How to Diversify Without Selling Stocks You Already Own and Love

Commit to starting new positions at a smaller size

Author's Avatar
Jan 29, 2017
Article's Main Image

Someone emailed me this question:

“I'd like to hear you talk about position sizing, cash size and the buying or selling pressures they put on an investor (in a general sense, I'll just add my specific current situation as a potential starting point). My retirement account right now has three positions and a 32% cash balance, the largest holding has recently appreciated to the point where it's 36% of the account.

I've considered selling a bit of my biggest holding, but I'd still rather have $X of this stock than $X of USD, especially while I already have an appreciable cash balance.

It doesn't seem like fixing my too large (but still undervalued) position by making my cash balance too large, or putting it in some other company that I feel is less undervalued and stable than this stock is a good idea. How is an investor meant to square this circle? I guess just turning over more rocks until I can find a place to deploy some of the cash I would get from paring back my largest position.

In a theoretical sense, though, what if this stock went up another 20% while I continued to not find any more satisfactory investments? At some point it seems as though the numbers would indicate that the most prudent action would be to sell this and have too much cash, or deploy some/all of the proceeds in an investment I'm not thoroughly convinced is a great use of cash.”

I’m going to start by directing this to your specific situation. You said your account now has “three positions and a 32% cash balance.” I’ll assume some positions – like the largest – have risen faster than others. Since you are concerned about your biggest position now being 36%, I’ll assume you’re not comfortable with a position being on average about 35% of your portfolio. In other words, you only own three stocks right now, but you’d like to own more than three stocks if you were fully invested. In your case, it sounds more like you might want four to five stocks when you’re fully invested. If that’s the case, I’d suggest starting your positions out in the 20% to 25% range and then just letting them grow to whatever they grow to. In other words, if you find that you keep having this problem of one position being too big – don’t sell that position now. Instead, just make a change to your future process to start positions out even smaller.

Let’s pretend you started this position that is now 36% of your overall account out as a 25% position. If that’s the case, I wouldn’t sell any of this stock right now. Instead, I’d make a note to only put 20% of your portfolio into the next stock you buy. From now on, you’ll start out at 20% and then the same rise in the stock – of about 40% in this case, will only get this position to say a 28% position instead of a 36% position. This way, it will take longer for a stock to get to a level you’re uncomfortable with. If this problem keeps happening, keep lowering the level you start your positions out at. If you find that you are still too concentrated in your best performing stock, lower the level you start your positions at to just 15%.

A lot of people are going to disagree with this advice. If you think about it, my advice here is really counterintuitive. You’re saying you already have too much in one stock and I’m telling you that the next time you buy a stock, you should buy less of it than you had been planning to. That means you won’t be diversifying as quickly as you otherwise would be. The knee-jerk reaction here would be to sell positions that get too large. In fact, this is what Sequoia fund announced they’d be doing in the future. I remember reading the transcript of Sequoia’s investor day where they spent most of their time discussing their losses in Valeant (VRX, Financial). They decided part of their mistake was in letting that Valeant position get too big. They didn’t sell a stock that had gone up too much. Now, I agree this was a problem. But, I don’t think the problem is exactly what most people think it is. So, let me explain what I think some of the risks in Valeant came from.

One, Sequoia bought Valeant. This was a “jockey” stock. It’s an execution dependent business. It’s not a business an idiot can run – at least it’s not a business model an idiot could carry out. That’s tough, and I’ve talked about this before. It’s a difference between how I feel comfortable investing and how Sequoia is comfortable investing. There are a lot of smart value investors – Bruce Berkowitz (Trades, Portfolio), Glenn Greenberg (Trades, Portfolio), etc. – who also focus a lot on who the management is and what they are saying. I don’t do as much of that. I think I’ve mentioned before that Quan and I considered picking DreamWorks Animation (no longer public) for our newsletter. We decided not to for a pretty simple reason. We asked the question: What if we woke up tomorrow and read that Jeffery Katzenberg had been hit by a bus. Someone else would be running DreamWorks? Would we buy the stock? Would we even consider it? The answer was no. I’d consider buying Berkshire Hathaway (BRK.A, Financial) (BRK.B, Financial) even if Warren Buffett (Trades, Portfolio) wasn’t running it. But, I’d never consider buying an animation movie studio with the wrong guy in charge. That’s a completely execution dependent business. I’m not saying Katzenberg is the only person we trusted. But, the track record we would care about was the movies that CEO – whoever he would be – had been involved in producing. If the wrong CEO is at the top of a movie studio, we have no interest in buying it. That’s the first risk Sequoia took with Valeant. It was an execution dependent business model and you had to put faith in management.

The second risk I think Sequoia took with Valeant is they listened to the CEO. They heard about the company from management. Now, they do that with most of the companies they invest in. And I think some of their biggest ever successes have been betting on the “jockey” so to speak. So, I’m not necessarily saying Sequoia shouldn’t talk to management. I’m just saying that spending a lot of time talking to management is probably part of what went wrong here.

Then finally, I do think the increase in stock price was a big problem, but not in the sense of the stock getting too concentrated. That’s obviously a risk. But, I think there’s another problem with a stock whose price has risen while you own it – that stock feels like a winner. You bet on it when others weren’t betting on it, and then those other people came around to your way of thinking and confirmed you were right. If the stock had gone up before you bought it, you might not feel this way.

