The short version
IBL Ltd is Mauritius’ largest company by revenue. It is not one business — it is four different machines stacked on top of a structure that lets a family holding company direct roughly 300 companies while owning a fraction of them. This is a full walk through how that works, from the accounting mechanics up to who ends up owning the money.
The headline is Rs 120.8 billion of revenue. Of every rupee a customer hands over, about four tenths of one cent reaches an IBL shareholder as a dividend. That is not a criticism — it is what a conglomerate is, and the rest of this post traces exactly where the other 99.6 cents go.
I picked IBL because it is the biggest thing in the Mauritian economy that publishes a full set of accounts, and because a conglomerate is the hardest kind of company to understand. A single-business company has one story. This one has four, plus roughly 300 subsidiaries, associates and joint ventures underneath them, in sixteen countries.
I read it over two days, chapter by chapter, and this post follows the order I actually learned things in — mechanics first, because without them every number in the report means something other than what it appears to mean.
1. What the document is
The report is 396 printed pages. It is really two documents bound together, and they are not equally reliable.
| Pages 7–199 | Pages 208–387 | |
|---|---|---|
| What it is | Management’s account of itself — strategy, people, clusters, photographs | The financial statements and notes |
| Audited? | No | Yes — Deloitte, unqualified opinion, signed 26 September 2025 |
| Share of the book | 211 pages (54%) | 180 pages (46%), of which 170 are notes |
Deloitte’s own report is explicit about the split. Several narrative sections were “expected to be made available to us after that date”, and it expresses “no assurance conclusion thereon”. That is normal and it is disclosed — but it means the two halves carry different weight, and throughout this post I say which half a figure comes from.
The two fences
On page 7, in a paragraph almost nobody reads, IBL draws its reporting boundary twice.
“The financial reporting boundary aligns with our audited financial statements boundary, and includes the Group and its subsidiaries… Our integrated reporting boundary goes beyond this financial scope, with information on all of the Group’s local and international operations, including operating subsidiaries, associates and joint ventures.”
IBL Integrated Report 2025, p. 7Fig. 1 — The same group, drawn twice. Source: pp. 7, 210, 388; Note 12.
2. How the numbers can mislead you
This is the section I wish I had read first. Six mechanics, each of which makes a published number mean something different from what it looks like. None of them is a trick — every one is required by IFRS, and every one is disclosed. But if you do not know them, you will read the report wrong.
Mechanic 1 — Consolidation: three ways to own something
Fig. 2 — Consolidation flatters revenue; equity accounting hides it. IFRS 10 governs the first, IAS 28 the second. Both lines are correct; they just answer different questions.
| You own | Called | Revenue | Profit | Balance sheet |
|---|---|---|---|---|
| >50% — control | Subsidiary | 100% consolidated | 100%, minority stripped at the bottom | Every asset and liability, line by line |
| ~20–50% — influence | Associate | Nothing | One line: your share | One line: carrying value |
| Joint control | Joint venture | Nothing | One line: your share | One line: carrying value |
What this does to IBL’s headline. Alteo (27.64%), Princes Tuna (40.64%), Miwa Sugar (27.64%), Equator Energy (33%) and AfrAsia are all described in the narrative as IBL businesses. Not one rupee of their sales is in the Rs 120.8 billion. Their entire contribution is Rs 364,217,000 — one line, 0.3% the size of the revenue figure.
Mechanic 2 — The pyramid: control and ownership are different numbers
Fig. 3 — IBL uses this repeatedly. PhoenixBev is consolidated at 23.28% effective ownership, UBP at 33.14%, Bazalt Réunion at 29.83%, Seychelles Breweries at 12.66%, Harley’s Tanzania at 41.6%.
Every one of those companies puts 100% of its revenue into the Rs 120.8 billion, and every one of them is mostly owned by somebody else. The effective percentages are printed openly on page 388 — but they are printed in a table at the back, not next to the headline.
Mechanic 3 — Non-controlling interests: the money that is not yours
The consequence arrives at the very bottom of the income statement, and it is the single most important line in the report:
| Rs’000 | FY2025 | FY2024 | Change |
|---|---|---|---|
| Attributable to owners of the Company | 3,006,785 | 2,974,223 | +1.1% |
| Attributable to non-controlling interests | 1,974,632 | 2,616,835 | −24.5% |
| Profit for the year | 4,981,417 | 5,591,058 | −10.9% |
Two fifths of the profit belongs to somebody else. And it shows up in three separate places, which is how you know it is real rather than an accounting artefact:
| Where | IBL’s shareholders | Minority shareholders |
|---|---|---|
| Share of profit (p. 210) | Rs 3,006,785k — 60% | Rs 1,974,632k — 40% |
| Share of equity (p. 208) | Rs 21,779,890k — 52.5% | Rs 21,044,547k — 47.5% |
| Cash dividends actually paid (p. 216) | Rs 516,971k | Rs 1,096,322k — 2.1× more |
The minorities took out more than twice the cash IBL’s own shareholders did. That is the price of the pyramid, paid every year, and you will not find it in any headline. It is the third row — the cash one — that convinced me this matters: profit shares are accounting, but dividends are money leaving the building.
