Geoff Gannon December 16, 2017

So: Am I Keeping Stocks Forever Now – Or Not?

There is a discrepancy between two posts I wrote. One is this post at Focused Compounding. Another is yesterday’s post here on this blog. A reader pointed this out:


In your post on NACCO from 15 December 2017, you state: “I don’t trade around a position. I buy all my shares at one point and sell all my shares at another.”


However in your post from 29 May 2017, your verdict is: “Geoff will never voluntarily exit a position entirely. Once he owns a stock, he’ll keep owning at least some of that stock forever unless that company is taken over or goes bankrupt. He will simplify things down to a true “buy and hold” approach. No thought will be given to selling a stock ever again.”


Don’t you think that these two statements are contradictory? Do you have a true “buy and hold” approach?


I don’t have a true buy and hold approach. I’m not a buy and hold investor. I’m always 100% open to selling a stock because I no longer like something about the business, the balance sheet, the management, the capital allocation, etc.


However, I’m not really open to selling a stock because it’s gotten too expensive.


Keep in mind: I find new stocks to buy over time. There are dry spells. Recently, I went almost two years without buying anything. But, my tendency to feed new ideas into the portfolio – in big initial position sizes – means that old ideas tend to become a smaller part of my portfolio over time.


So, even if a stock does become more expensive, I’d still be selling some shares of that stock over time just to fund new purchases. The stock could rise as a percent of my portfolio, but I’d still have sold shares in it. Two good examples are Frost and BWXT. Frost is a more than 25% position now (so, it’s slightly bigger in percentage terms than when I first bought it even though I’ve sold shares) and BWXT is close to a 15% position now while it was only originally part of a 20% position that got broke up (I bought Babcock & Wilcox stock pre-spinoff). I’ve sold about a third of BWXT and Frost, and yet they’re both just about the same percentage of my portfolio as when I first bought them.


Having said that, it should tend to be the case the the “stale” ideas in my portfolio will tend to get sold down and the “fresh” ideas in my portfolio will tend to be the biggest positions. The one exception to this would be if something in my portfolio was rising in price at a really unusual – probably very momentum driven – way.


It’ll be an interesting test of my resolve not to sell based on price if and when that ever happens.


So, I’m always open to selling a stock because I no longer like that stock (however, this has historically been a very rare reason for me selling). What has changed in my approach is that I no longer like to choose which stock to sell when buying a new stock.


As far as how I’ll eventually exit NACCO: I don’t know if it’ll be a sale of all of the position at once or a gradual reduction of something like one-third of the position each year (as I buy new stuff) along the lines of: 2017: 50% position, 2018: 34%, 2019: 22%, 2020: 15%, 2021: 10%, 2022: 7%, 2023: 5%, 2024: 3%, etc…


I’d prefer that my exit from the stock looks like that percentage position size series above. However, that will only happen if I never decide that NACCO – as a business – is too risky to hold on to. If I keep liking the stock, that gradual sell-down is what you’ll see. If I decide I made a mistake buying the stock, or something unexpected happens with capital allocation, management, etc. that really bothers me – you’ll see the position go to zero overnight.


I don’t know which I’ll do. My preference would certainly be to sell a stock only to replace it with another stock.


So, say my portfolio is now:


NACCO: 50%

Frost: 28%

BWXT: 14%

Natoco: 7%


And I decide – hypothetically, that I would like to have a new position in Howden Joinery. At a minimum, I would probably want that position to be 20% (as if it was part of a five stock portfolio) and at maximum I would probably want that stock to be a 33% position (as if it was part of a 3 stock portfolio).


In the minimum position size (20%) case, the resulting portfolio might look like:


NACCO: 40%

Frost: 22%

Howden: 20%

BWXT: 11%

Natoco: 6%


And, in the maximum position size (33%) case, the resulting portfolio might look like:


NACCO: 34%

Howden: 33%

Frost: 19%

BWXT: 9%

Natoco: 5%


However, there are 2 reasons why this literal application of the rule I laid out in the Focused Compounding post – that is, that I’d sell my existing positions down in in exact fractional proportion to my new position to fund that new position – would not be implemented.


One, this would be difficult to apply with illiquid and very small positions. So, at some point, I’d just eliminate a small position. To fund my NACCO purchase (in October) I sold about a third of Frost and a third of BWXT to have a position that was half my portfolio (the rest came from cash). However, I didn’t touch Natoco, because that stock was too difficult to trade quickly enough to fund a new purchase.


So, Natoco is likely to just be sold off from 7% of my portfolio to 0% at some point.


