Pond Street Ledger

Who Pays the Dividend: How Corporate Actions Reach a Stock Token

A share does things while you hold it. It pays dividends, splits, gets taken over and votes. An explainer on how each of those events is supposed to travel from the share register to a token balance, and the places the signal gets lost.

✓ 1593.efrogs.eth2026-10-038 min
TVL$1.05b+2.7% 7d

The question

A tokenized equity is marketed on price: the token tracks the share, the share moves, the token moves. That is the easy half. The hard half is everything else a share does while it sits in an account. It pays dividends on a schedule set by a board. It splits, reverse splits, spins off a subsidiary, gets acquired for cash or for stock in another company, issues rights to existing holders, and asks for a vote at the annual meeting. Each of those is a corporate action, and each one has to be detected, interpreted and pushed down to whoever holds the economic interest. In a traditional chain of custody that work is done by a transfer agent, a custodian, a broker and a back office, all of whom have done it ten thousand times. Onchain, somebody has to re-do it in software. This explainer walks the path an action takes from the issuer to a token balance, and marks where it tends to fall over.

The two models: pass-through and adjust-the-token

Broadly there are two ways to handle an action, and almost every design is a variant of one of them. The first is pass-through. The issuer or its agent receives the cash or the new shares, and distributes a corresponding amount to token holders, usually as a stablecoin payment or as additional tokens. The second is adjustment. Nothing is distributed. Instead the token's terms change so that the token continues to represent the same economic claim, which in practice means either changing the ratio of tokens to shares or folding the value back into the token so it accrues rather than pays out. A dividend handled by adjustment is simply a token whose reference value includes reinvested distributions.

The tradeoff is concrete. Pass-through is legible: a holder sees cash arrive and can reconcile it against the issuer's announcement. It is also expensive and slow, because the issuer has to identify every holder at a moment in time and pay each of them, including the ones holding through a pooled contract that does not care to be paid. Adjustment is cheap and needs no holder list at all, because changing a ratio changes everyone's claim simultaneously. What is given up is transparency and tax clarity: the holder received value but never received a payment, and the accounting for that varies by jurisdiction and is usually the holder's problem.

The record date problem

Every corporate action turns on a snapshot. Holders of record at the close of a particular day get the dividend, the split or the vote. In the traditional system the record date is an administrative fact produced by a settlement system that knows exactly who held what at a given instant. On a public chain, the equivalent is a block height, and somebody has to decide which one.

That sounds trivial and is not. The chain's record is the token, not the share. The share sits with a custodian in a structure whose own record date is set in a different timezone by a different system. If the token's snapshot block does not correspond to the moment the custodian's position was fixed, somebody is owed money and somebody has been overpaid. Worse, the token may be in places where a snapshot is meaningless: inside a lending pool where it has been posted as collateral, inside a liquidity pool where it has been partly sold, inside a bridge contract, or wrapped inside another token entirely. The snapshot sees a smart contract address holding a thousand tokens. It does not see the four hundred people with a claim on that contract.

Where it actually breaks: pooled and borrowed positions

This is the failure mode worth understanding, because it is structural rather than a bug. When a stock token is deposited into a money market and borrowed by somebody else, two parties have an economic relationship to one token. The borrower holds it and will be the address in the snapshot. The lender gave it up and expects to be made whole. Traditional securities lending solved this a long time ago with manufactured dividends: the borrower is contractually obliged to pay the lender an amount equal to the distribution. Onchain, that obligation has to be written into the lending protocol, and if the protocol was built for assets that never pay anything, it was not written at all.

The same problem appears in automated market makers. A liquidity pool holding a stock token against a stablecoin will receive the distribution at the pool address if it receives it at all, which either accrues to the pool as an uncounted asset, gets swept by whoever can call a function, or sits unclaimed. This is why issuers often prefer adjustment over pass-through for tokens they expect to be used as DeFi collateral: changing the ratio propagates automatically to every holder including contracts, while a payment has to find its way to a recipient who may be a piece of code with no instruction for receiving it.

Mergers, delistings and the forced exit

Dividends are the routine case. The interesting cases are the ones that end the token. If the underlying company is acquired for cash, the share stops existing and the custodian receives money. The token must then be redeemed, and the design question is whether holders are paid automatically or have to come and claim, and what happens to tokens that are locked in contracts or held by addresses that never return. If the acquisition is for stock in another company, the issuer either has to launch a token for the acquirer or liquidate and pay cash, and the choice is usually made by whichever is permitted rather than whichever holders would prefer. Delistings, trading halts and bankruptcies each produce a period where the underlying has no reliable price while the token continues to trade against whatever liquidity remains.

This is also where the token standard's administrative powers earn their keep. A token that can be frozen, burned or force-transferred by the issuer is uncomfortable to hold and necessary to wind down cleanly. A token with no such powers cannot be redeemed against holders who do not act, which means the structure is left holding assets against claims it cannot extinguish. The tradeoff is explicit: holder autonomy against the issuer's ability to finish the job.

Voting, which mostly does not happen

Most tokenized equity structures do not pass through voting rights, and the disclosure usually says so plainly. The token represents an economic interest; the legal holder of the share is the custodian or the issuing vehicle, and it either abstains, votes with management, or votes as instructed by a party that is not you. For an index-like exposure that is uncontroversial. For a company in a contested situation it is the difference between owning a share and owning a derivative on one. A reader deciding what a token is should check whether the word vote appears anywhere in the terms, and if it does, whether there is an actual mechanism or only a statement of intent.

What to check before you hold one through an action

Three documents answer most of it. The token terms say whether distributions are passed through or accrued, and who bears the cost. The reserve or custody disclosure says who holds the share and therefore who receives the action in the first place. The protocol documentation of wherever the token is deposited says whether that venue has any mechanism for distributions at all, which for a general-purpose lending market or AMM is frequently no. Then check the mundane thing: what the issuer does about withholding tax. A dividend reaching a holder through a foreign custody vehicle is taxed somewhere before it arrives, and the gross figure in the announcement is not the figure that lands.

Why this is the real test

Price tracking is a plumbing problem that can be solved with an arbitrageur and enough liquidity. Corporate actions are an operations problem, and operations problems are only visible when they fail. The market for tokenized equities is young enough that most tokens have not yet been through a messy merger, a rights issue or a bankruptcy. The structures that handle the first few of those cleanly, with holders in pooled positions made whole and no unclaimed residue, will be the ones institutions treat as securities infrastructure rather than as tracking products. The ones that do not will discover their holder base through complaint tickets.