Each year the fund splits its budget B across five critical-technology sectors using a preferential-attachment weighting controlled by the focus exponent f:
w_i = C_i^f / Σ_j C_j^f
investment_i = B · w_i
f = 0 spreads the budget evenly regardless of current strength (egalitarian); f = 3 concentrates it heavily on whichever sector is already strongest (winner-take-most), the same "rich-get-richer" dynamic seen in real technology-investment portfolios.
Capability C_i (0–100) grows with diminishing returns and decays from foreign competition and obsolescence:
ΔC_i = k·√(investment_i) − d·C_i (per year)
Each year, with the chosen shock probability, an export-control or supply-cutoff event strikes the weakest sector — the one a state is most import-dependent on — cutting its capability by 30%. This is why spreading investment too thin, or concentrating it too narrowly, both raise long-run risk.
Two headline metrics are tracked:
- Sovereignty Index — the average capability across all five sectors (0–100); a state that is strong everywhere scores high, one that is strong in only one sector does not.
- Concentration (HHI) — the Herfindahl–Hirschman Index of the allocation weights, Σw_i², expressed as a percentage from 20% (perfectly even across 5 sectors) to 100% (all-in on one). High concentration accelerates growth in the chosen sector but leaves the rest exposed to the next shock.
This 2D build renders the identical year-by-year model as the 3D version (same growth/decay constants, same preferential-attachment weighting, same weakest-sector shock rule) on a rotatable capability wheel plus scrolling history charts, instead of extruded 3D towers.
Real-world relevance: this mirrors the logic behind government instruments such as the EU's Sovereign Tech Fund, national semiconductor and AI-compute stockpiling programs, and CHIPS-Act-style industrial policy — capital is finite, growth is concave, and dependency risk is a direct consequence of the allocation strategy, not a separate variable.