The Oldest New Problem in Finance: Proof of Stake and the Return of the Bearer Share
A bearer share is a corporate equity instrument in which ownership is determined entirely by physical possession. No registry. No named holder. Whoever holds the certificate exercises the rights — dividends, voting, liquidation preference — and can transfer those rights simply by handing the paper to someone else. For about a century, bearer shares were a workhorse instrument of international commerce. Then, beginning in the early 2000s, regulators killed them.
The Financial Action Task Force flagged bearer shares as a primary vector of financial opacity between 2003 and 2005. The British Virgin Islands effectively abolished them in 2004. Luxembourg and Ireland followed in 2014. The United Kingdom banned new issuances under the Small Business, Enterprise and Employment Act 2015, requiring all existing bearer shares to be converted to registered holdings. Panama — the last major holdout, famously exposed by the Panama Papers — immobilised its bearer shares under Law No. 47 in 2015, after the OECD described the instrument as fundamentally incompatible with beneficial-ownership transparency.
The reasons for abolition were everywhere the same: bearer shares enabled hidden concentration of ownership, anonymous self-dealing, opaque governance capture, and wholesale evasion of tax, sanctions, and anti-money-laundering regimes. Every jurisdiction that conducted a post-mortem reached the same conclusion — the liquidity benefits of bearer instruments did not justify their governance and regulatory costs.
I want to argue that Proof of Stake has rebuilt the bearer share from scratch. Not metaphorically. Structurally. And that the governance consequences of this reconstruction are not speculative — they are already observable, already measured, and already producing exactly the pathologies that motivated bearer-share abolition in the first place.
The Property-Rights Structure
In a PoS protocol, the staked token confers three economic rights. First, a residual claim on protocol revenue — transaction fees, maximal extractable value (MEV), and new token issuance — functionally identical to a dividend. Second, a governance vote on protocol parameters, treasury allocation, and upgrade decisions — functionally identical to a shareholder vote. Third, a security bond, forfeitable via slashing on misbehaviour — a capital-adequacy requirement that doubles as a performance guarantee.
Ownership of these rights is determined by control of a private key. The token is transferred by signing a transaction. Once confirmed, the new controller exercises all three rights. There is no transfer agent, no registrar, no identity check. Title travels with possession, and possession — control of the key — is the sole evidence of title.
The only structural difference from a classical bearer share is the medium: a cryptographic key rather than a paper certificate. In every economic dimension that matters — residual claim, governance vote, anonymous transferability, absence of an ownership registry — the mapping is one-to-one. A PoS token is a bearer share, denominated in a different substrate.
The law-and-economics literature on bearer shares emphasised two properties that made them uniquely dangerous relative to other forms of equity: transferability without record and governance without identity. Both properties are present in PoS — and amplified. Transferring a physical bearer certificate required physical proximity and at least some logistical friction. Transferring a PoS token requires a signed transaction that settles in seconds, to any address on earth, with no counterparty, no registrar, and no record linking the transfer to a named human being. The marginal cost of creating a new pseudonymous address is zero, which means the marginal cost of fragmenting a position to disguise concentration is zero. This was never true of physical bearer shares, where certificate production, storage, and transport imposed real costs on position fragmentation.
Voting by Pseudonymous Blocks
PoS governance is explicitly plutocratic: one token, one vote, weighted by stake. This is not, in itself, unusual — one-share-one-vote is the default in most corporate-law jurisdictions. But the combination of plutocratic voting with pseudonymous ownership produces a governance surface qualitatively worse than either feature in isolation.
In a registered-share corporation, even with plutocratic suffrage, the identity of large holders is observable. In the United States, beneficial owners crossing the 5% threshold must file a Schedule 13D or 13G with the SEC, disclosing their position, intentions, and identity. Activist campaigns require public disclosure before they begin. A blockholder who votes self-interestedly is visible and, in principle, accountable. The entire edifice of minority-shareholder protection — fiduciary duties, derivative suits, proxy-voting regulation, insider-trading rules — depends on knowing who owns what.
In a bearer-share regime, none of this applies. And PoS is a bearer-share regime. A single economic actor can distribute tokens across hundreds of pseudonymous addresses and vote them as a coordinated block while appearing, on chain, to be an unrelated crowd. On-chain analytics can cluster some addresses heuristically, but sophisticated actors using cross-chain bridges, privacy protocols, or custodial intermediaries can obscure beneficial ownership to a degree that would be flatly illegal in any public-equity market.
The Nakamoto coefficient — the minimum number of independent entities that must collude to compromise a network — is already worryingly low on most PoS chains. But the effective Nakamoto coefficient, accounting for hidden common ownership across ostensibly independent addresses, may be lower still. By construction, it is unobservable. The governance surface of a PoS network is, in the precise sense that motivated bearer-share abolition, a Potemkin democracy: what looks like distributed consensus may be a single whale voting through sock puppets.
