There's a fashion conglomerate you already know without knowing its name. Every time you've walked into an Arezzo store, bought a Schutz bag, a FARM Rio dress, an Animale piece, a Reserva shirt or a Hering t-shirt, you were buying from a single company: Azzas 2154, traded on B3 (Brazil's stock exchange) under the ticker AZZA3. It's Latin America's largest fashion group — R$ 11.8 billion in revenue and R$ 770.7 million in recurring profit in 2025. And yet the market is pricing the stock at less than half of what its own balance sheet says it's worth. This piece breaks down why, and points to the full AZZA3 analysis we published today.
Is it worth buying? The short answer, up front
The question that actually matters gets a straight answer before anything else: our analysis scores AZZA3 at 6.0 out of 10, with a verdict of UNDER REVIEW, an estimated fair value of R$ 26.00 (43% upside from the current R$ 18.21) and a range of R$ 16 to R$ 34. That's neither a "buy now" nor a "run" — it's a value thesis with a condition attached. The stock looks cheap against the value already stored up — leading brands, book equity, profit already delivered — but the trend (revenue and margin direction) is pointing down. The discount only closes if two things hold up: the truce between the founders proves durable, and the second-half 2026 results show the business has stopped deteriorating.
The trade-off, no sugarcoating:
This works for contrarian value investors who can stomach headline-driven volatility and have a two-to-three-year horizon — buying peak pessimism while dividends pay for the wait. It's not for anyone who needs predictability (the company itself pulled its own guidance), wants growth (revenue fell in early 2026), or won't be tracking quarterly results and the ongoing ownership drama. The observable triggers come with dates: the second-quarter (August) and third-quarter (November) 2026 results, and every new twist in the arbitration between the founders.
Why a profitable company trades like it's in crisis
Before getting into the causes, it helps to understand the one number that sums up the discount. The P/BV ratio — price divided by book value, i.e. how much the market charges per real of accounting net worth per share — sits at 0.46x. In cash terms: Azzas' book equity works out to roughly R$ 39.60 per share, but the market is only paying R$ 18.21. It's the equivalent of pricing a mortgage-free R$ 400,000 house at R$ 184,000. That almost never happens to a company that's profitable and pays dividends — and Azzas does both. So the market is saying, loud and clear, that it sees a problem. There are actually two, and they compound each other.
Cause #1 — the founders' war. Azzas was born in August 2024 from the merger of Arezzo&Co (controlled by Alexandre Birman) and Grupo Soma (controlled by Roberto Jatahy). By 2026, the power-sharing arrangement built into that merger had collapsed: Birman reversed, through a material-fact filing, an integration of business units the board had already approved; Jatahy answered with an injunction request and a lawsuit alleging breach of fiduciary duty; and the fight landed at CAM-B3 arbitration (B3's Market Arbitration Chamber) in May 2026. A truce came in June — a new shareholders' agreement and a reshuffled board — but it's only weeks old.
Cause #2 — the operation got worse at the same time. In the first quarter of 2026 (the most recent results released), revenue fell 8%, recurring EBITDA dropped 23%, and recurring profit collapsed 45.7% to R$ 63.9 million. Leverage climbed from 1.28x to 1.40x. And in March 2026 the company officially withdrew its guidance — in practice, admitting it can't forecast its own year. Note that the decline started before the ownership crisis peaked, so the feud alone doesn't explain everything. Middle-class consumers, squeezed by high interest rates, were also simply spending less.
The detail that separates a serious analysis from social-media gossip is this one: a fight between controlling shareholders destroys real value, it isn't a tabloid story. While the two owners argue over who's in charge, strategic decisions — what to do with Hering, how much to invest in each brand, whether to cut costs — get held hostage. A fashion company runs on people and brand equity; good executives don't stick around when the chain of command is split. It's no coincidence that more than eight senior executives have left in the past 18 months. That's also why the management score inside the full analysis is just 4.5/10 — the weak link in the thesis sits at the top, in the boardroom fight covered in the governance section.
