The legendary investor Warren Buffett has a stoic, almost religious, objection to the concept of the stock split.
The Class A shares of Berkshire Hathaway trade at over $700,000 per share because Buffet believes a stock split is an empty gesture that changes nothing about the value of the underlying business. In Buffett's world, if you cut your pizza into eight slices instead of four, you don’t have more pizza. You just have more dishes to wash.
While a stock split might be a “non-event” from a valuation perspective, it is a highly regulated activity overseen by the SEC and enforced by exchanges.
When a company announces a split, it must file an 8-K and provide advance notice before the change takes effect. This critical window allows transfer agents to adjust share ledgers, brokerages to update internal systems, and data providers like Bloomberg to update their feeds so a $500 stock doesn’t appear to have crashed to $50 overnight on a 10-for-1 split.
Splits aren't the only corporate actions that require this level of coordination. Dividends introduce similar complexity.
On the ex-dividend date, a stock’s price drops by the amount of the distribution. Some funds, particularly high-yield income funds, take this to an extreme. They pay frequent distributions that are largely return of capital, effectively giving investors back their own principal rather than investment income. The share count stays constant, but the NAV erodes steadily over time.
Tracking the performance of these funds requires a clear distinction between price return and total return.
Suppose you own 100 shares of a high-yield ETF at $100 per share ($10,000 investment). Each month, the fund pays a $5 distribution, 90% of which is return of capital. After 12 months, you’ve received $60 per share in cash ($6,000 total), but the NAV has fallen from $100 to $46. The price return is negative $54, but the total return is $10,600 ($4,600 in remaining NAV plus $6,000 in distributions), or a positive 6%.
These are exactly the kind of problems blockchains are supposed to solve.
A single shared ledger, updated atomically and visible to everyone at the same time. If everyone is reading from the same onchain record, corporate actions like splits and dividends will propagate instantly across the system, eliminating the frantic reconciliation that happens today between siloed intermediaries.
That promise is why Robinhood (@RobinhoodApp) CEO Vlad Tenev’s announcement for a tokenized stock strategy in June 2025 was met with such enthusiasm.
After six months, the Robinhood tokens are live and the data is flowing. But unfortunately, a few problems are starting to surface.
The Good
Robinhood’s announcement served as a catalyst for the market.
Other issuers moved quickly to launch competing products, with **Backed Finance **(since acquired by Kraken) rolling out xStocks (@xStocksFi) on Solana, followed by **Ondo Global Markets **(@OndoFinance) with its tokenized equities offering.
Tokenized stocks had a bona fide breakout year. In the second half of 2025 alone, the asset class grew 128%, pushing total asset value to nearly $1 billion.
Robinhood’s tokenized U.S. stocks and ETFs are available to European customers. Each token is issued on Arbitrum, fully backed by shares held by Robinhood, and tradable 24/5 with zero commissions. The data is available on RWA.xyz.
But accurately capturing Robinhood’s tokenized stock metrics proved more complicated than expected.
The Bad
Most blockchain data platforms index tokens under the assumption that they follow standard conventions. For ERC-20 tokens, that means tracking mints and burns, accumulating supply from zero, and calculating market cap as supply multiplied by price.
This works well for thousands of tokens on Ethereum and any other EVM networks. But ERC-20 was never designed for securities that undergo corporate actions. The standard has no native concept of stock splits, reverse splits, or distribution-driven rebasing.
So Robinhood had to use custom contracts to handle these events correctly for its end users. These tokens function as intended within the Robinhood app, but their mechanics are opaque to data platforms and incompatible with DeFi protocols, which both assume ERC-20 compliance.
When we compared token supplies calculated using standard ERC-20 logic against the onchain data, the discrepancies were impossible to ignore. Some tokens were off by 10x. Others by 100x.
Nearly all of the errors were attributable to two categories: (1) distribution-driven NAV erosion and (2) reverse stock splits.
NAV Erosion from High-Yield ETF Distributions
These are high-yield option income ETFs that pay frequent distributions, with 90% or more classified as return of capital. Each payout returns cash to investors, but largely their own principal rather than investment income. Share counts remain constant, while NAVs steadily erode over time.
Robinhood’s contracts account for this by separating shares from tokens. A holder’s share count stays fixed, while an internal multiplier rebases the reported token supply downward as return-of-capital distributions accumulate, reflecting the underlying NAV erosion.
A data platform that follows a standard ERC-20 model, however, simply sums mints and burns. Under this approach, rebasing adjustments are not reflected, overstating the outstanding token supply, and, by extension, the reported market cap.
Reverse Stock Splits
The same issue appears in reverse stock splits. A reverse stock split consolidates shares to raise the per-share price, typically to meet exchange listing requirements. The share count divides down, but the price per share multiplies up, keeping total value unchanged.
