The Variable Maturity strategy has been around for some years. Larry Swedroe mentions it in his bond book and DFA Funds among others use a variant of it (to select individual bonds within a fund, I think). Basically it involves going out longer on the yield curve if one gets sufficiently compensated by higher yields.
Why would this be of interest? Well people including yours truly keep asking themselves what bond fund maturities are reasonable right now. This gives one an easy framework to decide for oneself. Basically one goes out further in time (intermediate bonds, maybe 5 year duration) if one gets at least 0.20% per year better yield (20 basis points per year). Otherwise one stays in shorter maturities. The 20 bp/yr number is not really critical. That is, the results are not terribly sensitive to the selection of the parameters in this sort of strategy.
Example: Currently the 5 year Treasury yields 1.56% and the 1 year Treasury yields 0.20%. This means one gets (1.56 - 0.20)/4 = 0.34%/year or 34 bp/year to extend out. Since this is well above 20 bp/year, one buys the intermediate Treasuries.
One could use this 5 year & 1 year Treasury yield curve signal as a template
to move between a fund combo like for example Vanguard Total Bond Market and Vanguard Short Term Index.
I think to use the VM strategy one should:
1) Be willing to do a monthly yield check
2) Be willing to switch between a short term and an intermediate term bond fund
3) Not be sensitive to tax implications of a once per year on average switch
So does the VM work well? Well I think the answer is yes for the Treasuries. Here is a table with 60 years of data summarized. These returns are real returns. It shows that from 1954 to present the 5 year Treasury gave 2.0% real return. The VM strategy gave a 2.5% real return.
Both strategies did great in the last 33 years from 1981 to now. But notice that the VM strategy did much better in the rising rate environment prior to 1981. For me that is a big concern that can maybe be addressed here -- a rising rate environment that might give lousy bond returns over a period of years. Also the VM strategy should help navigate a future inflationary period should we get that. For a retiree this is a big worry I think as one does not necessarily have 30 years to see many cycles.
The table assumes a 15 bp/year advantage to make a switch to intermediate bonds. After a switch one stays in the fund for 3 months before testing each month for a switch. This results in about 0.7 trades/year. So yes, you cannot set and forget with this method.
Here is a somewhat involved chart. The blue bars are the basis points per year spread between the 1 year and 5 year Treasuries. The red signal indicates trades with high meaning we are in 5 yr Treasuries and low in 1 year Treasuries. The black line (right axis) is the 5 year Treasury yield. As you can see we are in a very low territory right now.
I did this analysis for my own purposes. It is kind of involved for some but I thought I'd put it out here. Might help someone. At least it is a context for thinking about one's bond strategy.