Forget the glass towers in Nairobi and Kampala for a moment. Africa's largest retail asset manager does not have a licence, a logo, or an office. It meets on plastic chairs on a Sunday afternoon, passes a counting book...
Forget the glass towers in Nairobi and Kampala for a moment. Africa's largest retail asset manager does not have a licence, a logo, or an office. It meets on plastic chairs on a Sunday afternoon, passes a counting book around, and holds more of ordinary families' money than most fund managers will ever see.
In Kenya, the Kenya Association of Investment Groups has estimated around 300,000 chamas managing close to KSh 300 billion, and reporting collected by the UN's Africa Renewal put chama assets in the range of US$3.4 billion, with roughly one in three Kenyans belonging to a group (Africa Renewal; KAIG Chama Handbook). Across the border, Uganda's village savings and loan associations run on the same logic at village scale: groups of roughly 15 to 25 members saving small weekly amounts, lending to each other at agreed interest, and sharing out at the end of a cycle. CARE International, which helped formalise the VSLA model in 1991, counts more than 8 million members supported worldwide, most of them women (CARE International). A study of VSLAs in West Nile, Uganda found typical weekly contributions of under a dollar and median loans of around $20, and most groups running steadily for years (Texila Journal).
So the model works. Millions of families have paid school fees, bought iron sheets, and survived funerals because of it. And yet you probably know at least one group that died badly. Money unaccounted for. A treasurer who moved to another town. A member who left and demanded "her share" of a plot the group had bought together, and nobody could say what her share was. The group did not fail because members were dishonest. It failed because it outgrew its rules.
A merry-go-round is almost impossible to break. Everyone contributes, one person takes the pot, the cycle rotates, and at any moment the group's balance sheet is close to zero. There is nothing to steal, nothing to value, nothing to fight over. The design is its own protection.
The trouble starts the day the group succeeds. Someone proposes that instead of eating the money, you invest it. Buy a plot. Start a lending book. Put money in a money market fund. Now the group holds assets that grow, and everything that was safe about the merry-go-round becomes a trap. The balance is no longer zero. The question "what is my share worth?" now has a real answer, and nobody agreed on how to calculate it. The records that were fine for tracking KSh 2,000 contributions are not fine for tracking a plot worth two million.
Most chamas cross this line with the same handshake rules they started with. That is the collapse point. Not theft, usually. Ambiguity.
Three gaps do most of the killing:
No written constitution. The rules live in the chairperson's memory, and memory bends toward whoever is remembering. When a dispute comes, there is nothing to point to, so the dispute becomes personal.
No exit valuation rule. Nobody agreed, back when the money was small, what a leaving member is owed and when. So the first exit becomes a negotiation, the negotiation becomes a quarrel, and the quarrel takes the group down with it.
No record standard. One counting book, one person's handwriting, no second copy, no reconciliation. Even perfectly honest treasurers cannot prove they are honest, and a group where honesty cannot be proven eventually stops trusting itself.
The groups that survive the transition are not the ones with the richest members. They are the ones that wrote things down before things got serious. Seven rules, agreed early, carry most of the weight. Kenyan practice has increasingly pushed groups toward formal structure for exactly this reason, since an unregistered group with no documents has weak legal standing when things go wrong (Chamasoft).
1. A written constitution. One document, signed by every member, covering contributions, meetings, penalties, decision thresholds, and what happens on death, exit, and default. It does not need a lawyer to start. It needs to exist, and every member needs a copy. The test is simple: if your chairperson vanished tomorrow, could a stranger run the next meeting from the document alone?
2. A member register. Full names, national ID numbers, phone numbers, next of kin, date joined, and a running record of what each member has contributed. This sounds bureaucratic until a member dies and three relatives arrive claiming the balance. The register is the difference between a sad meeting and a court case.
3. An exit formula agreed before the money grows. Decide now, while the pot is still small enough that nobody is tempted to bend the answer: a leaving member receives their contributions plus their proportional share of realised gains, paid within an agreed window, with illiquid assets valued by an agreed method or bought out over time. The exact formula matters less than the timing. A formula agreed at KSh 50,000 is mathematics. The same conversation at KSh 5 million is war.
4. Dual signatories, always. No single person can move group money. Two signatures on the bank account, two people counting cash at every meeting, and mobile money held in an account the group can inspect, not in the treasurer's personal line. This is not an accusation against your treasurer. It is a gift to your treasurer, because it means she never has to defend herself alone.
5. An annual audit night. Once a year, the books are read aloud, line by line, contributions against records, bank statement against counting book, in front of every member. Make it an event. Cook food. The point is not to catch thieves; it is that a group that inspects itself annually rarely produces one.
6. A family-representation clause. When a member dies, the constitution should already say what happens: the named next of kin inherits the balance, or a nominated family member may take the seat with the group's approval, or the position is paid out per the exit formula. Choose any of these, but choose in writing, in advance. Groups without this clause do not just lose a member when death comes. They inherit a dispute.
7. Graduation criteria to a registered entity. Agree a threshold, in assets or ambition, at which the group formally registers: a self-help group registration, a cooperative, or a limited company holding the assets. In Kenya the shift of serious chamas into registered investment vehicles is exactly what turned kitchen-table savings into an asset class large enough for banks to chase (Africa Renewal). A plot of land should never sit in one member's name "for convenience". Convenience is how groups lose land.
Here is the larger reason to take this seriously, and it goes beyond the chama itself.
Everything that kills a chama also kills a family. Assets in one person's name for convenience. No written agreement about who gets what. No valuation method when someone wants out. No record anyone can check. One signatory. No plan for death. Read that list again and you have just read the autopsy of most family land disputes in East Africa.
Which means the chama is the cheapest governance school your family will ever attend. A member who has sat through an audit night knows what accountability feels like in the body, not just in theory. A member who helped write an exit formula understands why the family land needs a written succession plan before the parents die, not after. The habits transfer directly: the register becomes the family asset register, the audit night becomes the annual family council, the exit formula becomes the succession agreement.
You are not just protecting this year's contributions. You are rehearsing, at small stakes, the exact skills your family will need at large stakes. This is also why parents should want their grown children inside a well-run group before those children inherit anything. A daughter who has served as a chama treasurer under dual signatories and survived an audit night is a different heir from one who has only ever managed her own phone balance. The group teaches, cheaply and publicly, what families otherwise teach expensively and in court.
So here is the decision in front of you. Your group is meeting again soon, and the money is growing. You can keep running on goodwill and memory, and hope your group is the exception. Or you can stand up at the next meeting and propose the uncomfortable agenda: a written constitution, a register, an exit formula agreed now, two signatories, an audit night on the calendar, a death clause, and a registration threshold. It will feel like distrust for about an hour. It is the opposite. Decide before the next meeting which version of the group you belong to, because the money will not wait for you to choose.