The 2022 Shopify AGM put the spotlight on two controversial theories for driving corporate success: the founder-CEO and multi-class share structures.

When it went public in 2015, Shopify’s multi-class structure was fairly standard for an aspirational tech unicorn, if still unusual relative to capital markets as a whole. In recent years an increasing proportion of companies going public in the United States and Canada have opted for multi-class (also known as “dual-class”) share structures that provide superior voting rights to a designated share class over “subordinate” shareholders who buy in at or after the IPO. The designated superior shareholders are typically the founder(s) and early investors.

How did that work in practice? Prior to the 2022 AGM, Shopify’s founder and CEO, Tobi Lütke, together with fellow board member John Phillips, effectively dictated the outcome of all shareholder voting resolutions through their ownership of Class B Multiple Voting shares, which carry 10 votes each. Class B shares are 9.5% of outstanding equity but 51% of voting rights, with Lütke alone worth 34% of voting power. Both hold minimal Class A Subordinate voting shares. Class B shares can be converted on a 1:1 basis into Class A (but lose their multiple votes) for the purpose of being sold by the holder.

In most cases, the multi-class share structures are intended to be temporary, allowing the founders and early investors to maintain their influence over the company as it transitions into publicly listed status. In Shopify’s case, the company has become Canada’s greatest stock market success story of the high-tech era since Blackberry. And despite (or perhaps because of?) the successful transition, the board and management didn’t want to rock the boat quite yet.

Under Shopify’s former covenants, the multi-class structure would automatically collapse when Class B shares represent less than 5% of all outstanding equity (the “Dilution Sunset”). This event was deemed likely to occur in the coming months or years (we believe it was more likely to be sooner than later).

Why was the Dilution Sunset approaching? Because of the ballooning of the Class A share count since Shopify’s 2015 IPO, primarily through Class B shareholders converting to Class A and substantial equity-based employee compensation, all of which have reduced the Class B share capital proportionate to the aggregate outstanding equity, thereby making the collapse of the structure more likely as the 5% threshold nears.

At its June 7, 2022 annual shareholder meeting, Shopify proposed a drastic and novel rearrangement of its capital structure to pre-empt the Dilution Sunset and preserve the disproportionate voting power of the CEO.

The board, after negotiating with the CEO, recommended an arrangement under which Shopify would issue to the CEO a “Founder share” cementing disproportionate voting rights of 40% for life, provided that Mr. Lütke serves as at least one of: (i) board member, (ii) executive officer or (iii) consultant whose “primary engagement” is with Shopify.

Not only that, but Mr. Lütke would be permitted to reduce his economic stake by up to 70% of his current holdings, which currently equate to 6.3% of all outstanding equity, without giving up a single percentage point of his total voting power.

What was the board thinking?

It doesn’t take an expert in hostage negotiations to see that the independent board members were dealing from a position of relative weakness, or at the very least deference, compared to the already-powerful CEO.

For starters, as a matter of basic arithmetic, all Shopify board members owe their presence on the board to the support, or at least acquiescence, of the Class B multiple voting shareholders.

The special committee of independent directors also signposted early on that Mr. Lütke’s position as the individual “guiding the ship” was paramount. During negotiations lasting over the course of a year, the committee generally did not initiate any proposals to the CEO; rather, it waited for Lütke to propose terms. His first offer was to set and preserve the multiple voting power at 49.99%. Aside from one counter-proposal for 34% alongside a tighter service-based sunset and stricter divestment restrictions than those desired by Mr. Lütke, the committee did not push back. The final terms are as described above.

Clear from this is that the board places enormous faith in Mr. Lütke.

The history of corporate (North) America is replete with examples of one-company CEOs. The mythology of the self-made founder-CEO speaks to what for many are the most persuasive qualities of capitalism America-style. Shopify’s rise has seen it and Mr. Lütke compared to similar modern software and internet success stories with long-serving founder-CEOs, notably Microsoft (Bill Gates), Apple (Steve Jobs), Facebook (Mark Zuckerberg) and Amazon (Jeff Bezos). Of the aforementioned, all but Facebook have simple, equitable single class share structures, and an examination of their long-term record suggests that they were able to execute a vision that delivered enormous value to shareholders despite having one vote for each share.

