MSCI Artifact Tracker: Live Verification Feed for the November 11 Decision Window

Rupiah Stability Watch · 2026-10-26

MSCI Artifact Tracker: Live Verification Feed for the November 11 Decision Window

The premise

On November 11, 2026, MSCI announces its Semi-Annual Index Review — the first scheduled window since the October 15 consultation launch where an interim update on Indonesia's proposed reclassification from Emerging Market to Standalone Market status could appear. The market has priced a 120–180 bps risk premium into the rupiah since the consultation announcement.

The Pre-MSCI Decision Week: The Artifact Watch (this watch, Oct 26) established the framework: 21 remediation rows across four fiscal-transmission channels, their public verification artifacts, forced-selling volume/price thresholds, and a three-scenario decision matrix.

This piece operationalizes that framework. It is a living artifact tracker — a methodology and template for real-time monitoring of whether the accountable officers (BGN, BPOM, Bulog/Bapanas, Pertamina/MEMR) actually publish the verification artifacts named in the Remediation Verification Protocol (Oct 25) and the Unified Remediation Payoff (Oct 25), on the timeline keyed to the MSCI window.

The tracker is infrastructure, not analysis. It is written for the operational watch desk — MBG Watch, policy teams, portfolio managers — to use directly.

What the evidence supports

The 21 remediation rows with accountable officers and verification artifacts

The Unified Remediation Payoff (Oct 25) identifies 21 remediation rows across four channels, ordered by rupiah payoff per implementation month. The Pre-MSCI Decision Week piece (Oct 26) published the Artifact Feed Registry mapping each row to its public verification artifact, URL/endpoint, cadence, and format.

The table below consolidates both sources and adds the accountable officer, the conformance stage each artifact closes, and the re-scoring trigger from the Remediation Verification Protocol (Oct 25).

How to read the tracker: A row moves from Fail → Partial when its artifact publishes for the first time with machine-readable data covering ≥80% of the measured population. It moves Partial → Pass when three consecutive publications show zero critical findings. The watch desk tracks first publication date, coverage percentage, and critical-finding count per artifact.

Artifact feed methodology — how we monitor

The watch desk monitors nine primary feed families. Each has a primary endpoint, a fallback, and a machine-checkable health signal.

Ingestion pipeline (reference implementation):

• Language: Python 3.11+
• Scheduler: apscheduler — 5-min poll for real-time feeds; cron for daily/weekly/monthly
• Validation: jsonschema against published OpenAPI/JSON Schema per artifact
• Persistence: Append-only Parquet partitioned by date + artifact_id
• Alerting: Webhook to watch desk Slack/Telegram on: (a) schema validation failure, (b) timestamp staleness > 2× cadence, (c) coverage drop >10% vs. 7-day avg, (d) critical finding count > 0
• Dashboard: Static site (Next.js + Vercel) reading Parquet via DuckDB-WASM — zero backend

All feed endpoints are public, machine-readable, and require no authentication (API keys are public/rate-limited only). Any endpoint that gates behind login or paywall is treated as artifact absent until public access is restored.

"Artifact present" vs "artifact absent" — machine-checkable criteria

For each of the 21 rows, the tracker applies a uniform three-gate test. A row is artifact present only if all three gates pass.

Artifact absent = any gate fails. The tracker logs which gate failed, the error detail, and the timestamp of failure.

Special cases:

• Per-adjustment artifacts (Rank 20): G2 uses "≤ 30 days since last adjustment announcement" — tracked via BPH Migas press release RSS.
• On-publication artifacts (Rank 5, 13, 20 PDF): G2 = "published within 7 days of effective date per regulation"; G3 = PDF text-extracted fields match schema (uses pdfplumber + regex).
• Real-time feeds (Ranks 1, 8, 18, 21): G2 = "WebSocket/message received in last 5 min"; G3 = "message schema valid + required fields present".

Conformance re-scoring triggers — linking artifacts to the Protocol

The Remediation Verification Protocol (Oct 25) defines a re-scoring table for the MBG channel. The same logic extends to all four channels: any stage moving Fail → Partial or Partial → Pass triggers a re-score with a published artifact.

The tracker maps each artifact arrival to its conformance stage re-score:

Re-score output: The watch desk runs the six-stage conformance test (from the Measurement-Chain Standard, Oct 19) against the updated artifact state. The output is a channel conformance scorecard (6 stages × Pass/Partial/Fail) published to https://verifikasi.mbg.go.id/scorecard (or channel equivalent) per MSCI window.

Decision matrix update mechanism — how artifacts shift the three MSCI scenarios

The Pre-MSCI Decision Week piece (Oct 26) defined three MSCI announcement scenarios and a decision matrix for Households, Firms, and Policymakers. The tracker adds a dynamic update mechanism: each artifact arrival shifts the observable signposts that define the scenarios.

The three scenarios (recap)

Artifact-driven scenario shift rules

The tracker maintains a live artifact count A(t) = number of rows with artifact present (all three gates passing) at time t.