The house-of-brands model, explained from scratch
Azzas runs on a house-of-brands model: each label has its own look, customer base and price point, but they all share the same back office — logistics, credit, technology, sales channels. There are more than 20 brands across four main families: footwear and bags (Arezzo, Schutz, Anacapri, Alexandre Birman), premium women's fashion (FARM Rio, Animale, NV, Maria Filó), menswear (Reserva, Oficina, Foxton) and basics (Hering). Sales run through owned stores, roughly a thousand franchises, multi-brand retailers and e-commerce.
What makes this a good business is a 55% gross margin — extremely high for retail. It exists because strong brands carry real pricing power: someone who wants a FARM dress doesn't switch to a rival just because it's R$ 30 cheaper. The Achilles' heel is that clothes and shoes are among the first things a family cuts when the budget tightens. And Hering, the basics brand with the most price-sensitive customer base, is exactly the one most exposed to that squeeze.
A calendar quirk that changes how the numbers should be read: fashion retail has a strong fourth quarter (Christmas, summer, Black Friday) and a weak first quarter (clearance sales, no big holiday). That's why a quarter should be compared with the same quarter a year earlier, not with the one right before it. The 8% revenue drop in Q1 2026 is bad on a year-over-year basis — it isn't simply the normal start-of-year lull.
The numbers, read against the right baseline
This is where it's easy to get fooled, because the comparison base shifts every year. In 2023, the figures are Arezzo standalone, before the merger. In 2024, Grupo Soma is only consolidated for about five months, and the result carries the costs of executing the merger. 2025 is the first full year of the combined group — the base that actually matters for judging performance. And there's one more trap in the profit line, explained right under the table.
| Period | Net revenue (R$ mn) | Gross margin | EBITDA (R$ mn) | EBITDA margin | Profit (R$ mn) |
|---|---|---|---|---|---|
| 2023 (Arezzo only) | 4,847 | 54.2% | 781.7 | 16.1% | 399.4 |
| 2024 (merger in Aug) | 8,380 | 53.3% | 798.5 | 9.5% (merger costs) | 341.7 |
| 2025 (first full year) | 11,819 | 55.0% | 1,941.1 (rec.) | 16.4% (rec.) | 770.7 (rec.) |
| Q1 2026 (quarter) | 2,480 | 54.5% | 328.5 (rec.) | 13.2% (rec.) | 63.9 (rec.) |
| LTM through Mar/26 | 11,602 | ~54.9% | ~1,842 (rec.) | ~15.9% (rec.) | ~717 (rec.) |
Watch out for the 2025 accounting profit. Some sources show a full-year profit of R$ 911 million — but that number is inflated. About R$ 546 million of it comes from a deferred tax credit, an accounting entry that recognizes tax the company expects to avoid paying in the future; it flows through the income statement, but it isn't cash that actually came in, nor profit the operation itself generated. The figure that matters for shareholders is recurring profit: R$ 770.7 million. That's the number to use in any "how many times earnings" math for the stock.
Read that way, the numbers tell a story in three acts. Act One (2025): the first full year delivered a 55% gross margin, R$ 1.94 billion in recurring EBITDA and R$ 770.7 million in recurring profit, 30.5% above the comparable base — operationally, the merger wasn't a disaster. Act Two (the catch): Hering drags the group's margin down by roughly 2 percentage points (without it, EBITDA margin would be 18.4% instead of 16.4%), and financial expenses eat up R$ 782 million a year. Act Three (the warning, Q1 2026): revenue, EBITDA and profit all falling at once, with leverage climbing. The real test is the second half of 2026: if revenue and margin stabilize, the roughly R$ 700-million profit level holds; if Q1 2026 turns out to be the new trend, the cheap valuation shrinks right along with the profit itself.
