02 · Prove

One good curve proves nothing.

Analysis pulls a backtest apart — seven views, from monthly returns to start-date sensitivity — to show whether the result survives scrutiny, or just got lucky.

Golden Butterfly PublicUp to date · as of 19 Jul 2026
Monthly returns
+
JFMAMJJASONDYr
20261.4−0.60.82.60.51.20.1+6.1%
20250.91.6−1.80.71.30.42.2−0.40.61.50.8−0.7+7.3%
20241.80.51.1−2.10.91.40.31.7−0.90.62.8−0.5+7.7%
20232.9−1.61.20.4−0.82.41.6−1.4−2.0−0.94.12.7+8.6%
2022−3.2−1.51.3−3.60.0−4.42.9−2.2−4.91.92.6−1.0−11.9%
2021−0.71.41.12.30.61.20.91.5−1.92.5−0.51.8+10.6%
Simulation 2005-01 → 2026-07 · run 2026-07-19Costs on · taxes off · profits reinvestedBacktests show what would have happened — not what will.
1

Seven views, one verdict. Growth, positions, metrics, months, allocations, components, start dates — click through every angle on the same run.

2

The bad months stay visible. 2022 reads red, cell by cell — a strategy you'd actually have lived through, not a smoothed curve.

3

Luck gets its own trial. Start-Date Sensitivity re-asks the question from every starting month — it has its own section below.

The interrogation

Ask the result three hard questions.

Any curve can look good from the right angle. These are the views that catch a result lying — or prove it doesn't.

See it on a Library strategy →
Q1Where did the money actually come from?
Trade history
every position, opened → closed · nothing summarized away
SymbolNameGrossNetTotal P&L
SPYSpdr S&P 500 Etf Trust$128,313$127,173$146,293
IWMIshares Russell 2000 Etf$117,162$115,616$130,088
GLDSpdr Gold Trust$124,704$123,283$123,283
IEFIshares 7-10 Year Treasury Bond Etf−$3,928−$4,950$23,164
TLTIshares 20+ Year Treasury Bond Etf−$15,811−$17,128$17,490
Q2Is it one big bet in disguise?
Per-component metrics
each model graded separately
ModelCAGRMax DD
Large-cap equities+11.2%−48.1%
Small-cap value+9.7%−53.6%
Long-term bonds+4.4%−41.8%
Gold+7.9%−42.9%
Q3Does it hold up beside the benchmark?
Summary
strategy vs S&P 500 (benchmark) · full scoresheet
MetricStrategyBenchmark
CAGR+8.1%+10.9%
Volatility9.2%18.7%
Sharpe ratio0.690.49
Max drawdown−23.0%−55.2%

The luck test

Would a different month have changed everything?

Start-Date Sensitivity re-runs the question from every possible starting month — across 1, 3, 5, 7 and 10-year windows — and shows the whole spread: best, worst, and typical.

Start-Date Sensitivity
1 yr3 yrs5 yrs7 yrs10 yrs
CAGR range (5 yrs)
10.1pp
best minus worst start
Positive starts (5 yrs)
100%
of all starting months
Avg CAGR (5 yrs)
+7.7%
the typical outcome
Worst CAGR (5 yrs)
+4.2%
the unluckiest month
CAGR by start month
+
JFMAMJJASOND
2016
2017
2018
2019
2020
2021
Best startApr 2020+12.4% CAGR
Worst startJul 2021−0.8% CAGR
Most painful startNov 2021−18.6% max DD
5-yr windows · every completed starting month · Simulation 2005-01 → 2026-07Costs on · taxes off · profits reinvestedBacktests show what would have happened — not what will.
1

The spread is the verdict. A tight range across start months reads as method; a wide one reads as luck. Here it's 8.9 points — you decide if that's tight enough.

2

The floor matters most. Worst CAGR is the month you'd least want to have started — if you couldn't have lived with it, the average doesn't matter.

3

No competitor shows you this. One backtest is one draw from the deck. This view is the whole deck, face up.

The full kit

Seven views. Seven questions answered.

Every strategy carries all of them, computed on every run — no premium tier for the truth.

1 / 7

Capital Growth

How did money grow — and what did costs, taxes and interest take along the way?

inside: Growth · P&L breakdown · Summary
2 / 7

Position History

Every position, opened to closed — where the result actually came from.

inside: Trade history · By symbol · Dividends & splits
3 / 7

Performance Metrics

The full scoresheet, always beside the benchmark's.

inside: Summary · By year
4 / 7

Monthly Returns

The good and the bad months, cell by cell — nothing smoothed away.

inside: Monthly returns · Equity · Drawdown
5 / 7

Allocation History

Where the money sat, month by month — cash included.

inside: Current allocation + the full timeline
6 / 7

Component Analytics

Each model graded separately — no hiding inside the blend.

inside: Per-component metrics
7 / 7

Start-Date Sensitivity Featured above ↑

Would a different starting month have changed the story? The luck test — the view no competitor gives you.

inside: CAGR by start month · Starting point evolution · Summary statistics

Luck runs out. Evidence doesn't.

Every Library strategy carries this analysis, open — no account needed.

Fincanva provides no financial advice. Backtests show what would have happened — not what will.