How Do You Know When to Sell a Stock?

Author's Avatar
Feb 22, 2012
Someone who reads my articles asked me this question:


Your latest piece on coming up with an intrinsic value is interesting. Just out of curiosity, what methodology do you yourself use in terms of selling? I too look for things that are worth more than they are selling for without necessarily coming up with a hard IV figure.


But to some extent if you don't know where you're going any road will take you there. So obviously there needs to be some idea of how much is enough. Just curious what you do.


Regards,


Ken


Unlike a lot of people, I don’t set a price where I "will”, “must”, or “should” sell a stock. I look at the problem differently. I figure the more common problem isn’t holding a stock too long. It’s not holding a stock long enough. So, I set a price where I’m “allowed” to sell the stock.


The price I set depends on the stock. For different kinds of stocks, I set different kinds of allowable sale prices.


Here’s an example. I won’t use the stock’s name. It’s not really relevant. Just the company’s earnings and net cash are.


10-year average operating margin: 23%


Sales Per Share: $1.76


Net Cash Per Share: $4.73


Here I’ll just multiply sales per share times the 10-year average operating margin to get about 40 cents of average pre-tax earnings ($1.76 * 0.23 = $0.40). This matches up pretty well to what the company actually achieved over the last 10 to 15 years. So, we’ll use that as the operating business’s normal earnings. 40 cents before taxes is about 25 cents after taxes. If we assume a 10 times after-tax earnings multiple for that kind of business (my judgment is that this is a high quality, no-growth business) we get a value of $2.50 for the operating business. The company has $4.73 in net cash. So, $4.73 plus $2.50 equals $7.23 a share. Let’s round that off and call it $7.25. I would allow myself to sell the stock at $7.25 a share.


Now, I’d look at the high end for the business. Here, if it’s not a growth company it really shouldn’t ever be worth more than 15 times earnings (in my view). So, that’s $0.25 times 15 equals $3.75 a share for the operating business. Then we add the same $4.73 a share in cash and we get $8.48 a share. Which we’ll call $8.50 a share.


That’s where I’d allow myself to sell the stock. Below $7.25, I need to have found a much, much better stock to warrant selling this one. And above $8.50 a share, I need to have some added justification for holding the stock. Probably the only reason here would be that I misjudged the company’s normal earnings and growth prospects. So, for example, if they start reporting much higher operating earnings than 40 cents a share I should at least look at the possibility I’ve misjudged either average earnings or growth. Although if this is a strong time in the cycle for this kind of business, I won’t pay much attention to earnings increases. If it’s a lousy time for a business in this industry and yet they are reporting better earnings than I ever thought possible – then I’d need to rethink my analysis.


The other factor would be growth. If I judged this stock to be an absolutely zero growth business and they are suddenly growing sales, earnings, etc. at a faster than nominal GDP clip for a few years in a row – I would need to reassess that idea. This wouldn’t affect the top end of my sales range though. I’m not a big believer in paying up for growth. I like getting growth for free. I’d rather pay for current assets, preset day earnings, consistency in sales/earnings, and a “wide moat” rather than say a great organization that will find wonderful sales opportunities to exploit in the future. That’s great. But it’s not what I’m best at judging. So, in this case, all that would happen if I really bought into the idea of this as a true “growth stock” is that I’d shrink the allowable sales range in my mind. I’d still be thinking I should sell above $8.50. But I’d start to wonder if maybe I shouldn’t sell between $7.25 and $8.50.


Where I’m willing to sell always shifts depending on the underlying numbers. So, for a net cash company – or at least a company that traded below net cash when I bought it – the numbers that matter are what net cash per share is today and what I think normal operating earnings are.


Sure, there are tons of other issues that go into valuing the stock. For example: will the stock/bond portfolio that makes up this net cash grow or shrink? What’s the chance they’ll make some stupid acquisition? And so on. But I don’t believe in incorporating that mathematically into an intrinsic value estimate. Instead, I think you consider that before you buy the stock. And you do it more qualitatively than quantitatively.


So, I really only think in terms of a numerical intrinsic value when I’m asking myself: “is there still enough upside left in this stock?”. Before I buy, I look at intrinsic value when I know I like the company a lot. I’m comfortable with the current assets. I think the company’s got pricing power. Whatever. Then I have to ask myself how does this stock compare in terms of possible future returns to the stock I already own and the other stocks I might soon buy. That’s when I look at intrinsic value. When I need to compare alternatives.


It’s the same thing with selling. I almost never sell because of a price to intrinsic value comparison alone. I either sell because I change my mind about the company or I want to buy something else. When it comes to changing my mind about the company, if it’s an issue of safety – I just go ahead and sell regardless of price. If I’ve come to doubt the stock’s safety I bail out immediately. But if it’s a change in my perception of the growth potential of the company and maybe to some extent its quality – although often business quality falls under the heading of safety – then I consult my intrinsic value estimate.


