Last week, a quiet update appeared in Bybit's product changelog: USDC-funded trading for more than 1,000 US stocks and ETFs. The headline reads like another exchange chasing the tokenized-equity narrative that has consumed 2025. But when I pulled the integration apart the way I would a vesting contract in 2017, the architecture told a different story. This is not tokenization. It is a brokerage account wearing a stablecoin's clothes, and the most important line of code in the entire product belongs to a company most Bybit users have never heard of: Alpaca.
Let me explain why that distinction matters more than the product announcement suggests, and why the most technically revealing detail here is the one Bybit chose not to put on-chain at all.
Context: What Bybit Actually Launched
The Defiant reported that Bybit has launched USDC-funded trading in over 1,000 US stocks and ETFs, with trade execution handled by Alpaca and the underlying shares held by Alpaca. The product is positioned as a brokerage alternative for eligible users, and it sits alongside Bybit's existing xStocks offering, which takes the tokenized-equity route. The original piece is a short news brief with a paywalled body, so several critical details, including fee structure, settlement path, and custody isolation, remain undisclosed.
That scarcity of detail is itself informative. When a product's marketing emphasizes what it enables rather than how it settles, the settlement layer is usually where the discomfort lives.
For readers outside the infrastructure weeds: Alpaca is a US-registered broker-dealer that offers B2B APIs. It is plumbing. Exchanges, fintech apps, and neobanks plug into Alpaca the way a wallet plugs into an RPC endpoint. Alpaca executes the order, clears it, and holds the resulting shares in custody. Bybit supplies the user interface, the USDC on-ramp, and the distribution.
This matters because the product contains almost none of what the crypto industry means when it says "on-chain." There is no tokenized share, no smart contract governing settlement, no composable position you can collateralize in a lending market. The USDC is converted to dollars, the dollars buy shares through a traditional broker, and the shares sit in a traditional custodial account. What the user sees is a balance. What exists underneath is a brokerage statement.
That is not a criticism by default. It is a classification. And classification is where analysis has to start, because two products that look identical on a mobile screen can carry completely different risk profiles depending on where the assets legally reside.
Core: The Two Rails and Why They Cannot Be Fused
To understand Bybit's decision, you have to hold two architectures side by side.
The first rail is tokenization, represented internally by xStocks. Here the underlying share is represented by a token that lives on a blockchain, issued through a structure like Backed Finance. The token can theoretically move, be composed, and settle without a brokerage intermediary. Its appeal is cryptographic. Its problem is regulatory: a token that tracks a US equity looks a great deal like a security to the US Securities and Exchange Commission, and the entire category has been living in a provisional gray zone while the agency decides how to treat it.
The second rail is the brokerage alternative, which is what Bybit just launched with Alpaca. Here nothing is tokenized. The user's USDC becomes dollars, the dollars become shares, and the shares stay inside Alpaca's custodial perimeter. The appeal is regulatory: because Bybit is not issuing a security, it sidesteps the central legal question that haunts xStocks. The cost is that the asset never touches a blockchain, so every property the crypto crowd associates with ownership, including self-custody, verifiability, and composability, is simply absent.
Bybit running both rails simultaneously is the most revealing part of this story. It is a hedge, and the hedge implies that Bybit's own compliance team cannot confidently predict which regulatory path survives. Rather than bet, the exchange built both and let the market and the regulators sort it out.
Now, the trade-offs, examined forensically.
Counterparty concentration. On the tokenized rail, the trust surface is spread across an issuer, a chain, and a custody arrangement. On the brokerage rail, the trust surface collapses into a single point: Alpaca. If Alpaca fails, is acquired, or changes its terms of service, every Bybit user holding these positions is exposed. The Defiant's brief does not disclose whether the shares are held in a segregated account, whether they fall under SIPC protection, or whether Bybit has any contractual guarantee of continuity. Those are not minor footnotes. They are the difference between a customer asset and a claim on a broker.
No verifiability. I spent three months in 2017 auditing ERC-20 contracts line by line, and the habit that stuck is this: I do not trust a balance sheet I cannot independently verify. With the brokerage rail, a user cannot confirm on-chain that the shares exist. They cannot confirm that the shares are not rehypothecated. They cannot confirm that their specific position is segregated from Alpaca's proprietary book. They trust a quarterly statement and an API response. That may be acceptable for a traditional brokerage customer who has spent decades operating under that model. It is a significant downgrade for a crypto user who has been trained to verify rather than believe.
