At its heart, diversification is a simple concept and one with which you’re probably already familiar, but understanding the basics is important.
I would say that at least 80% of your investment results will come from a dedicated and consistent application of the basics and only 20% from anything that looks even remotely clever.
But back to that question: Just how many different shares should you hold in your portfolio?
Diversification by numbers
The most basic kind of diversification is where you put money into many different companies rather than just one. Diversifying into many companies is sensible because the stock market and the future are uncertain, and putting all your eggs in one (or just a few) basket(s) really is a bad idea for most people most of the time.
Opinions vary on this question of how many, but I prefer to stick within Ben Graham’s guidelines, where he said:
“There should be adequate though not excessive diversification. This might mean a minimum of ten different issues and a maximum of about thirty”
My personal preference is to hold more rather than less, so I target 30 stocks. When I started out as a stock picker, I targeted just ten holdings because I wanted to concentrate on my “best ideas”. I also wanted to research the stocks deeply, and that requires time, so owning fewer stocks left me time to have a life.
However, over the years, I have gradually moved to a less concentrated portfolio.
Diversify so that you can sleep at night
I moved to 30 holdings partly because I didn’t feel comfortable with so much money invested into each idea. If you hold ten stocks, the starting position size is 10%, but it can easily reach 20% if a particular investment does really well.
20% in one company makes me very uncomfortable, especially when that company’s shares could halve in value, which would cost the portfolio 10% from just that one investment.
With 10% or more in one company, I tend to get nervous. I’ll start watching the news for that company, and when some bad news comes out (trust me – every company gets bad press at some point, whether it deserves to or not), I’m much more likely to make an emotionally driven, low-quality investment decision.
I don’t want to make emotionally driven, low-quality investment decisions, and I don’t want to be stressed out by my investments; I want to enjoy the process, so that’s why I switched to holding more stocks with less money in each one. If I hold 30 stocks, then I invest only slightly more than 3% of the portfolio into each, and if one of them doubles, then it’s still only 6% or 7% of the total amount.
Diversify to spend less time on your investments
Somewhat surprisingly, diversification creates a paradox. The paradox of diversification is that the amount of research an investor needs to do will go down as the number of companies in the portfolio goes up.
That may seem odd, but think about it like this:
If you had to put all your money into one company, how much research would you do before investing? A lot, of course. Once you’d invested, how much time would you spend monitoring that company so that you knew exactly what was going on? You’d probably spend quite a bit of time keeping up to date – I know I would.
What if, instead of investing in just one company, you invested in a FTSE 100 tracker? This tracker holds 100 companies which is a lot; it’s far too many to keep an eye on in any detail. So how many people who invest in FTSE 100 trackers know much about the companies that they’re invested in? How many could name even half the companies in their fund? I guess the answer would be about zero.
The paradox of diversification holds because companies, in aggregate, just tend to grow over time at or above the rate of inflation. As a group, they also pay a consistent and growing dividend, which is partly why passive investing is hard to beat. But when you narrow your list down to one or two companies (or five or even ten), the odds of a massive blowup become significant, which is why a concentrated portfolio takes a lot more work than a diversified one.
When I say that I hold 30 companies, some people are surprised and comment on how much time it must take me to keep track of them all. But I usually point out that the DOW 30 also has 30 companies in it, and most people who track the DOW index do so for decades and don’t even know half of the companies in that index.
I also found that very deep research didn’t seem to add much value (at least for me) compared to just applying a few basic steps. These basic steps are applied with fanatic discipline, but they are still basic and far less time-consuming than very deep research (or, to paraphrase Ben Graham – I don’t bother counting every last bathtub that comes out of the factory). So by focusing on the few things that really matter, I can hold more stocks but spend less time researching them.
The limits of diversification
I like wide diversification, but I wouldn’t want to own everything in the FTSE 100. After all, you can’t beat the market if you are the market, so some concentration is required.
