Initial Public Offerings and Optimal Corporate Governance

Do companies adopt optimal governance arrangements when they go public?  This question has been a hotly debated topic in corporate law and governance and one that I examine in a recent paper.

At the time of an initial public offering (IPO), a company offers a package of governance arrangements to the outside investors.  The arrangements include dual versus single class structure, staggered or un-staggered board, an exclusive forum provision (with respect to either corporate law or federal securities law claims), and robust or narrow shareholder rights with respect to nominating directors, calling special shareholder meetings, or having access to the company’s proxy.  Presumably, when a company offers a set of governance features that the investors find unattractive, investors will discount the stock.  Given that the companies want to maximize the proceeds from the initial public offering, they would do better by switching to governance provisions that the investors would find more attractive.  If, for instance, having a dual-class stock structure would lower the price the outside investors would be willing to pay for the stock, the companies would be better off adopting a single-class stock structure and increasing the proceeds from the offering.

Based on this relatively simple, yet powerful, narrative, scholars and practitioners have argued that even controversial governance arrangements — such as a dual-class capital structure, staggered board, or mandatory individual arbitration provision — should not be banned but instead be left up to companies when they go public.  An underlying premise of the argument seems to be that each firm’s package of optimal governance features may differ (or the regulator (such as the SEC) may have difficulty determining which features are suboptimal), and if outside investors find these features unattractive, they can simply not purchase the stock or pay less for it.  The initial public offering market, in short, can function as a “market check” against a suboptimal or inefficient governance package.

This theory of optimal governance at initial public offerings has, however, been subject to various criticisms.  Some have argued that the IPO market does not function as well as, say, secondary stock markets, where information presumably gets impounded more easily into the stock price.  Unlike companies whose stock has been trading on the market for a long time and have been making numerous public disclosures over the years, companies that are going public for the first time are much less well known to the outside investors.  Furthermore, well-documented empirical evidence that the IPO market is prone to underpricing in the short-run and overpricing in the long-run also indicates that the market isn’t functioning as well as other financial markets.  As a result, one could argue that there is no guarantee that the IPO process will correctly value the offered governance package.  There also is empirical evidence that shows relatively widespread adoption of certain governance features, such as anti-takeover provisions (including staggered boards), at IPOs even when such features may be suboptimal for the adopting firms.

While the debate remains unresolved, the recent surge of dual-class stock IPOs, and the proposal to allow companies to adopt mandatory arbitration provisions for federal securities class actions, have brought the issue back to the fore.  The arguments over dual class stock have been particularly heated.  Opponents have argued that a dual class structure, by making outside investors’ voting rights irrelevant (sometimes through no-vote stock), epitomizes suboptimal governance, especially when such structure does not have a pre-determined termination date (“perpetual dual class stock”).  The Council of Institutional Investors (CII), for instance, has requested the New York Stock Exchange and NASDAQ to impose a mandatory sunset provision on firms that go public using dual class stock and has also asked entities that manage popular stock indices, such as S&P 500 and Russell 2000, to exclude from the indices certain no vote stock and dual class stock without a sunset provision.  Proponents, on the other hand, have argued that a dual class structure can actually enhance firm value by either allowing the founders to pursue their “idiosyncratic vision,” to stay “innovative,” or to facilitate founder-controller’s long-term commitment.

What is also interesting about companies going public with a dual class structure is that, despite the concerns and criticism, the structure has been gaining popularity among IPO firms and that there seems to be a substantial amount of variation among firms over the degree of separation between control and cash flow rights.  In 2021, for instance, among all the firms going public, about 31.7 percent of the firms had a dual class structure, and the fraction of tech firm IPOs involving dual class stock is even higher at 46.2 percent.[1]  Also, while most firms allow public (outside) shareholders to have at least some voting power (e.g., one vote per share), others, such as Snap and Alphabet (through its issuance of Class C non-voting stock), have taken away that power altogether (for certain shareholders).

Various antitakeover provisions also seem to have gotten more popular among IPO firms.  For instance, according to one study, about 90 percent of sample IPO firms (without a controlling shareholder) adopted a staggered board in 2020.[2]  Even if we were to assume that the IPO market is functioning relatively well, the fact that firms vary over adoption of dual class stock and other anti-takeover provisions suggests that the optimal governance arrangement may vary with the company.  In addition, unlike other governance provisions, such as whether shareholders can call a special meeting or can act through written consent, the dual class structure seems to be salient for the outside investors.  Investors, particularly institutional investors (as evidenced by the CII’s proposal), seem to care a lot about the dual class structure, or at least more than they care about other governance features.

With the help of game-theory, my paper argues that, even if the IPO market rationally values the stock and each firm’s governance arrangements, when different firms have different optimal governance structures, and the outside investors do not know the types of firms that they face, it becomes likely that the governance package offered by the firms (or at least a subset of them) will be suboptimal.  The analysis reveals that the traditional theory that the firms will adopt the optimal governance arrangement at the time of the IPO relies on at least one of two assumptions: One type of governance structure (e.g., one-share-one-vote arrangement) is optimal across the board; or outside investors know which governance arrangement is optimal for which firm.  When these assumptions break down, it becomes likely that the governance package chosen by (at least some of) IPO firms will be suboptimal.

After presenting the basic thesis, the paper also examines various mechanisms through which firms can be induced to adopt the optimal governance mechanism at their IPOs.  These mechanisms are divided into two categories.  On the one hand, firms themselves may be able to signal their true valuation to the market.  On the other, to the extent that their IPO share prices are over-valued, investors may bring claims against the firms after their IPOs, and the firms may be found liable (for material misrepresentation or omission).  With respect to the first, private ordering mechanisms, my paper examines how a firm can credibly signal to the market by employing a set of reputable agents to conduct the IPO, by relying more on internal capital markets (and less on external financing), and by deliberating underpricing its IPO.  The paper analyzes the potential cost associated with each mechanism.  One important theme that arises is that whether these mechanisms can enhance the efficiency of the IPO market depends a lot on how much “skin in the game” is being retained by a founder-controller (along with other pre-IPO shareholders) and the fraction of the firm’s equity being sold to the market.

The findings lead to various positive and normative implications.  On the positive side, the paper shows why it may be possible to observe IPO firms adopting relatively homogeneous governance features even if those features may not be optimal for all firms. The paper also shows under what circumstances we may be more confident that a particular firm’s choice would more likely be optimal.  On the normative side, the paper examines various policy proposals.  In particular, it looks at the sunset proposal, which would allow or require outside investors to revisit a firm’s initially chosen governance structure after the firm’s IPO and analyzes conditions under which such  revisiting may be optimal.  Although a full discussion is given in the paper, a few points may be worth a brief mention.  The first is that a sunset provision can create a tradeoff between ex ante and ex post efficiency: While it may allow investors to eliminate or mitigate inefficient governance post-IPO (ex post efficiency), it may lessen the firm’s incentive to adopt the optimal governance structure at its IPO (ex ante efficiency).  The second point is that, when firms have a choice to adopt a sunset provision, they may hesitate to do so at their IPOs for fear of sending an adverse signal to the market.  The third point is that the initial governance choice at IPO can create a “lock-in” effect, thereby making it difficult (if not impossible) to give the parties an incentive to revisit the structure post IPO by utilizing an optional sunset provision.


[1] This is based on the data collected by Professor Jay Ritter, available at:

[2] See Davis Polk, IPO Governance Survey (2020).

This post comes to us from Albert H. Choi, the Paul G. Kauper Professor of Law at the University of Michigan Law School. It is based on his recent paper, “Initial Public Offering and Optimal Corporate Governance,” available here.