A dense, quantified and sourced document answering a single question: is this market worth your time and capital. Demand structure, incumbent players, real access conditions, documented scenarios.
Not a fifteen-page summary copied from public reports. A working document, with the figures, the sources and the assumptions set out in plain sight.
We are opening continents one at a time, starting with Africa. Each study is built from the ground up before going on sale.
The cyclical peak has passed, order books have never been fuller, and three groups alone carry the equivalent of 87% of the sector's annual output. Where the openings are, and on what terms.
Households borrow half as much again, developers do not borrow at all, and arrears have doubled. This market is held back not by demand but by inventory and by the withdrawal of funding from the supply side.
The largest construction market in French-speaking West Africa, and one of the slowest growing. Its growth comes neither from bank credit nor from solvent demand: it comes from public procurement.
The flows are Moroccan, the means are not. Tanger Med handles more containers than every national ports agency port combined, yet 93.6% of them merely pass through. What the statistics measure, and the 29 places where they mislead.
Interested in a country or sector not yet covered? Tell us, it weighs on our priorities.
Suggest a study →The same standard as our monthly analyses: sources you can cite, series you can check, projections whose assumptions we own.
National statistical institutes, central banks, sector ministries, international organisations. Every figure is tied to its source and date, in the appendix.
Long series are rarely homogeneous: base changes, classification breaks, differing frequencies. We harmonise them before reasoning on them, and we say so when a series is fragile.
A central path, bounded by a high and a low scenario. The assumptions behind each are written out one by one, so you can challenge them if you disagree. Where the public series are too short to support a statistical projection, we say so and work through documented scenarios rather than display borrowed precision.