Seven Months in the Shadow of Hurricane Melissa: A Complete Analysis of Jamaica’s ATM Network Performance, December 2024 to April 2026

• Hurricane Melissa triggered seven consecutive months of BOJ uptime non-compliance.
• Scotia Bank permanently lost 11 machines and is now refreshing 137 more.
• Rural uptime recovery first achieved in April 2026, six months post-storm.
• JMMB’s 17.1-hour February 2026 recovery time was the series’ sharpest anomaly.
• JN Bank showed persistent multi-region uptime deficiency throughout the series.
• St James recorded a 58.3% uptime low — the worst parish reading in the dataset.

Seventeen months of Bank of Jamaica ABM performance data, from December 2024 through April 2026, are really two stories compressed into one dataset. The first is a story of a network operating broadly within regulatory norms, with most institutions meeting the BOJ’s minimum standards of 90 per cent operational and 95 per cent uptime across three geographic regions. The second story begins in October 2025, when Hurricane Melissa made landfall and transformed the ABM bulletin from a routine performance scorecard into a chronicle of infrastructure crisis, emergency response, and the slow, uneven work of recovery.

This analysis traces both stories across the complete series, identifies the institutions and parishes that defined the extremes of performance, and assesses what the dataset reveals about the structural resilience — and vulnerabilities — of Jamaica’s cash access infrastructure.

Before the Storm: December 2024 Through September 2025

The ten months from the December 2024 bulletin through to September 2025 established the pre-storm baseline. Jamaica’s ABM network during this period comprised approximately 900 or more machines operated by a range of commercial banks and financial institutions, spanning the Kingston Metropolitan Area, other urban centres, and rural parishes. System-wide uptime during this period was broadly above the BOJ’s 95 per cent minimum standard, with occasional institutional dips that attracted data-note commentary but no sustained systemic failures.

The pre-storm data established some important patterns that would become more legible in retrospect. Scotia Bank, as Jamaica’s largest ABM operator, consistently contributed the biggest individual fleet to the national total. NCB, JN Bank, JMMB, Sagicor, and the smaller financial institutions each maintained distinct performance profiles: some consistently above minimum thresholds, some with occasional institutional dips in specific regions, and some — notably JN Bank — with a pattern of uptime volatility that predated the storm and would be exacerbated by it.

The rural-urban performance gap was also visible in the pre-storm data. Rural ABM uptime was typically the most variable of the three regional categories, reflecting the greater vulnerability of rural sites to power fluctuations, internet connectivity issues, and the longer response times that characterise maintenance work outside of the main urban centres. That variability would be dramatically amplified after October 2025.

The Watershed: Hurricane Melissa and October 2025

The October 2025 bulletin marked the point of rupture. Hurricane Melissa made landfall on Jamaica and inflicted catastrophic damage on the island’s telecommunications infrastructure — the connectivity backbone on which ATMs depend for transaction processing, monitoring, and fault alerting. The immediate consequences for the ABM network were severe: system-wide uptime fell below the BOJ’s 95 per cent minimum for the first time in the modern bulletin series, and the recovery time metrics that indicate how quickly out-of-service machines are returned to operation surged well above normal levels in affected regions.

For the ABM network, the storm created three distinct categories of damage. The first was physical destruction: machines flood-damaged, structurally compromised by wind, or rendered inoperable by direct site damage. The second was connectivity loss: machines physically intact but unable to communicate with transaction processing networks due to downed fibre, destroyed cell towers, or power outages affecting the broader telecoms grid. The third category — the most insidious — was intermittent connectivity: machines that appeared to function some of the time but experienced repeated dropouts that dragged uptime percentages far below the operational percentage, creating the confusing statistical picture that would characterise the bulletin data for months to come.

Permanent Losses: Scotia Bank’s Eleven Machines

The December 2025 bulletin brought confirmation of what had been feared since the storm: Scotia Bank had permanently lost eleven automated banking machines. These were not machines awaiting repair or pending replacement — they were write-offs, machines too severely damaged by flood water, structural collapse, or sustained electrical fault to be economically restored. The network total fell from its pre-Melissa high into the high 800s, a reduction that reflected the permanent decommissioning of infrastructure that had served Jamaican communities for years.

