There’s a sentence I hear more than any other, and it always arrives sounding like discipline.
I’ll sell it when it gets back to what I paid for it.
It gets said calmly, usually by someone who is perfectly sensible about money in every other room of their life. They have a budget. They know what their fees are. And they are holding something they no longer believe in, waiting for it to reach a number that means nothing to anyone but them.
Because that’s what a purchase price is. It isn’t a fact about the investment. It’s a fact about you, a record of one afternoon in your own history, the day you happened to click the button. The company doesn’t know it. The market has never heard of it and wouldn’t care if it had. It’s a number that exists in one place, which is your head, and it’s doing most of the steering.
The number that isn’t real to anyone else
What you paid for something feels like data. To the market, it isn’t data at all. It’s a memory, and memories make poor inputs for decisions about tomorrow.
What ten thousand accounts showed
In 1998 a finance professor named Terrance Odean got hold of something people had been arguing about without data. Ten thousand accounts at a discount brokerage, seven years of trading records, every buy and sell.
He measured a simple thing. When people had a stock sitting at a gain, how often did they sell it? When they had one sitting at a loss, how often did they sell that?
They sold the winners at about one and a half times the rate they sold the losers. Not occasionally. Across seven years and tens of thousands of trades, steadily, as if following a rule nobody had written down.
Then he did the part that turns an oddity into a problem. He followed what happened next. Over the following year, the winners those investors sold went on to outperform the losers they hung onto by about 3.4 percentage points, measured against the market.
So they weren’t just picking arbitrarily. They were reliably selling the better thing and keeping the worse one, and paying for the privilege.
Hersh Shefrin and Meir Statman had named this behavior back in 1985, with a title that says the whole thing out loud: the disposition to sell winners too early and ride losers too long.
Why the loser gets to stay
The last edition of this newsletter was about how a decision you never make can be the most expensive one you’ll ever not make. (If you missed it: The Regret You Can’t Feel Yet.) This is the same machinery with a position on.
Sell a winner and you book a win. There’s a receipt. You were right, in writing, and the feeling is available immediately.
Sell a loser and you convert something bearable into something permanent. Up until the moment you sell, it isn’t really a loss. It’s a position. It’s a thing that might come back. The paper number goes up and down and none of it is final, and there’s a lot of comfort in that. Hit sell and the loss stops being weather and becomes a decision, yours, dated, with your name on it.
So the loser gets to stay. And we don’t call it flinching. We call it patience, or conviction, or being long term, and every one of those words is a genuine virtue that happens to be doing a second job here.
“Sell a winner and you book being right. Sell a loser and you make the mistake permanent. Holding isn’t always patience.”
But isn’t this just taxes
This is the smart objection, and it deserves a real answer, because in a taxable account there is a genuine tax logic to realizing losses and deferring gains. If that’s what people were doing, it would be sound rather than psychological.
Two things say otherwise.
The first is December. In Odean’s data the pattern doesn’t fade at year end. It flips. In December, losses got realized at a higher rate than gains, the only month that happened. Which means the ability was there the whole time. For eleven months people could not bring themselves to close a losing position, and in the one month the tax code hands them a reason, they suddenly could.
The second is Taiwan. Capital gains there weren’t taxed, realized or not. There was no tax reason to prefer one sale over another. Researchers went through five years of trades on the Taiwan exchange, over a billion of them, from something like four million traders. The pattern was still there, and roughly 84 percent of investors showed it.
No tax to manage, same behavior. Whatever ledger people are settling, it isn’t the government’s.
I should be straight about one thing. That this happens is about as well established as anything in behavioral finance, turning up in Finland, in Taiwan, in mutual fund managers, even in how people price their houses. Why it happens is still being argued over, and some recent work complicates the tidy version I’ve just given you. The behavior is solid. The story underneath it is a live question.
The question that deletes the number
There’s one question that gets people out, and it works by making the purchase price irrelevant rather than by arguing with it.
Would I buy this today, at today’s price, knowing what I know now?
That’s it. Notice it doesn’t ask what you paid, whether you’re up or down, or whether you were right. It asks about the only thing that can still be changed, which is what you own tomorrow morning.
If the answer is yes, then you’re holding it for a reason, and the reason has nothing to do with getting back to even. Good.
If the answer is no, then holding isn’t patience. It’s a purchase you’re quietly making again every morning, at today’s price, without ever deciding to.
I’m not telling anyone to sell anything. What to hold is a real question with real numbers behind it, and there are tax consequences to any of it that belong with a tax professional and someone who can see your whole picture. The point is narrower. Make sure the number steering the decision is a number that exists outside your own head.
The 7 Behavior Gaps
We put the money mistakes that quietly cost us the most into a short, free field guide: the seven gaps, with a one-page fix for each. No cost, sent straight to your inbox. Get the free guide →
Educational only. No obligation.
What it’s actually for
The thing I keep coming back to is that none of this is a discipline problem. Everybody in Odean’s data was trying. They weren’t lazy or reckless. They were protecting something, and it wasn’t their money.
It was the story where they were right. Selling the winner confirms it. Selling the loser edits it. So the portfolio slowly becomes a record of decisions we were willing to admit to, which is a strange thing to retire on.
Your purchase price is a souvenir. It’s fine to keep one. Just don’t let it vote.
Go deeper
Weighing when to put in your papers with the NYPD? That decision runs on the same reversibility question this piece is about. Run your own tier, VSF, and 457(b) numbers in about two minutes with the free NYPD calculator →
Stay Informed
Get analysis like this delivered to your inbox: the behavioral research behind better money decisions, in plain English.
No spam. Unsubscribe anytime.
Sirmium Capital | Fiduciary wealth management for 9/11 families, first responders, and veterans.
Disclaimer: This content is for educational purposes only and does not constitute personalized investment, legal, or tax advice. Sirmium Capital is a registered investment adviser.