Mr. Bezos maintained a substantial stake – enough to exercise effective control through an equitable structure – over many years; Mr. Bezos held approximately 25% in 2007, 19% in 2013 and today still holds approximately 13% despite having stepped down as CEO in 2021. Even Zuckerberg owns around 13% of all outstanding Meta Platforms (fka Facebook) stock, even though he’d need less than that to still guarantee effective control.

Shopify’s proposal to Mr. Lütke allows him to maintain effective control even while selling down up to 70% of his current 6% stake.

The Theory

The debate over whether multi-class structures should be permitted is contested. Proponents of either side can point to individual companies that have performed especially well or poorly, or indeed to singular events like the Succession-style blow-up in 2021 at Rogers Communication; in any case, performance assessments are relative, with some such companies, particularly in Canada, using a dual class structure to embed a founding family’s control of the business for multiple decades.

Like fractal geometry, the debate doesn’t end at the largest scale of the topic – i.e. should multi-class structures be permitted or not – but diverges into sub-debates around areas of best practice for application of the structure: what should be the ratio of voting superiority between the special class and the subordinate class? What form of sunset (expiry date) should be adopted – should it be service-based (e.g. the founder-CEO leaves their role), time-based (such as the seventh anniversary of the IPO) or dilution-based (where the multiple-voting shares automatically dissolve into the common stock class when they represent less than, for instance, 5% of all shares outstanding)? Should the sunset be a mix of all these?

While Glass Lewis in 2022 adopted a voting policy against multi-class structures without a reasonable time-based sunset (in our opinion, seven years or less from IPO), we are sympathetic to criticism of time-based sunsets as overly arbitrary.

In Canada, among the most thoughtful arguments against a blanket negative approach to dual class companies are those made by the Institute for Governance of Private and Public Organizations (“IGOPP”). The IGOPP has argued that companies with dual class structures are just as likely, if not more so, to deliver long-term growth as one-share-one-vote companies (see “Policy Paper No. 11: The Case for Dual-Class of Shares”, 2019.)

IGOPP believes there is “much merit to dual-class companies and family firms, provided holders of shares with inferior voting rights are well protected. [IGOPP’s emphasis].” IGOPP’s recommendations for ensuring the protection of subordinate shareholder rights include:

  • That voting strength of superior shares be capped at a ratio of 4:1, such that 20% of the equity would be required for absolute control (50%+ of votes). IGOPP implies that a ratio of greater than 10:1 is unacceptable and states that classes of shares with no voting rights should be prohibited.
  • Public disclosure of separate vote tallies for each class of shares.
  • That subordinate shareholders be entitled to elect one-third of board members.
  • Dilution sunset that would be triggered by the controlling shareholder’s voting power dropping below a percentage generally considered as a “blocking minority” – 33% (which would be 11.1% with a 4:1 ratio).

Shopify’s proposed Founder share and accompanying “governance update” does not include any of the above safeguards.

Glass Lewis View

In reviewing the proposed Founder share arrangement, we were struck by the degree to which the original dual-class structure (i.e. the Class A 1 vote / Class B 10 votes) appears to have aligned with its intended aims, with shareholder expectations and with those of other best practice standard-setters.

Shopify’s performance since listing publicly has been remarkable. So, to the extent that its dual-class structure insulated its early investors and leadership from short-term pressures, it fulfilled its purpose. Shareholders of Class A subordinate shares have been taken along for a glorious ride as Shopify became one of the main success stories in an extended bull market in which software and e-commerce companies outperformed. The company benefitted from a favourable consumer perception, peaking during pandemic-induced lockdowns, as the ‘anti-Amazon’ since its platform enables small businesses to sell to their customers directly, rather than having to compete in a single online marketplace and outsource order fulfilment.

As is often the way in the early years of a public company, particularly for a tech concern with a seemingly rocket-fuelled stock, common shareholders probably didn’t mind that their votes at general meetings were effectively meaningless.

Glass Lewis believes in the principle of one share, one vote, but our policy guidelines clearly earmark the period in each public corporation’s lifecycle in which an inequitable voting structure might be desirable: the early years. We agree that freeing up the founding leadership to concentrate on growing the business and trying to realize its vision is a strategy that works for some companies.

