Case Study 3 of 4
LIBOR→SOFR (ARR) Regulatory Transition
Owning the benchmark-rate migration backlog for a global bank's commercial lending business across 3 geographies — with zero disruptions to a live loan book.
LoanIQ
ACBS
Tableau
Jira
TL;DR
- Every commercial loan referencing LIBOR had to move to an alternate reference rate before a hard regulatory deadline — in three geographies, on a live loan book, across LoanIQ and ACBS.
- The key call: build the full impact map before touching a release plan, and let deadline pressure shape the sequencing, never shrink the discovery.
- Result: zero disruptions across all three geographies and 20% fewer post-release issues.
1Problem & who it's for
LIBOR — the benchmark interest rate underpinning trillions of dollars in global lending — was being phased out after regulators determined it was vulnerable to manipulation and no longer reliably representative of real bank funding costs. Every commercial loan referencing LIBOR needed to move to an Alternate Reference Rate (SOFR in the U.S., with regional equivalents elsewhere) before a hard regulatory deadline.
This wasn't a feature launch with room for iteration — it was a live loan book that had to keep functioning correctly for every existing customer relationship while the rate mechanism underneath it changed. Any workflow, API, or downstream integration still assuming LIBOR after the cutover risked miscalculated payments or compliance breaches. The people depending on it: the bank's commercial borrowers, the lending operations teams servicing their loans, and a leadership team answerable to three regulators on three timelines.
2Evidence & inputs
- A full impact map: every loan product, servicing workflow, API dependency, integration and customer journey touching LIBOR anywhere in LoanIQ or ACBS, documented through BRDs and FRDs.
- Business-process and root-cause analysis, which surfaced legacy rate dependencies that weren't obvious from current documentation. Several integrations assumed LIBOR formatting conventions in ways that wouldn't have surfaced as a bug until the actual cutover.
- Each region's regulatory deadline and local lending-product nuances, gathered through cross-geography stakeholder alignment, which set how much runway each rollout had.
3Workflow, controls & guardrails
The bar wasn't "migrate the rate". It was "migrate the rate with zero disruption to a single active loan", and the controls were built around that bar:
- Impact map as the control list. Nothing was released until every item mapped to it had a planned change and a test.
- UAT depth set by the cost of failure, not by a standard coverage target: a live-loan defect is a customer's payment calculated wrong.
- Release readiness coordinated across business, compliance, operations and engineering for each geography.
- Real-time visibility: Tableau dashboards tracking transition progress, impacted-loan coverage and release readiness per geography, for leadership.
4Options considered
5Decision, trade-offs & non-goals
Decision criteria. Regulatory deadline first, customer impact second, technical convenience last.
Trade-offs accepted.
- A longer discovery phase before any release plan, under deadline pressure, in exchange for finding the dependencies nobody knew about.
- A slower rollout from deeper UAT than a typical release cycle, in exchange for the 20% reduction in post-release issues.
Non-goals.
- Product changes beyond the rate transition. Improvements spotted along the way went to the backlog for after the cutover, so the migration's scope stayed fixed.
6What I delivered
- BRDs and FRDs for the rate-transition changes across LoanIQ and ACBS
- The impact map of workflows, APIs, customer journeys and integration dependencies
- The product backlog and user stories for the regulatory platform changes, with backlog refinement
- UAT and release-readiness plans across business, compliance, operations and engineering
- Tableau dashboards for transition progress, impacted-loan coverage and release readiness
A diagram of the approach, not a product screen. The real artefacts are the client's and stay confidential.
7Outcome & metrics
- Zero disruptions across 3 geographies during the LIBOR-to-ARR transition
- 20% reduction in post-release issues through rigorous UAT planning
Figures as recorded at the time. Client-identifying detail is withheld under NDA.
8Retrospective
The single decision that most reduced risk was building the impact map before touching a release plan. On a deadline-driven regulatory programme it's tempting to start migrating the loudest, most obvious dependencies first, but the legacy assumptions that only surfaced through root-cause analysis were exactly the ones that would have caused a live-loan incident if found late.
The broader lesson for any hard-deadline regulatory mandate: the deadline should shape sequencing, not shrink the discovery phase that finds what you don't yet know you're missing.