The Evidence: Concentration, Cartels, and Vote Markets
This is not theory. The concentration is measurable.
As of early 2026, approximately 35.9 million ETH is staked on Ethereum’s Beacon Chain, representing roughly 29% of total supply. Staking remains concentrated among a small number of entities. Lido, the largest liquid staking provider, commands approximately 24% of all staked ETH — down from a peak above 32% in 2023, but still representing a share that, in any traditional corporate context, would trigger mandatory disclosure requirements and regulatory scrutiny. Coinbase holds approximately 11.7%, Binance approximately 9.1%. According to available Dune Analytics data, the top ten staking entities control over 60% of total staked ETH. This is not decentralisation. This is an oligopoly operating behind a pseudonymous veil.
The validator-cartel problem mirrors, almost exactly, the board-capture problem in bearer-share corporate law. A small number of staking providers aggregate stake from thousands of depositors but make unilateral decisions about which blocks to propose, which MEV strategies to pursue, and whether to censor transactions. The depositor has formal economic ownership but zero practical governance power — the staking provider exercises that power as a de facto nominee. This replicates the nominee/bearer pathology in traditional finance: registered holder (staking provider) exercises power formally belonging to the beneficial owner, with the beneficial owner stripped of practical control.
Worse, PoS has now built explicit infrastructure for vote-buying — something bearer-share jurisdictions tried to prevent.
The “Curve Wars” are the clearest illustration. Curve Finance’s vote-escrow model grants governance power to CRV token holders who lock their tokens for up to four years. Convex Finance aggregated locked CRV at scale, accumulating voting control over Curve’s gauge system — the mechanism that directs CRV emissions to liquidity pools. Platforms like Votium and Hidden Hand then emerged as programmatic bribery markets: any protocol that wanted CRV emissions directed toward its pool could simply pay Convex voters (holders of vote-locked CVX, or vlCVX) to vote accordingly. Stablecoin issuers like Abracadabra committed over $10 million per voting cycle in bribes. Votium alone processed tens of millions in aggregate bribe payments. The entire system was transparent, on-chain, and market-priced.
Call it what it is. This is a programmatic vote-buying market for bearer-share governance rights. The word “bribe” is used without embarrassment because the instruments involved have no identity-linked disclosure regime that would make vote-buying illegal in the first place. In a registered-share corporation, paying shareholders to vote a particular way on a proxy proposal is a federal crime. In PoS governance, it is a product category.
Flash-Loan Governance: What Bearer Shares Never Enabled
One feature of PoS governance has no analogue even in the historical bearer-share regime.
On April 17, 2022, an attacker exploited Beanstalk Farms — an Ethereum-based stablecoin protocol — by taking a flash loan of over $1 billion from Aave. The attacker used the borrowed funds to acquire a supermajority (two-thirds) of Beanstalk’s governance tokens, passed a malicious governance proposal draining $182 million in protocol funds, repaid the flash loan, and laundered the proceeds through Tornado Cash. The entire attack — borrowing, voting, executing, and repaying — occurred within a single transaction. The attacker exercised total governance control with zero long-term economic exposure.
A physical bearer share, for all its problems, at least required possession for some duration. You had to hold the certificate to vote it, and holding it implied some economic exposure, however brief. Flash-loan governance eliminates even this minimal constraint. It enables governance power with literally zero stake — the economic equivalent of voting someone else’s shares, in someone else’s name, for the duration of a single heartbeat, and walking away with the treasury.
Beanstalk’s team acknowledged the vulnerability. The protocol’s smart contracts were paused, governance privileges were revoked, and the project was materially destroyed. The lesson is general: any PoS governance system that allows vote-and-execute within a single block or transaction window is vulnerable to governance capture at zero capital cost.
The SEC’s Awkward Position
In May 2025, the SEC’s Division of Corporation Finance issued a statement concluding that certain protocol staking activities do not constitute securities offerings under the Howey test. The Division’s reasoning was that staking services are “administrative or ministerial” rather than “entrepreneurial or managerial” — the staker’s rewards derive from their own network-securing activities, not from the efforts of a third party.
In August 2025, the SEC extended this reasoning to liquid staking, concluding that staking receipt tokens (like Lido’s stETH) are not securities because they function as receipts evidencing ownership of deposited assets, not as investment contracts.
The analytical move here is interesting and, I think, revealing. The SEC’s framework asks whether a service provider is making entrepreneurial decisions on behalf of the staker. It does not ask whether the instrument itself is a bearer equity instrument conferring residual cash-flow rights and governance power with no identity-linked ownership register. The Howey test is a test for investment contracts, not for bearer shares as such. A bearer share can fail the Howey test — the holder is not relying on the “efforts of others” for their return — while still being exactly the kind of instrument that every developed jurisdiction has independently concluded should not exist.
