PORTFOLIO COMPARISON / ONE VARIABLE

Portfolio Start-Date Risk Backtest

Repeat the same portfolio backtest across multiple starting years to test whether its conclusion remains stable.

THE QUESTION

Would the same portfolio decision look convincing if the investor had started in a different year?

BASE PORTFOLIOOne fixed portfolio and one selected start year
CONTROLLED CHANGEThe same rules across rolling start years
ONLY CHANGED VARIABLEStarting date

The base rules, data treatment, and calculation assumptions remain visible throughout the comparison. Review the portfolio backtest methodology used to interpret this result.

WHAT TO COMPARE

Read the trade-off as a system.

  1. 01Outcome range
  2. 02Worst start
  3. 03Median result
  4. 04Conclusion reversal rate

WHEN TO LOOK

Locate where the difference appeared.

Pre-crisis startsMarket peaksRecovery startsInflationary starts

A useful comparison shows whether an advantage was broad or created by one unusually favorable period. For a direct test of this risk, compare the same rules across nominal and inflation-adjusted outcomes.

OBSERVED RESULT / REAL MARKET DATA

What the comparison actually produced.

Five-year annualized returns ranged from 5.28% to 12.89% even though the portfolio rules never changed. A single ten-year result therefore conceals a 7.61-point start-date range.

WINDOWSixty rolling five-year windows beginning Aug 2016–Jul 2021
CAPITAL$10,000 initial investment in 60% VTI / 40% BND
CALCULATIONThe same annually rebalanced portfolio is restarted monthly and measured for the following five years using adjusted prices.
Five-year startEnding valueAnnualized returnMaximum drawdown
Best: Nov 2016$18,31312.89%−21.76%
Median window$14,9778.42%−21.76%
Worst: Oct 2017$12,9285.28%−21.76%

TEST ANOTHER ALLOCATION OR RULE

Change one variable in AI Backtest.

Build your own portfolio comparison →