Stanley Druckenmiller, in a recent interview, commented on his investing strategy:
“Put all your eggs in one basket and watch it very carefully….Every investor has 3 or 4 big winners and you usually know which they are. Where you get into trouble is when you do something you’re not terribly focused on. When you put 50,60, 70% or more of your assets into one asset class, trust me, you’re focused. And you are more risk averse than when you have 5-6% in something and have a blow-up.”
He went on to say that he never uses VAR models.
“I watch my P&L every day. If you are watching your P&L and your antenna are up, that’s more useful than any risk model”.
Now there is only one Stan Druckenmiller. He has never had a down calendar year (he has been down over a 12 month period). But the notion of having more than 70% of your assets in a single asset class seems really scary to me. Of course, as a macro investor he benefits from operating in markets like the currency markets where there is 24 hour trading and extremely high liquidity which means that he can quickly exit a position which starts to go wrong.
A poll on Twitter about position sizing had several participants talk of their largest positions being 70-80% of their portfolios. COming after Druckenmiller’s comments, this had me wondering if I have been doing this wrong, all this time. This is the second time I have looked at the subject, as I created a module in my Analyst Academy course talking about optimal position sizes. Rather than recap that content here, given that it’s a part of a course which students have paid for, I wanted to cover some other aspects in this article, and explore why Druckenmiller and others can tolerate this concentration when it seems extreme to me.
I have confined the discussion to equity portfolios where there is limited liquidity and markets and individual positions can be highly volatile. Although I cannot imagine having 70% of a professional portfolio in a single stock, what should the concentration limits be and how many stocks should you own? Obviously this varies enormously across individual investors but let’s look at a few considerations.
HOW MANY STOCKS?
Clearly, concentration limits dictate how many stocks you have in your portfolio. If you have a single 70% position, it would presumably be odd to have 100 positions overall, so let’s first think about this conceptually in terms of equal weighted positions – clearly, every investor will want to size up their highest conviction bets, but the ability to do so is contingent on the number of positions overall and the risk tolerance.
In the course, I go into detail on my views of the “right” number of stocks and position concentration and explain why this is dependent on the individual, the type of portfolio and the risk tolerance of any end-investors. Ensemble Capital believes that around 25 stocks is the level at which an additional stock provides little additional diversification benefit. I have been involved professionally both with more concentrated professional portfolios and with wider ones. Private investors with limited time may not want to have this many, but 25-35 stocks is a popular level for many successful investors (for example, Terry Smith) who run what are generally regarded as relatively high concentration portfolios.
This bent towards a 30-odd stock portfolio has many proponents. It’s notable that Jeremy Hosking, founder of the eponymous Hosking & Partners, an investor with a strong and long track record, takes the opposite tack. Each manager will run a portfolio of c.150 stocks and the blended portfolio will contain 400 stocks or more. This has been a highly successful strategy. Hosking is an exception. Here is what Ensemble Capital says:
“Owning 150 stocks or 350 stocks dramatically dilutes any ability you might have to beat the market without adding much in the way of diversification because you’ve already captured most of the benefits with your first 25 stocks. Yet this is exactly what most active managers actually do.”
Their contention is that your best ideas should perform best and obviously there is a limit to how many good ideas you can have. My experience in special situations investing was that the Pareto principle operated – a relatively small number of positions delivered 70-80% of the alpha in a year, but at the start of the year, you obviously don’t know which stocks will be the most successful. So by having a broader portfolio, you give yourself a better chance of exposure to the themes which become the most popular.
Lucy MacDonald, former CIO Global Equities at Allianz and a highly experienced investor, thinks that with a concentrated global portfolio of 30 stocks, your alpha should be delivered by a spread of positions. The chart shows our estimates of the contributors of the top 5 stocks to Fundsmith’s performance in the ten years since inception, excluding the first two months of 2010 (excluded as the results for a two month period are more likely to be more random).
The number of positions has been in a range of roughly 20-30, with the lowest seen being 22, but we don’t have data for every year. The average is probably close to 25 and the top 5 is probably close to the Pareto 20%. The rule does not quite work as you can see from the chart as the spread most years has been more even with the top 5 contributing c.45% and the remaining roughly 20 socks contributing 55%. But one year, the top 5 contributed almost the entire portfolio performance, helped by a takeover, and in another, the defensives contributed more than the entire modest portfolio gain with the remainder being negative. Overall, we estimate that the average is c.72%. Note that this is only a rough estimate as the individual performance data have not been disclosed each year.