The loss of eleven machines sounds like a modest number against a total fleet of nearly 900. But in specific communities — where a single machine may represent the only convenient cash access point for hundreds of residents — each permanent loss had an outsized local impact. For the bank itself, the eleven write-offs contributed to a broader strategic reassessment that would eventually manifest in the announcement that gave the April 2026 bulletin its headline story.

The Long Recovery: November 2025 Through March 2026

The five bulletins from November 2025 through March 2026 chronicled a recovery that was real but painfully slow. Month after month, system-wide uptime remained below the BOJ’s 95 per cent minimum, making the post-Melissa period the longest sustained compliance breach in the history of the bulletin series. The operational percentage — the proportion of machines physically available — recovered more quickly than uptime, because operational machines can be maintained and restocked even when their connectivity is intermittent. But uptime, which measures the proportion of time that operational machines are actually functioning and transacting, remained suppressed by the underlying telecom damage that emergency repairs alone could not fully address.

Several data points from this period warrant specific attention. The February 2026 bulletin contained the series’ most alarming single institutional reading: JMMB’s rural machines registered an average recovery time of 17.1 hours, against a national average of approximately 1.7 hours for the same month. A 17-hour average recovery time means that when a JMMB rural machine failed, it remained out of service for the better part of a working day before being restored — a scenario that points either to a concentrated fault event affecting multiple machines simultaneously, or to a severe shortage of available technicians in the affected rural area. The figure resolved sharply in March 2026 to 5.7 hours, consistent with a one-off event rather than a structural dysfunction, but the February reading remains the series’ starkest illustration of how quickly rural cash access can deteriorate when response infrastructure is stretched.

Persistent Underperformers: JN Bank Across the Series

Among the institutions whose performance data runs consistently through the series, JN Bank stands out as the most persistently volatile underperformer on uptime metrics. Across multiple bulletins in the post-Melissa period, JN Bank recorded uptime below the 95 per cent minimum in more than one regional category simultaneously — and in the April 2026 bulletin, the bank recorded below-minimum uptime in all three regions at once: KMA at 85.9 per cent, other urban at 91.7 per cent, and rural at 88.2 per cent.

The persistence of JN Bank’s underperformance across geographically distinct regions points to something beyond the localised telecom damage that Melissa inflicted. A bank whose KMA machines — located in Jamaica’s most densely connected urban environment — are also failing to meet uptime standards faces challenges that cannot be attributed solely to rural fibre outages or remote-site power failures. The data suggests structural factors: possibly ageing hardware, possibly insufficient maintenance scheduling, possibly a connectivity architecture that did not build in adequate redundancy before the storm exposed its vulnerabilities. Whatever the cause, the full series has documented a multi-month, multi-region pattern that the BOJ’s publication mechanism makes impossible to ignore.

The Fleet Renewal: Scotia Bank’s 137-Machine Programme

The dominant narrative of the April 2026 bulletin was Scotia Bank’s disclosure of a 137-machine fleet refresh programme — the replacement of 137 older machines with entirely new models, running in parallel with the bank’s continuing recovery from Melissa’s telecom damage. In the context of the full series, the announcement is best understood not as an emergency response but as a strategic decision that the emergency made inevitable.

When a major natural event exposes the age and vulnerability of existing infrastructure, it creates an opportunity for operators to choose between repair and replacement. For machines at the end of their useful life — already requiring more frequent maintenance, already carrying ageing connectivity hardware that makes them more vulnerable to network disruptions — the economics of repair may not be compelling when set against the cost of replacement with newer, more reliable models. Scotia Bank’s 137-machine programme represents a frank acknowledgement that a significant fraction of its fleet had reached that crossover point.

The programme is the largest single fleet renewal disclosed in the history of the bulletin series. Its effects will be visible in the monthly data over the second half of 2026: as new machines come online, Scotia’s operational and uptime percentages should improve, and if the new hardware includes improved connectivity options — dual SIM cards, satellite backup, enhanced power management — the bank’s resilience to future weather events will also be substantially enhanced.

The Parish Picture: From Perfection to Crisis

The parish-level data embedded within each bulletin provides a granular view of geographic performance that the regional aggregates can obscure. Read across the full series, the parish data reveals a Jamaica in which cash access reliability varies enormously depending on where you live — a variation that widened dramatically after Hurricane Melissa and has only partially normalised by the end of the series.

