WAEMU bond indices: methodology
Rules version 1.1, data through 20 November 2026. This text describes the calculation as it runs.
Purpose and scope
brvmdata.com computes three independent, rules-based total-return indices of XOF-denominated bonds of the West African Economic and Monetary Union (WAEMU), together with their sub-indices:
- the sovereign index: fixed-rate government bonds of the eight WAEMU States, from both venues: the regional auction market run by UMOA-Titres (OATs) and bonds of the States listed on the BRVM (public offerings);
- the Treasury-bill index: the States' Treasury bills (BATs) issued through UMOA-Titres;
- the corporate index: XOF bonds listed on the BRVM whose issuer is not a State: companies, banks and other financial issuers (including securitisation funds) and regional institutions.
The indices describe the market. They are not forecasts, ratings or recommendations. Rules version 1.1 (2026-11-20 data restatement; the rules are those of 1.0).
Sources
| Data | Source | Use |
|---|---|---|
| Auction results (each tranche: instrument, coupon, settlement and maturity dates, amount retained, weighted average yield and price, redemption terms) | UMOA-Titres, from 2014 | auction-market universe, par outstanding, sovereign yield curves |
| Monthly list of live securities (ISIN, State, type, outstanding, coupon, amortisation mode, grace period) | UMOA-Titres monthly reports, from March 2024 | ISINs, par anchoring, reconstruction check, coverage |
| Secondary-market trades (ISIN, date, average price, yield, volume) | UMOA-Titres weekly report and daily sheet, from May 2021 | prices of auction-market securities |
| BRVM sessions (traded volume and value, closing price) | bfin.brvm.org, from September 1998 | prices of listed bonds, business-day calendar |
| Official bulletin of the BRVM (residual nominal, price, accrued coupon, next coupon date, amortisation type) | BRVM, from 2018 | prices, payment calendar, observed amortisation |
| Technical sheets and amortisation schedules | DC/BR | exact cash flows of listed bonds |
| Issue approvals (visas) | AMF-UMOA, from 1998 | issue amount and value date of listed bonds without a DC/BR schedule |
Universe and eligibility
Sovereign index. Fixed-rate OATs issued by auction through UMOA-Titres and fixed-rate bonds of the eight States listed on the BRVM. A bond is eligible at a rebalance date when it has settled on or before that date, has a residual maturity of at least one year, has par outstanding and can be priced. Sub-indices: by State (8), by residual-maturity band at the rebalance date (1–3, 3–5, 5–7, 7–10 and 10 years or more, measured to final maturity) and by venue (UMOA-Titres, BRVM).
Treasury-bill index. BATs of the eight States that have settled on or before the rebalance date and mature after it. Sub-indices by State.
Corporate index. Fixed-rate XOF bonds listed on the BRVM whose issuer is not a WAEMU State, with a residual maturity of at least one year, listed on or before the rebalance date (first appearance in the BRVM session data or bulletin). Sub-indices: corporates and financials (companies, banks, insurers, finance companies, securitisation funds) and regional institutions (BOAD, BIDC-EBID, CRRH-UEMOA and the other multilateral or development institutions listed on the BRVM).
Treatment of particular instruments (written rules).
- Sukuk (sovereign or not) are excluded: they are lease-based certificates issued through a dedicated fund, whose distributions are rents rather than bond coupons.
- Floating-rate bonds (coupon set as a base rate plus a margin) are excluded from every index.
- Securitisation funds (FCTC) listed on the BRVM are included in the corporate index, in the corporates and financials sub-index, when a fixed coupon and an amortisation schedule are published (DC/BR). Senior and subordinated tranches are separate bonds.
- Guaranteed issues are classified by their legal issuer, not by the guarantor.
- Sub-sovereign and State-owned issuers (ports, utilities, public companies) are corporates. Benin's public debt agency (Caisse autonome d'amortissement) acted for the State and is classified as sovereign.
- A listed bond is in an index only when its cash flows (DC/BR schedule, or terms rebuilt as below) and its issue amount are known; listed bonds without them are counted as not covered in the coverage table.