Fee opacity. Bybit has not published the spread or commission structure in the available reporting. In a brokerage-alternative model, the platform typically earns through a spread, a commission, or a payment-for-order-flow-style arrangement. Without disclosure, the user cannot calculate the true cost of the trade. This is precisely the kind of detail that a tokenized product, with visible on-chain execution, would make harder to hide.
Replication risk. This is the quiet killer. Because Alpaca's API is standardized and available to any qualified integrator, the technical barrier to launching an identical product is near zero. Any exchange with a compliance team and a USDC treasury can replicate Bybit's offering within a quarter. There is no proprietary technology here, no network effect embedded in the code, no moat. The defensibility of this product is brand and distribution, not engineering.
Here is where the standard crypto commentary gets the story backwards. The reflexive take is that this is"another step toward mass tokenization of real-world assets." But read the architecture again. This product is the opposite. It is a deliberate de-tokenization of real-world assets for crypto users. Bybit is not bringing stocks on-chain. It is bringing crypto users off-chain, into Alpaca's brokerage perimeter, and papering over the transition with a USDC on-ramp. The blockchain is used for exactly one thing: moving the stablecoin in. After that, every subsequent step is traditional finance.
Contrarian: The Stablecoin Is the Real Product
If the shares are held by Alpaca and execution runs through Alpaca's rails, then the only genuinely crypto-native component of this entire product is the USDC. And that reframes the value capture entirely.
USDC is issued by Circle, a US-regulated entity with transparent reserves. Choosing USDC over USDT for a product that plugs into a US broker-dealer is not a coincidence. It is a compliance decision. Alpaca's legal team would have a far easier time accepting USDC as the funding medium than a stablecoin with a more contested regulatory posture. The stablecoin choice is doing quiet compliance work that the marketing materials will never mention.
But the deeper point is this: the pattern, if it scales, converts USDC from a crypto settlement instrument into a general-purpose dollar rail for purchasing traditional assets. Every brokerage-alternative product that adopts USDC as the funding medium adds real utility to the stablecoin, independent of speculation or yield farming. That is a demand-side catalyst that has nothing to do with token prices and everything to do with how dollars move across borders. A user in a jurisdiction where opening a US brokerage account is difficult can now fund an account with USDC and buy exposure to the S&P 500. The stablecoin becomes the bridge, and the bridge collects the traffic.
Which raises the contrarian question the coverage has not asked: who actually captures the value here? Bybit captures distribution and a spread it has not disclosed. Alpaca captures B2B service fees regardless of which exchange wins. Circle captures incremental demand for its stablecoin. The user captures equity exposure but surrenders self-custody, verifiability, and portability. Notice that the party that took the most regulatory risk, Bybit, is also the party with the thinnest moat. The party with the strongest position, Alpaca, took the least headline risk. This is a recurring pattern in crypto-adjacent infrastructure: the visible product is rarely where the durable margin lives. The shovel-sellers quietly outearn the miners.
Takeaway: Watch the Custody Clause, Not the Launch Tweet
The launch itself is unremarkable. What matters is the disclosure that will or will not follow.
If I were advising a user considering this product, I would ask one question before anything else: are the shares held in a segregated, bankruptcy-remote account, and is that arrangement documented in the customer agreement? If the answer is yes, this is a reasonable, if unexciting, brokerage alternative with a stablecoin on-ramp. If the answer is no, or if the answer is vague, then the entire product is a credit exposure to Alpaca dressed as an equity position, and the distinction between that and a tokenized product with on-chain custody becomes very large indeed.
The broader signal to track is whether other exchanges replicate this within the next two quarters. If OKX, Bitget, or Coinbase announce similar Alpaca-backed offerings, the industry has effectively conceded that the compliant path to equities runs through traditional broker-dealers, not blockchains, at least for now. And the tokenized-equity rail, xStocks and its peers, becomes a long-dated option on a regulatory future that may not arrive on schedule.
Rooted in the past, secure for the future: the tools of 2017 audit work, reading the actual settlement logic rather than the announcement, still tell you more than any product page. The question is not whether Bybit launched tokenized stocks. It did not. The question is what the custody clause says, and whether anyone will read it before the next market drawdown forces them to.