Shohei Index
Shohei Index ResearchUpdated 2026-08-30

When to sell, and the five signals that say look

The exit side: what a signal is allowed to claim, the two triggers nobody writes about, and why a prompt with no evidence behind it is never emitted.

Key finding

Every signal here is a prompt carrying the fact that triggered it — never an instruction, never a forecast, and never emitted at all when there is no evidence behind it.

Most card advice stops at the buy. But a profit only exists once you sell, and “hold forever” is not a strategy — it is the absence of one. In a prospect market there is at least a scheduled moment to think about: the debut converts an open question into a settled one, and the price is built around that conversion. Here there is no such moment. The career question is answered, and what is left is the market around a specific card and the rules you set for yourself.

So the review triggers this app emits are about those two things, and nothing else. Two are market facts, three are position facts, and every one of them is informational: an observed number, never a prediction.

The five review triggers

SignalWhat it says
Above the trailing rangeThe current estimate sits materially above everything in the card's recent confirmed sold range. A statement about where the price sits today — not a claim about where it goes next.
Unusually strong liquidityThis exact card is trading often right now, so an exit would be cheap to execute. Liquidity windows close; the trigger only reports that one is open.
Duplicate copies heldYou hold more than one copy of the same exact identity. Duplicates add exposure without adding collection coverage — a deliberate keep-or-trim decision beats drift.
Your target price reachedYou set a number; the current estimate has met it. Your rule, not ours — the trigger exists so the decision you already made actually gets made.
Bigger position than you meantNothing to do with the card. If one holding has grown past the share of your money you said you were willing to put on any single position, that is a decision you already made and are now not keeping.

Those last two are the ones nobody writes about, and they are the two that fire quietly. A target is easiest to ignore on the day it is hit, and a position only becomes oversized by going well — which is exactly when it is hardest to trim.

A signal is a prompt, not an instruction

Every one of these carries the fact that triggered it — the sold range and where the estimate sits against it, the sale count in the window, the number of copies you hold, the percentage of your portfolio — and each is ranked now, soon or watch. None of them sells anything, and none of them tells you what you must do.

The rule underneath that matters more than the ranking: a signal with no evidence behind it is never emitted at all. There is no “consider taking profits” with nothing behind it, because a prompt you cannot check is just an opinion wearing a badge — and a tool that produces those trains you to ignore the ones that mean something.

Why the duplicate trigger earns its place here

In a market with thousands of players, holding two of the same card is unusual enough to be a deliberate act. In a one-player market it is the default failure: everything on the shelf is the same person, the identities differ by a card number and a photograph, and a collection assembled a listing at a time drifts into three copies of one identity and none of the four others around it. Duplicates are exposure without coverage, and the trigger exists to make that visible rather than to condemn it.

The two things that are not exit signals

  • A bad week. A slump is a fact about baseball and a card is priced on a market. If the sold range has not moved, nothing has happened to your position.
  • A price you do not like. “I'll sell when it gets back to what I paid” is a statement about your cost basis, which the market has never heard of.

How we apply this

Every prompt carries the fact that triggered it: a sold range, a sale count, a copy count, a percentage of a portfolio with the after-fee figure attached. A signal with no evidence behind it is not emitted. And nothing here models a retirement clock or any other future event — the timing question that a prospect market answers with a call-up date has no equivalent answer here, and we would rather leave it unanswered than invent one.

What an exit actually pays you is a separate calculation, and it is the one people skip: see what you actually clear when you sell, or run your own numbers in the net proceeds calculator.

Basis & limits

What this is built on. The rules and figures this project's own identity and valuation engines enforce, plus the domain research behind them.

Where it stops. This is an explainer, not a study: it carries no sample size and makes no forecast. Figures that move in the real world — grading fees, print runs, marketplace behaviour — can date it; the updated line above marks the last material revision.

Methods are documented on the methodology page; sources and their limits on trust & data sources.

More on doing the deal

All doing the deal guides →

Keep reading

See this applied to real cards: every card page shows live listings, sold evidence and what is verified about that exact identity, and the market board ranks what is mispriced right now.