Every investing app looks roughly the same. An order field, a buy button, a sell button, and a number: the price of the last trade.
The number looks definitive. It has decimal places. It updates while you watch. And because it sits next to your holdings, it quietly becomes the thing you believe you own.
But nobody has promised to buy from you at that number. It says one thing only: the last time two people agreed, they agreed here.
Whether they will agree there again is a separate question, and it is never asked until the moment you need the answer.
Seven times the number stopped meaning anything
Tulip mania. In the Dutch Republic, a single rare bulb traded for the price of a canal house. When the market broke in February 1637, prices collapsed within weeks. It is worth noting that modern historians have found far fewer ruined fortunes than the legend claims — mostly, contracts were simply never honoured. Which is its own version of the point: the prices were real on paper and unreal in settlement.
Terra–Luna, May 2022. A stablecoin designed to hold one dollar stopped holding it, and the token backing it fell to effectively zero in under a week. Roughly $40 billion evaporated inside that one ecosystem. The mechanism that was supposed to restore the peg worked exactly as designed — and the design was the problem.
Celsius, June 2022. More than a million users woke up to a notice saying withdrawals were paused. Their balances were still visible. They were simply no longer reachable. The wealth was on the screen; the key was somewhere else.
Three Arrows Capital, July 2022. A leveraged fund failed, and the lenders who had financed it failed in sequence behind it. Creditors are still working through the estate years later. Nothing about the collapse was visible in the reported numbers until it was complete.
FTX, November 2022. Customer balances turned into an error message. Around $8 billion in client funds was not where the statements said it was. For years, those statements had been accurate in every way except the one that mattered.
Bitconnect, January 2018. A scheme promising implausible returns shut its lending platform and a market value of roughly $2.5 billion disappeared in a day. The returns had been reported monthly. They had never existed.
Turkey, September 2026. The regulator placed 131 funds run by seven management companies into liquidation. More than 450,000 investors found their money locked. The liquidation window was first set at three months, then extended to six. Those investors will not learn what their holdings were worth until the process ends.
They are not the same story — and that is the point
It is tempting to file all seven under "bubble" and move on. That would be a mistake, because they belong to three different families.
Two of them were fraud: money reported as present that was not present.
Three of them were leverage and design failure: money that genuinely existed but had been promised twice, or was held together by a mechanism that only worked while nobody tested it.
Two of them were liquidity mismatch: money that was really there, in assets that were really owned, at prices nobody could actually transact at in size.
Three different diseases. Very different moral stories. Very different court cases.
And here is the uncomfortable part:
Until the final week, all three look identical on the screen.
A fraudulent balance, an over-levered balance and an illiquid balance all display as a number with two decimal places, updated daily, rising pleasantly. The screen does not have a field for which kind of number this is. You cannot tell the families apart by looking at your account. You can only tell them apart by understanding how the number is produced.
That is why "does it look like it's going up" is not a risk assessment.
Liquidity is a property of the exit, not of the asset
When we call something liquid, we may mean either of two things:
- This asset can be bought and sold.
- I can leave, at a time I choose, in the size I hold, at something close to the number on the screen.
The first is true of almost everything. The second is rare — and the second is the one that matters.
Liquidity is not a characteristic of the asset. It is a characteristic of the exit. And the exit is not a fixed width: it narrows precisely when it is needed. The moment everyone wants out is, by definition, the moment nobody wants in.
This asymmetry is not a flaw in the system. It is the system. If the way in were as wide as the way out, panic would not exist as a phenomenon.
If you set the price, your return is a bookkeeping entry
Now the more uncomfortable part.
If very little of an asset actually circulates — if the freely tradable portion is small — then it does not take much money to set its price. Modest buying moves it a long way.
So picture a large pool holding a large position in exactly that kind of asset. The pool's value is calculated from the last traded price. And the party that made the last trade is, more often than not, the pool itself.
A loop forms:
Buy → price rises → the pool's stated value rises → performance looks excellent → new money arrives → the new money buys → price rises again.
Every individual step in that loop can be legal, documented and technically correct. But what the loop produces is not a return.
A price you created with your own buying is not performance. It is a bookkeeping entry.
And a bookkeeping entry is never tested until someone tries to convert it into money.
Test day
The test arrives when somebody asks for their money back.
The pool now has to sell. But the thing holding that price up was its own buying — so selling runs the same machine in reverse. The price falls. The pool's stated value falls. The remaining investors see it and also ask to leave. More selling is required. The price falls further.
This is a redemption spiral, and once it starts there is exactly one way to stop it: close the door.
When the door closes, the truth appears. The investor was never the owner of the number they had been watching for months. They were someone who will find out, at the end of a liquidation, what that number was.
Four things this says to an individual investor
1 · A return and the ability to realise it are different things. "It returned X% last year" carries a hidden clause: and you could have taken the X%. That clause is usually true. On the rare occasions it is false, everything else is false at the same time.
2 · Exceptional returns come from somewhere. When you see performance that stands out, the question is not "how did it make so much." It is "at whose expense was this made?" If there is no answer, the answer is usually nobody's yet — which means it has not happened.
3 · Diversification is counted in categories, not in line items. You can hold five different pools. If all five own the same kind of asset, bought the same way, exiting through the same narrow door, you do not hold five positions. You hold one. It will jam in one piece.
4 · Regulated does not mean safe. Regulation guarantees that rules exist, not that outcomes will be good. Institutions typically arrive after the loss, because their job is not to prevent the loss but to decide how it gets shared.
And one more thing: "I was already out"
It is the most common sentence after any collapse, and it invites the most dangerous lesson.
Leaving in time can be skill. More often it is a fact about where you were standing in the queue. When a door narrows, who got through tells you less than who did not.
So the lesson to draw from I got out is not "clearly I make good decisions." It is "clearly I never tested the door."
The honest question is this: if I had still been inside that day, would I have got out — or was my turn simply earlier?
In short
- The number on the screen is not a commitment. It is the record of the last trade.
- Fraud, leverage and illiquidity are three different diseases that look identical on a screen. You cannot tell them apart by watching; only by understanding how the number is made.
- Liquidity belongs to the exit, not to the asset — and the exit narrows exactly when it is needed.
- If your own buying sets the price, what you have produced is a bookkeeping entry, not a return.
- Diversification is measured in categories that do not jam together, not in the number of holdings.
- Getting out in time may be a fact about the queue rather than about you.
Try this
For every holding you own, write down the answer to one question:
"If I decided to sell this, how many days until the money is in my account — and how much longer does that take if everyone decides to sell on the same day?"
Anything where you cannot answer the second half is riskier for you than the screen suggests. This is not a signal to sell. It is a fact about how much room that holding should be allowed to occupy.
And the best time to ask the question is while you are still uninterested in the answer.
Figures are as reported at the time of writing and are given for illustration, not as a record of any particular case. I hold none of the instruments mentioned.
0 Comments