Mechanic 4 — Internal sales: the group selling to itself
BrandActiv sells into Winner’s. Both are IBL subsidiaries. If you simply added up every subsidiary’s sales you would count that twice, so IFRS makes you strip it out:
| Rs’000 | FY2025 |
|---|---|
| Sum of the four clusters + corporate | 127,730,153 |
| Consolidation adjustments — sales the group made to itself | (6,946,195) |
| Group revenue | 120,783,958 |
Rs 6.9 billion of trade happens inside the group. That is a real measure of how connected these businesses are — and it has a consequence for the reader: because those transactions are eliminated, the prices at which one IBL company sells to another are invisible from outside. Which cluster books the profit on an internal deal is a management decision you cannot audit from the accounts. I come back to this in the frozen-warehouse example later.
Mechanic 5 — Restatements: last year’s numbers are not last year’s numbers
Every FY2024 figure in this report is marked (Restated). Two separate things caused it.
| What | Why | Effect |
|---|---|---|
| Note 43 — put and call options over NCI | Options granted to the minority shareholders of Mambo Retail (Naivas), Elgon Healthcare and Afix Scaff “had not been recognised in prior periods” — a prior-period error under IAS 8 | FY2024 profit restated from Rs 5,867,523k to Rs 5,591,058k; a Rs 4.38bn liability added at 1 July 2024 |
| AfrAsia reclassified | Sold down after year end, so it becomes a discontinued operation — and IFRS 5 requires the comparative year to be restated onto the same basis | Both years re-presented with AfrAsia on a separate line |
So the −11% everyone will quote is measured against a number that was itself corrected downward this year. Against what IBL actually published last year (Rs 5.87bn), the fall reads −15%. Neither is wrong; they answer different questions. But you have to know which one you are holding.
The mechanic worth learning here is the put option. When you buy control of a company and grant the remaining shareholders the right to sell you their stake later, that right is a present obligation to pay cash. IFRS makes you put it on the balance sheet as a financial liability at the present value of the redemption amount. It is remeasured every year, and the movement goes through profit — which is why the CFO now warns that “other gains and losses” will fluctuate. A structural choice — buying control while leaving minorities an exit — has installed permanent volatility in the P&L.
Mechanic 6 — Which profit number? They are not interchangeable
| Measure | FY2025 | Move | What it excludes |
|---|---|---|---|
| Revenue | Rs 120.8bn | +19% | Nothing — but it is a perimeter, not a possession |
| EBITDA | Rs 12.8bn | +28% | Depreciation, amortisation, interest, tax — and working capital entirely. Not an IFRS measure |
| Operating profit | Rs 7.4bn | +36% | Interest, tax, one-offs, associates |
| Profit for the year | Rs 5.0bn | −11% | Nothing — but 40% of it is not IBL’s |
| Attributable to owners | Rs 3.0bn | +1.1% | Nothing. This is the shareholder’s number |
Four measures of the same year moving between +36% and −11%. Every one is accurate. The gap between the top two and the bottom two is the entire subject of this post.
And one more, which the CFO flags himself: FY2024 contained Rs 1.6 billion of one-off disposal gains that FY2025’s Rs 58 million did not. Strip those out and profit rose 24%; strip them from the owners’ share and it rose 88%. The test for whether to accept an “excluding exceptionals” figure is simple: will the excluded item recur? For a group that recycles its portfolio every year, disposal gains are arguably part of the business model — and indeed AfrAsia’s gain lands in FY2026. Hold both numbers.
3. The shape of the group
With the mechanics in place, here is what is actually inside the perimeter.
Fig. 4 — The whole group on one screen. Look at the two bar rows: Retail’s turnover bar is three times Consumer Brands’, and its profit bar is shorter. Services has the smallest turnover of the four and the longest profit bar.
Four clusters, twelve subsectors, 39,903 people. Geographically there are three layers that are easy to conflate:
| Layer | Count | What it measures |
|---|---|---|
| Territories on the presence map | 20 | Reach — where IBL can serve a customer or source a product |
| Countries with registered entities | 16 | Commitment — where capital has gone into a structure. 242 of ~300 companies are Mauritian |
| Markets with disclosed revenue | 8 | Scale — Note 39(iii) |
| Markets above 10% of revenue | 3 | Concentration |
Fig. 5 — The dashed outline is last year. Two markets did nearly all the work; one all but disappeared.