Obviously, the fewer stocks, the more concentrated the performance will be. And if you have confidence in your stock-picking ability, isn’t a narrower portfolio better? And should you not size up your favourite bets? My view, which I explain in more detail in the paid courses, is that this is a highly personal matter – you should have as many or as few stocks in your portfolio as suits you, provided that you have enough to give you diversification. On this subject, there is a significant amount of academic debate.
OPINION 1: 10 STOCKS
One of the first serious studies on the number of stocks required for a diversified portfolio was a 1968 study by John Evans and Stephen Archer. Diversification and the Reduction of Dispersion: An Empirical Analysis concluded that only 10 random stocks were sufficient to replicate the market as a whole. It’s such an old study, and the universe has changed so much, that it would be sensible to assume that it has been superseded.
OPINION 2: >30 STOCKS
In How Many Stocks Make a Diversified Portfolio?, Meir Statman concluded that a well-diversified portfolio of randomly chosen stocks must include at least 30 stocks, which contradicted the earlier study and what the author suggested was a then widely accepted notion that the benefits of diversification are virtually exhausted when a portfolio contains approximately 10 stocks.
OPINION 3: 50 STOCKS
In this paper, co-authored by Burton G. Malkiel, the widely acclaimed author of a Random Walk down Wall Street, four academics studied the volatility of common stocks and concluded that from 1962 to 1997 there had been a noticeable increase in firm-level volatility relative to market volatility, and that the number of stocks needed to achieve a given level of diversification had increased.
OPINION 4: 12-18 STOCKS
In the book, Investment Analysis and Portfolio Management, Frank Reilly and Keith Brown concluded that about 90% of the maximum benefit of diversification was derived from portfolios of 12 to 18 stocks. Of course, 10% is a meaningful miss – most people would not want to take a 1 in 10 chance with their retirement portfolio, for example.
In the course, I explain a series of procedures to follow which allow an investor to restrict the number of portfolios held, but still achieve adequate diversification. It’s pretty obvious that if you have a 15 stock portfolio which consists entirely of US (or worse, UK) stocks, you will not achieve the same benefits as if you had a global portfolio. Similarly, it’s generally accepted that having a spread of sectors is effective, and many professional investors use this as a method of achieving diversification by allowing no more than a certain percentage over or underweight from market weightings.
I recently discussed this with a client with a £500m+ portfolio who was introducing new sector weighting restrictions – c.5% plus or minus from the benchmark. I think this is an ineffective way of achieving diversification – I can understand why you might perhaps want to limit the overweight, but a zero weighting in an individual sector might be eminently sensible – for example in banks in 2008.
And a limit on permissible sector overweights implies an assumption that sectoral definitions are meaningful, sensible and achieve diversification. But is this really the case, as we saw with the creation of the new Communication Services sector? Previously, these stocks were presumably “mis-classified” in other sectors. Particularly in today’s environment, when companies can cross-over traditional sectors – are Facebook and Google tech stocks or advertising media businesses? – we should perhaps be less focused on sectoral allocations than formerly. It’s worth listening to Patrick O’Shaughnessey’s interview with Dennis Lynch, Head of $100bn+ AUM Counterpoint Global, where he discusses how the misclassification of companies often creates significant opportunities.
POSITION SIZING AND RISK
There are a huge number of great quotes on position sizing by a lot of great investors and we use a few of them in the course. Of course, they offer a range of viewpoints, from Druckenmiller’s concentration of bets to Hosking’s highly diversified portfolio. Many professional investors equate position sizing and risk – Steve Cohen talks of three main risks. Here is an extract from his recent appearance on the Stray Reflections podcast:
“ Listen, you’re going to lose money. You’re going to take risks you’re going to lose money. I think the three things are liquidity, leverage, and concentration. Those are the three rules. If you’re in illiquid stuff, that’s a problem. If you’re using too much leverage, that’s a problem. And if you’re too concentrated, that’s a problem.”
So for Cohen, the three principal issues are lack of liquidity, leverage, and concentration, and too much of any of them will cause an investor problems.