Weighting and rebalancing
- Rebalance date: the last BRVM session of each month. Eligibility, membership and holdings are set at the close of that session and kept until the next rebalance.
- Each bond is held in proportion to its par amount outstanding at the rebalance date (weighting by issuance; no country or issuer cap). Its weight is therefore its share of the index market value (par × full price) at the rebalance, and drifts with returns during the month.
- New issues and reopenings enter at the first rebalance after their settlement date.
- Coupons and principal (amortisation and redemption) paid during the month are held as cash earning nothing until the next rebalance, where they are reinvested across the new holdings.
Par amount outstanding
- UMOA-Titres securities. A line is one State, one instrument, one maturity date and (OATs) one coupon; reopenings add to the line, buyback legs reported with an auction subtract from it, and amortisation reduces it. The ISIN is taken from the auction record when present, else from a unique match (State, type, maturity, coupon) in the monthly lists. From March 2024 the par used by the indices is the outstanding printed in the latest monthly list that is public at the rebalance date (taken as public 15 days after its date), rolled forward with the auction legs settled since and with scheduled amortisation; a line that is not in a public list it should appear in is not held. Before March 2024 the par is the reconstruction from auction legs. The table
bix_reconreports, by State, month and instrument, the reconstruction alone against the published total: the UMOA-Titres monthly lists (referenceumoa_titres) and, from December 2020 to January 2024, the BCEAO's lists of live bills and bonds (referencebceao); a month without a monthly list takes its reconstruction error inbix_qualityfrom the BCEAO list of that month, where there is one. The BCEAO comparison is a check only: it does not change the par. - Listed bonds. Period outstanding from the DC/BR amortisation schedule. Without a schedule: the amount subscribed in the AMF-UMOA visa (the amount requested when the subscribed amount is not printed) and the amortisation rebuilt as below.
Cash flows and conventions
| Instrument | Coupon | Accrued interest | Yield |
|---|---|---|---|
| OAT (UMOA-Titres) | coupon rate / frequency, paid on the anniversaries of the maturity date (annual unless the monthly list says semi-annual) | actual/actual within the coupon period | annual compounding, actual/365, from the valuation date |
| BAT (UMOA-Titres) | none (discount) | none | auctions publish a money-market yield, actual/360 (price = 1 / (1 + y × days/360)); the indices convert it to annual compounding, actual/365 |
| Listed bond with DC/BR schedule | gross interest of each period as printed in the schedule | linear in actual days between payment dates | annual compounding, actual/365 |
| Listed bond without schedule | coupon rate / frequency (frequency from the bulletin, else annual) | actual/actual within the period | annual compounding, actual/365 |
These conventions were checked on the data, not assumed:
- OAT auction results: computing the price from the published weighted-average yield with the conventions above reproduces the published weighted-average price as a clean price (excluding accrued interest) within 0.05 point for 95% of 1,525 bullet tranches; the same computation as a dirty price fails.
- BAT auction results: the published discount rate and yield satisfy price = 1 − d × days/360 and yield = (1/price − 1) × 360/days for 95% of 1,413 tranches.
- UMOA-Titres secondary trades: the published OAT price is the clean price matching the published yield under the same conventions within 0.1 point for 99.6% of 4,503 trades.
- BRVM bulletin: prices are clean, in XOF per security, against a residual nominal per security that falls as the bond amortises (price in % = price / residual nominal × 100); the bulletin prints the accrued coupon separately.
- DC/BR schedules use issuer-specific coupon bases: of 4,024 schedule periods with interest, 61% pay exactly rate / frequency, 14% rate × actual days / 360, 8% rate × actual days / 365, and 15% are consistent with two bases. The indices use the printed amounts, so no basis is assumed for these bonds.
- OAT amortisation: bullet unless the auction terms or the monthly list give a grace period (constant annual amortisation after the grace period, "amortissement constant différé").