Why Hering weighs so heavily
Hering deserves its own section because it's the knot in this thesis. It's the group's most mass-market, basics-focused brand, with the most price-sensitive customer base — exactly the segment that suffers most from high interest rates. The result is that the basics vertical drags the entire group's margin down by roughly 2 percentage points. On a company that books R$ 11.8 billion in revenue, 2 points of EBITDA margin are worth more than R$ 200 million a year — that's not a rounding error.
Grupo Soma bought Hering for R$ 5.5 billion in 2021. Today the entire combined group is worth roughly R$ 3.8 billion on the stock market. There's a parallel subplot, too: according to press reports, a group linked to Hering's founding family hired investment bank BR Partners to explore buying the brand back — JP Morgan estimates the deal at R$ 600-800 million. The company's response was that Hering "is not for sale at this time." It's a brutal dilemma: selling cheap locks in a real destruction of value; not selling — and failing to turn the brand around — drags out the margin bleed. It's the hottest seat in the whole company right now.
So what's it actually worth? The valuation, assumptions on the table
AZZA3 is the textbook case where every multiple tells a different story. On book value, it looks absurdly cheap (0.46x). On 2026's depressed earnings, less so. The way through is to estimate a mid-cycle profit — neither the peak nor the trough — and cross-check it with independent methods. All of them use recurring profit, never the accounting figure inflated by the deferred tax credit.
| Method | Fair value | Core assumption |
|---|---|---|
| P/E on normalized earnings | R$ 26.00 | Mid-cycle recurring profit of ~R$ 720 mn × 7.5x P/E, discounted for governance risk |
| EV/EBITDA (normalized) | R$ 25.00 | ~R$ 1.85 bn recurring EBITDA × 4x (healthy peers trade at 6-8x), minus R$ 2.2 bn in net debt |
| P/BV × profitability | R$ 25.50 | R$ 39.60/share book value with ~9% recurring ROE, still below the cost of capital |
| Reference: sell-side consensus | R$ 34.81 | Median of 13 analysts · Itaú BBA R$ 36 and XP R$ 40 (buy) · Citi R$ 28 and Safra R$ 26 (neutral) — sanity check only |
The three methods we built ourselves converge on a R$ 25-26 range, and the full fair-value band runs from R$ 16 (floor) to R$ 34 (ceiling), centered on R$ 26.00 — 43% upside. Note that our number sits below the market's consensus (R$ 34.81): sell-side targets assume a stabilization that the Q1 2026 numbers haven't shown yet. At R$ 18.21, the stock trades below even the most conservative sell-side target (Safra's R$ 26) and close to its post-merger all-time low. Buying under R$ 17.50 builds in a real margin of safety even in the base case; above R$ 26.00, an investor is prepaying for a corporate peace and an operational stabilization that haven't happened yet. All the math, method by method, is in the valuation tab of the full analysis.
Three scenarios, three possible endings
| Scenario | Target price | Probability | What has to happen |
|---|---|---|---|
| Bull — peace and a Hering turnaround | R$ 36.00 | ~25% | Arbitration settled, revenue growing again, margin holding above 16%, Hering turned around or sold at fair value, Selic (Brazil's benchmark interest rate) falling |
| Base — a truce that holds | R$ 26.00 | ~50% | A cold but functional truce, revenue stops falling in H2 2026, recurring profit of R$ 600-750 mn, leverage stable |
| Bear — the feud reignites | R$ 14.00 | ~25% | Arbitration escalates, revenue keeps falling, profit under R$ 550 mn, dividend cut, Hering sold off cheap |
Notice the asymmetry: the base case already delivers +43% upside, the bull case nearly doubles the price, and even the bear case (R$ 14) is "only" -23% from today's price — because much of the damage is already in the price. But there's a warning that rarely gets mentioned: in the bear scenario, the book-value discount stops acting as a floor. A meaningful share of Azzas' equity is goodwill and brand value — intangible assets that, in a contentious breakup or a fire-sale of Hering, are worth less than the balance sheet claims. A cyclical company caught in a conflict tends to blow through its "rational" floor exactly during a panic.