This is a lot like – though not exactly the same as – a problem I see and think about a lot. I think of it as “risk habituation.” Basically, there are risks out there that people keep telling you about over and over again. At first you listen, but eventually you get used to hearing the warnings and not seeing anything bad happen. The example I use a lot is oil prices. But housing prices are another good example from the past. Interest rates could be an example for the future. Whenever I talked to other investors about oil prices back before they collapsed, people would say that, yes, they were maybe $100 and should be more like $70. But nothing was on the horizon that was going to bring them down fast. Then these people didn’t really talk about the possibility that if oil prices could easily rise $30 above what they thought the intrinsic value of oil was, they could also fall $30 below the level they thought was right. Very often, the people I talked to about oil prices didn’t actually have a different opinion than I did about oil. But oil prices had been so high for so long that they stopped listening as closely to what their own gut was telling them about the cost of finding and developing a marginal source of oil.

Whatever Sequoia’s concerns were about Valeant, they weren’t new. Sequoia has been right for a long time about Valeant even while other people were predicting that what the business was doing wasn’t sustainable in the long-run. Those critics weren’t really wrong about Valeant. But they were early. That happens all the time. I mean, if you had asked me three years ago to give you my best guess as to what the Fed funds rate would be at the end of 2016 and what level the Dow would be at, I wouldn’t have picked 0.50% to 0.75% and 20,000. I would have picked a higher interest rate and a lower stock price. I don’t think I was wrong to think it’s really risky to bet on either low interest rates staying that way forever or the Shiller P/E staying as high as it is now forever. There’s a danger that you look at how you were wrong about the timing and you say, well I must be wrong about some underlying reality here. But I don’t think there’s really any reason to do that in a lot of these cases.

Valeant’s model was always sustainable to a point. It was eliminating R&D, raising prices and adding to debt. That’s a very powerful strategy for fueling a cycle of really rapid earning per share growth. It will work – and it did work – for years. The same thing was always true for oil prices. They take time to adjust. When Quan and I were researching Carnival (CCL, Financial) and the other cruise lines, we saw that they were reducing fuel consumption, but we also saw how high oil prices had to be for how long to cause them to do this. If oil prices rise for what they think will be only a few years, they aren’t going to change business practices. Likewise, you aren’t going to see a lot of development of oil in the U.S. unless people actually believe oil prices are going to stay high for a while. So, for the first three years or so that oil prices are high, it seems like there is no balancing force that is either causing declining demand or increasing supply. But, if prices stay high for more like six years, then you’ll start to see big changes in long-term decision making. And then – as is always the case with these things – there will be a lingering over adjustment. Even three years after producers and consumers have started making adjustments, prices may only just then be reaching the correct, “normal” value for the future.

Well, this happens with stocks. It happens all the time. Let’s take the example of Frost (CFR, Financial). I owned this stock two years ago. It didn’t go up. I owned it last year. It was up close to 40% last year. What happened last year? The Fed raised rates a bit and said they’d raise rates a lot more. Oil prices – Frost is a big energy lender to Texas oil producers – weren’t in free fall. They rose a bit. That’s fine. But, let’s be clear about two things here. One, interest rates are still lower than what I thought “normal” would be in the future. And oil prices are now at the lower end of about what I thought “normal” would be. In other words, neither interest rates nor oil prices recovered as quickly as I expected them to when I bought Frost. Or, at least, they didn’t surprise me. This is about what I thought needed to happen for interest rates and oil prices to get back to normal. So, I thought they were both really below normal last year. Now they are still not above normal. In other words, there hasn’t really been anything surprising about interest rates or oil prices. Yet instead of Frost making slow progress each year – or rising by 20% the first year and 20% the next year or something – it had a huge part of its gains in just the last few months.

The danger here is that I would feel more confident in Frost now than I was last year. I think that danger is especially big when you are talking about a stock you owned a lot of for a long time and it kept going up, up, up like Valeant did for Sequoia.

I would warn you to be extra careful about your largest position, not because it’s too big a position, but because you might start feeling more confident about it now than you were in the past.

But, don’t – whatever you do – just sell the stock because it went up. Warren Buffett (Trades, Portfolio) once quoted Peter Lynch on this point. He called selling your winners to buy your losers “cutting the flowers to water the weeds.” I agree with that point. In the very long-term – we’re talking three to 15 years here, not three to 15 months – the best stocks to own are probably those that have done best for you in the past. There’s nothing wrong with holding on to the biggest position you have. Don’t trim it now. But, think seriously about lowering the initial amount you invest when you first buy a stock. That will help you avoid this situation in the future.

The level you pick is a personal one. I’d have no trouble starting at 20% to 25%. But, honestly, I have no problem if my best performing stock ends up being 50% of my portfolio. If your tolerance for concentration is let’s say 33% being the highest you ever want to see a good performer reach, then you probably want to start your positions off at more like 15%. But don’t change your position size now. Change your process for the future.

Disclosure: Long CFR

Start a free seven-day trial of Premium Membership to GuruFocus.Â