№ 010Deep Dive202611 min read

CryptoMarket Structure

Crypto Is Becoming Part of the Financial System, Quietly

By Akram Ayyash

How custody, settlement, and collateral, not speculation, are pulling crypto into global finance

The most significant cryptocurrency developments of the past year weren't announced at conferences or debated on social media. They happened in back offices, regulatory filings, and infrastructure contracts. Between January 2025 and January 2026, cryptocurrency shifted from speculative asset class toward embedded financial infrastructure in ways that received remarkably little attention.

The drivers aren't ideological. They're operational. In a 5% rate world, trapped cash is expensive. Settlement delays have become an actual P&L line item. The need for intraday liquidity, the optimization of collateral velocity, and the compression of settlement cycles now carry real economic weight. Traditional financial institutions responded not by "adopting crypto" in any philosophical sense, but by integrating blockchain rails wherever they reduced friction and freed capital.

I’ll focus on four functions where integration is already happening: custody (where accounting rules changed incentives), settlement (where repo markets became the killer app), collateral (where tokenized Treasuries unlock balance-sheet efficiency), and payments (where stablecoins compress time and reduce prefunding).

These four functions matter because they're the core plumbing of capital markets. When custody, settlement, collateral, and payments change, everything downstream changes with them.

The Accounting Unlock That Changed Everything

For years, the integration of digital assets into the regulated financial system was throttled by a single, obscure accounting rule. Staff Accounting Bulletin 121, issued by the SEC in March 2022, effectively prohibited crypto custody for regulated banks.

The mechanics were punishing. SAB 121 required any entity safeguarding crypto assets to record a liability on their own balance sheet, measured at fair value. If a bank held $10 billion in Bitcoin for clients, it had to recognize a $10 billion liability.

Prudential capital rules then required the bank to hold Tier 1 capital against that liability. For volatile assets, this often meant a near dollar-for-dollar capital charge. To custody $10 billion in client crypto, a bank might need to set aside $10 billion in shareholder equity.

The math was fatal. Return on equity collapsed. The business became commercially nonviable. Global systemically important banks with trillions in traditional custody assets sat on the sidelines while crypto-native custodians captured the market.

On January 23, 2025, days after Gary Gensler resigned as SEC Chair, the agency issued SAB 122, rescinding the restrictive guidance. The change returned crypto custody to standard accounting principles: if a custodian demonstrates bankruptcy remoteness (client assets legally segregated and protected from creditors), they no longer need to gross up their balance sheet.

The OCC followed with Interpretive Letter 1183, reaffirming that national banks could custody digital assets and participate in blockchain networks without prior regulatory approval. Letters 1184, 1186, and 1188 clarified execution, gas fee holdings, and riskless principal transactions. In December, the OCC granted conditional national trust bank charters to Circle, Ripple, BitGo, Fidelity Digital Assets, and Paxos simultaneously.

The GENIUS Act, signed July 18, 2025, established the first federal framework for payment stablecoins: 100% reserve backing in cash, short-term Treasuries, or high-quality liquid assets, with explicit prohibition on forcing banks to treat custodied crypto as balance sheet liabilities.

The framework isn't without tradeoffs. Reserve restrictions and compliance burdens create barriers to entry that favor large, well-capitalized issuers. Smaller stablecoin projects face a steeper path to regulatory approval. Product variety may narrow as the market consolidates around compliant structures.

But for banks weighing crypto custody and stablecoin services, the incentive set shifted materially. A binding constraint was removed. Once custody economics normalized, banks could finally evaluate crypto services like any other balance-sheet business. What followed was predictable.

Custody Migrates to Systemically Important Banks

BNY Mellon, the world's largest custodian with $57.8 trillion in assets under custody, moved fastest. In April 2025, BNY launched Digital Asset Data Insights, broadcasting fund accounting data to Ethereum with BlackRock's tokenized BUIDL fund as anchor client. In July, BNY became primary custodian for Ripple's RLUSD reserves. By November, the bank had launched a dedicated stablecoin reserves fund.

