Safilo's Buyback Machine: Why a Thin-Float Eyewear Maker Keeps Buying Itself

Safilo, an Italian eyewear manufacturer you've almost certainly worn but may never have heard of, now owns 6.28% of its own company.

That is not a typo. Safilo has been systematically buying back its own shares on the Milan stock exchange for three consecutive years. The latest update, at the start of September, reports another 400,000 shares repurchased in late August, pushing the treasury pile to just over 6% of outstanding capital. It is part of a €20 million buyback program that runs through the end of the year.

The basic point is that this is not a company saying "our stock is cheap" and pressing a button once. This is a deliberate, repeatable machine — a way for the controlling shareholder to concentrate ownership, shrink the free float, and keep a reserve of shares on the shelf for employee plans, acquisitions, or whatever comes next. And it is happening while the underlying business is losing sales momentum.

That tension — buying back shares while revenue declines — is the useful thing to understand about Safilo right now.

What treasury shares actually are (and are not)

A company holding its own shares sounds like a circle. Legally, the circle has edges. Treasury shares do not vote. They do not receive dividends. They are not counted in earnings-per-share calculations. On the balance sheet, they are recorded as a reduction in equity — not an asset.

So when Safilo says it holds 6.28% of its share capital, the right way to think about it is that those shares have been pulled out of circulation. The remaining shares — the ones actual investors hold — now represent a slightly larger slice of the company's economics.

But "slightly" matters. At €1.89 per share, those repurchases are adding up slowly.

Three years of the same trade

The 2026 program was authorized in April and launched in June, capping out at 10 million shares for a maximum of €20 million. By late August the company had repurchased roughly 1.6 million shares under the program — a steady pace for a program that runs through year end.

Before that, the 2025 program bought back roughly 11.5 million shares for €16.8 million (completed December 2025). The 2024 program repurchased a similar number for €11.8 million at an average price near €1.07 per share.

Across the three programs, that is roughly €42 million of buybacks. But the share price itself has moved — from under €1.10 in 2024 to nearly €1.90 by August 2026 — so the company has been buying fewer shares per euro as it goes. (Or you can say the shares were cheaper two years ago. The company would prefer the second reading.)

The whole program is executed through an independent broker — Kepler Cheuvreux — which handles the timing and keeps the company from looking like it is directing trades at a specific moment. It is the standard Italian-listed-company setup: shareholders authorize the pot, a broker pulls the trigger, the company publishes weekly receipts.

Who actually controls the company

Here is where the plumbing gets interesting. HAL Holding N.V. owns 49.5% of Safilo — 206 million shares. The top 25 shareholders together control 72% of the company. That leaves a free float of roughly 28%.

Now subtract the treasury shares. If treasury shares are 6.28% of the total outstanding, the actual float available for everyday trading is closer to 22%. A company with €983 million in annual revenue and nearly €1 billion in market value trades on a genuinely thin float.

The controlling shareholder — HAL Holding — does not pay for the buybacks. Safilo does. So every euro the company spends on repurchases slightly increases HAL Holding's effective stake in the remaining economics, without HAL writing a check. It is a capital-management tool that benefits the largest owner disproportionately, because the float is small to begin with.

This is not unusual for European family-controlled companies. It is just worth noticing that "shareholder return" and "return to the controlling shareholder" are not the same thing when 49.5% of the votes already go one way.

The business while the buyback runs

Safilo makes sunglasses and prescription eyewear under both its own brands — Carrera, Smith, Polaroid, Serengeti, Spy+ — and a licensed portfolio that includes Gucci, Dior, Celine, Tommy Hilfiger, Marc Jacobs, Kate Spade, and Victoria Beckham, among others. About 95% of those licenses are secured through 2030. The business generates roughly €1 billion in annual revenue from design, manufacturing in Italy, and global distribution.

The recent financial picture is mixed in a way that makes the buyback more notable.

Sales are declining. First-half 2026 revenue fell 1.9% at constant exchange rates, and the second quarter alone dropped 4.5%. Every region except the Middle East saw declines. Europe, the largest market, was down 3.2% for the half. North America fell 6.1% in the quarter. Asia-Pacific, which had been a growth star in 2025, plunged 17.2% in the first half of 2026.

But margins surged. The reported H1 gross margin jumped to 67.2%, up 6.1 percentage points. The reported adjusted EBITDA margin hit 16.8%. Group net profit rose 47%.

The reason is a one-time event: a €22.2 million refund of U.S. tariffs that Safilo had previously overpaid. Twenty million of that went straight through the income statement as a cost reduction, adding 8 percentage points to Q2 gross margin alone.

Strip the refund out and the underlying H1 gross margin improvement is 2.3 percentage points. The underlying EBITDA margin is 12.9% — still up from the prior year, but nowhere near the headline number. Management has said any further tariff refunds in the second half will be "fairly residual."

The economic point is that the buyback is being funded partly against a margin story that included a significant one-time credit. The cash flow number — €36.4 million in H1 free cash flow, or €46.9 million "normalized" — is real. But the profit expansion that makes the stock look like it is turning a corner is partly a government check, not a structural improvement.

What Safilo is spending the money on, besides itself

While buying back its own stock, Safilo has also been buying other companies' stock. In July it completed acquisitions of Spy+ and Serengeti from Bollé Brands for $24.6 million — those two brands generated about $39 million in combined 2025 sales. The deal was funded entirely from internal resources. Management said the new brands will be "dilutive to gross margin" but not materially so — the honest call of a company buying revenue and brand portfolio breadth, not margin expansion.

The structural question for a watcher

The useful frame here is not whether buybacks are good or bad — they are a tool, and the judgment depends on what else is happening.

Safilo's pattern is this: a cash-generative business with thin growth, a controlling shareholder at 49.5%, and a management team that has chosen to recycle cash into share repurchases and small acquisitions rather than dividends or major expansion. The buybacks mechanically improve per-share metrics. They shrink the float. They give management a stockpile for future use. And they quietly concentrate the economics in favor of the shareholder who already controls the company.

The risk side is straightforward. Revenue is declining, and the margin expansion that makes the current financials look strong is partly one-time. The buyback program gives the company flexibility — shares held in treasury are not a permanent decision, and they can be reissued for employee plans or acquisitions — but the money spent on repurchases is money not sitting on the balance sheet.

A watcher should care about three things going forward: whether the underlying (non-refund) margin improvement proves structural, whether H2 sales actually recover as management expects, and whether the buyback program continues at this pace once the current one expires in November. If sales stabilize while the company keeps pulling shares off the market, the per-share economics improve. If the revenue decline deepens, the buyback becomes less a value-creation tool and more a control-preserving one.

The stock traded around €1.88 in late August, roughly 19% below its peak earlier this year. The stock is not doing anything dramatic right now. The interesting part is the machine running underneath it.