feat(#978b): Source A + detrend in rate-sensitivity backtest #1025
No reviewers
Labels
No labels
admin
analytics
auth
automation
bug
business
chore
ci
compliance
data
data-moat
docs
duplicate
dx
enhancement
Fable 5 ревью
feedback/max
generative
GG-форсайт
needs-discussion
needs-human
observability
pause-bots
performance
priority/p0
priority/p1
priority/p2
priority/p3
scope/backend
scope/db
scope/devops
scope/frontend
scope/qa
scrapers
security
site-finder
stage/1
stage/2
status/blocked
status/done
status/needs-analysis
status/needs-fix
status/qa
status/ready
status/review
status/wip
tech-debt
tradein
ux
week ревью 1
wontfix
вторичка
ИРД
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference: lekss361/gendesign#1025
Loading…
Add table
Reference in a new issue
No description provided.
Delete branch "feat/978b-backtest-sourceA-detrend"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Cleaner OOS validation: --source A (survivorship-free corp_room_month) + --detrend. DEFINITIVE (vs прод): NO variant beats coin-flip — B-raw 0.148, B-detrended 0.148, A too thin. §9.6 негативный вердикт = РЕАЛЬНОЕ 'нет сигнала', НЕ survivorship-артефакт. monthly key_rate→sales не OOS-предсказуем на данных ЕКБ. Refs #978 #951
Cleaner OOS validation of the §9.6 negative verdict: - --source {B,A,both}: Source A = objective_corpus_room_month SUM(deals) monthly aggregate (survivorship-FREE, ~13mo); Source B path unchanged. - --detrend: linearly detrend ln(units) before differencing so a spurious survivorship trend can't drive β (_detrend_log helper). - cross-source verdict compares B-raw / B-detrended / A. DEFINITIVE RESULT (run vs prod 2026-06-03): NO variant beats coin-flip OOS — B-raw 0.148, B-detrended 0.148, A too thin (13mo<18). The §9.6 negative verdict is a REAL 'no signal', NOT a survivorship artifact (detrend + clean Source A agree). monthly key_rate→sales is not OOS-predictive on available EKB data → §9.6 stays advisory; forecast should lean on measured signals (supply/absorption). Tests 38→60.