Building a counterparty screening policy in Kenya

A written policy turns ad hoc Google searches into something auditors and credit committees can actually follow.

Most teams already screen counterparties — they just have not written down how. One analyst checks courts, another skims news, a third trusts the registry printout. When audit asks for the file six months later, the story changes depending on who you ask.

A screening policy does not need to be fifty pages. It needs to answer four questions clearly: who gets screened, what sources are in scope, when screens refresh, and who decides escalations.

Define the population

Start with tiers. Tier 1 might be new borrowers, government-facing suppliers, and any counterparty above a payment threshold. Tier 2 covers renewals and low-value vendors. Tier 3 is watch-list only. Without tiers, every merchant onboarding gets the same three-day manual search — and high-risk files wait in the same queue as stationery suppliers.

Scope sources by category, not by heroics

List source categories your policy requires: courts, official gazette, procurement restrictions, sector regulators, adverse media, director networks. Avoid naming every portal in the policy itself — categories survive when websites change. The operational tool (internal checklist or API) maps categories to live feeds.

Set refresh cadence

  • Onboarding — full screen before first payment or contract signature.
  • Annual renewal — re-screen active Tier 1 relationships.
  • Event-driven — material adverse media or gazette notice triggers out-of-cycle review.

Point-in-time checks age badly. A clean screen in January does not cover a debarment published in March unless you refresh or monitor.

Document outcomes, not vibes

Each file should store: legal name searched, match confidence, flags returned, source citations, analyst decision, and approver sign-off. “Looks fine” is not a control. PDF reports with embedded citations beat email threads when regulators visit.

Start small, automate the repeat work

Pilot the policy on one portfolio — SME lending, NGO vendors, or procurement panel — before rolling enterprise-wide. Automate the repeatable fetch; reserve humans for interpreting material flags and legal escalation.

Template your workflow. Compare manual research vs API screening on our comparison page, or see how BRIA assembles cited reports.

← All insights