Mauritius, Kenya and Réunion are 95.2% of group revenue. The other seventeen flags share 4.8%. And look at the growth column: Kenya and Réunion together produced 84% of the year’s revenue increase, while three Indian Ocean islands lost Rs 3.0 billion between them.
4. Where the money is made
Now the question the whole exercise was for. The report prints two pie charts on facing pages and never puts them side by side. Here they are as bars.
Fig. 6 — The cluster that gives you scale and the cluster that gives you margin are different clusters. That divergence is most of the argument for being a conglomerate.
| Cluster | Revenue | % rev | Op. profit | % profit | Margin |
|---|---|---|---|---|---|
| Retail | 64,643,733 | 53.5% | 2,019,794 | 24.6% | 3.1% |
| Consumer Brands | 25,901,452 | 21.4% | 2,090,785 | 25.5% | 8.1% |
| Industrials | 19,322,178 | 16.0% | 1,399,638 | 17.1% | 7.2% |
| Services | 17,614,565 | 14.6% | 2,684,884 | 32.8% | 15.2% |
| Corporate Services | 248,225 | 0.2% | (332,153) | — | — |
| Consolidation adj. | (6,946,195) | — | (414,668) | — | — |
| Group | 120,783,958 | 100% | 7,448,280 | 100% | 6.2% |
The DuPont decomposition — why a 3% margin is fine
My first instinct was that Retail, on a 3.1% margin, must be the weak business. That is wrong, and working out why was the most useful thing I did.
A margin tells you what you earn per rupee of sales. It says nothing about how much capital you tied up, or how fast you can do it again. The tool that puts those together is the DuPont identity, and it has three parts.
Fig. 7 — Retail and Consumer Brands reach almost identical returns on assets (8.4% and 8.6%) by opposite routes. Then the third column separates them completely.
The third column is the one worth slowing down on, because it is easy to misread. Leverage here is assets divided by equity, not assets divided by liabilities. The denominator is the sliver of the balance sheet the shareholders actually funded:
| Retail segment, Rs’000 | Who funds it | |
|---|---|---|
| Segment assets | 24,152,975 | — |
| Segment liabilities | (22,927,758) | Suppliers on 60-day terms, landlords through lease liabilities, banks |
| Net assets | 1,225,217 | The shareholders |
24,152,975 ÷ 1,225,217 = 19.71×. Each rupee of equity is carrying nearly twenty rupees of assets.
The reason it has to be equity and not liabilities is that the three terms are built to cancel:
Fig. 8 — Use assets ÷ liabilities instead and nothing cancels; the product means nothing.
And the reason the equity sliver is so thin is simply how a supermarket works. It takes cash at the till today and pays the supplier in sixty days — so for sixty days it is trading on the supplier’s money. Add capitalised store leases and there is barely any shareholder capital left in the business. That is negative working capital.
The CFO says the same thing in words on page 111, which is how I knew I had not invented it:
“even with lower profit margins, our Retail cluster remains efficient in generating high returns on capital, given its attractive balance sheet structure, which benefits from relatively lower fixed assets and a structurally positive cash cycle.”
Group CFO’s report, p. 111Profit per person — the sharpest cut
| Cluster | Employees | % of staff | % of profit | Profit per employee |
|---|---|---|---|---|
| Consumer Brands & Distribution | 3,427 | 8.6% | 25.5% | Rs 610,000 |
| Services | 6,735 | 16.9% | 32.8% | Rs 398,600 |
| Retail | 15,233 | 38.2% | 24.6% | Rs 132,600 |
| Industrials | 14,508 | 36.4% | 17.1% | Rs 96,500 |
A Consumer Brands employee generates 6.3 times the operating profit of an Industrials employee. Put more starkly: Consumer Brands earns more operating profit from 3,427 people (Rs 2,091m) than Retail earns from 15,233 people (Rs 2,020m) — on 22% of the headcount.
Three quarters of the workforce sits in Retail and Industrials and produces 42% of the operating profit. The other quarter produces 58%. Once you look at what each cluster owns, the reason is not subtle:
| Cluster | What it owns | The economics |
|---|---|---|
| Industrials | Quarries, yards, factories, canneries, cane | Heavy assets and heavy labour. You pay twice. Exposed to sugar and tuna prices and construction cycles |
| Retail | Store leases, inventory | Thin margin, enormous volume, funded by suppliers |
| Services | Hotels, property, licences, an insurance book | Heavy assets, high margin, slow turn |
| Consumer Brands | Brands, agencies, distribution rights, shelf relationships | You own the right to sell, not the means of production. An agency costs nothing to carry and scales with volume |
The further you are from making the physical thing, the more you earn per person. Industrials cans the tuna and crushes the rock. Consumer Brands holds the Coca-Cola bottling franchise for Réunion and the L’Oréal agency. One needs 14,508 people; the other 3,427; they earn roughly the same money.
