Pulling up an annual report and trusting whichever ROE number is printed in the summary is one of the most common mistakes retail investors make on the Casablanca Stock Exchange. The ratio is genuinely important, but only in specific places — and reading it the same way at a bank and at a telecom produces the wrong answer almost every time.
Picture a Moroccan retail investor opening two tabs side by side on a Friday evening: Attijariwafa Bank's annual report and Maroc Telecom's. Both are blue chips, both sit heavily on the MASI, both pay dividends. The investor writes down the ROE printed at the top of each — say 18 percent for ATW and 22 percent for IAM — and concludes that the telecom is running a better business.
That conclusion is wrong in a particular way. Not because the investor is careless. Because the ratio does not survive translation between the two industries. Once you understand why, ROE stops being a universal score and becomes what it actually is: a sharp tool for one specific type of company, and a misleading one elsewhere.
Return on Equity is net income divided by shareholders' equity. If a bank reports 3 200 MMAD of net profit on an opening equity base of 18 000 MMAD, that is an ROE of about 17.8 percent. The formula takes about four seconds to compute and is not where the interesting reading happens.
What matters is the question the formula is really asking: for every unit of shareholder capital this company has committed to the business, how much fresh profit did it generate this year? That framing is what makes the ratio useful — and what makes it depend entirely on whether the equity base is doing productive work in the first place.
A bank is a balance-sheet business. Deposits flow in on the liability side, loans go out on the asset side, and between them sits a slice of equity whose job is to absorb losses when borrowers default. Regulators size that slice carefully — it is the loss-absorbing cushion, and if it runs too thin, Bank Al-Maghrib will be unhappy. Every dirham of that equity slice is there on purpose.
When you then compute "net profit generated per dirham of loss-absorbing capital", you are measuring something real. ROE tells you how productive the bank's capital is, and because the capital base is regulatory, tight, and meaningful, the ratio carries signal rather than noise. This is why bank analysts read ROE every morning and why the Moroccan universal banks tend to cluster in a recognisable range: roughly 12 to 20 percent for a healthy universal bank in a normal rate environment, compressing when policy rates drop and expanding when they rise.
Attijariwafa Bank has historically sat in the upper half of that band, which is part of why the stock typically trades at a premium to tangible book. Banque Centrale Populaire is a subtler case: the consolidated equity base is mechanically larger because the regional Banques Populaires feed retail deposits up through the central institution, dragging the denominator. A BCP ROE in the low teens is not proof of underperformance. It is the structural signature of the cooperative model, and if you read it without that footnote you will punish BCP for something that is not a defect.
Now return to that second tab. Maroc Telecom's balance sheet is dominated by infrastructure: towers, fibre, data centres, switching equipment, spectrum licences. Those assets were bought with a mix of debt and equity, and the equity line is now essentially the accounting residual of decades of capex decisions rather than a carefully sized loss-absorbing cushion. Nothing forces IAM's equity base to be any particular size relative to the business it runs.
This is where ROE loses its grip. A telecom can post a high ROE because it is genuinely running the network well. It can also post a high ROE because it has been returning so much cash through dividends and occasional buybacks that the equity line has been gradually eroded — same ratio, different story. The reader cannot tell the two apart from the number alone. For a dividend-heavy stock like IAM, this is a real risk, because IAM's payout policy deliberately thins the equity line year after year.
The more honest indicators for a telecom are consolidated revenue trajectory, net income trajectory, and free cash flow conversion — the last one especially, because the dividend is paid out of free cash rather than accounting profit. ROE is not wrong to glance at, but it is not the number you base a decision on.
Comparing ROE across sectors. A 20 percent bank ROE and a 20 percent telecom ROE are not peers. The only useful comparison is within the same sector, and even then with a structural note for companies whose capital base is unusual. ATW versus BCP is a defensible comparison if you remember the cooperative drag on BCP. ATW versus IAM is a category error dressed up as analysis.
Reading one year as if it were a trajectory. A single period can be distorted by a provisioning spike, a disposal gain, a tax case, or a restructuring charge. You want three to five years of ROE for a bank, not one. If the number drifts in one direction with no corresponding change in the business, something in the income statement is doing the work quietly.
Ignoring leverage. A company can manufacture a higher ROE simply by loading up on debt and shrinking equity relative to assets. The ratio rises, the operating performance does not. For regulated Moroccan banks this is bounded by capital rules, so the trick is partially closed off. For unregulated issuers, it is wide open, and a flattering ROE can sit on top of a balance sheet that is quietly riskier than last year.
On the Dalil stock pages, ROE appears only when the value can be extracted unambiguously from a filing's consolidated statements at group-share scope. If a given year's report discloses ROE under a mixed scope, or the parser cannot assign it to the right period with confidence, the row is hidden rather than filled with a placeholder. The reasoning is on the methodology page: a missing row is honest and costs nothing; a fabricated row is dangerous and lasts forever.
When the next Moroccan earnings season lands, try a small habit. Open an ATW or BCP filing, find ROE in the management commentary, and ask yourself two questions before anything else: is it moving in the same direction as PNB over the last three years, and is the capital structure this year comparable to last year's. If both answers are yes, you can actually use the number. If either is no, treat it as a clue rather than a verdict. That is the whole craft.