Chapter 4

Owners and Stewards

Figures converted from Swiss francs at historical FX rates — see data/company.json.fx_rates for the rate table. Ratios, margins, multiples, ownership percentages, and share counts are unitless and unchanged.

Sonova's two co-founding families still own roughly 17% of the company, worth about $2.7 billion — a large, patient stake that a value investor should weigh. But those families sit outside the boardroom, and the people who actually run and govern Sonova hold almost none of it: the full board and executive team own about 0.10% between them. Pay is modest and tied to returns on capital. The record of what management did with surplus cash is the weaker point — it bought back stock and a consumer-audio business near the top of the cycle.

This chapter answers two of the reader's standing questions directly — how much insiders own, and what management is paid — and then tests the part that matters most for a company trading near half its peak: whether the people allocating the capital can be trusted to do it well.

Who owns Sonova

The register still carries its founders. Beda Diethelm and the family of Hans-Ulrich Rihs — both shareholders since before the November 1994 IPO of what was then Phonak Holding — together hold 10,396,572 shares, about 17.4% of the company [1]. At $257 a share that stake is worth roughly $2.7 billion. For an investor who prizes owners with real money at risk, this is the single most important ownership fact in the file, and one an outside screen for "no controlling shareholder" would miss.

Source: significant shareholders and share-ownership tables, FY2025/26 Annual Report [2], [3].

The catch is that the founders are passive. The filing is explicit: Diethelm and Rihs "can trade freely," there are no shareholders' agreements binding them, and neither family holds a board seat or an executive role [4]. Their stake aligns them with outside shareholders economically, but they exert no visible control over strategy or capital allocation. Sonova is not, in any operating sense, a founder-run company.

The people who do run it own very little. At 31 March 2026 the entire active Board of Directors held 53,378 shares — worth about $14 million across ten directors, with the new chair, Gilbert Achermann, the largest holder at 22,520 shares [5]. The eight Group Executives held 7,259 shares outright, plus 39,787 options and 5,764 performance share units; CEO Eric Bernard's direct holding was 4,685 shares [6]. Directly held board and executive shares come to roughly 60,600 — about 0.10% of the company, worth some $15.6 million against a market value near $15.3 billion.

Source: derived from the significant-shareholders and share-ownership tables, FY2025/26 Annual Report, at the 23 Jul 2026 close of $257 [7], [8].

Sonova does impose share-ownership guidelines on its directors and executives, and reports that in March 2026 one director and three executives were "marginally below" target after the share-price fall — a live sign the guidelines have teeth, even if the absolute holdings are small [9]. The honest read for a skin-in-the-game investor: the founders supply the alignment, but as passive holders; the fiduciaries running the business are agents, not owners.

What management is paid

Pay is not the problem here. For the 2025/26 year, CEO Eric Bernard — appointed mid-year, on 15 September 2025 — received total compensation of $3.3 million, against a full-year base salary set at $1.25 million [10]. All eight Group Executives together were paid $10.6 million, comfortably inside the $21.0 million ceiling shareholders had approved [11]. For a company earning north of $680 million a year, this is a restrained package.

Source: Compensation Report, FY2025/26 Annual Report; say-on-pay is the 2025 AGM binding vote on executive compensation [12], [13].

Two features matter more than the absolute amounts. First, the variable half genuinely varies. Group Executives' cash bonus paid out at 88.2% of target in 2025/26, but the five-year record runs from 85.7% down to 35.3% in the weak 2022/23 year — the plan pays less when the business underperforms [14]. Second, the long-term equity is gated on return on capital: performance options vest against a hard ROCE target, and the 2022 tranche vested at only 96.5% because the 24.0% target was narrowly missed (actual 23.79%) [15]. A ROCE gate ties the executives' payout to the same capital-efficiency the shareholder cares about.

Source: Compensation Report §6.2.2, FY2025/26 Annual Report [16].

Shareholders have signalled they are content: at the 2025 AGM they approved board compensation with 94.41% of votes, executive compensation with 89.97%, and the compensation report itself with 92.20% [17]. Pay is not where a skeptic should spend time.

What management did with the cash

Capital allocation is where the record turns mixed. Sonova is a strong cash generator, and over the five years to March 2026 it returned roughly $2.65 billion to shareholders — about $1.43 billion in dividends and $1.22 billion in buybacks — while cutting net debt and lifting the dividend to a record $5.88. On the surface, disciplined. The timing is the issue.

The buybacks were front-loaded into the two years when the stock was most expensive. Sonova spent $733 million in 2021/22 repurchasing 2,012,438 shares at an average near $364, and $487 million in 2022/23 on 1,532,910 shares at an average near $318 [18], [19]. Both averages sit 25% to 40% above the $257 the shares trade at today. Then the buying stopped: the $1.9 billion 2022–2025 programme was left roughly 70% unused and expired on 18 April 2025, and no shares were repurchased in 2025/26 — precisely the window in which the stock was cheapest [20].

Source: buyback volumes and cash from FY2022/23 Annual Report cash-flow and share-capital notes; current price is the 23 Jul 2026 close [21], [22].

The pattern shows up in the shape of the returns over time. Buybacks were a large slice of the payout in FY2022 and FY2023 and then vanished, leaving the dividend to carry the whole of shareholder returns through the de-rating.

Source: consolidated cash-flow statements, FY2021/22–FY2025/26 Annual Reports [23], [24].

The dividend record has its own asterisk. The per-share distribution was not a clean climb through the downturn: it rose to $5.02 for FY2023, was cut to $4.76 for FY2024, and only then rebuilt to $4.99 and a record $5.88 [25]. The commitment to the payout is real, but it bent once when profits dipped.

The clearest allocation misstep is a round trip that other chapters have flagged in passing. In March 2022 Sonova paid a purchase consideration of $352 million for Sennheiser's Consumer Division, a headphones-and-hearables business [26]. Four years later, on 23 March 2026, it announced it would divest that business; the exit and related impairments produced a loss after tax from discontinued operations of $133 million in FY2026 alone [27]. Capital was deployed into a diversification at the top of the consumer-electronics cycle and is being withdrawn at a loss, having consumed management attention along the way.

Set against that, the core reinvestment record is sound. Research and development runs near 6% of sales, capital expenditure below 3%, net debt fell from $1,632 million at the FY2023 peak to $1,243 million, and return on capital employed — while down from 24.1% to 19.0% through the margin trough — remains high for a medical-device maker [28]. The money spent inside the business has compounded; the money spent on its own shares and on Sennheiser was spent at the wrong price.

The read for this investor

For a reader who wants founders with skin in the game and a large margin of safety, Sonova lands in an unusual middle. The alignment box is half-ticked: two founding families still own about 17% and roughly $2.7 billion of stock, which is a genuine positive an ownership screen would flag — but they are passive, and the executives and directors steering the company own almost nothing. Pay is modest and ties to return on capital, which limits the risk of value being extracted through the compensation line.

The capital-allocation record is the honest caution. It is not reckless — leverage is low, the dividend is defended, the core is well funded — but the two big discretionary calls of the cycle were mistimed, and a management that bought its own shares 40% above today's price is not one that has shown it buys value when value is on offer. What would change this read is behaviour, not disclosure: a fresh buyback authorisation actually executed at today's depressed prices, or a clean, on-terms Sennheiser exit, would show the newly installed CEO and chair allocating differently from their predecessors. Until then, the alignment case rests on the founders' patience rather than on the stewards' record.