So, I view the intrinsic value “band” as being a range where I’m allowed to sell if I’m so inclined. And, in fact, the band can be very, very wide depending on what alternative I’m looking at. So, for the example of a company with $4.73 in net cash and an estimated operating business value of anywhere from let’s say $2.50 a share to $4 a share – I might have a range that looked like this:


Less than $4.75: Can’t sell for any reason.


$4.75 to $7.25: Can sell if I’ve found something much, much better


$7.25 to $8.50: Can sell if I’m so inclined


Over $8.50: Should consider selling


This makes the process sound more complicated than it is. Basically, if this particular stock were over $8.50 a share – I’d know it would probably be at the top of my portfolio in terms of a good stock to sell when I wanted cash. And when it’s under $4.75 a share – I know it’s really unsellable. After all, it’s a profitable business trading below cash. If you’re willing to sell those, what aren’t you willing to sell?


Then, probably any new idea I liked a lot could justify selling in the $7.25 to $8.50 range. Once the stock was over its net cash plus 10 times earnings value, it’s a possible sell for me if I’ve got something better planned for the cash. And then between net cash and net cash plus 10 times earnings I’d be open to selling the stock if there was absolutely nothing else I could sell.


I’d be inclined to say this stock should be near the bottom of my possible sales pile. But I suppose it’s possible I could have a lot of good ideas all at once – maybe the market is crashing or something – and as long as it was above net cash at the time I sold it, I could forgive myself for selling. But I can’t really defend selling a profitable company for less than net cash – especially since it’s a company I’m familiar with. There should always be an extra margin of safety in switching from the horse you know to the horse you’ve just met. You should always err on the side of the familiar friend.


But now we see how hard it is to talk about intrinsic values and sell prices. Obviously, even in the case I just described, the values move. In this case, it’s the net cash that keeps changing – mostly to the upside over time. Likewise, if I bought a stock because I thought it was worth book value and then book value increased over the years my “sale allowed” window would shift upwards with the stock.


I try not to get fixated on a number. And never on a level of return. I don’t – for instance sell a stock because it’s up 30% or 130%. That doesn’t matter to me. Depending on the stock, I could set an amount like 10 to 15 times the 3-year average free cash flow. Say I owned a stock like Omnicom (OMC, Financial) or Dun & Bradstreet (DNB, Financial). Those are companies I expect to earn the same or more in the future – rarely less. I’d probably use a price to free cash flow ratio for those guys. I don’t – and a lot of people disagree with me on this one – think that enterprise value is necessarily super important at a company like OMC or DNB. Especially DNB. Because if you buy DNB, you are buying it with certain ideas in mind about capital allocation not because of its value to a private owner. So, I tend to think of stocks like OMC, DNB, etc. more like bonds with an initial return and a growth rate.


The initial return – let’s call that the free cash flow yield – is kind of like my yield. And the growth rate is really a quality indicator for me. In stocks like that, I would have trouble convincing myself it was okay to sell the stock at a free cash flow yield over 10% and I’d have a hard time convincing myself the stock should really be held even when it had a free cash flow yield of less than 5%. So, let’s call that a “sale allowed” window of 10 to 20 times free cash flow.


I’ve owned other things that were different. I owned an insurance company and valued it on its tangible book value. I was willing to sell that stock in the 1 to 1.25 times tangible book range. I was totally unwilling to sell at below 0.8 times tangible book. In the 0.8 to 1 times book value range, I could probably be convinced to sell if I found something absolutely wonderful to buy.


From day-to-day these values don’t change much. I usually check in with the numbers that go into my thinking – tangible book, net cash, etc. – only when a company files its 10-Q. And with many of the companies I own the changes from quarter to quarter are small.


What ends up happening is more of an informal ranking of what I own. I know what is in the “I’ll never sell” range and what is in the “I should think about selling this range”. And I have some idea of what stocks are approaching either bound of that range. As a result, I don’t usually – day to day – have to think about the intrinsic value of a specific stock I own. Instead I just know there are a couple stocks I’d sell before selling that one. Or that it’s the first stock I’d sell if I had to sell something. I try to always know that order. If you asked me now which stock I’d sell if I had to pick one – I know which one that is. The one stock I’d never sell is often a little harder to know. Sometimes I think there are a couple stocks I’d never sell but then some wonderful new idea comes along and really tests my resolve.


I’m not sure if any of that is useful. But that’s the way I think about selling. Again, all this comes with the caveat that I sell without regard to price when I think something I once believed safe is now risky. That’s the one unbreakable rule I have when it comes to selling a stock. If I misjudged a stock’s safety – I sell.


All the other rules are bendable.


Ask Geoff a Question about Selling Stocks


Check out the Ben Graham Net-Net Newsletter


Check out the Buffett/Munger Bargains Newsletter