This has always struck me as an easy question, so the amount of handwringing over it has always left me nonplused. To be fair, many of the top value investors, people I admire, have said it is a tough question, so perhaps everything I write in this post will be wrong. But, judge for youself.
It’s my impression that people use the wrong mindset when deciding to sell. They seem to drop their value investing roots and start to try to guess what the market is going to do in the short term. If you try to do that, then yes, it’s hard to decide when to sell. It’s my opinion, backed by a lot of evidence, that no one can determine short term market movements. If so, if you try to do it all you are doing is speculating. It’s no more correct to speculate on the sell side than on the buy side.
Okay, so how do you decide to sell? Well, the most important thing to understand is that money has no memory. For example, say you buy a stock, some bad news comes out, and it drops 20%. Let’s say it is really bad news - a fundamental shift in the outlook for the company. In this case it makes no sense to just hang on, hoping to eventually make up the 20%. You sell, take the loss, and put the money into a better company. Your $800 doesn’t know it used to be $1000, and that company is not going to increase in value just because you invested in it and took a paper loss. Always, always, always put your money in an investment with the best rate of return.
So, we use that simple rule to decide when to sell. The rest of this argument is going to be phrased as if you are buying another stock, however, the argument still works if you are going to take the money and put it in a sock drawer or buy a boat with it. Why? Because money has no memory. It doesn’t know if it was used to buy a new stock, or a new boat, or whatever.
Okay, since money has no memory, don’t bother looking at your account and saying “I have 10,000 shares of XYZ”. You do have them, but that knowledge doesn’t affect the sell decision. You certainly don’t say “I bought 10,000 shares of XYZ at $10.00″. That also doesn’t affect the sell decision.
Pretend you have $100,000 to invest. In the entire universe of stocks, what stock or stocks would you buy? The very best stock you know of.
How you make that decision depends on your version of investing, and I won’t go into that here. But obviously you will be looking at FCF, multiple vs yield, your expectations of the future, and risk/reward.
Let’s say ABC is selling at $5 and you think it is worth $10. XYZ is selling at $10 and you think it is worth $13. Both have equal risks and uncertainty. In this case, forgetting for the moment that you own XYZ, you’d definitely buy ABC. Well, how do you get that 100K to invest in ABC. Well look! You have $100K worth of XYZ in your account. Sell it and buy ABC.
That’s extremely straightforward. Just own whatever you think are the best bargains out there. If that means you have to sell something, so be it. Every time you sit down to review your portfolio, figure out what you want it to look like, forgetting what you own now and the paper profit/loss on those. Then, do whatever you need to do (buy/sell) to make your portfolio look like it should.
A few complications: The US government taxes you differently based on your holding period. 15% vs 30% or whatever your effective tax rate is. There are dividends, currently taxed at 15% unless you have an MLP or somesuch. Etc. Natually, you have to factor those into the equation, but they are pretty simple. In this case you could argue the problem gets a lot harder, because you are deciding between selling today at a known price vs next year at an unknown price. Here I pretty much assume no changes in prices, which is a bit conservative given that over time markets rise. You could look at beta, but I think that’s a bit of voodoo.
Another issue related to taxes. Say you bought XYZ, and it dropped 20%. You love this stock, and the price drop was irrational. Normally you’d back up the truck, right? Well, let’s say ABC is equally depressed, and you love ABC just as much as XYZ. In this case, sell XYZ, take the tax write off, and buy ABC. You have the same dollar amount in stock as before, the future is equally bright in ABC vs XYZ, but now you have a juicy tax write off!
The final thing I might think about is delaying the tax bill. If I am figuring out what to sell and it is Nov 30th, I may just wait until the next fiscal year to sell my stock, merely to push the tax burden out a year. But this only works if you aren’t filing quarterly. If you are managing serious money, you are probably filing quarterly.
This post implies a lot of selling/buying. I hope that isn’t the takeaway point. If you bought Berkshire in the 80s and love it, why sell for a few extra percentage points? You know Buffett, presumably you agree with his investment and business philosophies, why go to somebody untried? But that should be part of the decision of “what do I want my portfolio to look like”? Me, I want some solid, indisputable performers like Berkshire, and then some higher risk, but world changing returns like ATP. That way if my risky buys don’t work out, I still have a comfortable nest egg. If they do work out, I’m not just comfortable, I’m very well off. But, here we are veering into buy decisions, which is a different topic.
So, my advice is pretty simple. Figure out the optimal portfolio for you, and just make it happen, not worrying about what you currently have in your portfolio. But don’t give Uncle Sam too much extra money if you can avoid it.
Disclosure: I own ATP, I don’t own Berkshire. They were just chosen as examples of different risk/reward profiles.