Dual-Class Stock and Controlled Companies: Voting Power, Sunsets, and the Limits of Founder Control

Document type: Article Practice area: Corporate — Corporate Governance Jurisdiction: United States (Delaware, exchange listing standards) Last reviewed: 5 September 2026

Separating money from control

The default rule of corporate law is one share, one vote. It is a default, not a mandate. A Delaware corporation may issue classes of stock with whatever voting rights its charter specifies, including no votes at all, and including ten votes per share.

A dual-class structure exploits that flexibility to separate economic ownership from voting control. The founders hold Class B shares carrying ten votes each. The public holds Class A shares carrying one vote each. The founders sell most of their economic interest over time and retain a voting majority indefinitely.

The arithmetic is stark. A founder holding 12% of the economic interest through ten-vote shares can hold 57% of the voting power. A founder holding 6% can still hold a voting plurality large enough to control every outcome in practice.

Why companies do it. The stated rationale is protection from short-termism: a founder who need not worry about a quarterly miss or an activist campaign can invest in long-horizon projects. The evidence on whether this produces better outcomes is genuinely mixed, and it is not the purpose of this article to resolve it. What matters practically is that the structure is available, that investors have accepted it in exchange for access to certain offerings, and that it creates a set of legal problems that recur.

Why investors object. Voting power is the ultimate check. A stockholder base that cannot replace the board cannot discipline management, cannot accept a premium offer the founder opposes, and cannot force a change of strategy. The governance protections that ordinarily substitute for the market — independent directors, say-on-pay, proxy access — all operate through votes the controller wins.

Building the structure

The classes

A typical structure:

Class A common stock, one vote per share, sold in the offering and traded publicly;
Class B common stock, ten votes per share, held by founders and pre-IPO holders, not publicly traded, convertible into Class A at any time at the holder's election and automatically on transfer;
Sometimes a Class C, non-voting, used for equity compensation and acquisitions so that issuances do not dilute the founder's voting power.

The Class C innovation deserves note. Where a company issues equity for acquisitions and compensation, a two-class structure dilutes the founder's percentage over time. Introducing a non-voting class means the founder's voting percentage is unaffected by issuance, which extends control indefinitely without further purchases.

Transfer restrictions and automatic conversion

The high-vote shares are only valuable if they stay with the founder. The charter therefore provides that Class B converts automatically into Class A upon any transfer, subject to permitted transfer exceptions:

Transfers to family members and trusts for their benefit;
Transfers to entities controlled by the holder;
Transfers to a partner in a fund that holds Class B, in a distribution;
Transfers by will or intestacy.

The permitted transfer definitions are where the estate planning happens, and they are worth close attention. A definition that permits transfer to "any entity controlled by the holder" allows a founder to place shares in a family office or foundation without conversion. A definition that permits transfer to a "family member" defined broadly can extend high-vote control into a second generation.

Sunset provisions

A sunset converts the high-vote shares to one vote per share on the occurrence of a specified event. Sunsets have become the main negotiating point between issuers and investors.

Sunset type	Trigger	Effect
Time-based	A fixed number of years after the IPO, typically seven to ten	Conversion of all Class B, sometimes subject to a stockholder vote to extend
Ownership-based	Founder's holdings fall below a stated percentage of the original amount or of the outstanding shares	Conversion of all Class B
Transfer-based	Any transfer outside permitted transfers	Conversion of the transferred shares only
Death or incapacity	The founder dies or becomes incapacitated	Conversion, sometimes after a transition period
Departure	The founder ceases to serve as an officer or director	Conversion, sometimes after a period
Event-based	A specified corporate event	Varies

The death and departure sunsets are the ones investors care most about, because the entire justification for the structure is the founder's specific vision. A structure that passes ten-vote shares to heirs who have never worked at the company is difficult to defend on any theory.