The founders' feud, without the euphemisms
It's worth explaining the CAM-B3 arbitration that shows up in every piece of news about Azzas. It's the Market Arbitration Chamber run by B3, the private venue Novo Mercado-listed companies use to settle ownership disputes instead of going through Brazil's regular courts — faster and confidential, but not painless. Roberto Jatahy's role at the group is, literally, one of the things being arbitrated. While the case runs, the question of who actually gets to decide what hangs over the company.
On one side is Alexandre Birman (CEO, his family holds around 21% of the capital), son of Arezzo's founder, with a centralizing style — it was his unilateral call to reverse the integration of business units that set off the crisis. On the other side is Roberto Jatahy (Chief Brand Officer, roughly 10% with his sisters), the founder of Grupo Soma, with a more collaborative style. Two owners used to calling the shots on their own, now forced to share one company. The June 2026 truce rebuilt the shareholders' agreement and elected a new board chaired by independent director Sylvia Leão Wanderley — but with both founders sitting on the same board and only two formally independent directors out of nine members, the counterweight is thin. That's why every new chapter in the arbitration moves the stock 3% to 6%.
Dividends: here, the income really counts
Unlike many value stories, at Azzas the dividend is a real part of the return. In 2025 the company returned R$ 667 million to shareholders (R$ 500 mn in dividends plus R$ 167 mn in buybacks), which added up to R$ 2.48 per share over 12 months — a 13.6% dividend yield on today's depressed price. Because the stock is so cheap, every real distributed translates into a high percentage return. And the buyback is still active, purchasing shares below 0.5x book value — a signal that management itself sees the stock as cheap.
The catch: don't count on 2025's payout size repeating. With 2026 earnings coming in well below that level, the distributable base shrinks. A realistic expectation for the next 12 months is a dividend of R$ 0.90 to R$ 1.80 per share (a yield of roughly 5% to 10%). Still meaningful — the company keeps generating cash, has trimmed capex and keeps buying back stock — but far from 13.6%. Here, the dividend is the "cherry that pays for the wait" while the arbitration plays out, not the central reason for the thesis.
The verdict, and what to watch
Azzas 2154 is a rare case of excellent assets sitting inside a broken ownership structure. The brands are category leaders, the 55% gross margin belongs in the big leagues, the group booked R$ 770 million in recurring profit in 2025, and it returned R$ 667 million to shareholders. But the merger that created this giant also created a two-owner arrangement that doesn't get along — and 2026 exposed that in full, with lawsuits, arbitration, and the business deteriorating right alongside it. The stock is cheap on stored-up value; the open question is about the trend.
Verdict: UNDER REVIEW — score 6.0/10. A value thesis with a corporate catalyst, suited for a small position, a strong stomach and active monitoring. Estimated fair value of R$ 26.00 (range R$ 16 to R$ 34). Buying today means buying peak pessimism with dividends paying for the wait; investors who prefer more certainty can wait for Q2 2026 and confirmation that the truce is holding, probably paying 20-30% more in exchange for a lot less headline risk.
What to watch, with dates: the Q2 2026 results (August 2026) — did revenue stabilize? did recurring EBITDA margin clear 14%?; the Q3 2026 results (November) as confirmation; and any material filing tied to the CAM-B3 arbitration or the Hering saga, both of which are immediate re-rating triggers.
This piece summarizes the full analysis. The detailed report — fundamentals, risks, governance, scenarios and brand-by-brand valuation — is in the complete AZZA3 analysis. There's also a PDF report, produced by the site with the help of artificial intelligence. Nothing here is investment advice or an official document from Azzas 2154 S.A.; it's a reading of public filings through July 21, 2026, and the decision to invest is always yours.
