Most of the electoral risk analysis we are shown answers the wrong question. It works hard on who is likely to win, assigns probabilities, and stops. For a company with an operating asset or a permit application in the country, the identity of the winner is often the least actionable part of the picture.
What actually reaches the profit and loss account is procedural: the administrative freeze before the vote, the fiscal correction after it, and the personnel churn that follows both. Those are considerably more predictable than the result, and they run on a clock that starts long before the campaign.
Four phases, one calendar
Phase one: pre-campaign, roughly eighteen to six months out
Fiscal policy loosens. Public procurement accelerates and payment terms improve. Regulators become noticeably more accommodating on enforcement. Subsidy and tariff decisions that would be unpopular get deferred.
This is the most favourable window in the cycle for getting things approved, and it is routinely wasted. If you have a permit, licence renewal, land allocation or tariff application in the pipeline, this is when to push it.
Phase two: campaign, roughly six months to polling day
The administration slows to a stop. Officials with political futures avoid signing anything that could be characterised as a favour to a foreign company. Files do not get rejected; they simply do not move. Inter-ministerial coordination, always the slowest part of any approval, effectively ceases.
Plan for a freeze of six to nine months on anything requiring discretionary approval. Most timeline overruns we see in the region trace back to a company that budgeted for the election day and not for the quarter either side of it.
Phase three: result and transition
This is the phase that dominates the coverage and, for most commercial operators, it is short. The material questions are whether the result is contested, whether contestation moves from courts to streets, and whether it reaches the specific locations where you have people and assets. Security planning here should be geographically precise. National-level analysis is close to useless for deciding whether to keep a site open.
Phase four: post-election, the first twelve months
This is where the balance sheet damage usually happens, and where almost nobody is still paying attention. The fiscal loosening of phase one has to be paid for. Expect some combination of subsidy withdrawal, tariff and levy adjustment, tax enforcement campaigns, slower payment of public arrears, and review of contracts and licences awarded by the previous administration.
Add to that the personnel effect. A reshuffle replaces the officials your team spent three years building working relationships with. Institutional memory of your file resets. In a system where a great deal depends on a specific director-general knowing who you are, that is a genuine operational cost.
Indicators worth tracking
A monitoring brief on an election cycle should be short and should watch things that are observable. In our experience the following carry the most information:
- Budget execution rate against plan, particularly capital spending
- Central bank net foreign assets and any tightening of repatriation practice
- Public arrears to suppliers, which is often the earliest fiscal stress signal
- Disputes over electoral commission composition, which predict contestation better than polling
- Whether the opposition is consolidating into a single candidacy or fragmenting
- Timing and scope of the pre-election cabinet reshuffle
- Any move to amend constitutional term or eligibility provisions
What to do with it
The framework produces three concrete instructions rather than a probability. Front-load discretionary approvals into phase one. Build a nine-month administrative freeze into any timeline that crosses phases two and three. And provision for a fiscal correction and a relationship rebuild in phase four, regardless of who wins.
You do not need to know the winner to plan well. You need to know the calendar, and to take it as seriously as you take the financial one.