Another issue is cost; the more stocks you own, the higher your trading costs are likely to be.
If you have £50,000 to invest and you hold each company for one year (although one year is closer to speculation than investing, so this is just an example), then a portfolio of 100 companies will have just £500 in each company. That may sound like a lot, but each trade costs around £10, so the act of buying and selling each company generates £20 of costs (not to mention stamp duty and the bid/ask spread), which is 4% of the total. A 4% per year drag on returns is enough to pretty much wipe out any returns from the stock market after inflation.
Sticking with 20 holdings instead of 100 would increase the size of each investment to £2,500, and so the £20 cost of buying and selling would be just 0.8%, leaving most of your inflation-adjusted returns intact. A longer holding period would reduce costs even further.
In general, more diversification is better, but only to the point where the benefits (lower risk, lower stress, lower effort) are outweighed by the costs (constraints on out-performance and trading costs). For me, that balance is best struck with around 30 holdings.
Interesting article though I think you miss an important point on diversification.
It’s not really the number of companies that you own that signifies your diversification but how uncorrelated their business interests are. Having 10 companies in 10 different uncorrelated sectors is likely better diversification than having 30 companies with all in the E&P sector.
There’s also diversification across investment factors that ought to be borne in mind. 30 value stocks might under-perform the market for several years in a row before bouncing back. Diversifying across other stock market factors like momentum and size can actually increase your returns while lowering your risk. This isn’t a well understood point at all by most investors.
Hi Ed, of course you’re right.
Diversification across various factors is important and the number of stocks alone is just one of them. Personally I don’t really think about diversifying across factors like ‘growth’ or ‘value’ or ‘momentum’; I think I get a mix of those from the way that my stock screen works. Instead I explicitly diversify across things like industry and geography and I’ll be putting up another post about those factors in a few days.
John
It depends what you want to achieve: if you want to beat FTSE All Shares you will hold anything between 10 and 45 stocks, with a bias to mid-cap and small cap.
I have sold 80% of my portfolio over the summer with the view to buy a proper house, which I didn’t find yet. I sold the lot: Diageo, Nestle, all the consumer staples, but I kept some Rolls Royce and the whole Rightmove stake. This one helped me to participate in the cyclical run this year along with Kingsfisher.
I even sold my oldest investment which I thought I will never sell – BAT which I bought in 2001. I kept BT on the defensive side, and still have to sort put the Vodafone saga – where I researched Verizon but not very happy to participate in a high leverage business, but I am not in a forced sale either. Played a bit the banks (Barclays) got a small loss never touch them again.
I do have some tobacco left, Imperial and Japan Tobacco. The last one I still believe in it, the first one I would like to sell it at a bit more as I can’t see more future in it because: Europeans gave up smoking and investing in Asia is very capital intensive and the competition BAT and Japan Tobacco are already there. Recently I sold the Astra Zeneca and Glaxo, but I bought some more Roche. I did also take the free money given by the UK Government as Royal Mail shares and together with my wife we have subscribed £750. Never miss a free lunch!
I always hold Japan stock in the last 4 years, it is little know this offers diversification in moments of crisis. People run to the hills / YEN and by doing so push these stocks up. I use the Baillie Gifford fund for this.
I did a lot of research on the European market where things are cheaper. Mediaset Espana is my second biggest holding now, at around 17% of my portfolio, and I only bought this year. It is cyclical nonetheless, but it has now 50% on the TV advertising there and the spending on marketing is growing in Spain. It did buy the competition on the cheap and it has little borrowing. The competition is week and highly leveraged, so I can only see Mediaset increasing the market share as it has the money to bid for big sports tournaments.
For me having 10 – 12 stocks and a fund is not a big worry. I don’t have a big plan yet for what’s left in my portfolio, although I am 43 in January I haven’t started yet my pension planning. I will design a plan as I go along.
The main issue is that for me I can afford to have such a concentrated portfolio, my clients can’t. As a result they will own 100+ shares.
