Diversifying to Reduce Risk

While answering a question at StackExchange regarding backtesting of investing strategies, I came upon a web site, AssetPlay, which lets you test a variety of different asset class mixes over the last 36 years or so (currently 1972-2008). While tinkering to find the returns from 1980-2008 for different S&P and 5 year t-Notes I discovered the following returns;

This shows mixes from 100% S&P index, and adds 10% increments of 5 yr T-Notes, to show the different Compound Annual Growth Rates (CAGR) and Standard Deviations (SD).  I find a few things curious about these returns. First, the addition of the T-Notes reduces volatility dramatically, yet, the first two 10% increments actually increase growth. So the 70/30 mix lagged by only 2/100 of a percent, yet the volatility was fully 5% lower. Of course, this isn’t new, it’s the basis of the Efficient Frontier, the theory that proposes the optimal mix of assets based on risk. While I still find the concept fascinating, I’d warn that part of the success shown here is based on the fact that the 5 year note by itself had a return over 9% during this time. I’m not in possession of a crystal ball, but I’m willing to bet the next  decade won’t be so kind to the bond market. Regardless of bond returns, I suspect the stock/bond mix will continue to mitigate risk will little impact to one’s overall return.