I decided that one way to evaluate this "rule of thumb" would be compare it to expert asset allocation models from three top-rated target date fund managers: BlackRock, JP Morgan, and Vanguard.
I averaged their bond allocation recommendations for each age cohort and graphed them in comparison to the "age in bonds" chestnut.
Visually, you can see that "age in bonds" produces a bond allocation that is too high at EVERY age group.
The age group for which this rule errs the least is the group you might expect (and likely the group for which it was originally devised), namely newly retired or nearly retired investors between age 65 and 70.
The age groups for which this rule is the most catastrophic are the youngest groups, roughly aged 20 to 40. These are investors with such a long time horizon, high savings rates, and generally low equity exposures that they should be 90% in stocks. This difference for a 40 year old investor who SHOULD be 10% in bonds but instead is 40% in bonds is approximately 150bps per year in lost returns.
I plotted various "age minus" rules to see which rule produced the least average allocation error for investors in their accumulation phase and it turns out that "age minus 18" is the least wrong for this group.
This rule is pretty easy to remember I suppose (18 is the voting age in the U.S.) but it still produces a considerable amount of allocation error for young and old age cohorts.
I set out to devise an approach that was significantly more accurate without being significantly hard to remember. Eventually I settled on the following:
- Age 40 and under, hold 10% in fixed income
- Age 45 to 60, hold "age minus 18" in fixed income
- Age 65 and over, hold 65% in fixed income/list]
Here's what this rule would look like (see the green line).
In actuality, as investors approach and enter retirement they probably have enough information about their accumulated savings and their spending needs to come up with a more circumstance-specific allocation.
But we are doing a great disservice to younger investors by perpetuating outdated bond allocation guidance for them. Telling an "average" 35 year old investor that they should have 1/3 of their retirement portfolio in fixed income flies in the face of every tenet of modern portfolio theory.
Obviously a simple rule is never going to be perfectly right for everyone. Heck, it might not be PERFECTLY right for anyone. But if we're going to dole out generalized advice, "age in bonds" has outlived its productive life and deserves to be retired.