The other reason I wouldn’t implement this rule is if I decided I wanted to sell a specific stock for some reason having to do with that old position other than buying a new position. In other words, I still might make a sell decision instead of a buy decision.


Let’s say I decide that – after about a year of owning it – I re-consider NACCO and decide I don’t like the capital allocation, I don’t like what the management is doing, I’m more concerned about the future of the company’s customers than I was at first, etc. Then, I’d just sell it down from 50% to 0% in one decision.


What I was talking about in that post about being a “collector” of stocks is something different.


In the past, if my portfolio looked like this:


NACCO: 50%

Frost: 28%

BWXT: 14%

Natoco: 7%


And I wanted a new 20% position, I’d decide between selling either BWXT and Natoco or just all of Frost or something like that. I’d make a decision about which stock to sell to fund a new purchase.


I’m not going to do that anymore. So, if you see my sell NACCO and just NACCO – it’ll mean I wanted to get out of NACCO for some reason.


If I decide to buy a new stock, I will fund it – as best as I think is practical – through selling down all the stocks I already own in equal proportion.


Now, what about the language I just used “as best as I think is practical”. There are 2 reasons why selling a position as part of a proportional sale of my existing portfolio would be impractical:


1) I’ve owned the stock for less than one year (in other words, it is impractical to sell from a tax perspective)

2) The stock is illiquid (in other words, it is impractical to sell from a trading perspective)


Finally, there is one reason why it would be impractical to keep a position:


1) Below a certain portion of my portfolio – it might start to become expensive to sell pieces of a position.


I don’t use a discount broker. Now, if you look at my actual trading behavior, although I use a very high fee, high commission, etc. broker – I don’t have high fees/commissions etc. as a percentage of my portfolio compared to people who use discount brokers. This is because I trade much, much less than they do.


One, I sometimes have lower portfolio turnover than people who use discount brokers. And two: I place much bigger orders – as a percent of my portfolio – than people who use discount brokers. I will place single buy or sell orders that are 20% to 50% of my portfolio. So, the actual number of trades that are executed each year on my behalf is very, very small.


But, this kind of approach would break down if I slowly sold off a very small position into oblivion. So, if I had a 5% position this year that became a 4% position next year a 3% position in 2019, a 2% position in 2020, a 1.5% position in 2021, a 1% position in 2022, a 0.67% position in 2023, etc. That would eventually become wasteful. If I had a $1 billion portfolio, trading a stock exponentially downward by like 0.80 or 0.75 or 0.67 a year forever wouldn’t create trading costs that mattered. But, I don’t have a $1 billion portfolio. So, eventually that would create meaningful trading costs relative to the position for me.


It certainly won’t at 5% of the portfolio or above (and honestly, I could go a lot lower without worrying about this).


But, as a rule, I’d say that you’re right in thinking that positions which are about 5% or higher will be sold down proportionately to fund new positions – as I discussed in the post on “collecting” stocks.


Once a stock gets to about 5% or below – I might eliminate it at some point. I haven’t had to face this situation yet. Other than Natoco, it could be years before this becomes an issue for me.


As far as never selling a stock…


I never intended that to mean I wouldn’t sell a stock because I no longer liked that stock. I am still retaining the power to sell a stock as a sell decision. What I am doing away with is my power to pick between which stocks to sell to fund a new position.


So, in the past, if I owned both FICO (FICO) and Omnicom (OMC) and wanted a new position – I’d choose to either sell all of FICO or all of Omnicom.


In the future, I’d just sell a fifth of my FICO position and a fifth of my Omnicom position to create a new position that was a fifth of my portfolio.


Like I said in my post on whether I’d sell NACCO – I’ll re-consider that stock in about a year. If I don’t like that stock for some reason, I’ll sell all of it.


But, if I just want to buy a new position, I’m not going to sell all of NACCO to do that.


I was unclear in my post on NACCO. I said I’d sell all of the position. I didn’t mean I’d sell all of it to fund a new purchase. I meant only that if I disliked the stock for some reason – I’d sell it all at once.


Whenever I make a new purchase, I’ll fund it through selling all of the stocks in my portfolio in the same fractional terms as the new position I want is stated relative to the total portfolio. So, for a 20% position, I’d sell one-fifth of everything I currently own.


Finally, I don’t know if I mentioned this before or not – but, I’ve always though of Natoco (the Japanese net-net) as being separate from this new approach. Natoco is an old, leftover position that is more difficult to trade and which I’ve long said (it’s been years now) is slated for the chopping block. When I see a moment where I’d like to exit Natoco, I’ll probably exit Natoco in full.