License or reprint this article


Boiled Shrimp and the Sunk-Cost Fallacy

By Andrew Feinberg
April 11, 2006


I made an investment mistake last night. I ate the shrimp.

5:10 PM

I'm going to tell you a story about boiled shrimp, but it’s really a story about bad investing. In my kitchen last night, the "sunk-cost fallacy" appeared, dragging its chains behind it like Marley’s ghost. I, a professional investor who should know better, did not respond well.

The sunk-cost fallacy is one of the pillars of behavioral finance, the only academic discipline that actually explains why investors do the irrational things they do. (Standard economics doesn’t comprehend irrationality. If you want to become a better investor, read some behavioral finance. A good starting place is Why Smart People Make Big Money Mistakes and How to Correct Them: Lessons from the New Science of Behavioral Economics, by Gary Belsky and Thomas Gilovich.)

Back to sunk cost. Last night, as I searched the fridge for the pound and a half of Cajun catfish I planned to cook, I found a one-pound bag of raw shrimp. Cost: $7.99. Now, I bought these shrimp earlier in the day in only the most literal sense. I hadn't planned to. They must have belonged to the woman behind me in the line at Fairway. But I paid for them and, alas, I took them home.

So now I’m pissed and, I assume, the other woman at Fairway is pissed. I had a brief thought of running an "I have your shrimp" ad on Craigslist but thought better of it. (The world is getting smaller, but it's not that small yet.) Given that I was very pressed for time, a rational response would have been to throw the shrimp in the garbage. But the rat brain part of me, the part that doesn’t understand behavioral finance, focused on the ridiculous waste of $8. Who did I think I was, Rockefeller? (Actually, what I thought I was was a guy who billed $250 an hour for his time and who had a tight deadline that night to finish his taxes. Cooking the catfish was already a stretch. Adding the shrimp to the party was culinary insanity.)

Well, rat brain put the shrimp in salted water, hung out until it boiled then turned down the heat for three minutes or so. Then I dumped the shrimp in cool water and peeled and deveined them.

So I spent roughly half an hour to salvage my $8 of sunk costs. So I valued my time at $16 per hour, a far cry from what I make. If you think I'm some Julia Child who feels most alive in the kitchen, you would be very wrong. I did not do this for fun. I did it because I had accidentally invested $8 in boiled shrimp futures, and I was damned if I was going to have to write off my investment. (For $20, I could have had a pound of shrimp delivered to my door.)

Now, back to your portfolio. When the sunk-cost fallacy meets loss aversion (a nearly universal human trait), some horrible things happen to investment returns. I know dozens of friends, clients and acquaintances who underperform the market largely because they can't part with their losers. If they buy a stock at $50 and it falls to $30, the last thing that they think about is selling. They focus on their sunk cost of $20 per share and all they want to do is get even. The fact that the stock may have plunged from $50 to $30 for a very good reason doesn't interest them. They can't sell because they have a large sunk cost and because they believe that, by not selling, they are somehow avoiding the loss.

Now, when this $50 stock reaches $5, they really, really can’t sell. Now it's even more painful than before. The sunk costs are greater. The wounds to the ego are greater. And besides, math illusion enters the picture. Almost everyone thinks it is easier for a $5 stock to double than it is for a $25 stock to do the same. (That’s actually not true -- low-priced stocks underperform every other kind of stock -- but people intuitively believe it anyway.)

But the reason economists call it the sunk-cost fallacy is because your future attempt to enhance the value of your portfolio should be independent of what has occurred in the past. It is perfectly irrelevant that JDS Uniphase (JDSU) once traded for $150. The only interesting question is whether the stock is a buy or sell here at $3.59 per share. (Focusing on the $150 price is called anchoring, and it is yet another bane of investors.)

Consider this noninvesting, nonshrimp example. You buy two nonrefundable movie tickets in advance. In between your purchase and the time of the show, you discuss the movie with 12 people you know well. Each tells you that you will loathe the movie, that it is the cinematic equivalent of Chinese water torture. Would you go see it anyway because you’ve already spent $24? A behavioral finance expert would say you shouldn’t. Why compound your financial loss by spending two or more hours of your life being unhappy? Why not do something more enjoyable instead? Why throw good money -- in this case your enjoyment of life, which is what money sometimes helps you buy -- after bad?

I’ll leave you with one other thought about losing stocks -- which I will explore in depth at another time. It's about the right kind of ego involvement. I would bet a lot of money that most good professional investors have a lot less ego invested in the fate of any particular stock than you do. That is, they're not emotionally invested in their mistakes. On the other hand, these above-average professionals have more ego involvement in the results of their overall portfolio than you do. And they understand that when your self-esteem is wrapped up in the fate of Lucent (LU), JDS Uniphase, Sun Microsystems (SUNW) or Cisco (CSCO), you're more likely to make the wrong move rather than the right one.

So, please, throw out those shrimp.

Positions: Long JDSU in selected client accounts.