The October unrest has a national price tag. It is not any firm's exposure.

On 16 October, Business Mauritius put the activity losses from the week's riots and curfews at Rs 270–470 million.

It is a careful number. It is not one any firm can use.

It measures a modelled 20–35% fall in direct value added across the economy on the most affected days. It says nothing about what a particular shop, warehouse or hotel stands to lose next time.

So we asked a narrower question. What would a firm need to know about itself to judge how exposed it is to civil unrest in Mauritius, and can it find that out before the next episode? We traced the public record to 19 October: Cabinet decisions and curfew orders, the Civil Code, insurers' own product pages, company filings from comparable episodes in South Africa, New Caledonia and the Pacific, and the research on how businesses recover from disruption.

The first answer sits in the Civil Code. Article 1983-46 relieves the insurer of liability for losses caused by riots or popular movements “sauf convention contraire”: unless the contract provides otherwise. In Mauritius, riot cover is an exception that has to be bought. On every Mauritian insurer page we could read, explicit riot cover appeared only on motor products. That is a finding about those pages, not about every policy written. After the 1999 riots, two businesses that sued the State over their uninsured losses lost, twelve to fifteen years after the event.

Insurance that does respond pays slowly. In New Caledonia, about 14% of expected indemnities had been paid six months after the 2024 riots, with business interruption the slowest. In South Africa, one listed retailer had recognised 30% of its claim two months on. Liquidity is part of protection, not an alternative to it.

Not a counsel of despair

Most of what decided outcomes elsewhere can be measured in advance. Where losses were heavy, value was concentrated: the South African retailer with the largest net loss after insurance, about R650 million, was the one that lost two distribution centres, while those whose hubs stayed secure restocked quickly. Firms that could restart elsewhere traded again within weeks, even where rebuilding took years. And the early Mauritian signals to 19 October, with no wave of cancellations, travel advisories urging caution rather than advising against travel, and no fall in the rupee, are consistent with a short-lived national effect, if the unrest does not recur.

Two things surprised us

Our view, for what it is worth

The national estimate is being asked to do a job it was not built for. A board with exposure to Mauritius does not need a better economy-wide number. It needs a handful of facts about its own firm, and most of them sit in documents it already holds. We would start with four questions.

These are questions, not advice; the answers, and the decision, stay with the firm. Our study turns them into a scored diagnostic, but it is a screening instrument. No resilience score we reviewed has been validated against real outcomes, and Mauritian recovery evidence does not yet exist.

If the public record misses something you know, such as a published riot wording, a support scheme or early recovery figures, we would like to hear it.