And every strategic move in the year points the same way: sold most of a bank (AfrAsia 30.29% → 7.89%), agreed to sell 80% of a Caterpillar dealership (Scomat), bought a brewery (Seychelles Breweries 54.4%), won the Coca-Cola bottling franchise for Réunion, took on L’Oréal’s luxury range, expanded healthcare distribution across four countries.
But profit per head is not profit per rupee of capital. Retail is worst on people and best on capital, by an order of magnitude. Which business is “better” depends entirely on which resource is scarce — and in FY2025 IBL was short of both: a tight labour market at home and a deleveraging plan on the balance sheet. Which is exactly why those two are the clusters it pushed hardest.
5. How the money flows
This is the picture I most wanted when I started and could not find anywhere in the report. Every figure in it is from the audited statements on pages 210 and 216.
Fig. 9 — The whole machine on one screen. Note how small the blue bar is by the time you reach it — and that the red bar (non-controlling interests) is larger than everything that follows it.
A few things only become visible when you draw it this way.
| Step | Rs’000 | % of revenue | What it is |
|---|---|---|---|
| Revenue | 120,783,958 | 100% | What customers paid |
| Cost of sales | (89,746,867) | 74.3% | Goods and direct costs — the biggest single leak by far |
| Operating expenses | (25,972,883) | 21.5% | Includes staff costs of Rs 12,273,242 — 10.2% of revenue |
| Finance costs | (3,573,084) | 3.0% | On Rs 45.0bn of borrowings |
| Tax | (1,373,171) | 1.1% | 15% + a 2% CSR fund + a new 2% climate levy |
| Non-controlling interests | (1,974,632) | 1.6% | Profit that was never IBL’s |
| Retained in the business | (2,489,814) | 2.1% | Funding next year’s growth |
| Dividends to IBL’s shareholders | 516,971 | 0.43% | What actually left the building for the owners |
Three observations. First, NCI takes more of the profit than tax does — Rs 1.97bn against Rs 1.37bn. Second, the payout ratio is only 17.2% of the owners’ earnings, because the group is deleveraging. Third, and this is the honest counterweight: the Rs 2.49bn retained is not lost — it is the shareholders’ money reinvested, and it is what funds the store openings and acquisitions that make next year’s Rs 120.8bn bigger.
And the cash version, which differs
Profit is an opinion; cash is a fact. The cash flow statement tells a different story about the same year:
| Rs’000 | FY2025 | FY2024 |
|---|---|---|
| EBITDA | 12,799,000 | 10,015,000 |
| Net cash from operating activities | 6,812,587 | 9,071,395 |
| Cash conversion | 53.2% | 90.6% |
EBITDA up 28%, operating cash down 25%. The cash flow statement explains it: a Rs 3.4 billion swing on provisions and Rs 1.4 billion of inventory build — exactly what a group opening Naivas stores, taking on a new luxury cosmetics range and consolidating a Réunion quarrying business would do. EBITDA ignores working capital entirely. Growth has to be funded before it pays.
And where the cash went:
| Rs’000 | FY2025 |
|---|---|
| Net cash from operating activities | 6,812,587 |
| Capital expenditure (PPE, intangibles, investment property) | (4,986,027) |
| Free cash flow | 1,826,560 |
| Dividends paid (owners + NCI) | (1,613,293) |
| Left over | 213,267 |
| Meanwhile, spent on deals: advance towards an acquisition Rs 4,046,393 + acquiring subsidiaries Rs 3,901,639 = Rs 7,947,032, funded by Rs 6,815,504 of new borrowing | |
Free cash flow just covered the dividends. Every rupee of acquisition spending was debt-funded. That is normal for an acquisitive group in a build year — and it is exactly why the deleveraging plan exists and why the dividend rose only 2%.
6. Beyond Borders
IBL’s expansion strategy has a name and a date: Beyond Borders, launched 2021. It has six stated targets for 2030, of which exactly one is numerically testable.
| 2030 target | Testable? |
|---|---|
| Generate more than 60% of revenue outside Mauritius | Yes — at 54.2% today |
| Drive consistent, above-market growth in key markets | Directional — no baseline given |
| Become a top 3 player in each chosen sector | Directional |
| Deliver strong returns on capital | Directional |
| Be a leader in sustainable development in the region | Directional |
| Contribute strongly to the Mauritian economy | Directional |
Foreign revenue is currently 1.18× Mauritian revenue and must reach 1.5×. Given Réunion alone added Rs 7.5 billion in a single year, that is within reach — but it depends on the region, not on Mauritius.
