Why It Is Not Done
and Whether It Should Be The banking sector earns more when the exchange rate spreads further between institutional and retail clients. The corporate client gets the best rate. The household pays the residual. Four policy instruments exist to narrow that gap. None has been used. This note explains why -- and what it would take to change that.
The foreign exchange market in Mauritius was liberalised in 1994 under Article VIII of the IMF Articles of Agreement. From that point, commercial banks were permitted to set their own FX spreads above and below the interbank reference rate published by the Bank of Mauritius. The spread offered to any given client is a function of that client's transaction volume, relationship value and bargaining power. A conglomerate processing USD 500 million in annual FX transactions receives a spread that may be 0.5 to 1.5 percent better than the rate offered to a household converting Rs 50,000 for an overseas payment. The difference is commercially rational. The bank earns more per transaction from the retail client than from the corporate one, but the corporate volume compensates. Both parties transact. Both are satisfied with the arrangement. The household that cannot negotiate is not a party to the satisfaction. It simply pays.
The practice is legal in Mauritius and in every liberalised economy in the world. No Mauritian statute requires uniform FX pricing across client categories. The Bank of Mauritius Act empowers the central bank to set the reference rate and to regulate exchange dealers, but it does not mandate a uniform spread. The Banking Act requires fair treatment of customers but does not define fairness in terms of rate parity between corporate and retail clients. The Financial Services Commission oversees capital market conduct but FX spot transactions fall outside its primary mandate. The regulatory framework as it stands contains no instrument that would require a bank to offer the same FX rate to a garment factory worker converting rupees for a family remittance as it offers to the conglomerate importing that factory's raw materials.
Whether it should remain legal is a different question. The argument for the current arrangement is that differential pricing reflects differential cost of service: large transactions are cheaper to process per dollar than small ones, and the relationship infrastructure supporting a major corporate FX client is more expensive to maintain than a retail desk. The argument against is that the differential does not merely reflect cost of service. It reflects structural market power: the conglomerate can threaten to move its FX flow to a competing institution; the household cannot. The rate the household receives is therefore not the outcome of a negotiation. It is the outcome of an absence of negotiating capacity. In markets where one party has no alternative, the price is not determined by supply and demand. It is determined by the tolerance of the party with no choice.
The household converting rupees for a remittance has no leverage, no alternative and no negotiating capacity. The rate it receives is not a market outcome. It is a measure of how much can be extracted from a party with no choice.
The Meridian · Policy Note · April 2026Four policy instruments are available to the Mauritian government and the Bank of Mauritius. They are listed in ascending order of structural impact and descending order of political feasibility.
The most recent large-scale attempt to impose a unified FX rate in an otherwise liberalised emerging market economy was Nigeria's multiple exchange rate unification in 2023. The Central Bank of Nigeria collapsed its official, investor and exporter windows into a single rate in June of that year, ending a system in which different actors accessed different official rates. The immediate consequence was a sharp naira depreciation of approximately 40 percent as the artificially suppressed official rate converged toward the market rate. Inflation accelerated. The parallel market, which had operated throughout the multiple rate period, did not disappear -- it adjusted to the new differential between the unified official rate and the rate at which dollar holders would actually transact outside the official system. The unification corrected one distortion and produced several others. Nigeria is a larger and more complex economy than Mauritius, and the parallel applies imperfectly. But the directional lesson is clear: in any economy where FX is scarce and the incentive to arbitrage is structural, imposing a single rate does not eliminate the spread. It moves it underground.
Mauritius's adoption of IMF Article VIII in September 1994 -- the full liberalisation of current account transactions -- was a foundational commitment that enabled the island's subsequent development as an international financial centre. The offshore banking sector, the global business company framework and the DTAA treaty network that now contribute significantly to fiscal revenue all depend on the credibility of that liberalisation commitment. Any policy that a rating agency or the IMF reads as a reversal of Article VIII -- including a mandatory uniform FX rate -- would immediately be assessed against the island's sovereign credit profile. At a moment when public debt is at 86.5 percent of GDP and the fiscal deficit is at 9.3 percent of GDP, a credit downgrade would raise the cost of sovereign borrowing at precisely the point when Mauritius has least capacity to absorb it.
The policy note does not end at the limit of the politically feasible. It ends at the limit of the structurally necessary and states clearly where the gap between those two limits lies. The uniform FX rate is not achievable without producing worse outcomes than it corrects. The FX spread cap and the STC essential goods rate are achievable and would produce meaningful household relief. The transparency requirement is immediately achievable and should be implemented without delay. The honest answer to the question of why none of these have been implemented is not technical. It is political.
The Bank of Mauritius has the regulatory authority to require FX spread transparency through a directive to all licensed exchange dealers. No parliamentary amendment is required. The directive could be issued in the next monetary policy cycle. It has not been issued because the banking sector's institutional relationship with the government -- as the primary underwriter of sovereign debt, the processor of state payroll and the dominant infrastructure of the financial system -- gives it leverage over any reform that reduces its non-interest income. MCB earned Rs 12.8 billion in non-interest income in FY2025. A meaningful portion of that income is FX spread revenue. The bank that funds the government's deficit is also the bank whose spread income the government would need to regulate. That is not a coincidence. It is the structural definition of the political obstacle.
Is differential FX pricing legal? Yes. Universally so in liberalised economies.
Should it be regulated more tightly? Yes. Spread transparency and a cap on the institutional-retail differential are both technically feasible and socially justified.
Can a single uniform rate be imposed? No. It would produce a parallel market, a credit downgrade and capital flight in the current fiscal environment.
Why has nothing been done? Because the institution that would be regulated funds the deficit of the government that would regulate it. That structural dependency is the most precise available definition of capture.
The foreign exchange market in Mauritius is not broken in the technical sense. Transactions clear. The rate is published. The system functions. What it does not do is distribute the cost of currency risk equitably across the population that bears it. The household that earns in rupees, buys in rupees and converts to foreign currency only occasionally pays the widest spread in the market. The conglomerate that earns partly in foreign currency, imports at scale and has institutional banking relationships pays the narrowest. The difference between those two spreads is not a fee for service. It is a structural transfer from the population with no negotiating power to the institutions with maximum negotiating power. It happens every day, on every retail FX transaction, invisibly and legally.
The corrective does not require a uniform rate. It requires a transparency directive that the Bank of Mauritius can issue tomorrow, a spread cap that the Banking Act can accommodate with amendment, and a political decision to treat the household's FX cost as a public policy concern rather than a private market outcome. None of those things are technically difficult. All of them are politically costly. That gap between the technically possible and the politically delivered is not unique to Mauritius. It is the gap in which the Import Dependency Trap, the conglomerate profit extraction and the human capital drain all operate. The FX spread is not the cause of those conditions. It is one more expression of the same underlying arrangement: those with scale capture the gain, and those without it absorb the cost.
This policy note accompanies the April 2026 trilogy: The Egg Mauritius Never Grew (Import Dependency Trap), Who Booked the Profit (conglomerate extraction mechanisms) and Private Schools, Public Debt (human capital displacement). The four pieces form the most complete structural analysis of the Mauritian political economy published in a single edition.
April 2026 · Policy Note · themeridian.info