A Framework for Selling a Stock

Determining a good strategy for when to sell a stock is both important and difficult. In simple terms, your returns are going to come from two primary sources: 1) the reappraisal of an undervalued holding to its intrinsic value and 2) growth in intrinsic value. Many investors sell their holdings if the price appreciates to fair value. Others, like Buffett, Russo, Greenberg, etc. hold their stocks for the long-term and look for gains from growth in intrinsic value.

Both of these approaches work and have generated a great deal of wealth.

In September, I posted a 1999 interview with hedge fund manager Morris Mark. Mark began his career at First Manhattan, an investment advisory firm founded by Sandy Gottesman, a large shareholder ($2 billion) and board member of Berkshire Hathaway.

Here’s what Mark said about selling Coca-Cola.

“We sold it three years ago because the valuation it was trading at in relation to our anticipated rate of earnings growth on a near to long term basis seemed to be well discounted versus our rate of return objectives. Generally, we would sell a position when something bothers us, otherwise the sale is likely related to valuation.”

This strikes me as an immanently rational framework which I would tweak only slightly:

Sell a stock when, after due consideration of taxes and opportunity costs, its anticipated rate of total return is well discounted versus your rate of return objectives.

This framework works for all types of investing and allows for the fact that different investors have different hurdle rates. Of course, considerable judgment is involved in making this determination, but these are the type of issues that should have been considered carefully before making a purchase. If you can’t figure out how much intrinsic value will grow in the long-term, the stock should probably be sold when it is no longer undervalued.

I would love to hear your comments about when you sell a stock.


4 thoughts on “A Framework for Selling a Stock

  1. kungfu panda

    for a truly great company, sometimes it’ll give you pleasant surprise (successfully entered a new market, launched a new product, etc), then its intrinsic value may change quite a bit,i am just wondering, how should we discount that probability when we sell?? hummm

  2. Andrew Schneck

    My favorite description I’ve heard about selling was from Mohnish Pabrai..

    He talked about holding stocks until they reach their intrinsic value. When asked about what he does when he finds a stock that would triple his money when he held stocks that would only double his money (basically, higher expected return from a new investment vs. older holdings), his response was that he likes his wife better than a new mistress. In other words, he was saying to let your hand play out. He wouldn’t sell one he’s comfortable with for a new position just because they looked younger or sexier.

    Although I don’t agree with him (I’m looking for highest expected return & all for tweaking my portfolio), I like his description- i thought you might enjoy it.

  3. Hopton

    A company with a very strong competitive advantage is tough for me to sell, especially if I’m facing a large tax bill. I know, I know… more humility and don’t let the tax tail wag the investment dog.

    I like Klarman’s quote…something about picking your regrets (as there will always be some). If valuation get to stretched it feels like I’m staying too late at the party.

  4. Greg Speicher Post author

    kungfu panda, I think with a truly great company you need to give it plenty of room to grow. There are times, however, when even the rosiest scenario is well discounted in the price of the stock. For example, I believe this was the case when Coca-cola reached nose-bleed multiples; Buffett later suggested that it should have been sold.

    Please see the prior post Lessons from Buffett’s Decision not to Sell Coke: “I talked when I should have walked” http://gregspeicher.com/?p=841.


Leave a Reply

Your email address will not be published. Required fields are marked *