Here is the answer to the question Part III ended on, and it is worth stating plainly because most readers, including many who work in African business, have never had it explained.
There is no market for Ugandan shillings against Rwandan francs.
Not a thin market. Not an expensive one. For practical purposes, at any commercial scale, none. No bank in Kampala quotes a live price in francs, and no bank in Kigali quotes shillings. The two currencies are not traded against each other because there is not enough two-way flow to make a market worth running, and there is not enough flow partly because there is no market. So when value has to move between two countries that share a border, a language of trade, a regional bloc and a customs union, it does not move between them. It goes out and comes back.
The mechanism is called correspondent banking, and it works like this. A bank in Kampala does not have a branch in Kigali, so it keeps money on deposit with a larger bank that has relationships in both places. That larger bank is usually not African. It is in London, New York, Frankfurt or Johannesburg. The Kampala bank holds an account there. So does a bank in Kigali. To send value between them, the shillings are sold for dollars, the dollars move between two accounts sitting on the books of a foreign institution, and the dollars are then sold for francs.
Take into consideration the cost of that. Two currency conversions, each involving a spread earned by whoever quotes it. Correspondent charges, and usually correspondent charges again from the second bank through which the first bank goes. Cut-off times in the time zone of a foreign country, which is the reason why payments initiated on a Friday afternoon in East Africa have to wait till Monday. And finally, a settlement chain where no one in the pipeline knows where the money really is.
Then it got worse, for reasons that had nothing to do with Africa.
After the financial crisis, and after a run of enormous penalties imposed on global banks for money-laundering and sanctions failures, compliance stopped being a cost centre and became an existential risk. The rational response for a large Western bank was to look at its correspondent relationships and ask which ones generated modest revenue while carrying jurisdictional risk that a regulator might one day punish. Small African banks scored badly on that test. So the relationships were closed, in numbers, across the 2010s. The industry term is de-risking, which is a striking phrase when you consider that the risk was not eliminated. It was moved onto the people at the end of the chain.
The consequences are measurable in the one place they are hardest to defend. Sub-Saharan Africa is the most expensive region on earth to send money to, and has been for as long as the World Bank has published the numbers. Corridors within the region are worse still, and some of the most expensive corridors in the world are between neighbouring African countries.
Thus the reserve management of the colony's currency board was carried out in London, where the interest on these reserves accrued. A hundred years later, the African banks maintain their settlement balances in London and New York, paying for the privilege. The means have changed. Geography has not.
This is the problem that everything in the next three sections is trying to solve.
