You held it long after the reason to hold it died.

Conviction is about the token. Refusal is about you.

Both produce the same behaviour, which is why conviction is the word that turns up when you need cover. Say what would have to happen for you to close this. If the answer describes the token, its liquidity, its volume, its deployer, that is a thesis. If the answer describes a price, back to entry, back to green, that is your cost basis talking. The token does not know what you paid.

The tell is that the reason keeps changing while the position stays the same. You bought for a 3x on a launch. At minus 30% you are holding for a bounce. At minus 70% it is basically free at this point. At minus 95% it is a lottery ticket you never decided to buy.

What the thesis dying actually looks like

Memecoin theses are thin on purpose. You are holding two things: liquidity deep enough to get out of, and attention strong enough to produce the next buyer. Both die fast, both die visibly, and neither requires you to predict where price goes next.

• Liquidity is pulled, or the deployer sells into the book. Whatever the chart does after that, what you bought no longer exists in the form you bought it.
• Volume decays and stays decayed. Not a quiet twenty minutes. Hours where the only flow is exits.
• Attention rotated. The callers, the chat, the timeline are on a different ticker. Memecoins run on who is looking, and you can see when nobody is.
• Your catalyst resolved. The migration finished, the listing landed, the account posted. Once the event is behind the chart, so is the reason.

The drawdown is the cheap part

Start with arithmetic, which nobody argues with. Down 40% needs a 67% gain to get back to flat. Down 80% needs 400%. Down 95% needs 1,900%. What you are waiting for is not a recovery, it is a bigger run than the one you already missed.

Then add the cost your PnL never shows you. Two SOL parked for eleven days in a chart with no volume is eleven days of two SOL doing nothing, on top of the drawdown. That never appears as a red trade. It appears as an account smaller than your win rate suggests.

Not every red position is a bag

Duration on its own proves nothing. Sitting through a 60% drawdown on a token that still trades and still has the thing you bought is a trade going badly, not this.

Two things separate them. Whether you could state the invalidation before you entered, because one invented at minus 60% is a rationalisation with better grammar. And whether you were still making a decision at all, because bag holding is the absence of one. If you cannot point to a moment you re-evaluated the position, you did not hold it, you just failed to sell it.

What changes it

The difference sits in where the exit condition is written and when. An invalidation set before entry is stated in the token's terms: volume dying for a stretch, the deployer wallet moving, liquidity changing shape. Those are things you can watch happen on a chart. Getting back to entry is not an observable condition. It is the cost basis again, restated as a plan. Size works on the same problem from the other end, because a position large enough to change how the red number feels turns closing into something you avoid rather than something you decide.

The other thing that moves is looking at the position as if you did not own it. Market cap now, volume now, and whether this is something you would open today with fresh SOL. That framing strips out the one input that has been keeping you in, which is what you paid.

Dossier does not warn you while it is happening. It reads the swap after you close it, rebuilds the chart from 15 minutes before entry to 60 minutes after exit, grades the trade A-F, and names the mistake when this is the one. On that chart, the last stretch of real volume and the moment you sold sit on the same timeline. A gap with timestamps on it is different from a vague sense that you held too long.