When to Sell a Stock: A Complete Framework

0 Shares

Updated: August 2, 2026 | By Jenna Lofton, StockHitter.com

Jenna’s Bottom Line

Let’s play a quick game. You have two stocks. One is up. One is down. You need cash, so you have to sell something.

If your gut says “sell the winner, keep the loser, it’ll come back,” congratulations, you have fallen for one of the most well-documented, studied, and expensive biases in all of investing. You’re in extremely good company. That company just underperforms by 3.4 percent a year.

Key Takeaways

A landmark 1998 study analyzing 162,948 real trades found that winning stocks investors sold went on to outperform the losing stocks they kept by 3.4 percentage points over the following year. You are, statistically, selling the wrong one.
The “disposition effect” describes our documented tendency to sell winners too early to lock in a good feeling, and hold losers too long hoping to avoid admitting we were wrong.
The right sell decision comes down to two honest questions: is the original thesis still true, and would you buy this stock today at this price knowing everything you now know?
The “endowment effect” makes it worse. Owning a stock makes you irrationally attached to it, requiring a higher price to sell than you’d ever pay to buy the identical stock fresh.
None of the real reasons to sell involve staring at the price chart. They involve the business itself changing in ways that matter.

Table of Contents

The Bias That Costs You 3.4 Percent a Year, Give or Take

In 1985, two researchers named Hersh Shefrin and Meir Statman gave this behavior a name: the disposition effect. The tendency to sell winning positions far more readily than losing ones, even when the winner is the better business and the loser is the one whose thesis actually fell apart.

For thirteen years that was a clever academic observation with no real number attached. Then in 1998, researcher Terrance Odean got his hands on the trading records of 10,000 real brokerage accounts from 1987 to 1993.

Across 162,948 actual trades, Odean’s published research found that retail investors sold their winning stocks at roughly 1.5 times the rate they sold their losers. Then he checked what happened next. The winners they sold went on to outperform the losers they kept by 3.4 percentage points over the following twelve months.

Read that again slowly, because it’s the funniest and most expensive sentence in behavioral finance. Investors weren’t just selling their winners early. They were reliably selling the better stock and keeping the worse one, on purpose, over and over, for years.

Why Your Brain Does This to You

The disposition effect runs on loss aversion. Selling a loser means admitting, in a very final and paperwork-official way, that you were wrong. Selling a winner just means you get to feel like a genius.

There’s a second bias making this worse, and it’s called the endowment effect. Simply owning something makes you value it more than you would if you were considering buying it fresh. Applied to your portfolio, this means you’ll happily hold a stock you’d never actually purchase today at today’s price, purely because you already own it and selling feels like a breakup.

Put these two biases together and you get an investor who sells good businesses too early to feel good, and holds bad businesses too long to avoid feeling bad. Neither decision has anything to do with what the business is actually worth. Both decisions are about managing your own feelings, which is an expensive hobby to fund with your retirement account.

The Two Questions That Replace a Thousand Feelings

Here’s the good news. You can shortcut past almost all of this bias with two honest questions, asked every time you’re deciding whether to hold or sell.

Question one: is the original thesis still intact? Not “has the price gone down, which feels bad.” The actual business reason you bought it in the first place. Is it still true?
Question two: would you buy this stock today, at today’s price, knowing everything you currently know? Not “would you buy it back at your original cost basis.” Today’s actual price. Today’s actual facts.

Two yeses means hold, possibly add. One no means sell, regardless of whether the stock happens to be up or down from where you bought it. Your cost basis is not a business fact. It’s a number your brokerage remembers so you don’t have to feel it every time you check the app.

Five Signs the Thesis Actually Broke

The two-question test works best when you already know what a genuinely broken thesis looks like, rather than confusing it with a stock price having a bad month.

Growth decelerating across multiple quarters, not one messy quarter with an obvious one-time explanation.
The competitive moat genuinely eroding, meaning a real competitor is taking real market share, not just a scary headline about a competitor existing.
Management credibility damaged through missed guidance, walked-back promises, or communication that stops matching what the business is actually doing.
The balance sheet materially weaker, with debt growing faster than earnings and no clear plan to reverse that.
The original reason you bought it no longer exists, which sounds obvious until you actually try to write down why you bought a stock and realize you can’t remember.

Notice what’s missing from that list: the stock price. None of these five signs require you to look at a chart. They require you to look at the business, which is the entire point.

What This Looks Like With AI Infrastructure Stocks

This framework gets tested constantly with volatile growth names, which is exactly where the disposition effect does the most damage.

Palantir (PLTR) has experienced multiple corrections of 20 percent or more within its broader uptrend. Investors applying the two-question test during those drops would have asked: is U.S. commercial revenue growth still accelerating? Is the platform still winning new government and enterprise contracts? If yes and yes, the price drop is noise, not a signal.

Contrast that with a stock where the actual growth rate has been decelerating for several quarters while the price also happens to be falling. That’s not the same situation wearing a different stock ticker. That’s the thesis genuinely breaking, and the price decline is the market correctly noticing it before you did.

Experience Transparency

I held a regional bank stock down 35 percent from my cost basis for nearly two years, telling myself I’d sell once it “got back to even.” That sentence should have been a five-alarm fire, because the stock does not know or care what I paid for it.

The actual thesis had broken well before I admitted it. Loan quality was deteriorating, not temporarily, structurally, and I was holding on purely to avoid the paperwork feeling of a realized loss.

I finally ran the two-question test honestly, got two clean no’s, and sold at a worse price than I could have gotten six months earlier. The lesson stuck permanently. My cost basis is not a business fact. It never was. It’s just a number I got emotionally attached to for no defensible reason.

When Selling a Winner Actually Makes Sense

The disposition effect gets so much attention that it’s worth being fair to the other side. Sometimes selling a winner is exactly right, and it has nothing to do with the disposition effect.

Position sizing is the most common legitimate reason. A stock that’s grown from a modest starting size to a large chunk of your portfolio, purely through price appreciation, can warrant trimming regardless of how the business is doing. For our full framework on this, see our guide to position sizing.

Valuation catching up to the thesis is the second. If a stock has run so far that the current price already assumes flawless execution for the next five years, trimming some exposure isn’t a bet against the business. It’s an honest acknowledgment that the easy money has already been made.

The key distinction: trimming because the math genuinely changed is a decision. Selling because a green number makes you feel like locking in a win is a feeling wearing a decision’s clothes.

Wall Street Reality Check

Financial media loves stories about the stock that “should have been sold” after a big run, usually written well after the crash with the benefit of perfect hindsight. What gets covered far less is the much larger and quieter cost of investors holding broken theses for years out of pure stubbornness.

There’s no headline for “investor holds mediocre regional bank stock for two years hoping to break even.” There should be. It happens constantly, it’s boring, and it’s arguably a bigger drag on real people’s retirement accounts than any single bad stock pick that made the news.

Bottom Line

The decision to sell a stock should never involve the price you paid for it. That number is for your taxes, not your judgment.

Ask whether the original thesis is still true, and whether you’d buy the stock today at today’s price knowing what you know now. Two yeses, hold. One no, sell.

The bias runs the opposite direction of good judgment almost every single time, which is exactly why a rule beats a feeling.