A Hedgehog Commentary

When To Sell

1/14/98 - A reader recently emailed the Hedgehog that he had found that he would often sell a stock only to have it immediately rise 20%. He noticed that his efforts to trade weren't gaining him much over buying and holding. He asked us for advice on how to decide when was the best time to sell. The Hedgehog knows the feeling, a while back I bought Southwest Airlines, knowing it was a well run company with a dominant slice of the market. I became impatient though and sold after a small gain, but then as the travel market picked up, Southwest shares soared.

The easy answer to our reader's question is that there is no way to know when to sell, at least not from a short term point of view. Many stocks, particularly the more volatile, thinly traded ones, can easily bounce up and down 10% within a few days. Other factors that can move a stock in the short term, are analysts upgrade/downgrades, a favorable write-up in Barrons or online at The Motley Fool, or a good presentation by the CEO on CNBC's morning program. Unless you are well connected, you can't know about any of this in advance. More importantly, it's very hard to know whether a company is going to meet or beat analysts short term earnings expectations. If they can't guess right with all their inside access, then you probably can't either. All in all, it would take a miracle for you to sell at the absolute peak price.

However, we can give you some advice on knowing when to sell, it's actually rather simple. Use the same criteria you used to decide when to buy. If you've read our company evaluations before, you know that we usually make our decisions based on a comparison of the company's valuation with its earnings growth potential. First we compare PE's with similar companies, then we look at PEG (PE divided by long term growth). The lower these numbers are, the better; how low is good enough to buy will depend on the company and industry. A small high tech company with a short history should have a lower PEG than a proven, brand name company such as GE. If the numbers look good, we'll then start asking whether there are problems justifying the low multiples. Finally, we'll ask ourselves whether we believe in the estimated growth rate (the analysts are often wrong and often have a conflict of interest). If these things all check out, maybe we'll buy.

The same criteria should apply to selling, and it should be easier to apply because as you've owned the stock, you've (hopefully) become more familiar with the company. First, would you buy the stock today at today's price based on today's knowledge. If the answer's yes, obviously keep it. If it's no, then there are more questions to answer. The first thing to decide is your timeline; how long are you willing to hold the shares? The shorter your timeline, the more concerned you have to be with valuation. If the current valuation of the stocks seems high (using the criteria you used to decide to buy the stock) and your timeline is short, you may want to sell. If you have a longer time frame, you may want to keep a stock that seems highly valued, if the future looks bright enough to make the long term upside even higher. For example, I own Cisco, I'm not sure that I'd buy at today's prices, but because I believe that this sector has a lot of promise over the next few years and Cisco is the leader, I'm keeping the stock.

Folks like Warren Buffett work under an even longer time frame. Buffett believes in buying and rarely ever selling. In that case, you might be willing to wait out drops on a currently overvalued stock, knowing in the long run that it will eventually go back up and that efforts to time the market probably aren't going to work out anyhow.

Of course, if the future doesn't look bright, if you no longer believe in the company, or if it's starting to drop more than you can stand, then you had better get out. The disk drive industry, for example, will probably come back again someday, but who knows when or in what form, so I recently decided to give up on my investment in Applied Magnetics and sell. Both of these examples (Cisco and APM) are an example of another theory which probably has at least some merit: let your winners ride and cut your losers.

Also, depending on your time frame, you may want to set limits on how far you'll let the price fall before you sell. This may also be a good idea if you're like me, overly optimistic about your stocks and continually expecting a turnaround when it may not be justified.

Most of what I've said of course only applies if you utilize valuation, growth potential, and a fairly long timeline in picking stock, if you work on momentum, market timing and day trading, well, good luck. The Hedgehog has pretty much given up on day trading (see our previous commentary on this topic) and doesn't believe in the technical market timing approaches to determine when to buy and sell. There are a number of these systems being touted that look at moving averages, trends, market psychology, and other voodoo. If you've been on the internet newsgroups, you've probably seen them being hawked before. I'd do a lot of testing before I actually let one of these things control what I do with my cash.

Commercial: If you want to read more about these issues, consider Buffettology (number 3 on the Business Top 10, The Motley Fool Investment Guide, and (for small cap investors) Midas Investing. If you are interested in market timing, you may want to check out Advanced Cycle Trading. All of these are available in our online bookstore.