« Need investment advice, but cannot afford it? | Main | Diversification: Do you really understand it? (Part 2 of 3) »

Diversification: Do you really understand it? (Part 3 of 3)

When we decided to write about diversification, we thought it would be safer to spread our views over three blogs.

In the first blog we argued that you should concentrate on ‘risks’ rather than ‘assets’ when you are looking to diversify.

In the second blog we showed that if all risks gave you the same reward; and all risks were unconnected to each other, then the more you spread your risk budget across different risks (ie diversify), the higher your expected return per unit of risk (‘information ratio’) would be. 

However, the two assumptions are quite unrealistic.  We showed that if the first assumption was false (so the second best risk was slightly poorer than the best risk; and the third slightly poorer than the second;…) then the benefits of diversification are quickly eroded – “diworsification”.

Now we are going to tackle the second assumption.

Part 3: Beware correlation

Diversification works best where the risks involved are completely unconnected.  In mathematical terms this is called being “uncorrelated”.  Correlation goes from -1 where assets always go in opposite directions to +1 where assets move completely in line with on another.  Uncorrelated risks have a correlation of 0.

Let’s compare a portfolio which was 100% invested in BP to one which was 50% in BP and 50% in Shell.  The second portfolio would have some diversification.  But that diversification would be compromised because of the correlations between the stocks.  If a fall in BP’s price was due to falling oil prices, growth in alternative fuels, global economic slow down, adverse currency fluctuations or regulatory changes in the industry then Shell’s price is likely to also be affected.  You would only really be diversifying the risks that are very specific to BP as company (e.g. a change of management or major accident).

So correlation undermines the effectiveness of diversification, as shown in the chart below.  The chart is based on the example we set out in the last blog.

This is not just a theoretical idea.  In the recent financial crisis, many investors who thought that they had well diversified portfolios found that all of their investments fell at the same time.  Unfortunately this is quite typical.  In times of financial stress, the correlations of risky assets tend to increase sharply (e.g. demand for liquidity at virtually any price in the recent financial crisis), removing most of the benefit of diversification at the very time you need it the most.

One last chart.  The effects of diworsification and correlation are cumulative.  In this chart we see the effects of all of the risks being 50% correlated with one another as well as each asset being only 90% as good a ‘bet’ as the last one.  In that scenario, diversification would peak after holding only three different assets.


Diversification is generally a good thing.  It can be a free lunch.  However we need to watch out for:

  • Diworsification

This is where if each new risk (or source of return) added is not quite as good as the ones you already have, diversification will erode your expected returns.

  • Correlation

This is where if your risks are linked to each other, then the risk-reducing benefits of diversification can be quickly eroded.

PrintView Printer Friendly Version

References (7)

References allow you to track sources for this article, as well as articles that were written in response to this article.

Reader Comments

There are no comments for this journal entry. To create a new comment, use the form below.
Member Account Required
You must have a member account on this website in order to post comments. Log in to your account to enable posting.