The tool behind it, which is the best thinking in the report
| Bucket | The test applied | The mandate |
|---|---|---|
| Mauritian champion | Local expertise, minimal potential for regional growth | Maximise operational strength and innovation — get better, not bigger |
| Regional player | Potential for regional scale | Create integrated ecosystems across countries — the focus of Beyond Borders |
| International expert | World-class expertise replicable internationally | Scale proven models globally |
Most portfolio frameworks sort businesses by how attractive their market is — the BCG matrix on growth and share, the GE screen on industry attractiveness. IBL sorts on how portable the operating model is.
That is the right axis for a group based in a country of 1.3 million people. The binding constraint is not whether a market looks attractive; it is whether what you know how to do travels. A supermarket operating model travels. A basalt quarry travels only as far as the rock. Mauritian corporate services travel only where the regulatory regime resembles Mauritius. And note that “Mauritian champion” is not a failure category — it is told to get better, which is a real mandate rather than neglect.
The transferable version: sort your assets by how portable their operating model is, then give each bucket a different mandate.
You can see the framework working in the year’s disposals. Scomat is a Caterpillar dealership — a Mauritian champion with no regional runway. So 80% of it goes to La Compagnie Financière de Belmont, which owns Caterpillar’s dealer for Madagascar, Seychelles and Mayotte and does have one, and IBL keeps 20% of the upside.
Phase 1 and Phase 2 — and the change most people will miss
| Phase 1 — laying the foundation | Phase 2 — FY2025 onward |
|---|---|
| 2019 Nairobi office created | Bazalt acquired in Réunion by UBP |
| Jun 2022 40% of Naivas, Kenya | BrandActiv expands to Réunion and Kenya ★ |
| Mar 2023 Equator Energy, East Africa | CMH and Bloomage expand to Kenya ★ |
| Jun 2023 a further 11% of Naivas | Edena wins the Coca-Cola franchise, Réunion ★ |
| Aug 2023 Make Distribution, Réunion | Seybrew acquired in Seychelles by PBL |
| Oct 2023 Harley’s, Kenya | Divestment from AfrAsia and Scomat |
★ = organic growth. Five of the nine Phase 2 items are.
Two structural changes are hidden in that table.
First, the acquirer changed. Bazalt was bought by UBP. Seybrew by PBL. Not by IBL Ltd. The parent has stopped doing the deals; its operating subsidiaries now do them, in their own sectors, with their own management. That is decentralised M&A, and it explains why the Deputy Group CEO’s brief is the corporate centre rather than deal origination.
Second, growth has shifted from buying to building. IBL’s own split of this year’s 19%:
| Rs’000 | Amount | Rate |
|---|---|---|
| FY2024 (restated) | 101,565,161 | |
| Organic growth | +13,353,000 | 13% |
| Inorganic growth | +5,866,000 | 6% |
| FY2025 | 120,783,958 | 19% |
And the 6% is worth inspecting, because it is not what it appears. It consists of: a full year of Run Market against only 10 months last year; a full year of Harley’s against 8 months; Bazalt consolidating from 1 July 2024; and one small clinic. IBL says it plainly — “all inorganic growth registered this year relates to transactions announced or completed before the start of the year”.
That is called annualisation, and it is a one-time effect. Owning something for twelve months instead of eight produces growth that is pure arithmetic and disappears once the comparative year is also a full year. Part of this 6% will not repeat in FY2026. 13% organic is the number that matters — anyone with a balance sheet can buy revenue; only an operator grows it.
The five-step template, evidenced twice
Fig. 10 — Two businesses, fourteen years apart, following an identical sequence. Sources: pp. 52–55 and the Beyond Borders timeline, pp. 48–49.
The first move is an office, not a cheque. IBL put people in Nairobi in 2018–19 and bought nothing in Kenya until 2022. The early presence “helped build familiarity with the East African market and laid the groundwork”, and it reappears in the risk report as a mitigation for regional instability — “strong presence on the ground, with a dedicated office and resources”.
Step 5 is the proof. Coca-Cola and Pernod Ricard do not hand territory franchises to companies with good intentions. They audit your bottling lines, your cold chain, your distribution reach and your quality systems, then grant exclusivity. Winning one is third-party verification that steps 3 and 4 genuinely happened — the most credible evidence in the report that “integration” is an activity and not a word.
What I could not find out: how they actually researched these markets. I searched all 915,000 characters for “market study”, “market entry”, “investment committee”, “screening criteria” and “hurdle rate”. None appear. The report gives the sequence and never the method. What the pattern does suggest — and this is my inference, not their disclosure — is a consistent preference for family-owned targets (Naivas, Harley’s), staged stakes (40% then 51%) and co-investment with a development finance institution (Proparco alongside the Harley’s deal).