- Listed bonds without a DC/BR schedule: payment month and day from the bulletin's next coupon date, maturity year from the title, amortisation from the bulletin type (in fine, or constant amortisation) with the grace period that best fits the residual nominal observed in the bulletins and in session values; bullet when no amortisation is ever observed.
- Coupons are taken gross of the withholding tax on securities income.
Prices
Every bond is priced at every BRVM session, in this order:
- Trade price. UMOA-Titres securities: the average price of a secondary-market trade of that day. Listed bonds: the bulletin price of a session with traded value of at least XOF 1,000,000 (BRVM session data before the bulletin archive starts; their closing price is rounded to whole percent, so the price implied by traded value and residual nominal is used when it agrees within one point).
- Trade challenge. A trade more than 3 points of par away from the evaluated price of the same day (step 3, computed before the trade) is set aside; it is accepted once a later trade of the same bond within 30 days deviates in the same direction. Accepted trades price the bond and reset its spread; counts are in
bix_meta.trade_filter_json. - Evaluated price. The yield of the bond is the issuer-State curve at the bond's weighted average life, plus the bond's spread to that curve measured at its last accepted trade. Without any trade, UMOA-Titres securities take no spread and listed bonds take the spread at issue (issued at par, so the issue yield is the coupon-implied yield). Regional institutions and issuers outside the eight States use the union curve. Before any curve exists (before 2014), a listed bond keeps the yield of its last trade (or of its issue).
- Curve. For each State and day, the weighted-average yields of the State's OAT and BAT auctions held in the previous 365 days (730 days, then the union curve, when the State held none) are grouped in residual-maturity buckets (0–0.5, 0.5–1.5, 1.5–2.5, 2.5–4, 4–6, 6–8.5 and over 8.5 years); each bucket keeps its most recent auction day (amount-weighted when several tranches fell that day); yields are linearly interpolated in residual maturity between bucket nodes and held flat beyond the ends. The union curve pools the eight States' auctions.
The price source of each bond and day is recorded; bix_levels.trade_priced_share is the share of the index market value priced by a trade that day and bix_constituents.price_source_share the share of a bond's days priced by a trade over the holding period.
Validation. At every trade, the evaluated price computed from information available before the trade is compared with the traded clean price. bix_pricing_validation publishes the median, mean and 90th-percentile absolute error, in points of par, by year and venue. Trades set aside by the challenge are included in this comparison.
Index calculation
At each rebalance date R the index holds each eligible bond's par outstanding. On a later session t of the holding period:
- Total return: level(t) = level(R) × [Σ par held × full price(t) + coupons and principal received since R] / [Σ par held × full price(R)]. Full price = clean price + accrued interest; amortisation reduces the par held and is received as cash at par.
- Price return: the same with clean prices, counting principal received but not coupons.
- Base: 100 at each index's base date (a month end).
- Analytics, weighted by par outstanding on day t: yield to maturity (annual compounding, actual/365), modified duration, average life, average coupon (not shown for bills), number of bonds and par outstanding of the holdings.
Start dates and coverage
Each index is computed from its base date (level 100), the first month end from which a coverage test passes at 12 consecutive month ends (or at every month end to date, at least three, near the present). The test, run at every month end:
- at least 5 eligible priced bonds (headline indices) or 3 (sub-indices), and
- coverage of at least 70% of the reference outstanding.
The reference outstanding depends on the instruments:
- UMOA-Titres securities in months with a monthly list for every State concerned: par held against the outstanding in the month's list (OATs of one year or more residual maturity for the sovereign family; all BATs for bills);
- UMOA-Titres securities in earlier months: an estimate equal to (1 − share of outstanding older than the State's auction record, measured on the age profile of the monthly lists) × share of live auction legs whose amount was published;
- BRVM-only indices (corporate, sovereign BRVM venue): share of listed bonds of the family with residual maturity of at least one year that the index can value (cash flows and issue amount known).