Drafting points that matter:

Define incapacity precisely, and identify who determines it — a physician, a committee of independent directors, or a court.
Address a transition period after death so that an estate is not forced to act immediately.
Consider a "low-vote" intermediate step rather than a cliff, converting ten votes to three, then to one.
Address the interaction with permitted transfers. A founder who transferred to a trust before death should not escape the death sunset because the trust did not die.
Extension mechanics. A time-based sunset that can be extended by a vote of the Class A holders is meaningfully different from one that can be extended by a vote of all classes together, which the founder controls.

The external constraints

Exchange listing standards

Both major exchanges permit dual-class structures at listing. What they prohibit is disparate reduction — a listed company generally may not reduce or restrict the voting rights of an existing class through a recapitalization, exchange offer, or charter amendment. In practice this means the structure must be in place before or at the IPO; a company cannot go public one-share-one-vote and later create high-vote shares for its founders.

Listing standards also provide controlled company exemptions. A company in which more than 50% of the voting power is held by an individual, group, or another company need not comply with requirements for:

A majority-independent board;
A fully independent compensation committee;
A fully independent nominating and governance committee.

Note what is not exempted: the audit committee independence requirements apply regardless. Every listed company needs a fully independent audit committee, and the controlled company exemption does not reach it.

A practical observation. Many controlled companies decline to use the exemptions, maintaining a majority-independent board and independent committees voluntarily. This is usually the right choice: the exemption saves nothing operationally, and giving it up removes an argument in every subsequent controller-transaction dispute.

Index providers

The most significant external constraint on dual-class structures came not from regulators but from index providers, which announced policies restricting or excluding companies with unequal voting structures from certain widely tracked indices. Exclusion matters commercially, because index inclusion drives a large and price-insensitive source of demand.

The policies have evolved and differ among providers, and some have been softened. The practical consequence for issuers is that index eligibility is now a real input into the structuring decision, and it is why sunsets became standard: a structure with a credible sunset is treated differently from a perpetual one.

Proxy advisory firms and investor policies

Large asset managers and the proxy advisory firms maintain policies on unequal voting structures, typically recommending against directors at companies with perpetual dual-class structures or supporting stockholder proposals to eliminate them. Because the controller wins these votes, the practical effect is reputational and cumulative rather than decisive — but a controlled company that ignores investor sentiment entirely finds its cost of capital and its ability to use equity as currency affected.

The internal constraint: fiduciary law

Here is the point that founders most often misunderstand. A dual-class structure changes who wins votes. It does not change the fiduciary duties owed to the minority.

The controller's duties

A stockholder with voting control owes fiduciary duties to the corporation and to the minority stockholders. The duties are not triggered by how the control was obtained, and holding control through high-vote shares rather than through economic ownership makes no difference.

The governing test for whether heightened scrutiny applies comes from Sinclair Oil Corp. v. Levien, 280 A.2d 717 (Del. 1971): the business judgment rule applies to a controller's conduct unless the controller receives a benefit "to the exclusion of and detriment to" the minority. Pro rata dividends are fine, even large ones that some stockholders dislike. A transaction in which the controller sits on the other side is not.

Where controllers actually get into trouble

Compensation. A founder-controller's pay package is approved by a compensation committee whose members the controller effectively appoints. Where the package is extraordinary, the transaction is challenged as a controller self-dealing transaction subject to entire fairness rather than business judgment review.
Related-party transactions. Leases from entities the founder owns, services from affiliates, purchases from family businesses.
Transactions with a controller-affiliated buyer, including a going-private transaction.
Charter amendments that entrench or extend the structure.
Corporate opportunities taken by the controller or entities it controls.
Differential consideration in a sale, where the controller receives something the minority does not — a consulting agreement, a rollover on better terms, or a premium for the high-vote shares.

The MFW path

Where a controller stands on both sides of a transaction, Kahn v. M&F Worldwide Corp., 88 A.3d 635 (Del. 2014) offers a route back to business judgment review: conditioning the transaction from the outset on both an independent, empowered special committee and an informed, uncoerced majority-of-the-minority vote.