The SEC’s position creates an irony worth stating plainly: the United States has effectively said that PoS tokens are not securities because the holder is doing the work themselves. But the bearer-share problem was never about reliance on others’ efforts. It was about anonymous ownership of governance power. The SEC has answered a question about the nature of the service while entirely ignoring the question about the nature of the instrument.
Commissioner Caroline Crenshaw’s dissent noted, correctly, that this guidance sits uneasily with prior enforcement actions. But even Crenshaw’s objection was framed in Howey terms. Nobody in the regulatory conversation is asking the prior question: should instruments that confer anonymous, stake-weighted governance power over multi-billion-dollar protocols — instruments structurally identical to bearer shares — be permitted to exist at all?
The Re-Intermediation Irony
History suggests two equilibria for bearer instruments. Either the state forces registration, as happened with bearer shares globally between 2004 and 2015, or intermediaries emerge who de facto re-register ownership, creating a shadow registry that performs the transparency function the instrument itself was designed to avoid.
In PoS, both are happening simultaneously.
On the regulatory side, compliance pressure is building. Custodial exchanges (Coinbase, Binance) already perform KYC on their staking clients, creating a partial beneficial-ownership registry for the fraction of staked assets that flows through regulated channels.
On the intermediation side, liquid staking providers like Lido have become de facto nominee holders. Lido’s stETH is now being wrapped into European exchange-traded products — WisdomTree launched a physical stETH ETP — and VanEck filed with the SEC for a Lido Staked ETH ETF in late 2025. Institutional demand is being routed through regulated wrappers that, by their nature, create the identity-linked ownership records that the underlying PoS instrument lacks. Figment, an institutional staking provider, reported that ETH staking demand from its institutional clients doubled after the SEC’s May 2025 clarification, and Figment was the largest gainer of new stakers in mid-2025, adding roughly 344,000 ETH in a single month.
This institutional on-ramp is, in effect, a voluntary dematerialisation programme — analogous to the custodial immobilisation regimes that jurisdictions like Panama and the Cayman Islands adopted as intermediate steps before outright bearer-share abolition. The difference is that in those jurisdictions, immobilisation was mandatory and applied to all outstanding bearer instruments. In PoS, it is voluntary and applies only to the fraction of stake that flows through regulated intermediaries. The remainder sits outside any transparency framework.
The irony is structural: a technology marketed as disintermediating finance has recreated the bearer-share problem, and is now being re-intermediated by exactly the kind of centralised custodians and regulated funds it was designed to bypass. The residual “decentralised” segment — the portion of stake held by pseudonymous addresses outside regulated channels — remains in the bearer regime, unregistered and opaque. The governance implications of this two-tier structure are unexplored and potentially severe: the regulated tier is visible and accountable; the unregulated tier is not; and the unregulated tier can vote.
What This Means
The analytical claim is narrow and, I think, difficult to dispute on the evidence.
PoS tokens confer residual cash-flow rights and governance votes. Ownership is determined by possession of a private key. There is no mandatory register of beneficial owners. Tokens can be transferred pseudonymously and voted in coordinated blocks without disclosure. Programmatic vote-buying markets exist and are actively used. Flash-loan governance attacks have occurred, enabling governance capture at zero economic exposure. Staking is concentrated among a small number of entities operating as de facto nominee holders with no fiduciary obligations to their depositors.
Every one of these features — hidden concentration, anonymous governance, vote-buying, nominee abuse — is the feature that motivated bearer-share abolition in the traditional corporate context. PoS adds capabilities that bearer shares never had: zero-cost pseudonym generation, smart-contract-coordinated collusion, and cross-jurisdictional opacity that makes enforcement functionally impossible at the protocol level.
The century of institutional learning about why bearer shares are dangerous applies in full. The only question is whether regulators will act on that learning, or whether the crypto-native re-intermediation of PoS — through custodial exchanges, regulated ETFs, and institutional staking wrappers — will perform the same function, more slowly and less completely, from the inside.
Either way, what is being governed on PoS networks today is bearer-share capitalism on a global, permissionless ledger. Anyone participating in PoS governance — or claiming that PoS governance is “decentralised” — should be required to explain how their system avoids the pathologies that led every major jurisdiction on earth to conclude that bearer shares should not exist.
The proponents’ usual defence — that “anyone can participate” — is precisely the defence that bearer-share advocates offered for a century. The lesson of bearer-share history is that formal openness and practical decentralisation are different things. When ownership is anonymous, formal openness is compatible with extreme hidden concentration. “Anyone can buy a bearer share” was true. It was also irrelevant to the governance pathology. The same applies to “anyone can stake.”
The burden of proof sits with PoS proponents, and the on-chain evidence does not currently support their claim. What it supports, rather clearly, is that PoS governance replicates the bearer-share regime in all material respects, amplifies its worst features through programmability and zero-cost pseudonymy, and has already produced the exact outcomes — concentration, cartelisation, vote markets, governance capture — that bearer-share abolition was designed to prevent.
So far, nobody has offered a serious rebuttal. The silence is itself informative.