Real-Data Calibration

All quantitative parameters in this prototype are now calibrated on three primary datasets supplied for the dissertation. Use the button at right to push the empirical parameter set into the live pricing / climate / rates engines.

Rainfall
REAL
Filerainfall_2000_to_2024.xlsx
ProvenanceWorld Bank CCKP / FAO
Observations25 years
Mean647 mm
Std. dev.30.3 mm
Drought freq.20%
κ_C2.00
σ_C2.00
c₀ (latest)-0.66
Maize Production
REAL
Filemaize_production_2000_to_2024.xlsx
ProvenanceUSDA / FAOSTAT
Observations24 years
Mean1.09 Mt
Std. dev.479 kt
Worst year2003 (500 kt)
ρ(z_rain, maize)-0.19
Slope (t / σ_rain)-94,540
RBZ T-Bill Auctions
REAL
Filetbills_auction_results_2000_to_2024.xlsx
ProvenanceReserve Bank of Zimbabwe
Auctions12
Monetary erasZWD · USD · ZWL · ZiG
r₀ (anchor)7.50%
κ_R0.80
σ_R500 bps
λ₀ (credit)8.3%
β_R0.020
Annual rainfall index (z-score)
Maize tonnage (Mt)
RBZ T-Bill yields by era
Rainfall z vs Maize (joint years)
Empirical Pearson ρ = -0.19 across 24 years.
Methodology footnote

User-uploaded: rainfall_2000_to_2024.xlsx (World Bank CCKP / FAO, annual mm, n=25)

User-uploaded: maize_production_2000_to_2024.xlsx (USDA/FAOSTAT, metric tonnes)

User-uploaded: tbills_auction_results_2000_to_2024.xlsx (RBZ auctions, mixed currency eras)

The short-rate r₀ is anchored on the last USD multi-currency T-Bill print (2023, 7.5%) to avoid contamination by ZWL/ZiG monetary-regime jumps. The κ_R/σ_R parameters use the same USD-era window. Sovereign hazard λ₀ is set from the historical default frequency (two restructurings in the 24-year sample), with rainfall-shortfall sensitivity β_R derived from the rainfall ↔ maize correlation.