In a dual-class company, the majority-of-the-minority condition means a vote of the Class A holders excluding the controller. That is a real constraint and a real cession of power, which is precisely why it earns deferential review.

For controllers accustomed to winning every vote, this is a difficult concession — and it is the entire point. The alternative is entire fairness, a fairness trial, and years of exposure.

Charter amendments that reallocate voting power

Williams v. Geier, 671 A.2d 1368 (Del. 1996) is the leading Delaware case on a recapitalization that changed voting power. The company adopted a tenure voting plan giving long-held shares ten votes, which had the effect of increasing the family's voting power. The plan was recommended by the board and approved by a majority of the outstanding shares.

The Supreme Court applied business judgment review, not entire fairness or enhanced scrutiny. Key to the holding: the amendment was approved by a fully informed stockholder vote; it applied to all shares on the same terms, with any holder able to obtain ten votes by holding; and the board had not acted to entrench itself against a specific threat.

Williams is frequently over-read. It does not hold that any voting reallocation is protected. It holds that a facially neutral amendment, adopted with full disclosure and approved by an informed vote, receives business judgment review. An amendment that benefits the controller specifically, or that is adopted in response to a threat, is analyzed differently.

Providence and Worcester Co. v. Baker, 378 A.2d 121 (Del. 1977) confirms the underlying statutory freedom: a charter may validly provide for scaled or limited voting rights, and such provisions are enforceable according to their terms. The charter is the contract, and Delaware enforces it.

The governance architecture a controlled company should adopt

Whether or not required, controlled companies benefit from building structures that make controller transactions defensible.

A majority-independent board. Give up the exemption. The cost is nothing; the benefit is that every subsequent transaction is evaluated by a board a court will respect.

A standing conflicts committee. Rather than forming a special committee whenever a controller transaction arises — which is slow and looks reactive — maintain a standing committee of independent directors with a charter covering related-party transactions above a threshold. It develops expertise, it has continuity, and it is in place before it is needed.

A related-party transaction policy with a real threshold, a defined approval process, and an annual review of recurring arrangements. Every lease, service agreement, and family-business supply arrangement should be on a schedule and reviewed.

Independent adviser access. The conflicts committee should be able to retain its own counsel and financial advisers at company expense, without further approval.

Executive session practice. Independent directors meeting regularly without the controller present, with a designated lead independent director.

A robust audit committee. Required in any event, and the one committee where the controller has no formal role.

Disclosure discipline. Related-party transactions, the controller's holdings, the sunset terms, and the governance structure should be disclosed clearly and consistently. Investors accept dual-class structures they understand; they punish surprises.

Variations on the theme

Dual-class is the most visible way to separate control from ownership, but it is not the only one, and a practitioner should recognize the alternatives because they raise overlapping issues.

Tenure voting. Shares acquire additional votes after being held for a period — the structure at issue in Williams v. Geier. It is facially neutral: any holder who holds long enough gets the extra votes. In practice it favors long-term insiders, because institutional turnover is high and index funds' shares move between accounts. Tenure voting has attracted renewed interest as a "fairer" alternative to dual-class, and it is permitted where the charter provides for it, subject to the practical difficulty of tracking holding periods through the nominee system.

Non-voting stock alone. A company can issue non-voting common stock without high-vote shares, achieving the same dilution-proofing without the ten-to-one asymmetry. This is the Class C structure without the Class B, and it appears where a founder already holds a majority of a single voting class and wants to issue equity without dilution.

Voting agreements and proxies. Rather than structuring the charter, holders can agree contractually to vote together, or grant an irrevocable proxy to a founder. This achieves control without a charter provision, and it can be used post-listing where a dual-class recapitalization is prohibited. The constraints are contractual: the agreement must be enforceable, must have a defined term, and must address transfers, death, and termination. Note that a voting agreement among holders can create a Section 13(d) group, with the disclosure consequences that follow.