The nadir came in the April 2026 bulletin, when St James — Jamaica’s second tourism hub and a parish with significant cash transaction volumes driven by hospitality sector employment — recorded a single-institution uptime figure of 58.3 per cent. In a parish that handles significant tourist withdrawal activity and daily wage transactions for workers in hotels, restaurants, and attractions, a machine that is out of service more than four times in ten represents a genuine economic impediment. The 58.3 per cent reading is the lowest parish-level uptime recorded anywhere in the seventeen-month series and serves as a sobering reminder that the post-Melissa recovery is still, in specific locations, far from complete.

At the other extreme, Sagicor delivered 100 per cent uptime in the rural category in April 2026, with an average recovery time of 0.0 hours. That reading — achieved in a month when the overall network was still running below minimum standards and JN Bank was reporting 88.2 per cent rural uptime — demonstrates conclusively that the infrastructure challenges of the post-Melissa environment are not a barrier to compliance for institutions that have invested in the right combination of hardware, connectivity redundancy, and maintenance protocols.

Rural Recovery: The Breakthrough in April 2026

The most significant positive data point in the post-Melissa period arrived in the April 2026 bulletin: rural uptime crossed above the 95 per cent minimum for the first time since the storm struck, reaching 95.2 per cent. For six consecutive months following Melissa, rural Jamaica had been below threshold — meaning rural communities bore the longest and most sustained service deficit of any geographic category in the series.

The April breakthrough did not arrive as a single dramatic improvement but as the cumulative result of incremental telecom restoration, machine decommissioning of the most chronically troubled units, and the gradual return of connectivity to communities where fibre routes had been severed by the storm. It does not erase the six-month service gap that preceded it, nor does it signal that the recovery is complete — system-wide uptime remained at 93.6 per cent in April, a seventh consecutive month below the 95 per cent national minimum. But it represents a turning point in the narrative of a rural cash access crisis that had persisted through one of the most difficult half-years in the modern history of Jamaica’s banking infrastructure.

What the Series Reveals About Infrastructure Resilience

Taken as a whole, the seventeen-month bulletin series offers a case study in infrastructure resilience that has implications well beyond the ATM sector. Jamaica’s experience with Melissa illustrates the cascading vulnerability that emerges when physical infrastructure (the machines), connectivity infrastructure (the telecoms network), and power infrastructure (grid and backup power) are simultaneously stressed by a single weather event. No single element of the system caused the seven-month compliance breach. It was the combination — and the fact that the telecoms recovery timelines were outside the control of the banks themselves — that made the post-Melissa period so difficult to resolve quickly.

The series also illuminates the importance of institution-specific choices that predate any disaster. Banks that had invested in connectivity redundancy, hardware modernisation, and proactive maintenance protocols before Melissa struck — institutions like Sagicor, which recorded strong performance throughout the post-storm period — were better positioned to maintain service continuity than institutions that had relied on the ambient infrastructure without building in adequate backup. The 137 machines that Scotia Bank is replacing were not all destroyed by the storm; some simply revealed, under the stress of Melissa’s aftermath, that they were no longer adequate for the demands being placed on them.

Looking to 2026 and Beyond

The May 2026 bulletin, when published, will be the first to capture the early contributions of Scotia Bank’s replacement programme. If the fleet refresh is as comprehensive and well-executed as its scale suggests, the improvement in Scotia’s metrics should be visible and could contribute to the first system-wide uptime reading above 95 per cent since September 2025. The institutions that have been most persistently deficient — JN Bank across all regions, NCB and JMMB in urban categories — will need to demonstrate improvement that goes beyond Melissa recovery to suggest their underlying maintenance and connectivity frameworks are being addressed.

For Jamaica’s banking regulator, the complete seventeen-month series provides an unprecedented evidence base for evaluating the resilience of the island’s cash access infrastructure and the adequacy of existing minimum standards in a climate environment that is producing more frequent and more severe weather events. The question of whether 95 per cent uptime — a standard calibrated for normal operating conditions — remains the appropriate benchmark in an era of intensifying Atlantic hurricane seasons is one the series raises without answering. It will be for the BOJ and Jamaica’s banking community to address that question before the next Melissa arrives.

Follow Jamaica Homes on Youtube @jamaicahomes and Instagram @jamaica_homes and on Facebook @jamaicahomes Send us a message or email us at onlinefeedback@jamaica-homes.com or editor@jamaica-homes.com