There is a token in my cold wallet that I would not buy today. I know this because I have asked the question directly, more than once, and the honest answer keeps coming back no. It is still there anyway. Every review cycle it finds a way to survive, and the reason changes each time, which is the tell. When a position keeps generating fresh justifications, the position itself is doing the arguing, and the thing arguing on its behalf is the plain fact that I already own it.
This has a name, the endowment effect, and it is one of the most replicated findings in behavioral economics. The classic demonstration is the mug experiment from Kahneman, Knetsch, and Thaler. Hand half a room coffee mugs at random, then ask the new owners the minimum they would sell for and ask everyone else the maximum they would pay. Same mug, same room, minutes of ownership. Owners typically demanded roughly twice what buyers offered. Nothing about the mug improved when it changed hands. Ownership alone moved the price.
The standard explanation is loss aversion. Giving up something you hold registers as a loss, and losses weigh heavier than equivalent gains, so your selling price inflates to compensate for pain that has nothing to do with the asset. There is a second layer that matters more for traders. Things you own get folded into your sense of who you are, and the longer and louder the ownership, the deeper the fold. Selling a mug barely touches identity. Selling a coin you have held through two drawdowns and defended in three group chats touches it a lot.
Why crypto is the worst case
In equities the effect is real but blunted. A stock has earnings, cash flow, guidance, a board you can be annoyed with. Your thesis has scheduled collisions with reality four times a year, and when the numbers stop supporting the story, the bad news at least reaches you in a standard format. Most tokens have no equivalent. The thesis is a narrative, and narratives are soft enough to bend around whatever you already hold.
Crypto then adds an accelerant that equities mostly lack. Buying a token frequently means joining something. There is a Discord, a feed, a vocabulary, sometimes the ticker sits in your bio. Selling closes a position and it also walks you out of a group, and the second part hurts in a way that has nothing to do with price. The culture even names this openly. Holding through anything is diamond hands and gets celebrated. Selling gets framed as betrayal or weak conviction. Nobody calls you a traitor for trimming an index fund.
The mechanics deepen it further. Staking and lockups mean you have done work to keep holding, and yield paid in the same token grows the position without a single fresh buy decision ever being made. Airdropped coins arrive feeling like house money and often never get underwritten at all. And your information diet is curated by your holdings, because you followed the accounts and joined the channels after you bought. The confirmation loop builds itself, quietly, in exactly one direction.
The would-you-buy-it-today test
The counter I use is old and I did not invent it. For every position, on a fixed schedule, ask one question. If my entire portfolio were cash this morning, would I buy this asset, at this price, at this size? What the question does mechanically is convert a selling decision into a buying decision. The endowment effect lives in the gap between what owners demand and what buyers will pay, so you force yourself to answer from the buyer's side of that gap, where your judgment is less contaminated by ownership.
Three answers are possible. Yes at this size means hold and move on. Yes but smaller means trim to whatever size you would genuinely initiate today, which is the answer I get most often on old winners. No means exit, and if you find that you cannot, write down the real reason the position is staying and date it. Reading last cycle's excuse next to this cycle's excuse is uncomfortable in a productive way.
A few details decide whether this works or turns into theater:
- Put the cadence on the calendar. Monthly if you trade actively, quarterly if you do not. Running the test only during drawdowns guarantees you run it at your most emotional and least honest.
- Hide your cost basis while you answer. It is the strongest contaminant in the room. The plan to sell when it gets back to even is the endowment effect wearing a breakeven costume, and the market does not know your entry price.
- Answer in writing, one or two sentences of thesis per holding. If the thesis names a person or a community instead of a mechanism, flag it and look harder next cycle.
- Ask at full size. Would I buy a little is an escape hatch. The question is whether you would buy the exact dollar amount you hold right now.
Where the test breaks down
The failure mode I see most is running it only on losers. Winners need it more. A position that has tripled has quietly grown into a portfolio weight nobody ever chose on purpose, and it carries the most identity, because it is the trade you tell people about. The second failure is letting tax become a permanent veto. Taxes are a legitimate input to timing, but if tax deferral is the only surviving reason for three consecutive cycles, the thesis is dead and the tax argument is only deciding when you admit it.
The third is outsourcing the answer. If your first move is to check what the community thinks before you can say yes or no, the identity half of the bias is answering for you. The test works solo, from your own notes, against something harder than sentiment. I keep it mechanical, every position in one flat list, cost basis hidden, thesis next to current evidence. Blockcircle's portfolio scorecards happen to give me that view in one place, but a spreadsheet does the same job if you fill it in honestly.
The token from the first paragraph is on its third consecutive no, and the exit is written into next review, sized and dated, because I have learned the answer and the action can sit apart for a long time if you let them. The test does one thing well. It keeps the gap between what I believe and what I hold on paper where I can see it, and gaps I can see tend to close, even if a quarter or two later than they should.