Friday, July 15, 2016

Stocks vs bonds in asset allocation - It pays to be long, but may not pay to be passive

A long history of the difference between equity and bond returns helps to frame discussion on asset allocation. The table shows the advantage of equities (S&P 500) over bonds (10-year Treasury) since 1928 on an annual basis. The average excess return is just over 6%. Any higher equity to bond allocation will do better over the very long-run given the bias to a greater risk premium; however, the variation in the differences is overwhelming.  One standard deviation is over 21%, so a one standard deviation event will range between plus 27.5% and -15%. The extremes are approximately 50% in either direction. The potential downside of holding stocks over bonds was over 50% in 2008, the worst single year of underperformance. 

Getting the stock/bond allocation wrong can wipe out any alpha generation. Assume that the alpha from hedge funds could be anywhere between 3-5% as measured by some academic research.  A 10%  allocation to hedge funds will add about 30-50 bps to performance from alpha, A 10% wrong allocation between stocks and bonds will cost 60 bps on the entire portfolio. You can add hedge funds and offset any of the risk from underweighting stocks versus bonds.

Nevertheless, it is more important to look at the variation between alpha and asset class allocations. There is less variation in alpha, perhaps 150-250 bps or on average 20 bps per 10% allocation. The standard deviation from being wrong on the stock/bond allocation is easily over 200 bps even on a 10% mismatch. The risk of being wrong on the stock/bond allocation is orders of magnitude higher than the risk on alpha production.

You need a lot of alpha to offset the risk from being wrong on your asset allocation decision. It thus makes sense to spend extra time on what should be the optimal stock bond allocation.

No comments: