When a pension fund manager wants to sell $5 million of Apple corporate bonds, she does not open a screen, see a live order book, and click sell. There is no such screen. She contacts one or more dealers — Goldman Sachs, JPMorgan, Bank of America — describes what she wants to sell, and waits for them to quote her a price. The dealers know roughly what the bonds are worth because they have been buying and selling similar bonds all day. She does not have access to that information in real time. The price she receives reflects that asymmetry.
This is the US corporate bond market in 2026: approximately $10.7 trillion in outstanding investment-grade and high-yield debt, essentially none of it trading on an exchange, with prices discovered through bilateral dealer negotiations and reported to FINRA's TRACE system after the trade has already occurred. Pre-trade transparency — knowing what price you can actually get before you commit to transacting — is limited or nonexistent for most bonds most of the time. The dealer earns the spread between where they buy and where they sell. The end investor absorbs it.
The trust problem the dealer solves is real: corporate bonds are heterogeneous, thinly traded, and genuinely difficult to price without market information that is costly to aggregate. The dealer provides immediacy and price discovery as a service. The question is whether that service needs to be as opaque and as expensive as it currently is — and whether a tokenized bond market could provide the same functions more cheaply and with greater transparency for the investors on both sides of every trade.
Why Corporate Bonds Don't Trade on Exchanges
The equity market's exchange model — continuous order matching, real-time price dissemination, standardized lot sizes — works because equities are highly standardized instruments. A share of Apple is fungible with every other share of Apple. The same issuer has one class of common stock, and every dollar of it is identical to every other dollar.
Corporate bonds are the opposite. A large issuer like Apple has dozens of individual bond series outstanding — different maturities, different coupon rates, different indenture terms, different subordination levels in the capital structure. The $1.5 billion 3.85% notes due 2043 are a different instrument from the $2 billion 2.375% notes due 2029, even though both are Apple's obligations. Each series trades separately, each has its own liquidity profile, and the average series trades infrequently enough that an exchange-style continuous order book would often have no bids and no offers at all.
This fragmentation is structural, not accidental. Every time a company issues debt — which large issuers do repeatedly, in varying sizes, across different maturities, to match the duration preferences of different investor classes — it adds another instrument to the secondary market that must be separately priced and separately traded. The US investment-grade corporate bond market has roughly 40,000 individual securities outstanding. The Nasdaq lists approximately 3,300 companies. The sheer number of distinct instruments in the bond market makes exchange-style trading impractical for most of them under the current model.
TRACE and the Limits of Post-Trade Transparency
The Financial Industry Regulatory Authority introduced the Trade Reporting and Compliance Engine — TRACE — in 2002, requiring dealers to report corporate bond trades within fifteen minutes of execution. It was a genuine improvement over the preceding era, when bond prices were almost entirely opaque. For the first time, investors could see what recent trades had printed at and use that information to assess whether the price they were being quoted was reasonable.
TRACE did not, however, create pre-trade transparency. Knowing that a bond last traded at 98.50 two hours ago tells an investor something. It does not tell her what price she can actually transact at right now, how much size is available at that price, or whether the dealer she is talking to has inventory. The information is backward-looking by design, and the fifteen-minute reporting window — subsequently compressed to one minute for most bonds, but with significant carve-outs for large trades — still means that by the time a trade is visible to the market, the dealer has already captured the spread.
The dealer's informational advantage is not incidental to the current model. It is the model. Dealers commit capital to hold bond inventory, bear the risk of price movements between purchase and sale, and earn the bid-ask spread as compensation for providing that immediacy. The spread is wider for illiquid bonds, wider for large trades, and wider when market conditions are uncertain — precisely the moments when an investor most needs liquidity and is least able to negotiate effectively. The asymmetry compounds in stress conditions.
A 2020 study by economists at the Federal Reserve and Cornell University estimated that retail investors in the corporate bond market pay transaction costs approximately ten times higher than institutional investors on similar trades — a direct consequence of having less information about fair value and less bargaining power with dealers. The opacity is not evenly distributed. It falls hardest on the investors least equipped to absorb it.
Electronic Trading Has Helped, Up to a Point
The past decade has seen meaningful progress in bond market electronification. Platforms including MarketAxess, Tradeweb, and Bloomberg's bond trading system have moved a growing share of corporate bond volume to electronic protocols, reducing the friction of the traditional phone-based request-for-quote process. In 2015, roughly 20% of investment-grade corporate bond volume was executed electronically. By 2025, that figure had risen to approximately 45%.
Electronic trading platforms have compressed bid-ask spreads for liquid, benchmark bonds — the large, recently-issued series that trade frequently and that dealers are willing to quote competitively. For the long tail of the corporate bond universe — the thousands of smaller, older, or off-the-run issues that make up the majority of outstanding securities — electronification has had limited effect. The structural problem of instrument fragmentation remains, and the dealer intermediation model persists wherever liquidity is thin.
What the electronic platforms have not done is change the fundamental architecture. They are faster and cheaper versions of the same bilateral dealer model — investors still request quotes, dealers still respond with prices that reflect their informational advantage, and settlement still runs through the same Depository Trust & Clearing Corporation infrastructure that clears equity trades, with a standard cycle that settles corporate bonds in T+1 as of 2024. The intermediaries are more efficient. They are still intermediaries.
What Tokenized Bonds Change
Several things happen when a corporate bond is issued as a token on a public or permissioned blockchain rather than as a traditional security held through the DTCC's book-entry system.
Settlement becomes atomic. A tokenized bond trade can settle at the moment of execution, with simultaneous transfer of the bond token and the payment token, with no counterparty risk in the interim. The T+1 settlement cycle exists because the current system requires time to match, confirm, and net trades across multiple intermediary systems. On-chain settlement does not require that process. Counterparty risk between trade execution and settlement — which is small but real for institutional trades and significant in volatile markets — goes to zero.
Ownership records become transparent and real-time. The current system for tracking corporate bond ownership runs through the DTCC's Cede & Co. nominee structure, which holds the vast majority of US securities in street name. Beneficial owners are recorded by their brokers, not on any public ledger. A tokenized bond's ownership is visible on the blockchain at all times, enabling issuers to know who actually holds their debt — something that is surprisingly difficult to determine under the current model — and enabling investors to verify their holdings independently.
On-chain order books become viable for actively traded instruments. Several projects — including Ondo Finance's tokenized Treasury products and Backed Finance's tokenized bond wrappers — have demonstrated that fixed income instruments can trade on decentralized exchange protocols with continuous price discovery rather than dealer quotes. For liquid benchmark bonds, an automated market maker or a central limit order book on a high-throughput blockchain could provide price discovery that is genuinely continuous and observable in real time, eliminating the pre-trade opacity that currently benefits dealers.
The Regulatory Problem That Shapes the Timeline
The primary obstacle to widespread tokenized corporate bond issuance in the United States is not technical. It is regulatory. Corporate bonds are securities under the Securities Act of 1933, and their issuance and transfer must comply with SEC registration requirements or satisfy an exemption. The existing DTC-based clearing infrastructure has decades of regulatory blessing built into it. Tokenized alternatives require either regulatory approval to use new settlement mechanisms or reliance on exemptions that limit the investor base.
The SEC's approach to tokenized securities has been cautious and inconsistent. Staff guidance has generally treated tokenized securities as subject to the full weight of existing securities law — which was not written with programmable settlement in mind. Several institutional blockchain bond projects, including experiments by Societe Generale and the European Investment Bank, have been structured under European regulatory frameworks precisely because EU rules have been more accommodating of tokenized issuance than US rules. The EU's Markets in Crypto-Assets Regulation and its DLT Pilot Regime create specific pathways for tokenized security settlement that have no direct US equivalent.
The SEC under its current leadership has signaled a more open posture toward digital asset markets than its predecessor, and there is active engagement with the question of how tokenized securities should be regulated. But the regulatory infrastructure that would allow a US corporation to issue a bond directly on a public blockchain, with settlement through an on-chain system rather than the DTCC, does not yet exist. The technical architecture is ahead of the legal architecture by several years.
The Dealer's Position
The large broker-dealers that dominate corporate bond market-making have not ignored tokenization. JPMorgan's Onyx platform has settled over $900 billion in repurchase agreements on blockchain infrastructure. Goldman Sachs issued a digital bond on its own private blockchain in 2022. Both firms have made the standard institutional gestures toward blockchain adoption — dedicated units, public statements about interest in the technology, selective pilots that demonstrate engagement without threatening core revenue.
What none of the major dealers has done is advocate for a bond market architecture that would eliminate the pre-trade opacity their businesses depend on. The spread income from corporate bond market-making is not a rounding error. It is a material revenue line for every major fixed income dealer. A genuinely transparent, on-chain corporate bond market with atomic settlement and continuous price discovery would compress those spreads in ways that are difficult to offset. The dealers will participate in tokenized bond infrastructure to the extent it expands their capabilities. They will not lead the transition that makes their informational advantages obsolete.
That transition will come, if it comes, from issuers who want cheaper distribution, investors who want genuine price transparency, and regulatory frameworks that permit the infrastructure to operate. The trust problem that dealers currently solve — providing immediacy and price discovery in a fragmented market — is real but not permanent. It is a function of instrument fragmentation, settlement friction, and information asymmetry, all of which programmable infrastructure addresses directly. The question is not whether that architecture is technically superior to the current one. It clearly is. The question is how long the regulatory and institutional inertia can sustain the market structure that exists because it predates the alternative.