The "magic" of diversification...or is it the Dark Art?

Don't tell me about diversification.  I am comfortable saying I've thought about this more than you have, so hear me out.

The concept of diversification, too often called the magic of diversification, amounts to a free lunch, you have been told.  This is wrong.  Diversification leads to financial crises.  Diversification is the core of the argument for giant financial institutions.  Frankly, its also the core of why GE needs to be as big as GE.

In a nutshell, here's what we've all be taught: There are idiosyncratic risks and there are systematic risks.  Idiosyncratic risks are diversifiable.  Systematic are not.  Systematic risk is compensated, meaning you earn a return for taking systematic risk.  Because idiosyncratic risk can be diversified "away", it isn't compensated.

What does this mean in practice?  Everyone wants to diversify away their risk.  However, diversification has fixed costs.  Therefore, diversification has returns to scale.  For example, the Vanguard Index fund can hold 500 stocks much less expensively than you or I can.  It's bigger.  The same is true for a bank.  We know the North Jersey Community Bank thinks they are better risk takers.  This is hard to prove.  What we certainly know is that NJCB cannot be as diversified as J. P. Morgan Chase because that would require NJCB to hold a sliver of every position that JPM holds, and they'd need all of JPM's fixed costs.



The same is true of companies.  Companies seek to produce stable earnings.  Why?  Stable earnings are more predictable, less risky and more valuable.  How do you produce stable earnings most easily?  Diversification.  Buy that incremental grandma fell warning system because it's diversifying to your earnings, and maybe you can cut some expenses.  Frankly, depending on the nature of the earnings, relative to other activities, the marginal acquisition could release capital.

So, virtually every manager, financial or otherwise, seeks stability of earnings whether that means financial diversification or product diversification.  The greater the diversification, the less it matters what in particular you do.  This ties back to my earlier discussion of corporate governance: Is a board's job to guide a company assuming the particular company is an investor's only holding, and therefore should be well diversified, or should the board assume the position is immaterial to the investor, and therefore should seek the highest risk adjusted return, knowing the risk is idiosyncratic?

The corporate or investment push toward diversification is incredibly strong.  The problem is that it makes sense!  But, like all good investment ideas, it makes the most sense if you are the first one there!  But you can't be.  So, in the end, the never ending push to diversification causes everyone to do everything.

The result, for the most part, is that we're left with one core risk in the market, that is not hedge-able, and not diversifiable.  Discount rates and market risk: What will someone pay today for a dollar in the future?  Where do we most clearly see this risk?  Equity market valuation measures, (P/E, P/B, pick your poison) and the risk free yield curve.  Everything else is a sliced and diced version of these risks, in the aggregate.