DATA & METHODOLOGY

Every figure on these pages is computed from the open datasets listed below or quoted with a direct link. Nothing is typed in without a source; verdicts and ranks are computed from the snapshot, never written by hand.

Refreshed at least with every IMF WEO release (April and October).

Datasets

Derived figures (5-year averages, indexing to 100, scatter residuals) are computed by nurion.org from World Bank data (CC BY 4.0) and IMF data; modifications: averaging, indexing, regression.

Similar indicators differ across sources: the World Bank and the IMF define debt differently and cut data on different dates; national statistics are fresher but methodologically different. We name the source on every figure and never mix series.

Curated facts and their sources

Wages are converted to US dollars at the annual average official rate of the observation year; NSO methodologies differ — figures are indicative.

Corruption Perceptions Index scores are shown verbatim (CC BY-ND): no recalculation, excluded from ranks.

The sector-method thresholds are the authors’ editorial choices, not statistical estimates. Each is stated explicitly:

• -8 — A credit shortfall of 8+ pts of GDP is structural, not noise, on a peer spread of ~50 pts.
• 0.6 — Monetizing below 60% of Georgia’s achieved level marks a clear gap against a real, not hypothetical, benchmark.
• 0.5 — Under half of Armenia’s share — the ICT takeoff has not happened yet.
• 10 — A 10+ pt gap to the median is beyond Findex survey noise.
• 2 — Manufacturing 2+ pts below the median with surplus rural labor reads as under-industrialization.

• Credit/income scatter positioning: Private-sector credit vs ln(GDP per capita, PPP) across the 7 countries; Kyrgyzstan’s residual from the fitted line.
• Account-ownership gap: Kyrgyzstan’s Findex vs the peer median; threshold 10 pts.
• Insurance penetration: Premiums to GDP per the regulator; threshold 1% of GDP.
• Receipt per arrival: Receipts per arrival vs Georgia’s achieved level; threshold 0.6 of it.
• Tourism receipts level: Kyrgyzstan’s receipts vs Georgia’s in the same year.
• ICT share vs Armenia: ICT share of service exports vs Armenia; threshold 0.5 of its share.
• Labor cost: NSO average wage as a cost proxy (indicative; methodologies differ).
• Manufacturing gap: Manufacturing share vs the peer median; threshold 2 pts below.
• Rural labor pool: Agricultural employment above the peer median.
• Ministry estimate: Technical potential of 142 TWh/yr and the utilization share are ministry figures — cited, not charted.
• Logistics friction: LPI below the peer median (World Bank survey years).
• Trade throughput: Trade openness above the peer median.

The weakness register

• GDP per capita is second-lowest in the set: shown as convergence runway together with the 2000→today slope.
• Control of corruption and rule of law sit in the bottom quartile globally: the WGI panel is shown in full, no cherry-picking.
• Three changes of power (2005, 2010, 2020) are annotated directly on the GDP charts — the economy passed each without a banking crisis.
• FDI is low and volatile (the Kumtor dispute, settled 2022): that is the under-invested-market thesis itself.
• Exports are concentrated in gold — the concentration is the diversification map on the Sectors tab.
• The paper current-account deficit is distorted by unrecorded re-exports — the footnote is mandatory wherever it appears.
• The 2022–2024 growth is partly re-export-driven: the 10-year average is printed next to the spike.

Not shown due to data gaps

• agri_share ARM: 14 years
• fx_rate UZB: 14 years
• industry_share ARM: 14 years
• manufacturing_share ARM: 14 years
• manufacturing_share TJK: 14 years
• npl MNG: 10 years
• npl TJK: 12 years
• reserves_months UZB: 12 years
• reserves_usd UZB: 12 years
• services_share ARM: 14 years
• tariff_rate TJK: 10 years
• tariff_rate UZB: 10 years
• tourism_arrivals TJK: 13 years
• tourism_receipts UZB: 12 years

DATA & METHODOLOGY

Download the full dataset (JSON)