7. The machinery of integration
So what does “synergy” physically consist of? The answer is duller and better than I expected.
| When | What happened |
|---|---|
| Oct 2023 | Harley’s acquired — Kenya, Tanzania, Uganda |
| Late 2023 | Mauritius go-live: warehouse management brought in-house, new ERP over sales and warehousing, and the warehouse physically redesigned |
| End 2024 | Tanzania and Uganda migrated onto the same system |
| Apr 2025 | Kenya rolled out across four locations; Kenyan warehouses redesigned to match |
| Result | One common ERP across four countries. 400+ colleagues on the same processes. |
Eighteen months from acquisition to one system across four countries, Mauritius first and then exported. Note the warehouses were physically rebuilt to match — harmonising a system without harmonising the operation underneath it gives you one screen describing four different realities.
The group targets CMM-I Level 3 — “Defined”, meaning one standard way that everyone follows — across eight IT governance domains, with a second maturity assessment being run by EY. You do not need world-class IT to run a regional hub. You need identical IT.
Alongside it sits GREAT, the IBL Academy, and the reason it exists is strategic rather than sentimental. If your strategy is “replicate a proven Mauritian operating model”, then the model has to be teachable. Know-how locked in the heads of experienced people in Port Louis cannot be exported to Nairobi. An academy is the mechanism that converts tacit knowledge into transferable training.
And what it is worth, drawn to scale
Fig. 11 — Goodwill is the promise; integration is the delivery.
A group can pay Rs 8.8 billion of premiums and realise none of it because the businesses never actually got joined together. Which means a goodwill impairment review is really a test of whether integration happened — and Deloitte made exactly that one of its four Key Audit Matters.
More than half of IBL’s goodwill is one Kenyan supermarket chain. Naivas carries Rs 4,585,674,000 of it, and the assumptions behind the annual test are published:
| Naivas — value-in-use assumptions | 2025 | 2024 |
|---|---|---|
| Discount rate (WACC) | 16.93% | 16.60% |
| Annual growth rate, years 1–5 | 6.60% | 10.80% |
| Terminal growth rate | 5.00% | 5.00% |
Management cut its own five-year growth forecast for the asset carrying half the group’s goodwill from 10.8% to 6.6%, and took no impairment. Goodwill on a balance sheet is a forecast, not a fact — Rs 8.8 billion of IBL’s Rs 142.8 billion of assets exists only because management projects cash flows that justify it.
The example that justifies the whole conglomerate
Buried in a paragraph about logistics is the clearest evidence I found that the group form creates value a standalone could not:
“the construction of a new frozen warehouse, with Bloomage leading the investment, Logidis and BrandActiv securing long-term contracts, and Commercial Engineering and Manser Saxon supporting with engineering expertise… a model IBL aims to replicate in other markets.”
Performance report, p. 123Five companies, three clusters, one asset. The property arm gets a building pre-let on day one — the single hardest thing in property development. The distribution businesses get cold storage built to their specification. The engineers get a contract. Nobody negotiated against anybody.
And it is also the honest limit of the argument: a related-party deal only creates value if the rents are at market. Priced right, everyone wins; priced wrong, one cluster’s profit has simply been moved into another’s. Those contracts are eliminated on consolidation — correctly — so from outside you cannot tell which.
8. Who controls it
We have seen the pyramid. Here is the other half of the structure, and it is the part that explains everything else.
Fig. 12 — The votes and the economics have been separated completely.
“GML Ltée held 1,510,666,650 RRS, representing 68.95% of the voting rights. These shares are not listed and the only right attached to these shares is the power to vote at general meetings. GML Ltée has no right to dividends or distribution or to any surplus from the Company in case of winding up.”
Statement of Compliance, p. 173Work out what a vote costs by each route: Rs 0.0033 through the restricted shares, Rs 30.00 through the market. A vote is roughly nine thousand times cheaper one way than the other.
I want to be careful here, because it would be easy to make this sound sinister and I do not think it is. Dual-class structures answer a real problem: a family that has built something over 190 years wants to invest across decades without a hostile bid interrupting it, and wants outside capital without surrendering direction. The trade offered is explicit — you get the economics and the growth, we keep the steering wheel. And GML Ltée takes no dividends: it is not extracting cash through the control shares. The family’s economics come through ordinary shares like everyone else’s. The structure buys direction, not money.
Whether it is a good trade is a judgement about the family’s competence — which is, I think, exactly why the report spends so much space on governance. Of fourteen directors, four are independent, and independents chair both the Finance, Audit & Risk Committee and the Strategic Committee. Those are the two seats that can most constrain management, and they are deliberately not held by the family. Five Lagesse family members sit on the board; the CEO is one of them.
9. What the model costs
Four bills that come with the strategy. All four are disclosed; none is in a headline.
A regulatory change in one country, worth about a billion rupees
Mauritius introduced wage relativity adjustments and a mandatory fourteenth-month salary during the year. The People section names the cost: “adding nearly a billion rupees in unplanned expenses for the Group”. Audited staff costs moved from Rs 10,933,584,000 to Rs 12,273,242,000 — about 13% of operating profit, absorbed in a single year.
