Thursday, June 23, 2016
The price paid for diversification - a simple case
I have recently seen more work on the poor performance of multi-asset and global macro funds. The simple analysis states that these strategies have underperformed the 60/40 stock bond mix of a domestic equity index (SPY) and long Treasuries (TLT) and have not done their jobs. If you conduct a breakdown of the impact of diversification, you can see there has been a significant drag from trying to spread risk even within asset classes. You can hold the stock bond mix constant but still have a significant diversification drag.
We have taken the simplest set of cases through comparing the 60/40 (SPY/TLT) mix with two simple diversification strategies. One, we substitute for the bond component the Barclays Aggregate index which holds corporates and mortgages and has a lower duration. Two, we switch the equity exposure to the MSCI world index (ACWI) which provides global stock diversification. Over the period from 2010 to the present, the impact is telling.
Global equities should reduce the exposure to the US business cycle but will increase exposure to non-US business cycles and currency changes. US stocks through their foreign operations are still exposed to the global business cycle and currencies, but the effect is muted. For fixed income, holding the Barclays Aggregate will reduce interest rate risk and should increase carry from credit spreads. Since 2010, having more duration and term premium exposure through holding TLT generated more return.
Many global macro and multi-asset funds which start with the premise of holding a more diverse portfolio have seen a large performance drag versus the SPY/TLT mix. This drag had to be made up through security selection or dynamic changes in allocation. Maya were not able to generate the added return. It may not always be the case that there will be this drag, but all investors have to be aware of the cost of diversification.