Hello Mustachians!
I don't know if this idea has been brought up before, but I had the thought today of "What if you didn't reallocate away from 100% equity as you age?" It seems like lots of financial resources tell you that you need more bonds as you get older, for stability, but I wanted to test that idea out using a FIRE sim. The FIRE sim I used for these tests is cFIREsim. Here are the initial conditions (as they differ from the default configuration), which are as close to my family's condition as I could get:
Portfolio: 50000
Spending: 20000
Retirement Year: 2020
Years to Model: 60
Inflation adjusted spending
Keep Allocation Constant
Equities: 100%
Fees: 0.05% (VTSAX)
Other Income/Saving 1: Amount: 100000 Start Year: 2014 End Year: 2020 (This is income until we retire)
Other Spending 1: Amount: 50000 Start Year: 2028 One Time (This is a downpayment on a house)
Other Spending 2: Amount: 17000 Start Year: 2028 End Year: 2043 (This is the 15 year mortgage on the house)
No Investigation
I know this model doesn't take into account a lot of things, but it seems accurate enough for me. Running this simulation gives you the results of never changing the allocation from 100% equity. Under the "Portfolio Inputs" section, you can choose "Gradual Allocation Change" and set it to "Change Allocation Gradually". For two alternate simulations, I chose to change the allocation to 75/25 stock/bond in between the years 2025-2035 and 2055-2065 to simulate shifting allocation when I retire and around when I'm 60, respectively.
All three of the above simulations give 100% success rate, but their ending balances show that keeping 100% equity leaves you the richest person in the graveyard. Now, since being the richest dead person isn't actually a good goal, and since we might get different results this way, I chose to run a second set of simulations. I kept all the same initial parameters, except I increased my spending rate until I got less than 100% success rate. My thinking in doing this was "Maybe stabilizing your investments with bonds actually helps you out when you aren't guaranteed to not run out of money."
You can follow along by increasing the spending of this model to $30000. This gives you 83% success rate for each of the three simulations. However, the median ending portfolio amounts are higher the later you reallocate, and highest for no reallocation at all. Increasing spending by even more gives you an actual change in the success rates, with the early reallocation having the least success and the later reallocation and no reallocation giving the same. Messing with even more variables to give a wider spread shows that the success rate is lowest for early reallocation, highest for no reallocation, and in the middle for late reallocation.
So what does this mean for the advice that you should change allocation as you age? I might be missing something huge that repudiates all of my findings, but if I'm not, it begs a serious reconsideration of this age-old rule of thumb.
Please let me know your thoughts on this and anywhere I might have made mistakes.