What makes this example worth keeping is that the number appears nowhere as a line item. It disappears into staff costs, into operating expenses, into operating profit. The only reason you can see it is that the narrative half named it. That is the strongest argument for why integrated reports bind the two halves together.
Rs 21.8 billion of currency exposure
| Currency, Rs’000 | Financial assets | Financial liabilities | Net |
|---|---|---|---|
| Kenyan Shilling | 3,413,917 | 13,623,218 | −10,209,301 |
| Euro | 1,310,908 | 9,228,006 | −7,917,098 |
| US Dollar | 963,866 | 4,865,588 | −3,901,722 |
| Net exposure | −21,793,824 |
IBL publishes its own sensitivity: a 10% move in the rupee is worth about Rs 1.1 billion through profit. And the exposure is the strategy’s shadow — the shilling is Naivas and Harley’s, the euro is Réunion, the dollar is the Maldives. You cannot buy Kenyan supermarkets and Réunionnais quarries without acquiring a shilling and euro balance sheet.
A covenant that broke
A covenant is a promise inside a loan agreement — keep your debt ratios below some level. Break one and the lender may demand immediate repayment, so under IFRS the whole loan reclassifies as current, even if the lender has no intention of calling it.
“The subsidiary did not comply with the debts covenant ratios when it was tested using the figures as at 30 June 2025; primarily due to the underperformance of Bazalt Reunion compared to the initial business plan, attributable mainly to the exceptional downturn which the construction sector is currently facing in Reunion Island.”
Note 22(b)(ii) — a Rs 2.5 billion borrowingFollow that across the document and it is one event described three times in three languages. The strategy section presents Bazalt as a success that lifted UBP’s turnover by nearly 50%. The risk register adds a brand-new entry this year at number 13: “Performance of capital investments”. And the notes say the acquisition missed its plan and breached a covenant. All three are true. Only reading all three gets you the picture.
And the risk register, which is the strategy’s shadow
| # | Risk | The choice that created it |
|---|---|---|
| 1 | Cybersecurity threats | One ERP across four countries. Integration converts four separate points of failure into one |
| 2 | Regional instability | 54% of revenue abroad; Rs 44bn from East African acquisitions |
| 3 | Forex fluctuations | Rs 21.8bn of net currency exposure |
| 5, 11, 14 | Government policies, talent scarcity, talent management | The Rs 1bn wage shock |
| 7 | Volatility of commodity prices | Sugar and tuna, inside Industrials |
| 13 | Performance of capital investments — NEW | Phase 2 of Beyond Borders |
And the risk that left the top 15 tells you as much as the one that joined it: “Capital investment abroad” was ranked 8th last year and is now unrated. The question has moved from should we put capital abroad at all? to is what we bought actually working? That is the Phase 1 to Phase 2 transition, written in the risk language.
10. What the shareholder got
Everything above describes a business that had a decent year. Here is what a shareholder experienced.
| FY2025 | |
|---|---|
| Share price, 30 June 2025 | Rs 30.00 |
| Highest during the year | Rs 42.05 |
| Lowest during the year | Rs 30.00 — it closed at its low |
| Capital depreciation | (Rs 10.00) — −25.0% |
| Dividend received | Rs 0.76 — +1.9% |
| Total holding period return | −23.1% |
Revenue +19%. Operating profit +36%. Shareholder return −23.1%.
| Valuation | |
|---|---|
| Market capitalisation | Rs 20.4bn |
| Equity attributable to ordinary shareholders | Rs 21.78bn → 0.94× book |
| Earnings attributable to owners | Rs 3.01bn → 6.8× earnings |
| The parent’s own investment portfolio, at fair value | Rs 40.67bn → the market pays ~50c per rupee |
This is the conglomerate discount, and the usual explanations all apply: you cannot get at the assets, there are layered costs at every level of the pyramid, the structure is complex, and seven of the group’s companies are separately listed so you can buy the pieces directly.
But the reason it will not close is structural:
| Liquidity, FY2025 | |
|---|---|
| Total shares traded all year | 5,910,672 |
| As a share of the company | 0.869% |
| Average daily volume | 24,125 shares — 0.004% |
A discount normally closes because somebody buys the company or forces a change. Neither can happen here. 69% of the votes are not for sale at any price, and at 24,125 shares a day there is not enough float for an institution to build a position. So it is not a mispricing waiting to be corrected — it is a structural feature of a controlled, illiquid stock. Anyone buying the shares is buying a claim on dividends, not a claim on direction, and should price it that way.
