Research note
Composite Backtest: analytics with unknown results
Timeline, quintiles, transitions, plumbing and lens IC are distinct analyses. No frozen per-view results or page-wide verdict are inventoried.
Backtest verdict: Unknown — no frozen result recorded in the inventory. Composite analytics do not provide a universal trade ledger or one page-wide passing verdict.
Open this research tool in Backtesting
The rule
The timeline reconstructs the composite and its inputs over the selected range/cadence. Four further views inspect quintiles, transitions, plumbing and lens IC. The opening range is three years at a fourteen-day cadence; other range chips retain their own sampling steps.
This is an analytical family rather than a single trading rule. A quintile forward return or transition observation has no implied entry/exit fill, short/cash allocation, position size, stop, target or transaction-cost model. Those trade-execution fields are not applicable to analytics and must not be borrowed from the Score replay.
Methodology
Evidence as of: the INVENTORY review dated November 5, 2026; this is the consolidation planning baseline, not a new observation window. Publication does not extend any study's sample.
Reconstructed indicator history and forward benchmark returns feed each endpoint's native analysis. Exact dates, sample N, input revisions, return basis and per-view outputs are not recorded in the inventory. Overlapping horizons and cadence can affect effective sample size; a correlation is not a net trading return.
The three-month and fifteen-year choices have existing analytics bounds failures. These remain visibly unavailable rather than being silently resampled or interpreted as zero performance. Timeline availability does not certify that every analytics endpoint accepted the same request. No backtest was recomputed for this post.
Results
| Study / variant | Inventory evidence | Window / basis | Strategy return / CAGR | Benchmark | Max drawdown | Closed trades / observation N | Significance / IC | Native verdict |
|---|---|---|---|---|---|---|---|---|
| Timeline | D12 / F19 / H23–H24 | not recorded | not recorded | not recorded | not recorded | not recorded | not recorded | Unknown — no frozen result |
| Quintiles / transitions | D12 / F19 / H23–H24 | not recorded | not recorded | not recorded | not recorded | not recorded | not recorded | Unknown — no frozen result |
| Plumbing / lens IC | D12 / F19 / H23–H24 | not recorded | not recorded | not recorded | not recorded | not recorded | not recorded | Unknown — no frozen result |
The linked workspace retains the complete native output and notes when available. Missing evidence is not a zero return, a successful empty result or a failed strategy.
Verdict and limitations
Unknown — no result/verdict recorded in this inventory. Green IC, a favorable bucket or a timeline curve is not a trading PASS. An invalid request is not a failed financial strategy. Keep the full native arrays, bounds errors and notes beside any later interpretation.
Sources and live research
INVENTORY D12, F19/H23–H24 and its route-bound notes. Sources to check: frontend/components/backtest/CompositeBacktestBoards.tsx, BacktestAnalytics.tsx, frontend/lib/composite-backtest-page.ts, backend/app/scoring/history.py and backtest_analytics.py. A numeric claim needs the exact selected view/window/cadence/output and its benchmark basis.