How To Diversify
|December 19, 2016||Posted by Oddmund Grøtte under Stock Analysis|
How many stocks do you need to diversify? The idea with diversification is to reduce volatility and to make sure you don’t have too many eggs in one basket. If you for example hold 10 stocks, then a big hit in just one of these will take a huge chunk of your returns.
Most academic studies divide portfolios into deciles. For example, an academic study of the performance of low P/E stocks might divide group of stocks into deciles. This often involves holding a portfolio of about 100 stocks! So, in order to get this premium that exists in “cheap” stocks, you need to buy all those 100 stocks (theoretically). Assuming you only buy 10 of those 100 stocks, you are narrowing your universe a lot. Most of the research I have done (and also read), concludes that most of the premium reaped by buying cheap stocks come from just a few (of those 100). These few stocks are of course very difficult to spot in advance.
A highly concentrated portfolio has a high probability to miss the big winners, while at the same time a big loser have a huge impact on performance. If you hold 10 stocks and one drops 40%, the other 9 stocks must perform really well to outweigh this loss.
Diversification is also good for your mental health. To sleep well at night and not worry about money are the most important thing so you can actually follow your strategy. Thinking about money is in most cases detrimental to performance because you make wrong irrational decisions based on money and fear.
Warren Buffett is a strong advocate of a very narrow portfolio. He says diversification is “protection against ignorance”. To some extent he is right, but it assumes you are a very good stock picker. He says diversification means you are taking on bigger risk because you might buy stocks you don’t know much about.
Personally, I prefer to make diversification into about 30 companies. I simply don’t trust my abilities as much as Buffett does. I’m also quite risk averse. Not matter how much research you do, some of it will be wrong. No one makes just good calls.
Here are some points consider when setting up a portfolio of stocks:
Diversify to different industries
Of course, there is no point in diversifying into 30 companies in the same industry. Your portfolio must also be a balance among different industries.
Simply pick stocks among a wide range of industries. If your portfolio is 30 stocks, then you should have at least 10 different industries.
I’m from Norway (but not living there). Oslo Stock Exchange is highly concentrated towards oil and fish (salmon). There simply isn’t many opportunities outside oil and fish. Hence, I should look at other regions outside Norway:
Diversify into different regions
In a well diversified portfolio you should have companies that derive income from different parts of the world. That doesn’t necessarily mean buying stocks in for example Hong Kong. There are lots of companies with sales in Hong Kong but not listed there.
As an example, FTSE 100 companies only generates about 20% of sales from the UK! As a rule of thumb I prefer companies that have sales in all parts of the world and are not too dependant on a single market.
Example of a company I like
I prefer companies that are diversified into several divisions and generate sales from different products and from different parts of the world. One company I like, but currently does not hold, is UK based Spectris Plc (tickercode SXS.L on yahoo!). The company makes highly specialized measuring instruments and controls.I would like to own it, but think it slightly overvalued at this time.
The annual report from 2015 describes very well what I mean by diversification:
There are two things worth noting: Sales is spread among 4 different divisions and quite equal weighted. And if you look at sales by region it is spread quite evenly all over the world. This company is not dependant on just a few products and certainly not a “home market”. This is something which I try to look at when I make investment divisions.