Topic 3: On diversification

Asset risk is usually thought of as being made up of two components: diversifiable risk and systemic risk. The former type of risk is asset-specific because it results from the individual characteristics of the asset: for instance, in the case of a company’s stock, the asset-specific risk refers to the consequences of a bad product launch, a recall or a loss of market share. This type of risk is diversifiable through portfolio selection: by combining together different assets, an investor can reduce the variability of his or her portfolio without necessarily sacrificing expected returns. Systemic risk, instead, refers to the potential downfall brought about by the fact that a particular asset is embedded in a wider system: in the case of a bond, such risk may be represented by an increase in interest rates or inflation that reduces the real rate of return.

Although it is not possible to diversify entirely systemic risk, the asset classes as well as geographic areas in which one can invest have grown significantly in recent years, shifting part of the risk that used to fall under systemic risk into diversifiable risk. For instance, while a few decades ago most investors would have been unable to diversify the risk of being exposed to their domestic market, globalization has opened up new markets and allowed investors to diversify to a higher degree.

Diversification seems a no-brainer, and perhaps rightly so: but what are the limits and risks of diversification? As described above, diversification reduces the variance of a portfolio and does so through what in statistics is called regression to mean: as the sample size increases, the likelihood that the mean of the observations take on an extreme value decreases. By combining different assets with the same expected return, an investor can lower the variance of his or her portfolio without necessarily impacting expected return. However, empirical evidence suggests that diversification may, after all, impact returns: Ivkovic, Sialm and Weisbenner (2008), for instance, find that “wealthy households holding highly concentrated portfolios perform significantly better than the wealthy households holding widely diversified portfolios”. Although a formal investigation of this phenomenon is beyond the scope of this entry, it is interesting nonetheless to speculate about the mechanisms that may come at play with diversification. Research from Citi estimates the number of stocks that need to be included in a portfolio for it to be effectively diversified as high as 25 to 30 (Mincer, 2008). In this sense, diversification involves holding a significant number of different assets. In addition to this, the same research also found that over the previous two decades, only 10% of the stocks of the S&P 500 achieved returns consistently higher than the index over a three-year period. However, the proportion of stocks that the same study found to have underperformed the index by at least 15% in any year is as high as a third. Even though the way in which the results are presented makes it hard to compare them, the research seems to indicate that by selecting a stock at random from the index, it is more likely that the investor select an underperformer than an overperformer. In this sense, it may well be that as they increase the number of different assets they hold, investors end up purchasing lower-quality assets, with a negative impact on the expected return of their portfolio. Even though diversification remains one of the basic principles of portfolio theory, it is useful to explore its potential hidden costs.

The recent financial crisis also seems to have rebalanced the concept of risk in investors’ minds: after experiencing almost a decade of relatively stable markets, where crises were mostly sector-specific, many investors had discounted the systemic portion of risk and, as a result of this, overestimated the benefits of diversification. One of the main shortcomings of diversification has been identified with the fact that correlation increases significantly:

“One of the most striking features of 2008 was the fact that correlations between most asset classes went up substantially: everything declined at the same time. One of the principal motivations behind diversifying is that all of your holdings will not decline at the same time. Declines in one class will be buffered by gains in another—or at least lesser losses in others. This effect has not provided much buffer in 2008.” (Consildine, 2009)

In this sense, it is important to realize that diversification is designed with normal market conditions in mind and this may significantly limit its effect under extraordinary market circumstances. A similar hypothesis had been advanced even before the current financial crises by Butler and Joaquin (2001) that had pointed out the limits to international diversification. In this sense, it may be beneficial to expand the way we think about diversification and adopt a more holistic, multi dimensional approach to in order to manage portfolio risks (note the plural) more effectively. Diversification has traditionally been intended in terms of variance, but an investor may be interested in other features of his or her portfolio. For instance, an investor may want to diversify his or her portfolio along the liquidity spectrum, to ensure performance is appropriately balanced with exit options.

Not only does diversification have more dimensions that the one described in textbooks, but it may be something different than we think altogether. The recent financial crisis seems to have made a permanent mark on the markets, that are increasingly dominated by irrationality and wild swings. If this is a sign that, in fact, the dynamics that drive asset prices are less deterministic than we think as argued by Taleb, then investors are really left at the mercy of markets and diversification may just be a recognition of this.


  • Butler, K. and Joaquin, D. (2001). Are the gains from international portfolio diversification exaggerated? The influence of downside risk in bear markets. EFMA 2002 London Meetings;
  • Consildine, G. (2009). Opportunities in a High Correlation World. Available online at ;
  • Ivkovic, Z. Sialm, C. and Wiesbenner, S. (2008). Portfolio concentration and the performance of individual investors, Journal of Financial and Quantitative Analysis, Vol 43, No 3 pp. 613-656;
  • Mincer, J. (2008). Why it’s wrong to hold too much of one stock. Wall Street Journal, 4/9/08;
  • Taleb, N. (2001). Fooled by Randomness: The Hidden Role of Chance in the Markets and in Life. W. W. Norton.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s