Pyramid and holding company structures. A founder controls Holdco, which holds a majority of Opco, which holds a majority of the listed entity. Each layer requires only a majority, so control at the top is achieved with a small fraction of the ultimate economic interest. Common outside the United States and used domestically in family enterprises and in certain regulated industries.

Partnership and up-C structures. A public corporation holds an interest in an operating partnership, with founders holding units exchangeable into stock and a high-vote share class tracking their partnership interest. These structures are driven by tax considerations — the tax receivable agreement is usually the point — but they produce control asymmetries similar to a dual-class structure and require the same governance analysis.

Golden shares and special approval rights. Rather than voting power generally, a class may carry specified consent rights: the right to appoint a number of directors, or to approve a merger, a charter amendment, or a change in the business. This is more targeted than general voting control and can be easier to justify to investors, because the reserved matters are visible and bounded.

Litigation patterns in controlled companies

The claims that recur, and what determines their outcomes.

Compensation challenges. A controller-executive's pay package approved by a committee the controller appointed. The plaintiff pleads that the committee lacked independence and that the package is not entirely fair. The outcome usually turns on the committee's actual process: whether it retained its own consultant, whether it benchmarked, whether it negotiated, and whether the package was submitted to a disinterested stockholder vote. A ratifying vote of the disinterested minority is powerful, and a company willing to seek one converts a fairness trial into a motion.

Related-party transaction challenges. Leases, service agreements, and supply arrangements with entities the controller owns. These are individually small and cumulatively significant, and they are the most common subject of books-and-records demands. Defensible where there is a policy, a threshold, an independent approval process, benchmarking to market terms, and annual review; indefensible where the transactions simply appear in the proxy each year with no visible process.

Freeze-out and going-private challenges. Covered by the MFW framework. The recurring failure is timing: conditions imposed after negotiations have begun.

Sale-of-control challenges. Where the controller sells its block at a premium the minority does not share. Delaware generally permits a controller to sell its shares at a premium, subject to duties not to sell to a known looter and not to structure the transaction to divert consideration that belongs to all holders. The line is crossed when the "premium" is really consideration for a corporate asset — for example, a large consulting agreement or non-compete payment that substitutes for merger consideration.

Entrenchment and structural challenges. Charter amendments extending control, sunset extensions, and defensive measures. Williams v. Geier provides the framework for facially neutral amendments approved by an informed vote; amendments that benefit the controller specifically are analyzed under the fairness standard.

Caremark and oversight claims. Not specific to controlled companies, but a controlled company's independent directors face a particular version of the problem: they are on notice that the controller has interests adverse to the minority, and a board that establishes no monitoring system for related-party dealings has an oversight exposure.

What the plaintiffs do first. A books-and-records demand seeking board materials on the challenged transaction. Controlled companies should expect these, should have a production protocol, and should recognize that a demand is usually the prelude to a complaint rather than an end in itself.

Worked example one: the Adeyemi IPO

The company. Solvent Loom is a design-software company founded by Folasade Adeyemi, who owns 41% pre-IPO alongside three venture funds and an employee pool.

The structure. At IPO, Solvent Loom recapitalizes into Class A (one vote), Class B (ten votes), and Class C (non-voting). Adeyemi and the funds receive Class B. The public buys Class A. Future equity grants and acquisition currency will be Class C.

Post-IPO, Adeyemi holds 26% of the economic interest and 61% of the voting power.

The negotiation with the underwriters and the anchor investors focuses on three points.

First, the sunset. Investors want a seven-year time-based sunset. Adeyemi wants none. The compromise: a ten-year sunset, extendable only by a majority vote of the Class A holders voting separately, plus an ownership sunset if her holdings fall below 10% of the outstanding shares, plus a death and incapacity sunset with a twelve-month transition, plus a departure sunset if she ceases to serve as an executive officer.

