PORTFOLIO BACKTEST / ASSUMPTION LAB
Backtest and compare portfolios, one variable at a time.
Describe a portfolio, change one allocation or rule, and see how the change affected return and drawdown.
No coding · No brokerage connection · No investment recommendations
PORTFOLIO VALUE
$23,8912020-062026-07
$24,481$19,567$14,654$9,740
PORTFOLIOCOMPARE NOTE / 01
Moving 10% from VTI to GLD changed ending value by $276 and maximum drawdown by +1.7 pts versus 100% VTI.
REAL DATA / CONTROLLED COMPARISON
What changed when 10% of VTI moved to gold?
$10,000 invested once from 2020-06-02 to 2026-07-31. Curves use the committed Marketstack adjusted-close snapshot—not a conceptual illustration.
100% VTI 90% VTI / 10% GLD
ACCOUNT VALUE · LUMP SUM$24,481$19,567$14,654$9,740
2020-062026-07
Adjusted close · annual rebalance · nominal USD · before fees and taxes · 1542 observations
PRODUCT PREVIEW / ILLUSTRATIVE WORKFLOW
From a portfolio question to a measurable trade-off.
This motion preview shows the intended workflow. It is not a recording of a live product.
“What changes when a 60/40 portfolio becomes 40/60?”
2020-06 → 2026-07$10,000 INITIALDIVIDENDS REINVESTEDANNUAL REBALANCE
MODERATE60% VTI · 40% BND→CONSERVATIVE40% VTI · 60% BND
Same capital. Same window. Same rebalancing. Only the stock–bond allocation changes.
Lower equity reduced the loss only slightly.
It also reduced the ending value.
RETURNDRAWDOWNPATHDEPENDENCE
THE CORE WORKFLOW
One base portfolio. Multiple controlled variants.
Generate several variants from one comparison, change a single variable in each, and compare which configuration performs best under the criteria you choose.
Generate controlled variants.
Branch from a base portfolio. Each variant changes one asset, allocation, contribution, or rebalancing decision.
See what drove the difference.
Trace an advantage to the periods and assets that created it—not just the summary metric that made it look convincing.
Test whether the conclusion holds.
Change the start date, market regime, allocation range, and assumptions before you decide a result is reliable.
ONE CORE / SIX SUPPORTING TOOLS
A complete comparison desk, built around the same question.
No disconnected calculators. Every tool adds context, evidence, or continuity to the comparison you are already working on.
01BEFORE THE TEST
“I want a long-term core portfolio with less severe drawdowns.”
STRUCTURING IDEA…VTI 60%BND 30%GLD 10%
Idea Builder
Turn an ordinary-language investment idea into explicit, testable assumptions.
02BEFORE THE TEST
DividendsReinvested✓
InflationIncluded✓
RebalancingAnnual✓
Data coverageExtended index✓
Assumption Inspector
See what the backtest assumes before the result has a chance to persuade you.
03UNDERSTAND
VTI VS 70/30 PATH
Difference Breakdown
See when two backtests diverged and which assets contributed to the difference.
04ORGANIZE
v1.0Base allocationNOW
v1.1Added 10% gold4M
v1.2Annual rebalance8M
Comparison Timeline
Track every change, rationale, and backtest result in one versioned comparison.
05AFTER THE TEST
RESULT CONSISTENCY82APR
→RESULT CONSISTENCY71JUL
One assumption weakened with new data.
Conclusion Monitor
Re-run saved comparisons and notify you when the original result weakens or reverses.
06SHARE & VERIFY
RESILIENT CORECOMPARISON REPORT / 07.2026
DATA v26.07ENGINE 0.8.4VERIFIED
Reproducible Report
Share the result together with its data version, rules, assumptions, and known limitations.
CONTINUOUS REVIEW
Markets change. Your conclusions can change with them.
Saved comparisons are re-evaluated as new data becomes available. Instead of another price alert, see whether the reasoning behind your portfolio still holds.
- Original advantage still exists
- The result has become less consistent
- An assumption is being challenged
- Portfolio has drifted from its rules
MONITORED PORTFOLIO / 04LAST RUN · JUL 28, 2026
PORTFOLIO
CONDITIONALGold as a drawdown buffer
60% VTI · 30% BND · 10% GLD
RESULT CONSISTENCY71 / 100
JANFEBMARAPRMAYJUNJUL
01
CONCLUSION UPDATE
Drawdown protection remains visible, but the advantage narrowed from 6.4 to 4.9 points after the latest data update.
02
ASSUMPTION CHECK
The result remains sensitive to the selected start year.
METHODOLOGY
Every result should be explainable and reproducible.
A backtest is only as credible as the assumptions underneath it. Every comparison records the details needed to understand what was actually tested.
01
Data sources and update dates
VERIFIED02
Asset and index substitutions
VERIFIED03
Dividend, inflation, and cost treatment
VERIFIED04
Contribution and rebalancing rules
VERIFIED05
Calculation engine version
VERIFIED06
Known limitations and data gaps
VERIFIEDCREATE YOUR FIRST COMPARISON
Which portfolio configuration holds up best under your criteria?
TRY AN EXAMPLE
EXPLORE COMPARISON EXAMPLES
See how one portfolio change affected the result.
Browse a real question, inspect the controlled change, and copy the example when you are ready to test your own assumptions.