While in 2022 we settled on a seven-year time-based sunset as a general policy for newly public dual-class companies, sunset mechanisms may achieve a similar result via a different route.

If Shopify’s original dilution sunset is indeed close to triggering – and the board’s clamouring to propose the Founder share at this AGM suggests that it is – then it is doing so uncannily close to the seven-year anniversary of Shopify as a public company.

One of our main concerns with multi-class structures is that they may foster misaligned incentives between the superior voters and dispersed common shareholders, and that these can be exacerbated over the long-term. This is especially true of companies where the subordinate class has zero votes per share or where the gulf in votes per share is especially large (as noted above, best practice proponents recommend not exceeding 4 votes per share in order to preserve a semblance of economic alignment between superior and subordinate shareholders).

As such, a well-crafted dilution sunset makes sense: the superior shareholders may decide to use their disproportionate control of the company to, for example, issue a deluge of stock options to employees as the company grows and competes for talent. They may be totally justified in doing so (we’ve heard that software engineers can be expensive and that they tend to expect equity compensation). But a dilution sunset at least keeps the experience of the superior shareholders tethered to that of ordinary shareholders, who have good reason to be concerned both about excessive dilution and about the spectre of a shareholder with a small economic stake exercising effective control.

What Happened

Shortly after the meeting, various news outlets began reporting that the arrangement had received sufficient shareholder approval. Only the following day did it become clear just how fine the margin was.

The meeting materials noted that to be approved, the arrangement would need the support of both 2/3 of all votes cast by Class A and Class B voting “as a single class” AND a majority of the votes cast by Class A and Class B voting “as a single class”, excluding Mr. Lütke and associates. Per Shopify’s reported voting results, the votes were cast as follows:

Two-Thirds Majority of the Votes Cast by Class A and Class B Shareholders

Majority of the Votes Cast by Class A and Class B Shareholders Excluding the Company’s Founder and Chief Executive Officer, Tobias Lütke, and His Associates and Affiliates

Crucially, the words “as a single class” did not translate into one-share one-vote, meaning that the multiple voting shares maintained their 10-votes-to-1 advantage. Even more crucially, as it turns out, the other major Class B shareholder, Klister Credit Corp (“Klister”), which is controlled jointly by Shopify’s pre-IPO director John Phillips and his wife Catherine, was allowed to vote along with ordinary shareholders for the purposes of measuring “disinterested” shareholder sentiment.

Since Shopify does not provide a breakdown of votes cast by class – and indeed the CEO rejected even this meagre governance enhancement as part of the Founder share arrangement – we have assumed that Klister voted its multiple voting shares in favor of the arrangement and have estimated the breakdown below:

All Votes Cast by Class A and Class B Shareholders Including Lütke – if shares were cast on a 1:1 basis

Votes Cast by Class A and Class B Shareholders Excluding Lütke and Affiliates (but including Klister Credit Corp, controlled by director John Phillips and his wife)

With the Dilution Sunset averted and the Founder share approved, Klister will now be free to reduce its stake.

Relatively little light was shed by the Company on the role of Klister in precipitating discussions around the arrangement (since any selling-down by Klister of its Class B shares pre-arrangement would have brought the Dilution Sunset much closer), other than a laconic assurance about Mr. Phillips’ intention to “remain a long-term shareholder”.

In any case, the vast majority of minority shareholders who opposed the new multi-class arrangement, only to see it edged over the line by the existing multi-class structure, may feel aggrieved by the outcome. Given that the new capital structure has significant implications for Klister, it is also questionable that the Phillips family was permitted to participate with a 10:1 advantage in the vote of “disinterested” shareholders.

While it doesn’t represent a majority of the voting stock, a 40% stake is generally enough to exercise effective control of a public company in which not every shareholder will vote. This is particularly true of companies with a large retail investor following (currently 29% of all Shopify shareholders, according to Capital IQ). As such, non-affiliated shareholders at Shopify will have relatively little recourse to influence governance at the company going forward.

***

For more on Glass Lewis’ approach to multi-class voting structures and post-IPO transitions, see our current proxy voting policy guidelines. Or contact us to learn about our approach to corporate governance  and proxy voting:

GROW@glasslewis.com (Institutional Investors) | ENGAGE@glasslewis.com (Public Companies)