The separate Class A vote on extension is the concession that mattered. Without it, the "sunset" would be extendable by a vote Adeyemi controls, which is not a sunset.

Second, the funds' Class B shares. The venture funds will distribute their shares to limited partners within a few years. The charter provides that Class B converts on transfer, with a permitted transfer exception for pro rata distributions to fund partners — but Solvent Loom's investors object, since that would put ten-vote shares in the hands of hundreds of passive limited partners.

The resolution: fund distributions are not permitted transfers, so the funds' shares convert to Class A on distribution. The funds accept this because their economic interest is unaffected and they do not intend to hold long-term. The effect is that Adeyemi's voting percentage rises over time as the funds' high-vote shares convert — a dynamic worth modeling before agreeing to it.

Third, the controlled company exemptions. Solvent Loom's counsel, Bertrand Mkhize, recommends declining all of them. The board will be majority-independent with independent compensation and nominating committees from listing. Adeyemi initially resists — she wants operating executives on the compensation committee — and Mkhize gives her the argument that persuades: "In three years you will want to do a transaction with a company you also invest in. On that day, the fact that you gave up the exemptions voluntarily will be worth more than anything else in the file."

Fourth, and not negotiated but flagged: index eligibility. Mkhize advises that the sunset package materially improves the company's position with index providers relative to a perpetual structure, and that this is a commercial rather than legal consideration.

Worked example two: the sunset that triggers

Three years later. Adeyemi is diagnosed with a serious illness and steps back from day-to-day management, remaining as Executive Chair. The departure sunset provides for conversion if she "ceases to serve as an executive officer of the Company."

Is Executive Chair an executive officer? The charter does not define it. The company's proxy statement lists her as an executive officer. Her employment agreement describes her role as executive. The compensation committee treats her as one.

The board faces a genuine problem. If Executive Chair is an executive officer, no sunset. If it is not, all Class B converts and Solvent Loom becomes a widely held company overnight.

How this should have been drafted. "Executive officer" should have been defined by reference to a determinable standard — the officers identified as executive officers in the company's annual report, or the officers subject to Section 16, or a list in the charter. The ambiguity exists because nobody expected the founder to move to a chair role.

How the board handles it. The independent directors, advised by their own counsel, conclude that Adeyemi remains an executive officer on the objective facts — she is designated as such in the company's filings, she has executive responsibilities under her agreement, and she is treated as one for compensation purposes. They document the analysis, disclose it, and amend the charter prospectively — with a separate Class A vote — to define the term.

Why the separate Class A vote matters. An interpretive decision that preserves the controller's power, made by directors the controller appointed, without minority input, is exactly the fact pattern that produces litigation. Submitting the clarifying amendment to a separate class vote converts a unilateral interpretation into a ratified one.

Worked example three: a controller transaction

Two years further on. Solvent Loom wants to acquire Pelham Rendering, a small company in which Adeyemi personally holds 34% through a family investment vehicle. The purchase price would be approximately $180 million.

The analysis. Adeyemi is on both sides. Under Sinclair Oil, she receives a benefit to the exclusion of the minority — the Class A holders get no share of the Pelham consideration. Entire fairness is the default standard.

The path to business judgment review is the MFW framework: the transaction must be conditioned, from the outset, on approval by an independent special committee empowered to say no and to retain its own advisers, and on approval by a majority of the disinterested Class A shares voting separately.

What Adeyemi has to give up. The second condition means the Class A holders — whose votes have never mattered on anything — get a veto on this transaction. That is a real cession, and Adeyemi resists it.

The alternative she is offered. Proceed without the conditions, accept entire fairness review, and defend the price at trial in three years, with discovery into her family vehicle's records and her communications.