Again, Robinhood’s contracts adjust token supply to reflect the reverse split, while third-parties following a standard ERC-20 model overstates both outstanding supply and reported market cap.
Total Robinhood Discrepancy
Across the 21 mismatched tokens we identified, reported supply was overstated by roughly 64,000 tokens, a 56% discrepancy. NAV erosion from high-yield ETFs accounted for approximately 90% of the gap, with reverse stock splits explaining the rest.
Any data platform relying on the ERC-20 calculated supply will materially overstate Robinhood’s tokenized stock market cap, often by multiples.
The Fix
Taxonomy of Tokenized Stocks: Models x Infrastructure
Tokenized stock issuers have taken different approaches to handling corporate actions. They broadly fall into two categories.
Rebasing Models
Rebasing models maintain spot-price parity: 1 token should always trade near the price of 1 underlying share. When a corporate action occurs, token balances adjust to preserve this relationship. Issuers using this approach fall into two camps based on their relationship with the underlying asset issuer.
*Rebasing (Third-Party): *Issuers operate independently of the company whose stock is being tokenized. **xStocks **(@xStocksFi) (Backed Finance / Kraken) and **Robinhood **(@RobinhoodApp) both take this approach. Tokens are backed by shares held in custody, but because there's no direct relationship with the underlying issuer, they replicate economic exposure without conferring legal ownership.
*Rebasing (Direct): *Issuers work with companies to tokenize their shares. **Superstate Opening Bell **(@SuperstateInc) and **Securitize **(@Securitize) operate as SEC-registered transfer agents and serve as the official shareholder registrar. Because the token is issued in coordination with the company, the token is the legal security, and holders have actual shareholder rights that third-party models cannot provide.
Both structures need multiplier infrastructure to reflect corporate actions onchain.
Solana's Token-2022 Scaled UI Amount extension provides this natively. Issuers update a multiplier that adjusts displayed balances without altering raw token amounts. A 2-for-1 stock split moves the multiplier from 1.0 to 2.0; wallets show double the balance while raw token counts remain unchanged. Because the standard is native to Solana, data platforms can query multiplier changes directly.
EVM networks have no equivalent standard. Issuers like xStocks and Robinhood** **have built their own multiplier mechanisms. Balances adjust correctly and wallets display spot-price parity, but these implementations are bespoke. Third-parties relying on standard ERC-20 calls have no way to detect when a multiplier changes or query its current value. Each issuer's approach must be understood individually.
This is why Superstate's Chris Ridmann and Robinhood's Gilbert Shih co-authored ERC-8056, a draft proposal to bring a standardized Scaled UI Amount extension to ERC-20 tokens. It would give data platforms a consistent interface for tracking corporate actions across issuers.
Until it's adopted, platforms like RWA.xyz must work directly with each issuer to understand their mechanics and integrate accordingly.
Total Return Models
**Ondo Global Markets **(@OndoFinance) sidesteps the issue by taking a different approach: its tokens are designed as total-return trackers rather than spot-price replicas. Token supply remains constant, and corporate actions are reflected in the token price on EVM and through Scaled UI balance on Solana. Dividends are reinvested into the underlying asset, increasing the value backing each token.
Like xStocks and Robinhood, Ondo is a third-party issuer operating independently of the companies whose stocks are tokenized, with tokens representing economic exposure rather than direct legal ownership. The difference is in how corporate actions are handled: price absorption rather than rebasing. This preserves economic equivalence, but token prices diverge from the underlying share price.
RWA.xyz’s Solution
To report correct supply for Robinhood’s tokenized stocks, we took a different approach. We read historical totalSupply values directly from the contracts using an archive node, rather than reconstructing supply by summing mint and burn events.
This approach produces accurate post-corporate action figures, but does come with trade-offs. While we report accurate supply and market cap for Robinhood products, we cannot yet support transfers, holder balances, and more than 50 additional metrics available for every asset we cover.
The important numbers are correct, but the data is less expressive than usual.
The Way Forward
It’s tempting to treat issues like this as evidence that tokenized assets aren’t ready for mainstream adoption. We disagree.
Traditional finance had decades to develop standards for corporate actions. Institutions like the DTCC, transfer agents, and clearing systems didn’t emerge fully formed. The plumbing we now take for granted began as a patchwork of ad hoc fixes that gradually evolved into standards.
Tokenized assets are going through that same process on a far more compressed timeline. In a new frontier like this, friction isn’t a sign of failure. Rather, it’s how new standards emerge. Through experimentation, competing implementations, and eventual convergence.
At RWA.xyz, we’re working directly with issuers and builders to ensure the infrastructure is being built correctly from the start, so the market can continue to scale. These challenges are a reminder of how early we still are, and how much opportunity remains ahead.
