I’ve always liked the above map of the market, which shows how different investment styles have performed each year going back to the early 90’s

A couple of observations stick out to me, while a third is not as obvious.  First, leadership in most years appears to be relatively random, with styles not often repeating as top performers in successive years.  Second, in spite of infrequent repeats at the top, U.S. bonds have tended to be the worst performers on multiple occasions (seven times) – nearly 40% of the periods studied.  

The data suggests that a diversified style based approach to investing makes sense.  It also suggests that U.S. bonds tend to underperform in all periods except recession years.  (2002 and 2003.) 

What is not as obvious is the observation that only U.S. bonds tend to have negative correlations with everything else.  In other words, the greatest diversification gains may simply be had by investing a portion of a portfolio in U.S. bonds. 

How much depends on what your income needs are and what volatility you’re willing to give up for safety, both on the upside and downside.  All other styles will improve diversification, but perhaps more in relative terms than absolute.  (In other words, they will still tend to move up or down together, but at different rates, rather than opposite of each other as could be expected with U.S. bonds.)