She takes the MFW path. The standing conflicts committee — which exists because the company built one at IPO — is empowered on day one, retains its own bankers and counsel, negotiates the price down to $154 million, and recommends. The Class A holders approve with 68% of the disinterested shares voting yes.

What made it work. The governance architecture adopted at IPO, when it cost nothing: a majority-independent board, a standing conflicts committee with a charter, and a company that had declined the controlled company exemptions. On the day the transaction arose, everything needed was already in place.

Unwinding a dual-class structure

Structures end — by sunset, by agreement, or by transaction — and the unwind has its own mechanics.

By sunset. The cleanest path, and the reason to draft the triggers precisely. On the trigger, Class B converts automatically into Class A on a one-for-one basis. The company must: confirm the trigger has occurred and document the determination; instruct the transfer agent to effect the conversion; file any required amendment or certificate; update the exchange, the index providers, and the transfer agent's records; and disclose the conversion and its effect on voting power. The share count does not change; only the voting power does, which surprises people who expect an economic event.

By voluntary conversion. A founder may elect to convert, and some have, whether for governance reasons, index eligibility, or to facilitate a transaction. The conversion is generally unilateral and irrevocable, and the company should confirm that no charter provision requires board or stockholder action.

By charter amendment. Eliminating the structure entirely requires a charter amendment, which typically requires a class vote of the Class B holders — meaning the founder must agree. Where the founder is willing, the amendment is straightforward. Where a company wants to collapse the structure over the founder's objection, it generally cannot.

In a sale of the company. Where the company is acquired, the classes usually receive identical consideration, and the charter should say so. This is a provision worth checking at the outset: a charter that permits differential consideration to the high-vote class creates a serious fiduciary problem in any sale, and investors will price it.

Practical consequences of the unwind. A company that becomes single-class overnight is suddenly subject to the full range of ordinary market pressures: activists become viable, the board becomes replaceable, and a hostile bid becomes possible. Boards that have operated under controller protection for a decade are frequently unprepared. The right time to build the defensive and engagement infrastructure — advance notice bylaws, a shelf rights plan, an investor engagement program, a stockholder outreach cadence — is the year before the sunset, not the week after.

Disclosure obligations specific to controlled companies

The disclosure package for a dual-class controlled company is larger than for an ordinary issuer, and inconsistency across documents is a recurring problem.

In the registration statement and annual report. A clear description of each class and its rights; the conversion and transfer provisions; every sunset trigger and its mechanics; the voting power held by the founder and by all insiders as a group; and a risk factor addressing the concentration of voting power, the inability of Class A holders to influence outcomes, the exchange exemptions relied upon, and index eligibility consequences.

In the proxy statement. Beneficial ownership tables reflecting both economic and voting percentages — presenting only economic percentages materially understates the controller's position and invites criticism; the controlled company exemption disclosure, stating which exemptions are relied on and which are not; the related-party transaction disclosure with the approval process described; and the compensation discussion, which will receive particular attention where the controller is also an executive.

In transaction documents. Where a controller transaction is submitted to a majority-of-the-minority vote, the disclosure must satisfy the MFW informed-vote condition — meaning the committee's process, the advisers' analyses and conflicts, the negotiation history, and the controller's interest must all be disclosed at a level permitting evaluation.

Consistency. The charter, the risk factors, the ownership tables, the governance section, and the exchange filings must describe the same structure in the same terms. Drift accumulates over years as documents are updated by different people, and a plaintiff comparing this year's proxy with the IPO prospectus will find it.

Practice notes

For companies going public with a dual-class structure.

The structure must be in place at or before listing. Disparate reduction after listing is prohibited by the exchanges.
Negotiate the sunset seriously. A perpetual structure has commercial costs, and death and incapacity sunsets are the ones investors weigh most.
Define every trigger precisely — "executive officer," "incapacity," "family member," "controlled entity" — and identify who determines each.
Model the voting percentage over time, accounting for fund distributions, permitted transfers, equity issuance, and any non-voting class.
Decline the controlled company exemptions. The cost is nothing and the benefit is durable.
Build the conflicts committee before you need it.