Hi Eugen, you’re approach of 10-12 stocks plus a fund (or funds) is what a lot of people do, as I’m sure you’re aware. I guess what matters is the size of the position in a single stock. I think about 10% is really pushing it and 5% or less will be far more psychologically comfortable for most people. So if you have 50% in a few funds and 50% in 10 direct stocks that’s probably a reasonable approach.
As for your clients having 100+ stocks that’s exactly what I’d expect. Most people have little interest in analysing companies and so active or passive funds is the way to go.
Efficient frontier theory applied to pre 1980s data tells us that above about 40 uncorrelated and diversified stocks you effectively own the index so you might as well buy a tracker as its cheaper. Under about 15 and you are too concentrated.
http://www.investopedia.com/ask/answers/05/optimalportfoliosize.asp
However this has been questioned recently. Using the latest data a figure of about 50 is now being suggested as the minumum number of holdings necessary to manage the concentration risk.
http://www.efficientfrontier.com/ef/900/15st.htm
Hi Andrew
Thanks for the links, although I would disagree with 50 holdings being required. The Efficient Frontier site is very good but does take a very statistical approach which I like, but also think misses half the picture. For passive investors 15 or 30 stocks is madness, as the author says “Fifteen stocks is not enough. Thirty is not enough. Even 200 is not enough. The only way to truly minimize the risks of stock ownership is by owning the whole market.”.
But ‘minimising risk’ is not what I’m after. I want the best risk adjusted returns for the amount of risk I’m willing to take, and when I say risk I don’t mean beta either (which misses out far more than half the story), but at least you can measure beta, and so far my model portfolio (which holds ‘just’ 30 stocks diversified across multiple factors) has a trailing 12 month beta of around 0.6 which is lower than I expected.
So what do you do when an individual stock with just normal growth moves to 6%. Sell some and loose the dividends and pay tax’s on the gains or let it run.? 3% being about $30,000
Hi Brian, yes I do actually have a rebalancing ‘rule’ that says if a stock goes over 6% of the total I sell half of it. That reduces the holding back to my default 3% (around 1/30th of the total portfolio) and frees up some cash to be reinvested elsewhere.
If there is tax to be paid I just pay it and any loss of dividends is only temporary as the realised cash will be reinvested fairly promptly.
Also, the half of the stock that was sold will in most cases be more ‘expensive’ than the stock replacing it, so the new dividend yield will in most cases be higher (the old stock is probably expensive as it must have grown at twice the rate of the overall portfolio to have reached 6%, and usually that growth comes from share price growth rather than fundamental growth of the underlying company).
Hi John,
Thank you for this great article, which just saved my day today. Having just read a book about Warren Buffett and his recommendation to not hold more than 10 stocks in a diversified portfolio, I was happy to read that other value investors are diversifying more.
With a portfolio with more than 20 different stocks I would just not fee comfortable. At least I am not confident enough with my investment decisions (despite all the homework) to put all eggs into one basket.
However, I understand that Buffett is smarter than all of us.
Best regards from Germany, Dirk
Hi Dirk, yes 10 stocks is for either the very intelligent or the very brave.
It was that sort of Buffett-quote that got me into trouble years ago when I used to hold just 10 deep value micro cap companies. A very bad idea, very expensive and very scary.
I agree with you entirely and I now hold 30 companies, mostly from the FTSE 100, and all with long and relatively successful track records. It may be a bit more work to look after 30 stocks than 10 (although I’m not sure it really is) but it sure is a much easier portfolio to live with!
Great article – just setting out on my journey to build my share portfolio – and had in my mind of getting to a £30k target with around 30 companies, so on the money really in terms of the approach suggested around diversification. Agree at £1,000 a company it does mean that for some of those 30 companies taking a bit more of a gamble (riskier sector / company) is possible, and if it fails is a bit of a pain, but not the end of the world – who knows on the flip side it might be the next Microsoft!! Thank you