Official and extended history. The official start (bix_indices.official_start) is the first month end from which every month end up to the launch date passes the test. Levels from the official start are the official history (bix_levels.history = 'official'). Levels between the base date and the official start are computed with the same rules and published as extended history (history = 'extended', start in bix_indices.extended_start): coverage was partial in at least one of those months. The series is continuous: there is no rebasing at the official start, so returns over any period are comparable. When the last month end before the launch fails the test, the official start is the first passing month end after the launch.
An index is computed every session, including months that fail the test; bix_quality publishes every month's bond count, par covered, reference par, coverage, trade-priced share, reconstruction error and test result. A sub-index left without any eligible bond at a rebalance holds cash for the month.
Point in time, back-calculation and restatement
- A level on date t uses only information public on or before t: auction results from their operation date, trades from their trade date, BRVM sessions from their session date, monthly lists 15 days after their date. Bond terms (DC/BR schedules, AMF visas, auction terms) are public from issuance.
- Levels dated before the launch date (
bix_indices.launch_date, 17 November 2026) are back-calculated and flagged (bix_levels.back_calculated = 1). - Published levels are not revised. A daily run recomputes from the last rebalance and appends the new sessions; it reports any difference with the stored levels (
bix_meta.restatement_check_max_rel_diff) without overwriting them. A full recalculation (restatement), for a rules change or a material data correction, is announced with a new rules version. - After the launch, a month that fails the coverage test is flagged in
bix_qualitybut never moves the official start. - 2026-11-20 restatement (data correction, listed bonds): four bonds now use their DC/BR schedules. EOS.O35, EOS.O36 and EOS.O37: DC/BR prints their start balance at half the issue, so the issue size is taken from the sheet's effective amount, which equals the principal it repays. FBOAD.O2: the sheet's header cells all read "Echéance"; its columns are now read off the sheet's own arithmetic (count × nominal = start balance, start − principal = end balance, annuity = interest + principal). EOM.O19 and EOM.O20 take the DC/BR sheets filed under their own symbols (
TNC_EOM.O19,TNC_EOM.O20); the official bulletin prints each one's title on the other's symbol, and bfin and DC/BR agree with each other. EOM.O17 and EOM.O18, listed earlier, had wrongly been matched to those two sheets: EOM.O17 now falls back to the AMF visa rule, and EOM.O18, with no other source for its schedule, leaves the index. Matching by symbol also corrects the issue amounts of EOS.O22–O29: same terms, two tranches each.
Known limits
- Auction data start in January 2014: OATs issued earlier are missing until they mature, which delays the sovereign start date; the estimate is disclosed month by month.
- Buyback-only and exchange operations are not in the auction record. Before March 2024 the reconstruction can overstate the par of the lines concerned; from March 2024 the monthly lists correct the par used by the indices.
- Listed bonds before 2012 often lack a DC/BR schedule; their cash flows are rebuilt from the listing terms and the AMF visa, and some cannot be valued at all (shown as not covered).
- BRVM bond trading is thin and prices are often round numbers; most index days are evaluated rather than traded (see the trade-priced share).
- The curve is built from primary auctions, whose yields include issuance terms; the carried spread from the last trade adjusts each bond to its own market level.