I really enjoyed Shane Parrish’s podcast with Joel Greenblatt. Here is what Greenblatt said about position sizing:
“If you have this great investment, unless it’s, you know, Tesla last year, that’s going to go 10 x and you put 2% in, uh, and it goes up 50 or 100%. You could have blown it. That could have been your worst investment of all time because you should have had a 10 or 20% position, you know, and really move the needle position.
Sizing is the most important thing – being too timid on the few good ideas that come your way is like the biggest mistake people make.
But of course, to take a large position, you have to be willing to be wrong, take big losses to wait for that big position to know when is there? The biggest positions I’ve had are not my best ones, the ones that I think will go up five or 10 times. I’m really looking down, not up when I take a big position.
So in other words I will size the position larger if I don’t think I can lose much money, It’s not like the thing that’s going to pay 10 or 20 times. Obviously if you have a $10 stock that’s sitting with $9 in cash and no debt and they have a little business attached, you can buy a lot of that one as long as you don’t think they’ll dissipate the cash.
As long as there’s a chance for a big upside, you’re looking for asymmetric returns. And one of the, one of the best things I think I ever wrote was in, you can be a stock market genius I wrote, If you don’t lose money, most of the other alternatives are good.
I think people look at the size of their position, that’s the wrong way to look at it. It’s looking at how much can you lose from that position, when I lose?
When I say lose crazy things happen, whether it’s COVID or 2008 Lehman goes down or something like that, we are over a short period of time stocks or other securities can trade at any price. That’s not what I’m talking about, how much I’m risking.
What I’m talking about is if a reasonable person won’t lose much money, if they can pick their spot to sell it over the next couple of years, that’s what I’m viewing is how much I have at risk if I can be patient. So what I want to sell it over the next couple of years, what’s the worst I really think I’m going to do, That’s how much I’m at risk for and then how much does the reward pay off relative to that?
So if something I could lose a dollar or two in, I could easily have a 20 position because I’m not risking that 20 I’m risking. ……. Better to look at, you know, how much do I think I could lose in this investment rather than that big hairy number?
I think this encapsulates this risk aspect of position sizing incredibly well. You can have a highly concentrated portfolio of stocks, or a much greater concentration in a few stocks, if they genuinely have limited downside and you have enough time to wait. But this can be an incredibly long time in markets.
We have recently seen a private equity approach by Clayton, Dubilier & Rice for Wm Morrison, the UK supermarket retailer. It’s a sector which has been out of favour for most of the last ten years. Discount retailers Aldi and Lidl have been growing, taking market share, and dragging prices down at one end; and Ocado has captured a tiny share, but in some of the most lucrative areas and has generated significant demand for online delivery, which has increased the incumbents’ costs – they have been forced to compete online to protect market share, at the expense of their margins.
I have had positions long and short in this and related sectors. In 2016/17, I went to a supermarket results meeting and asked the then Chairman why the stock was trading at a discount to the value of its property. He responded to the effect that the property might not be worth what they had paid for it – quite an interesting comment from a company chairman. Obviously, many of the retail outlets were acquired when values were higher and many of the out of town sites are no longer desirable, with limited alternative use for units that large. But the warehouses are highly attractive with plenty of alternative use and can easily be the core of a sale and leaseback deal. This is exactly the type of stock that Greenblatt might consider in his example above. But look at the share prices:
Morrison had a strong run in the first quarter of 2016 then traded sideways for 5 years while Sainsbury had a flurry of hope in 2018, but then collapsed and it has taken 5 years to recover its start-2016 price. This is perhaps a lesson in that even a stock with high asset backing may be cheap enough to warrant a large position but will not necessarily pay off in most investors’ time horizons. Obviously, Greenblatt would be picking better companies, but I illustrate this simply to show that concentration is highly risky.
In my view, the correct approach to sizing is whatever suits the individual. Druckenmiller’s concentration would make most investors lie awake at night, but it works for him. Few professional investors and no retail investor could manage Hosking’s 400 positions, but he has an excellent track record. Although today some might consider it concentrated, I believe that a 25-30 stock global portfolio for professional investors is likely to become the norm in a few years, and it would be my preference. For private investors, it can be tricky to come up with this many names, and greater concentration may be a concern for many, but there is a simple solution. Combine your individual stock picks with an index ETF – that will allow you to focus on the stocks that you like, without causing undue anxiety if things don’t pan out as planned.