The Anatomy of a Bubble: Five Stages That Keep Repeating
Key takeaway
- Bubbles follow a repeating five-stage sequence, and it starts with something genuinely new rather than with irrationality.
- The fuel is credit, not enthusiasm. Enthusiasm alone produces a rally; borrowed money is what turns a fall into a collapse.
- You cannot time the top. You can check how much of your position is borrowed and whether you could sit through a 60% fall — which is a different, answerable question.
Learning pathHold through a cycle: crowds, crashes and when to sellStep 3 of 8
Read before this:Second-Level Thinking: Why a Good Company Is Not a Good Buy
Based on Manias, Panics, and Crashes — Charles P. Kindleberger, 1978
The sequence, which is older than any market you have traded
Charles Kindleberger's Manias, Panics, and Crashes is an economic history rather than a trading book: tulips in 1637, the South Sea Company in 1720, railways, Florida land, 1929, and everything since. Its argument is that the episodes are not each unique. They share a structure, borrowed from the economist Hyman Minsky.
| Stage | What happens | What it feels like from inside |
|---|---|---|
| Displacement | Something real changes — a technology, a rate cut, a new market opens | Sensible. The early buyers are usually right |
| Boom | Prices rise; credit becomes available against the rising asset | Rewarding. Caution has been costing money for a while |
| Euphoria | New buyers arrive for the price move itself; valuation methods get replaced | Obvious. Doubters look like they do not understand it |
| Profit taking | Insiders and early holders sell into strength | A healthy pause. The first fall reads as a discount |
| Panic | Credit is withdrawn; forced sellers meet no bid | Fast. Much faster than the way up |
Credit is the mechanism, not the mood
Enthusiasm makes prices rise. Borrowed money is what makes the fall a collapse.
This is Kindleberger's most useful correction to the popular account. Manias are usually described in terms of psychology — greed, herding, fear of missing out — and psychology is present. But an unleveraged crowd that changes its mind produces a bear market. A leveraged crowd that changes its mind produces forced selling, and forced sellers do not have the option of waiting.
Why the top cannot be timed, stated precisely
This is worth being blunt about, because "I will get out before it breaks" is the plan almost everyone holds. It requires knowing when the pool of new buyers empties, which is not published anywhere. What history offers instead is the observation that the last stage of the rise is usually the steepest, so anyone waiting for the top gives back the most.
Four things you can check, unlike the top
- Is credit expanding against the asset itself? Margin debt, lending secured on the thing that is rising, buyers who need the price to keep rising to refinance. This is the single most informative of the four.
- Have the valuation measures changed? Not the numbers — the measures. When the standard metric is retired in favour of a new one that flatters the price, that is stage three arriving in writing.
- Are new buyers arriving for the move rather than the asset? The clearest version is people who cannot describe what they own but can describe how much it is up.
- Who is selling? Insiders and early holders distributing into strength is the fourth stage happening in public, in filings, before it happens in the price.
None of these is a timing signal, and treating them as one is how people spend three years short of something that keeps going. They are position-sizing information: reasons to hold less, borrow nothing, and know your own exit rule in advance.
What this changes about how you hold things
Trying to trade the mania
- Requires calling the top
- Needs leverage to be worth it
- One wrong week ends the account
Surviving the mania
- Size assuming a 60% fall
- Own no borrowed exposure to the story
- Write the exit before you need it
The honest use of this history is not prediction. It is calibration of the downside: knowing that assets in stage three routinely fall 60–80% from the peak, and asking whether your position survives that at its current size. If the answer requires the fall not to happen, that is the finding.
And the reason it keeps happening to new people is structural rather than moral — each cycle draws in a cohort that has not seen one. That is the same argument Mr. Market makes at the level of a single day.
Try this week
- For your largest holding, write what a 60% fall would do to your total account, and whether you would still hold it.
- Check whether any of your exposure is borrowed, directly or through a leveraged product.
- Name the metric you are valuing your most-loved position on, and check whether you were using the same one two years ago.
- Write your exit rule for that position now, while nothing is happening.
Common questions
What are the five stages of a bubble?
Displacement, boom, euphoria, profit-taking and panic — the sequence Hyman Minsky described and Charles Kindleberger applied to four centuries of financial history. It begins with a real change, not with irrational behaviour.
How can you tell if something is a bubble?
You cannot tell reliably, and anyone claiming otherwise is selling a timing service. What you can check is whether credit is expanding against the asset, whether the standard valuation measure has been replaced, and whether new buyers are arriving for the price move rather than the asset.
Why do bubbles need credit?
Because credit is what makes the unwinding forced. An unleveraged crowd changing its mind produces a bear market; leveraged holders face margin calls and have to sell into the fall regardless of what they believe.
How far do bubbles fall when they burst?
Historically, falls of 60% or more from the peak are common in the assets at the centre of a mania, and the recovery can take many years. The practical use of that number is to size positions as though it will happen, rather than to predict when.
Is 'this time is different' always wrong?
No — sometimes a genuine change is under way, which is precisely why the phrase persuades. What it reliably indicates is that the argument has moved from valuation to narrative, because the previous measures stopped supporting the price.