Why you miss the 100x winners: it is your exits, not your entries

You bought the right token. It doubled inside two hours, and you had already sold the whole position at a thirty percent gain on the way up, or a stop took you out on a one-minute wobble. This lesson asks how much of a runner's gain a trader gives back through the way they sell, and what to change. The answer comes from our own paper book on Robinhood Chain, replayed trade by trade against stored one-minute prices, losers included.

The paper book

A paper book is a set of trades we log at real prices with no money behind them, so we can test a rule without paying for it. From 29 September to 2 October 2026 our Robinhood Chain Momentum lane logged 83 trades. We stored a price for each one every minute, so we can replay the same 83 entries under different selling rules and see what each rule would have returned. The replay reproduces the stored book to the cent, so these figures are proven against the prices rather than estimated.

What selling at a target did

The book's rule was to sell everything at +30% and to cut the trade at -25%. A stop is an order that sells when the price falls a set distance below where you bought.

Under that rule the lane returned -1.7% per trade. It lost money even though 54% of the trades reached the +30% target.

Two things did the damage. The stops did not fill where they were set: a Robinhood Chain token can drop twenty points inside a minute, so a stop set at -25% often sold well below it. And the +30% target sold the runner. The trade that would have paid for the losers before it was closed for a third of a gain.

What letting the runner run did

Same 83 trades, a different rule: sell half at +20%, and let the other half trail the price, selling when it falls 20% below its own high. A trailing stop follows the price up and does not move down.

That rule returned +19.9% per trade, with a median of +5.3% and 58% of trades ending as winners. Without the best three trades the same rule returned +0.5% per trade. Read that last figure twice. The typical trade under the better rule was close to break-even, and the profit came from a few runners the rule did not sell early.

The rule worked because of what the trades did after entry. Of the 83, 46% doubled inside two hours and 15.7% went 5x. The median two-hour peak was 1.87x. The +30% target was selling tokens that, more often than not, went on to double.

What the Trade Bot ladder does

A ladder sells in steps at set targets instead of all at once. In the PumpPill Trade Bot, the default ladder sells fifty-five percent of the position when the price doubles, then ten percent each at 3x, 5x, 8x and 10x, and leaves five percent to ride. Each target is measured from the price you paid. On Robinhood Chain the ladder arms the moment your buy lands. On Solana it arms on the buy when Auto Sell is on for your account, and a held Solana token shows an "Arm Auto-Sell ladder" button that arms it in one tap.

You can change any step. Open Orders in the bot to see each armed step, cancel one, or cancel the whole ladder. A target can be a percent gain, a multiple, or a market cap, and you can add a stop-loss. If the paper book persuades you, the change to make is to the first rung: take less at the first target and let a trailing rung handle the rest, rather than selling out at a fixed gain.

The five percent that rides

The last rung of the default ladder leaves five percent of the position on with no target. That rung exists so that a token which keeps running still has you in it. It also rides down.

Our logs say how often. Of 917 Solana calls that reached 10x and are now more than thirty days old, 420 (45.8%) were still above entry at our last look, and the median one sat at 3.7% of its peak. The runners mostly round-trip. A rung with no target is a small bet that this one will not, and it should stay small.

What this does not tell you

Eighty-three trades over four days is a small sample from one week on one chain, a week that had several large Robinhood Chain runners in it. The gap between the two rules will move with each week of data.

The replay used stored one-minute prices with a fixed cost per trade. Your fill will differ, and on a thin pool it will differ more.

The ladder measures from your entry. It cannot see where the peak will be. A trailing stop set twenty percent below the high will sell on a twenty percent dip whether the token is finished or pausing.

The paper book's entries came from our Momentum lane. Your entries will be your own, and a selling rule cannot rescue a token that does not move. On Solana, more than half of the calls we logged never got 10% above entry.

What to do next

Open @PumpPill_Trade_Bot, make a small buy on a token you have already scanned, and open Orders to read the ladder it armed. Change the first rung before the price moves, so the decision is made while you are calm. The screens are at /bot, and the paper book's method is described at /research.