Every metric. Every source. Every cross-check.

Methodology

Three rules apply to every metric on the dashboard:

• Free sources only. No paid feeds in the displayed data layer. Paid sources may be cited as additional cross-references in the metric drawer.
• ≥ 2 sources per metric. One primary, at least one cross-check. Build fails if cross-check is missing (audit Gate 2).
• Disagreement protocol. Cross-check fetch runs within 5 min of primary. If they disagree by > 5%, metric flips to crosscheck_pending and renders the lower value with a warning badge.

Trend computation:

• MoM = today vs same period 30 days ago. Mandatory for every verified metric (audit Gate 2).
• YoY = today vs same period 365 days ago. Mandatory for every verified metric.
• vs 5Y avg = today vs trailing 5-year mean. Optional; populated as history accrues.
• 12-month sparkline = month-end values for the last 12 months. Empty during initial accrual; renders dashed line + "history pending."

Derived metrics (formula visible in metric drawer):

• india_risk_score = weighted of 6 driver scores (oil 25% · flows 20% · macro 20% · real 15% · freight 12% · sector 8%)
• absorption_ratio = max(DII MTD, 0) / |min(FII MTD, 0)|
• real_10y_yield = G-sec 10Y − CPI YoY
• ind_us_10y_spread = G-sec 10Y − FRED US 10Y
• Driver scores: weighted normalisations of underlying metrics; auto-shock if any input metric is in shock state

Update cadence — see Tech Architecture §04 for the full 9-slot schedule.

Source directory

Every source we draw from. Click to visit.