Wednesday, July 20, 2016

What do you really get from bond diversification in the 60/40 blend?

There are more stories about how the 60/40 stock bond mix is dying. There is no question that as yields decline further there is more bond risk, albeit this has been a concern around the world well before rates went negative. That said, it is good to explore what has been the real value of bonds which has the hedging power of negatively correlated asset. We can explore this through decomposing the volatility of the 60/40 stock/bond mix.

The first graph shows the volatility of SPY, TLT, and the 60/40 blend. Equities are generally the most volatile asset followed by bonds; however, the long bond has had periods of greater volatility. The 60/40 mix will always be below the two with the volatility decline being significantly below the weighted average of the two assets. Clearly, if both bonds and stocks are higher than the 60/40 blend, there is diversification benefit. 

If the correlation between the two was 1, the the 60/40 mix would be a liner combination of the two assets and be between the stock and bond volatility. The low volatility of a blend is achieved through the low correlation of bonds to stocks.

We have discussed in the past the diversification benefits of an asset. (See Decomposing Diversification.) The gain from lower correlation can be divided into a de-correlation effect and the hedging effect. Our graphs focus on the diversification benefits through showing the impact of changing correlation. 

The second graph shows the risk reduction between taking a 60/40 linear combination of SPY and TLT virus the 60/40 blend. The linear combination assumes that the correlation between SPY and TLT is 1 while the blend accounts for the actual correlation or diversification. There is significant risk reduction because of the non-correlation between stocks and bonds. The diversification benefit can be anywhere from 4% to 10%

However, we want to go one step deeper by looking at the diversification caused by de-correlation which is the value added from the correlation declining from 1 to zero and the added diversification benefit of bonds being a hedge. The hedge competent is the added risk reduction from the volatility of a stock bond mix where the correlation is zero and the actual volatility of the blend.  The third graph shows the risk reduction from moving between 1 and 0 correlation and then the added reduction from the hedge created by the negative correlation of bonds and stocks. This hedge component is more volatile because the correlation between stocks and bonds moves around. When the correlation is positive, there is no risk reduction from the hedge but there is still decor relation.

So what are the biggest concern for the 60/40 blend? One, will bonds generate return at low yields, and two, will the diversification benefit form bonds still exist. The first is hard to predict, but we can say that if bonds no longer serve as a hedge, the risk reduction could decline by more than 50%.

No comments: