency fund threshold in months. It’s what lets you hold when holding is difficult.
- Pick a mechanical rebalancing trigger. A drift percentage or a fixed date, decided in advance.
- Expect four years, not four months. 2009’s bottom took until early 2013 to recover fully.
- Assume “this time is different” will feel true. It has felt true every time, and the pattern repeated anyway.
You can’t predict the bottom and you don’t need to. You need a plan written by the calm version of you.
Sources & further reading#
- NBER Business Cycle Dating Committee announcements — how and when recessions are officially dated.
- Real-time prediction of the Great Recession and the Covid-19 recession (arXiv) — average NBER announcement lags of 6.3 to 11.7 months.
- Complete list of US recessions, 1854–2026 — 34 NBER-dated recessions and their durations.
- A Wealth of Common Sense, “Missing the Best & Worst Days” — why the best days cluster inside recessions and bear markets.
- Federal Reserve History, “The Great Recession” — the 2007–2009 decline and recovery timeline.