11. What I took from it
| # | The thing |
|---|---|
| 1 | A revenue figure is a consolidation perimeter, not a possession. Rs 120.8bn includes 100% of companies IBL owns a third of, and 0% of companies it part-owns. |
| 2 | Control and ownership are different numbers. 73.2% × 51% = 37.33%, consolidated at 100%. |
| 3 | Find the NCI line before you find anything else. 40% of the profit, 47.5% of the equity, and 2.1× the dividends. |
| 4 | Revenue tells you about scale; it does not tell you about money. Retail is 51% of one and 30% of the other. |
| 5 | A low margin is not a bad business. Margin × speed × other people’s money is the equation, and Retail wins it at 3.1%. |
| 6 | The further you are from making the physical thing, the more you earn per person. Rs 610,000 for an agency business; Rs 96,500 for a quarry. |
| 7 | Split growth into organic and inorganic, then check how much of the inorganic is just annualisation. |
| 8 | Sort assets by how portable their operating model is, not by how attractive the market looks. |
| 9 | Presence before capital. An office three years before the first cheque, twice. |
| 10 | Goodwill is the promise; the ERP is the delivery. Rs 8.8bn riding on Rs 300m of software. |
| 11 | EBITDA ignores working capital. Up 28% while operating cash fell 25%. |
| 12 | Read past page 199. The wage shock, the currency exposure, the broken covenant and the halved cash conversion are all in the notes. |
And if I had to put the whole operating model in five sentences: IBL buys controlling stakes in strong local operators in markets that resemble Mauritius, funds them partly with minority partners’ capital, and consolidates them fully. It then makes the acquired businesses run on one system, one standard and one talent pipeline, so that a proven Mauritian operating model becomes portable. Scale comes from low-margin retail; margin comes from asset-light distribution rights and a services book; the combination is the reason to hold them together. Roughly two fifths of the resulting profit belongs to those minority partners, and the group is directed by a family holding company that paid Rs 5 million for 69% of the votes and takes no dividends. The strategy is working operationally, and the market has no mechanism to pay for it.
12. How I did this, and what I could not see
What I read
The IBL Ltd Integrated Report 2025, for the financial year ended 30 June 2025, all 396 printed pages. I extracted the text so I could search it, rendered several pages as images where the charts do not extract cleanly (waterfalls and two-row timelines extract out of order and will silently give you wrong pairings), and worked chapter by chapter over two days. Every figure carries a page number or a note number. Where I have computed something myself — the per-employee figures, the DuPont decomposition, the money-flow waterfall, the vote-cost comparison — I have said so and shown the inputs.
The two halves
Pages 208 to 387 are audited by Deloitte, which gave a clean unqualified opinion on 26 September 2025 with four Key Audit Matters: valuation of properties, impairment of goodwill, valuation of unquoted investments, and the recognition of put and call options over non-controlling interests. Three of the four are valuation judgements, which is what a conglomerate’s audit is. The other 211 pages are not audited.
What I could not work out
Three things. How they research a market before entering it — the report gives the sequence and never the method. Which metrics the executive bonus is measured against — 55% of executive director pay is performance-based and the criteria are not disclosed. And how the group prices contracts between its own subsidiaries — those are eliminated on consolidation, correctly, so the internal pricing that decides how profit is divided between clusters is invisible from outside.
What this post is not
It is not investment advice and it is not a recommendation. IBL is a going concern with a clean audit opinion, roughly 40,000 employees, and a strategy I found more coherent than I expected. Nothing here suggests otherwise. I am a student reading a public document and trying to learn how a large group works; if I have got something wrong, I would like to know.
Also on this site
The BAI collapse in Mauritius, explained — the 2015 BAI / Bramer collapse decoded from the nTan forensic report, chapter by chapter. · New Mauritius Hotels valued against its listed peers — a trading comparables set for nine listed Mauritian hotel companies.
Sources
Principal references within the report: About this Report p. 7 · At a Glance p. 10 · 2025 Highlights p. 12 · International Presence p. 14 · Cluster Overview p. 16 · Chairman’s Message p. 20 · Group CEO interview p. 40 · value creation model p. 44 · Our Strategy pp. 46–49 · Regionalisation in Action pp. 52–55 · People pp. 64–73 · Technology & Transformation pp. 84–93 · Group CFO’s Report pp. 106–115 · cluster pages pp. 116–124 · Risk Management Report pp. 126–144 · Corporate Governance Report pp. 146–169 · Statement of Compliance p. 173 · Statutory Disclosures pp. 170–199 · Independent Auditor’s Report pp. 202–207 · Statements of Financial Position p. 208 · Statements of Profit or Loss p. 210 · Statements of Cash Flows p. 216 · Note 6(a) intangible assets · Note 12 investment in associates · Note 20(b) restricted redeemable shares · Note 21 assets held for sale and discontinued operations · Note 22 borrowings · Note 36 related party transactions · Note 37(a) financial instruments · Note 38 business combinations · Note 39 segmental information · Note 43 prior year adjustments · Industry Segments p. 388.