On economic complexity and industrial policy.

Industrial capital in Kenya is conventionally measured in throughput — units produced, revenue generated, jobs created. These are legitimate metrics. They are also the wrong ones for judging whether a given industrial position shifts the country’s long-term productive frontier. The better metric, borrowed from Hidalgo and Hausmann’s work on economic complexity, is whether the position raises the range of things the economy can produce at all.

An assembly line that packages cheaper versions of imported goods displaces imports. It does not raise Kenya’s complexity. A precision-manufacturing position that produces components the country has never produced — and that require tooling, skills, and supplier networks the country does not yet possess — raises it by one product, and that one product creates optionality on a dozen adjacent products over the following decade.

The matrix is directional, not quantitative. The intent is not to rank positions on a score but to separate positions that will compound the country’s productive base from positions that will not. Positions in the Compound quadrant — where the required capability is adjacent to what already exists, and the strategic adjacency opens further doors — are the positions where industrial capital compounds at rates standard IRR frameworks cannot observe.

The Leap quadrant — strategic adjacency without existing capability — is where well-intentioned industrial ambition consistently fails. Without supplier networks, tooling, and operator skill in adjacent domains, the leap cannot be financed, built, or sustained. The capital may close; the capability will not compound.

The firm underwrites industrial positions in the top-right quadrant. These positions are hard to find, expensive to build, and systematically under-priced by capital that measures only throughput. Their contribution is not the revenue they generate in year five. It is the adjacent capabilities they make financeable in year ten.