For controlled companies operating.

Maintain a related-party transaction schedule and review it annually.
Use the MFW framework for any controller transaction of consequence, and condition it at the outset — timing is the requirement most often failed.
Disclose the structure and the sunsets clearly, and do not let the disclosure drift from the charter.
Give the independent directors real information and real time. A committee that receives a proposal on Tuesday and approves it Thursday is not a defense.

For minority investors.

Read the sunset provisions, not the summary. The extension mechanism is where the substance is.
Check whether a separate class vote is required for amendments affecting the Class A holders.
Watch the permitted transfer definitions for provisions extending control to heirs and entities.
Track related-party transactions in the proxy statement, and compare them year over year.
Understand that fiduciary duties still apply. The controller wins votes; it does not get to self-deal.

The debate, briefly and fairly

Because practitioners are asked their view, it is worth stating the arguments on both sides without pretending the question is settled.

The case for dual-class. Founders with a specific vision are sometimes right when the market is wrong, and a founder who can be removed in a bad quarter cannot make a ten-year bet. Several of the most valuable companies of the last two decades were built by founders who retained control and made investments that public markets would not have funded. Investors buy the shares voluntarily, with full disclosure, at a price that reflects the structure; nobody is defrauded. And the alternative — a market in which every listed company is permanently exposed to activists and acquirers — has its own costs.

The case against. Voting power is the mechanism through which every other governance protection operates. A structure that removes it does not merely reduce accountability at the margin; it removes the ultimate remedy. Founders are not uniformly visionary, control tends to persist past the period in which it is justified, and the founder who was right at thirty-five may be wrong at sixty. Perpetual structures pass control to heirs with no claim to the founder's judgment. And the "voluntary purchase" argument weakens considerably when index funds must hold the shares to track a benchmark.

Where the practical consensus has landed. Not on prohibition, and not on unlimited permission, but on sunsets and governance architecture. A structure with credible time, ownership, death, incapacity, and departure sunsets, at a company that maintains an independent board and a working conflicts process, attracts far less opposition than a perpetual structure at a company that uses every available exemption. That is the negotiating space, and counsel advising an issuer should understand that the terms of the structure — not its existence — are what will be scrutinized.

What counsel should tell a founder. The structure buys freedom from votes. It does not buy freedom from duties, from disclosure, or from litigation. The transactions that matter most — your compensation, your affiliate dealings, your eventual exit — will be tested under a standard that gives no weight to your voting power. The governance you build when nothing is happening is what determines how those transactions come out.

Quick reference

The structure. Class A one vote, publicly traded. Class B ten votes, insider-held, converts on transfer. Class C non-voting, used for issuance so that the founder's voting percentage is dilution-proof.

The external constraints. Exchanges permit the structure at listing but prohibit disparate reduction afterward. Controlled company exemptions cover board and committee independence but never the audit committee. Index providers and investor policies operate commercially rather than legally, and are the reason sunsets became standard.

The internal constraint. Fiduciary duties are unaffected by how control was obtained. Business judgment applies unless the controller takes a benefit to the exclusion of and detriment to the minority; then entire fairness, unless the MFW conditions are satisfied from the outset.

The sunsets that matter most. Death, incapacity, and departure — because the justification for the structure is the founder's specific judgment. Then ownership and time, with attention to who votes on any extension.

The drafting points that decide cases. Precise definitions of "executive officer," "incapacity," "family member," and "permitted transfer." A separate Class A vote on extension. Equal consideration to all classes in a sale.

The governance to build at IPO. A majority-independent board, a standing conflicts committee with its own advisers, a related-party transaction policy with a threshold and an annual review, and a decision to decline the controlled company exemptions. All of it is cheap on the day of listing and decisive on the day a controller transaction arrives.