That Monster Called Diversification

Posted on May 28, 2010


Yes, this blog is targeted toward entrepreneurs.  But today I feel like writing about diversification in finance.  It might be every financial analyst’s most favorite word.

According to common theory, you can “diversify away” all asset-specific risk and remain with only the systematic (market) risk by including  a broad range of asset-classes in your portfolio.  The popular Capital Asset Pricing Model (CAPM) says that an asset that has a higher volatility than the market needs to yield a higher return than the market in order to justify the additional risk.  Since of course nobody knows what return to expect from a particular asset (a fact that doesn’t prevent investment bank analysts from putting wild expected return assumptions into their spreadsheets – that is when they are not icing each other), the safe choice is to invest broadly in the market and express your risk appetite by either using leverage or by investing part of your portfolio in “risk-free” treasuries.

So far so good.  This theory has convinced a lot of people and has lead to the immense popularity of S&P500 index funds.  But something seems wrong.  Here is what I am thinking:  Let’s say I am CEO of Bogus, Inc.  I go to an investment bank and ask them to arrange for an IPO of Bogus, Inc. shares. Those shares are now an asset that plays a role in the CAPM!  Fund investors will have to invest in my company if they want to be diversified! The fact that I am selling shares in pure bogus doesn’t matter. No matter how the shares of Bogus, Inc. do, they reduce a portfolio’s unsystematic risk and that’s all that the CAPM requires of them.

What I am trying to illustrate is that diversification, by eliminating non-systematic risk, can have the effect of increasing systematic risk instead.  Let’s use the dotcom bubble as an example.  A bunch of internet entrepreneurs produced an assortment of mainly bogus.  But every fund manager educated in the classic capital market theory was basically obliged to invest in this segment.  Otherwise she didn’t diversify correctly.  By doing so, she increased the asset class’s weight in the market, which lead in turn to even more investment in the name of diversification.  A self-fulfilling prophecy!  Combine this with the herd mentality of the non-diversified crowd and you have enough momentum to create a huge bubble.  All in the name of reducing risk.

Shall we apply this little theory to the latest crisis?  A lot of people knew that mortgage backed securities were nothing but bogus.  Yes, Goldman knew it.  So did the guys Michael Lewis describes in his latest bestseller.  But so did thousands of others!!  I distinctively remember an Economist cover screaming that housing prices are way to high.  I remember Steve Schwarzman repeating again and again that the end will come and leverage will disappear due to a crash in the housing market.  (And yes, Charles O. Prince The Third mumbled something about dancing.)  Pretty much everybody knew that it was coming.  What nobody knew was when.  And I suspect that the devil called diversification is partly to blame for that.  In a world where diversification rules among portfolio managers (who in turn control about 80% of all securities out there), one can’t get out of an asset class unless everybody else does so as well (otherwise, according to the CAPM, you would increase your risk unnecessarily).

Well and you know what happens when everybody leaves an asset class at the same time.