Critical-path artifacts (highest payoff/month, from Unified Remediation Payoff) carry 2× weight in the count:

• Row 1 (MBG Traceability API) — Rank 1
• Row 2 (LKPP Correction Tracker) — Rank 2
• Row 8 (BPOM API Traceability) — Rank 8
• Row 14 (BULOG Buffer Stock API) — Rank 14
• Row 18 (SKK Migas Cargo Tracker) — Rank 18

Weighted artifact count W(t) = A(t) + critical_path_present(t). The scenario thresholds above apply to W(t).

Forced-selling cascade thresholds (from MSCI Forced-Selling Mechanics, Oct 24)

The tracker monitors six market metrics in parallel. Two or more cascade thresholds breached simultaneously = cascade signal.

The tracker correlates artifact count W(t) with cascade thresholds: if W(t) ≥ 14 AND zero cascade thresholds breached → Status Quo path credible. If W(t) ≤ 6 AND ≥2 cascade thresholds breached → Cascade path active.

What the evidence does not support

• That MSCI will announce a final decision on November 11. The consultation runs to June 2027. November 11 is an interim update window only. An interim update could be: "feedback received, review ongoing" — which is neither Reclassify nor Defer nor No Change in final terms.
• That forced selling executes in a single session. The $4–6B capacity is mechanical capacity; actual flow depends on fund rebalancing calendars, liquidity provision, and Bank Indonesia's SBN purchase program. The cascade thresholds are signposts, not predictions.
• That the 120–180 bps premium is purely MSCI-driven. The premium compounds: MSCI consultation + Q3 portfolio outflows ($6.2B YTD) + oil >$108/bbl + El Niño haze risk + 2027 budget stress. Attribution is necessarily partial.
• That remediation artifacts will appear on schedule. The Remediation Verification Protocol (Oct 25) sets a timeline keyed to the MSCI window, but publication depends on ministerial sign-off, data readiness, and the Verification Cell's independence — none guaranteed.
• That the tracker's ingestion pipeline will capture every artifact without gaps. Public endpoints change, schemas drift, agencies delay. The tracker is a best-effort monitoring infrastructure; its output is evidence, not guarantee.

The least-harm path

The least-harm path for the rupiah through the November catalyst is artifact-led compression:

1. Artifacts publish → Verification Cell independence secured (≤3 ministry appointees) → MBG channel remediation unlocks → Rank 1–7 payoff realizable
2. BI maintains absorption capacity → SBN purchase program calibrated to $200–300M/day → forced selling absorbed without reserve depletion
3. Firms and households act on signposts, not fear → hedging and substitution decisions tied to observable thresholds → no panic-driven capital flight
4. MSCI interim update is "Defer" → consultation continues → June 2027 final decision with remediation evidence on record

This path requires no heroics — only that the artifact feeds publish, the Verification Cell independence gate passes, and BI's market operations function as designed. Each element is observable, verifiable, and within the control of Indonesian institutions.

The tracker makes that observability operational.

What I'm uncertain about (ranked by materiality)

1. BGN/LKPP publication commitment. The Protocol is operationally complete; its activation is a political decision. No technical barrier exists. The risk premium remains until publication begins.
2. Anomaly model retraining timeline (Row 2). The <3% false-positive target for procurement requires 4–6 weeks of live traffic for calibration. The 21-day trigger window assumes the model is pre-trained on 24-month history and only needs live calibration. If pre-training is incomplete, Row 2 slips 2–3 weeks.
3. Cross-channel auditor independence. The same KAP firm auditing MBG cold-chain sensors cannot audit Pertamina refinery throughput without conflict-of-interest waivers. The Protocol assumes separate auditors per channel with a shared methodology standard.
4. MSCI decision timing. The "mid-November" window is based on MSCI's published consultation calendar (Oct consultation, Nov decision). A delay to December shifts all timelines but does not change the tracker logic.
5. Feed endpoint stability. Several endpoints (e.g., verifikasi.mbg.go.id, audit.mbg.go.id, kejagung.go.id/mbg/cases) are notional — announced but not yet live. The tracker treats notional endpoints as "artifact absent" until they resolve to HTTP 200 + valid schema.
6. Data quality in early publications. First publications often have coverage gaps, schema drift, or stale timestamps. The tracker's three-gate test (schema, freshness, completeness) is designed to flag these honestly rather than count them as "present."

Appendix: Quick-reference card for the watch desk

Daily checklist (T-14 to T+5)

Weekly checklist

Per-MSCI-window checklist

Tracking links (live)

• Artifact Feed Dashboard: https://artifact-tracker.rupiah-stability-watch.myaga.org (to be stood up by Nov 1)
• Conformance Scorecard: https://verifikasi.mbg.go.id/scorecard (per MSCI window)
• Cascade Threshold Monitor: https://cascade.rupiah-stability-watch.myaga.org (real-time)
• Decision Matrix Calculator: https://matrix.rupiah-stability-watch.myaga.org (interactive)

This tracker is a living document. The watch desk updates W(t), cascade thresholds, and scenario weights in real time. The methodology is fixed; the data moves. Next scheduled revision: T-7 (Nov 4) — pre-November review positioning window.