GDP per capita, nominal
$2,558
~$5,800 at purchasing power parity
Five-year band
real, local currency · % a year over five years
-3%+10%
India -2Indonesia -1.6Bangladesh +1.5Philippines +7.7 · excluded
-50+5+10+15
Seven members carry no value and sit at no point on this axis — see the roster for what each one states instead.
The span is read from the framework, not fitted to these marks. No per-analog weights are published, so the cohort cannot be weighted; the engine states the cohort's own span rather than deriving a narrower one from the three analogs that happen to carry a number. [ke-cohort-span] · as of method v0.4 · Jul 2026
Spot KES/USD 129.3 (May–Jun 2026) — a unit restatement, not a currency view. [ke-5]
Structural position
What an analog is matched on: income, urbanization, credit depth and mortgage depth — never geography.
Urban share
30%
growing 3.7% a year
Credit to GDP
~30%
the binding constraint in the method's fourth layer
Mortgage to GDP
1.9%
rate cuts transmit through developer finance, not mortgages
One sentence states all four, and it is ours~$2,558 GDP per capita nominal (~$5,800 PPP), 30% urban growing 3.7% a year, credit/GDP ~30%, mortgage/GDP 1.9%.
Analog cohort
10 members · 3 carry a number. Members that contribute no value stay on the roster with what they state instead.
- Bangladeshadmitted+1.5% a year
- Indiaadmitted−2% a year
- Indonesiaadmitted−1.6% a year
- Egyptshadowcautionary FX episode; nominal gains masking real losses, no annualized real outcome stated
- Ghanashadowcautionary FX episode; nominal gains masking real losses, no annualized real outcome stated
- MoroccoboundMorocco −1.4%
- Nigeriashadowcautionary FX episode; nominal gains masking real losses, no annualized real outcome stated
- Philippinesexcluded+77% real 2010–18, powered by a BPO office-jobs boom with no Kenyan equivalent.
- South Africaboundstill sits below its 2010 real level after fifteen years
- Vietnamexcludednot importable
Coverage is why three members are shadows — BIS residential property statistics cover only Morocco and South Africa in Africa, a severe survivorship bias that is foregrounded rather than smoothed over.
Evidence coverage
5 of 9 series have a release. A series with no release is marked wanted, not quietly dropped.
- Hass composite property indexHassConsult · quarterly · askingwanted
- KBA house price indexKenya Bankers Association · quarterly · valuationcurrent · 2026-Q1
- Monetary Policy Committee statementsCentral Bank of Kenya · irregularcurrent · 2026-06
- Nairobi metropolitan residential researchCytonn Investments · semiannual · researchwanted
- Residential property price indexKenya National Bureau of Statistics · quarterly · askingcurrent · 2026-Q1
- Consumer price indices and inflation ratesKenya National Bureau of Statistics · monthlycurrent · 2026-06
- Kenya Economic UpdateWorld Bank · semiannualcurrent · 2026-07
- World Economic OutlookInternational Monetary Fund · semiannualwanted
- African Economic OutlookAfrican Development Bank · annualwanted
What this edition does not publish
Absences are printed, never implied. Each one names the thing that would have to exist first.
No ordinal rank
Ranking is retained only where forecasting skill has been measured, and no back-test has run. A ranked list with no measured skill is fabricated credibility.
No weights
Weights come from a judgment panel that is not built. A weight is a falsifiable claim, and this edition makes none.
No worst-case percentile
The only worst-case figures in the corpus belong to a superseded run, so there is nothing honest to stress against. The published band is not a worst case.
No movement to report
First edition — there is no earlier edition to diff, so an empty delta is printed rather than implied. Movement reporting starts with the second edition.