In a recent post I evaluated whether Betterment could beat a target date fund and compared returns of the Betterment allocation and determined that Betterment has had better returns then a fund similar to a target date fund.
After a little thought, I quickly realized my analysis was flawed. The analysis was subject to hindsight bias. You see, the historical performance data from Betterment was just that, historical. Betterment was founded in 2008, but the historical performance data started 3 years earlier in 2005. When Betterment launched that already knew the performance for this 3 year time period and could’ve built their portfolio in a way that made their portfolio look good. Your could say Betterment had a 3 year head start over the comparison 80/20 portfolio, Vanguard Life Strategy Growth Fun (VASGX).
Errors like this happen to investors all the times. Obviously it got me, and I am rather experienced and highly conscious of backtest errors and financial data.
To get an accurate analysis we need to use the period after Betterments portfolio went live, 2008 onward. Lets take a look:
Over this period of time, the Betterment portfolio still won, but not by as much. The Betterment portfolio was ahead by a mere $152.58. Since we all know that CAGR is the real tell tale of returns I have put the CAGR in a chart below. You’ll see the CAGR tightens up when using ‘live data’.
| Betterment 80/20 | VASGX | |
| 01/01/2005 – 6/30/2016 | 6.639% | 5.541% |
| 01/01/2008 – 6/30/2016
(accurate period; after Betterment was launched) |
4.125% | 3.991% |