Value investing opinion is divided on the optimal number of positions an investor should own. From Walter Schloss, who often held over 80 stocks at a time, to Charlie Munger’s early partnership days, where he was concentrated in a handful and sometimes only one*.
Buffett famously stated that a skilled, “know-something” investor should “place all their eggs in one basket and watch that basket”. But how many eggs are too few?
After far too much mulling on the subject, I have come to the goal of 10-18 stocks, most quantitatively scouted and then fundamentally researched, that I aim to hold for 5-7 years. I will go into more depth about how I arrived at this structure below.
I was prompted to write this piece because I recently read the quarterly letter of a fund manager who placed more than half their fund in one company. The stock drastically underperformed the market in Q1 and consequently their fund tumbled over 50% in value. This was clearly too few eggs.
This manager is highly skilled as an analyst, in my opinion, but left themselves open to an extreme event that will take a lot to recover from. The fact that Covid-19 was outside their control and anyone’s possible foresight is probably little consolation.
While these events are unpredictable, it is a given in investing and life that unpredictable events come to pass. In fact over an investing career, you should expect to see quite a few of them in varying degrees of severity.
Therefore, a portfolio that can be ruined by a single stock is a ticking time bomb and substantially at risk of blow-up. This violates one of the key tenets of portfolio management that I have discussed several times now- survivability.
Put simply, 40 years of diligent and consistent investing is almost certain to leave someone quite well off, even if they underperform the market. But going to zero and having to start again part way through the journey makes success so much harder. Not to mention the mental anguish to go along with a wipeout.
So how few are too few? And how many are enough for those who still want to be concentrated? Joel Greenblatt expanded on the idea in his excellent book “You Can Be A Stockmarket Genius”, where he states that an investor can eliminate 46% of the non-market risk of only owning one stock by owning two. 81% is cancelled by owning eight positions and by the time an investor owns 32 they have removed 96% of this risk- not all risk of course, but non-market risk.
Of course, aiming for 100% is pointless and self-defeating for an active investor and the market itself can still be very volatile in any case. Risk is a very personal thing, but I have no career risk and prefer to shoot for the slightly lumpy 15% than the smoother 10%.
If I found a loaded coin I knew was skewed 65/35 in my favour, I still wouldn’t bet my entire net worth on a single toss. However, if I was rational I would bet heavily if there was a series of twenty tosses to play. Even when the odds are heavily in my favour, things can go wrong with a small enough sample size. The larger the number of these promising bets I make, the more accurately my results will reflect the underlying probabilities.
The Kelly Criterion is often mentioned in value circles as a helpful way to size positions. The formula was developed by John Kelly, a scientist at Bell Labs, in the 1950s and has been popularised in investment circles by Ed Thorp and Mohnish Pabrai.
Given known odds and an edge, Kelly will calculate an optimal betting size. Ed Thorp used the criterion in gambling before turning his attention to the stockmarket. At Princeton-Newport partners he employed it with phenomenal success, outperforming the market by 5%pa and only registered three negative months in nearly two decades!
One of Thorps most legendary trades came when AT&T was being split up during the ’80s. Thorp, whose fund was worth around $15 million at the time, bought $330 million of the old AT&T while shorting the new “Baby Bells” to the tune of $332.5m. Thus, taking advantage of a risk-free arbitrage, gaining $2.5 million for his fund on conversion of the deal.
In relation to my own approach, I know I don’t have the computing power or the courage to pull off extreme bets like Thorp. I personally don’t think Kelly is applicable to a retail investment portfolio, because while it was designed for a casino setting where probabilities are calculable with precision, in the stockmarket we will never know the exact odds of an investment’s success or pay off (besides certain fixed income or arbitrage situations).
Also, in a casino scenario, games are played in series one after the other. A gambler knows how much they have left in the pot before starting the next bet and also what the next game is going to look like. The market doesn’t afford the same luxury, with thousands of potential games running in parallel. An investor needs to manage their best estimations while also being totally blind to the future opportunity set.
So then it becomes a balance- how many investments of an appealing calibre are available? While a simple screen could churn out many cheap stocks, it is unlikely they would all be equally appealing. Also, for a retail investor like myself, the brokerage and taxes would become crippling as many quant styles involve high turnover.
Many well researched strategies, such as Tobias Carlisle’s Acquirer’s Fund, turn over their portfolios quarterly. Institutional scale can render this a non-issue, but my Comsec international account, while seeming to be the best retail option in Australia, is still expensive and clunky to use.
I have chosen a goal turnover of 15-20% a year (based on 5-7 year holding period) on extensive research by David Dreman in the great “Contrarian Investment Strategies”. Amongst a wealth of data, Dreman finds that not only do value stocks outperform over a one year period on average, but they continue to excel over the next five years. This seems intuitive to me, as cycles can take many years to play out and and sentiment can continue to improve for a long time if a stock was truly considered toxic at purchase.
I have been successful at this so far, as I have only exited two positions in the two years I have been systematically tracking my performance. Periods of high volatility tend to offer more opportunities though, so I can’t rule out further activity while Covid-19 continues to wreak havoc.
I have read some managers describe their 40 stock portfolio as “concentrated”. This is far too many for my situation. My time is valuable as a one-person, part-time operation. If I do the work to get comfortable enough to buy something I am going to make it more than 2.5% of the portfolio. It’s just not enough to make a difference and 40 holdings plus new ideas is definitely too many for me to follow well each quarter.
Warren Buffett has also weighed in on this one, stating “I can’t be involved in 50 or 75 things. That’s a Noah’s Ark way of investing- you end up with a zoo that way. I like to meaningful amounts of money in a few things“.
A large number of positions seems to lend itself towards a quantitative strategy, where an investor seeks exposure to a basket based off a screening criteria. A look at the holdings of the modern day quant giants, such as AQR, Two Sigma or RenTec will show hundreds of positions. This works incredibly well for them, but for the reasons stated above, I am seeking concentrated exposure to fundamentally analysed names and believe I can beat the market this way.
Time will tell if I succeed at this, but I am enjoy the qualitative research and love the challenge. Elements of quantitative cheapness will remain the starting point to ensure I’m fishing in the right ponds.
So for now, I have settled on 10-18 stocks as about the right balance between nullifying single-name risk, concentrating in the best opportunities and managing portfolio friction and turnover.
Guy
PS. These are my own thoughts based on my circumstances. Your situation will be different and I encourage you to seek financial advice if you are unsure about how to construct your portfolio.
*Warren Buffett’s biography “The Snowball”, by Alice Shroeder, details Munger’s early partnership days. At one point he placed his whole fund (Wheeler, Munger an Co.) and leveraged himself into an arbitrage opportunity on British Columbia Power.
Superfluous Value Blog 