Pricing validation
| Year | Venue | Trades | Median | Mean | 90th percentile |
|---|---|---|---|---|---|
| 2026 | BRVM | 657 | 1.00 | 1.82 | 4.61 |
| 2026 | UMOA-Titres | 1,522 | 0.57 | 1.28 | 3.88 |
| 2025 | BRVM | 658 | 0.93 | 1.88 | 5.02 |
| 2025 | UMOA-Titres | 1,300 | 0.62 | 1.67 | 5.01 |
| 2024 | BRVM | 839 | 0.97 | 2.08 | 6.21 |
| 2024 | UMOA-Titres | 867 | 0.74 | 1.81 | 5.31 |
| 2023 | BRVM | 777 | 0.94 | 1.94 | 5.50 |
| 2023 | UMOA-Titres | 788 | 1.00 | 2.06 | 5.64 |
| 2022 | BRVM | 808 | 0.64 | 1.67 | 5.23 |
| 2022 | UMOA-Titres | 695 | 0.51 | 0.98 | 2.68 |
| 2021 | BRVM | 726 | 0.62 | 1.38 | 4.39 |
| 2021 | UMOA-Titres | 439 | 0.49 | 0.96 | 3.02 |
| 2020 | BRVM | 515 | 0.61 | 1.10 | 3.00 |
| 2019 | BRVM | 430 | 0.30 | 0.67 | 1.99 |
| 2018 | BRVM | 311 | 0.10 | 0.51 | 1.37 |
| 2017 | BRVM | 236 | 0.09 | 0.64 | 1.83 |
| 2016 | BRVM | 284 | 0.14 | 0.48 | 1.31 |
| 2015 | BRVM | 320 | 0.11 | 0.75 | 1.56 |
| 2014 | BRVM | 254 | 0.49 | 1.18 | 2.04 |
| 2013 | BRVM | 230 | 0.01 | 0.57 | 1.90 |
| 2012 | BRVM | 168 | 0.04 | 0.77 | 1.99 |
| 2011 | BRVM | 123 | 0.01 | 0.56 | 1.01 |
| 2010 | BRVM | 74 | 0.03 | 0.82 | 3.99 |
| 2009 | BRVM | 161 | 0.00 | 0.55 | 1.79 |
| 2008 | BRVM | 133 | 0.01 | 0.11 | 0.33 |
| 2007 | BRVM | 102 | 0.01 | 0.10 | 0.49 |
| 2006 | BRVM | 75 | 0.01 | 0.21 | 0.75 |
| 2005 | BRVM | 62 | 0.01 | 0.19 | 0.50 |
| 2004 | BRVM | 39 | 0.01 | 0.38 | 0.07 |
| 2003 | BRVM | 45 | 0.02 | 0.15 | 0.12 |
| 2002 | BRVM | 76 | 0.01 | 0.06 | 0.07 |
| 2001 | BRVM | 66 | 0.01 | 0.16 | 0.77 |
| 2000 | BRVM | 58 | 0.01 | 1.03 | 4.97 |
| 1999 | BRVM | 9 | 0.02 | 0.03 | 0.06 |
Outstanding reconstruction
| Country | Instrument | Rebuilt (XOF bn) | Published (XOF bn) | Error (%) | Lines rebuilt / published / matched |
|---|---|---|---|---|---|
| Burkina Faso | BAT | 212.4 | 212.4 | –0.01 | 16 / 16 / 16 |
| Burkina Faso | OAT | 2,814.4 | 2,816.8 | –0.08 | 171 / 171 / 171 |
| Benin | BAT | 54.9 | 54.9 | –0.01 | 3 / 3 / 3 |
| Benin | OAT | 948.7 | 948.7 | 0.00 | 27 / 27 / 27 |
| Côte d'Ivoire | BAT | 262.5 | 262.5 | 0.00 | 16 / 16 / 16 |
| Côte d'Ivoire | OAT | 7,791.3 | 7,208.0 | +8.09 | 198 / 198 / 197 |
| Guinea-Bissau | BAT | 105.6 | 105.6 | –0.01 | 13 / 13 / 13 |
| Guinea-Bissau | OAT | 461.3 | 461.3 | –0.01 | 39 / 39 / 39 |
| Mali | BAT | 345.9 | 345.9 | 0.00 | 25 / 25 / 25 |
| Mali | OAT | 2,310.2 | 2,327.7 | –0.75 | 124 / 123 / 123 |
| Niger | BAT | 71.5 | 71.5 | +0.01 | 5 / 5 / 5 |
| Niger | OAT | 2,059.1 | 2,090.9 | –1.52 | 42 / 41 / 41 |
| Senegal | BAT | 1,285.1 | 1,285.1 | 0.00 | 30 / 30 / 30 |
| Senegal | OAT | 4,021.1 | 4,067.4 | –1.14 | 142 / 143 / 142 |
| Togo | BAT | 18.5 | 18.5 | 0.00 | 2 / 2 / 2 |
| Togo | OAT | 1,646.5 | 1,611.5 | +2.17 | 57 / 56 / 56 |