Version 1.1 · 18 November 2026

How the Mansa Markets Stablecoin Premium Index is measured

The index measures what one USDT stablecoin dollar costs in 24 African currencies, covering 36 countries, and the premium of that price over the official exchange rate. It is a measurement of observed prices. It does not predict, explain or judge them.

1. What is sampled

Every hour, at seven minutes past, the collector reads the live USDT order books for each currency on 3 major peer-to-peer marketplaces: both sides, the offers to buy USDT and the offers to sell it. Every offer seen is stored, whether or not it is used.

2. Which offers count

An offer is removed before any price is computed when its merchant has:

  • a completion rate under 90%,
  • fewer than 20 completed orders, or
  • less than $200 of USDT available.

Offers more than 25% away from the median of their book are also removed as bait prices.

3. From offers to one rate

  1. Remaining offers are sorted by price.
  2. Every run of at least three independent offers whose prices agree within 1.5% is a candidate group.
  3. Each candidate is widened to every offer within 1.5% of its median, and the group with the most dollars on offer wins: the price where the money actually is, not the cheapest one.
  4. The published rate is the volume-weighted average price of the group, weighting each offer by the dollars it makes available.
  5. If no such group exists, no rate is published. The page shows the price range instead. A wrong number is worse than no number.
  6. If a second group of offers more than 10% away holds at least a quarter of the winning group's depth, the result is flagged as split rather than averaged across.

This is done separately for each marketplace and each side of the book. The headline buy and sell rates are the median across the marketplaces that quote both sides, so one marketplace going wrong cannot move the answer on its own. The mid rate is the average of the two.

4. The premium

premium % = (USDT mid rate − official rate) ÷ official rate × 100

The mid rate is the average of the buy and sell rates. A positive premium means a stablecoin dollar costs more local currency than the official rate implies.

5. The official rate

Where a central bank publishes a usable daily rate that we collect directly, that is the official rate and the page labels it “Central bank rate”. Elsewhere it is a market reference rate for the same day, labelled “Market reference rate”. The label is shown on every country page.

6. Flags on a row

  • Crossed: buy quotes sit above sell quotes on that market. It is real market information, so it is published as quoted rather than hidden.
  • Thin: few offers agree, so the rate is indicative.
  • Delayed: the newest measurement is more than two hours old.
  • No rate: the offers did not agree within 1.5% (see step 3).

7. Daily fixings and history

Each currency gets one daily fixing per UTC day: open, high, low and close of the hourly mid rate, plus the day-end official rate and premium. Hourly data starts on 15 August 2026; daily fixings are backfilled from it. Per-marketplace hourly detail is kept for 90 days, then reduced to one reading per day. The hourly headline series is kept in full.

8. Coverage

24 currencies: ZAR, KES, EGP, MAD, MZN, GNF, MRU, BWP, XAF, UGX, RWF, TZS, NGN, GMD, LRD, BIF, DZD, GHS, MGA, MUR, XOF, TND, CDF, ZMW. The West African CFA franc (XOF) covers 8 countries and the Central African CFA franc (XAF) 6, so those countries share one rate. Currencies without enough peer-to-peer liquidity to measure are left out rather than estimated.

9. Dollar-adjusted stock returns

For each stock index whose currency the index covers, returns are shown three ways over the same dates: in local currency (index close to index close), in US dollars at the official rate, and in dollars at the stablecoin rate. Each dollar return divides the index by the day's fixing (the official rate, or the USDT mid rate) at both ends of the window:

USD return % = ((index_end ÷ rate_end) ÷ (index_start ÷ rate_start) − 1) × 100

The gap is the official-rate dollar return minus the stablecoin-rate dollar return, in percentage points. Windows start no earlier than 15 August 2026, when the stablecoin history begins, and an index is left out if its latest close is more than ten days old.

10. Licence and citation

The index values, daily fixings and CSV files are published under the Creative Commons Attribution 4.0 licence (CC BY 4.0). You may use, share and adapt them, including commercially, provided you credit “Mansa Markets Stablecoin Premium Index” and link to the page you used. Suggested citation:

Mansa Markets Stablecoin Premium Index. https://www.mansamarkets.com/stablecoins/<country>. Accessed <date>.

For live data inside an application, use the MansaAPI FX endpoints.

11. Changelog

v1.1 · 18 November 2026
  • Added dollar-adjusted stock returns: each covered stock index shown in local currency, US dollars at the official rate and dollars at the stablecoin rate.
  • Nigeria's official rate is now the Central Bank of Nigeria rate (from 17 November 2026).
v1.0 · 17 November 2026
  • Index named and published as the Mansa Markets Stablecoin Premium Index.
  • Hourly measurement restored (it had run three times a day since 2 September 2026).
  • Daily fixings published for every currency, backfilled from hourly data to 15 August 2026.
  • Crossed books marked; rows older than two hours marked as delayed.

Observed market data, published for information. Not a quote, an offer to exchange, or financial advice.