BNY now provides fund services for over 80% of digital asset exchange-traded products in the US, Canada, and EMEA. This happened through incremental partnerships, not press conferences.

State Street integrated with JPMorgan's Digital Debt Service in August, completing a $100 million tokenized commercial paper transaction with same-day settlement. Citi confirmed crypto custody for 2026. US Bank resumed services after pausing during the 2023-2024 uncertainty.

The market is bifurcating. Traditional banks are capturing "slow money": pension funds, sovereign wealth funds, and endowments that require the legal protections and creditworthiness of a GSIB. Crypto-native custodians retain the high-velocity trading segment where speed and asset coverage matter more than balance sheet strength. For slow money, counterparty credit quality matters more than execution speed.

This bifurcation reflects a "build vs. buy" tension that will play out over the next cycle. Banks are partnering early (BNY with Fireblocks, State Street with Taurus, US Bank with NYDIG) to acquire capabilities quickly. But partnerships create vendor dependence and margin pressure. Over time, banks will likely internalize these capabilities to control costs and risk. The current partnership model appears transitional.

The end state isn't replacement. Trading velocity and product innovation will likely remain with crypto-native infrastructure. What's migrating is the institutional wrapper: the custody, fund services, and compliance infrastructure that slow money requires.

Repo Is the Killer App (Not a New Wallet)

Custody changes who holds the assets. Settlement changes how fast they move. The most profound integration of crypto into finance is happening not at the consumer layer, but in the deep plumbing of repo markets.

For readers less familiar with repo: repurchase agreements are short-term loans collateralized by securities, typically Treasuries. Repo is the daily liquidity and collateral backbone of modern markets. When it gets faster, everything downstream gets cheaper. Daily volumes exceed $4 trillion.

In traditional repo, settlement is frictional. The borrower delivers securities to the lender, moving assets between custodial accounts at central securities depositories. Trades typically settle T+1 because asset movement takes time and must align with Fedwire windows.

Intraday liquidity (lending money for just a few hours) has been operationally difficult because of settlement uncertainty. You can't confidently lend cash from 10 AM to 2 PM if you're not certain the collateral will arrive and return on schedule.

Broadridge's Distributed Ledger Repo platform changed this. By December 2025, DLR was processing approximately $9 trillion monthly, with average daily volume of $384 billion. That's roughly 3% of the US repo market, up 490% year-over-year.

The mechanism: instead of moving underlying Treasury bonds between accounts, DLR locks securities in a dedicated depository account and creates a digital representation on a private ledger. When cash payment confirms, token ownership transfers instantly. Smart contracts ensure delivery-versus-payment without settlement risk.

The economic value is capital efficiency. Institutions can now lend cash for specific hours, earning yield on previously dead time. A bank with $500 million excess at 11 AM but needing it back by 3 PM can deploy those funds. This optimizes liquidity coverage ratios, reduces buffer capital held against settlement fails, and mitigates Federal Reserve daylight overdraft fees.

JPMorgan's Kinexys platform has reached similar scale, surpassing $1.5 trillion in cumulative notional value. The bank's December announcement of a tokenized money market fund seeded with $100 million of JPMorgan capital marked the first such product from a GSIB.

But scale creates a new problem. If Broadridge's DLR, JPMorgan's Kinexys, and other settlement ledgers remain siloed, liquidity fragments. The next bottleneck isn't technology. It's market structure: cross-ledger settlement coordination, message standards, and inter-ledger collateral portability. Interoperability becomes the scaling constraint.

Collateral Mobility Is Really Balance Sheet Compression

Faster settlement enables something more valuable: collateral that moves. In a high-rate environment, idle cash is expensive. Institutions need assets that are safe, yield-bearing, and liquid enough to serve as margin. Tokenized Treasuries have emerged as the solution.

The market grew from approximately $1.7 billion to $7-10 billion during 2025. BlackRock's BUIDL fund leads with roughly $2.5 billion. These products serve as collateral in derivatives markets and yield-generating instruments for institutional treasuries.

The value proposition is balance sheet compression. An institutional trader invests idle capital in BUIDL, earns the risk-free rate, and simultaneously pledges that token as derivatives collateral. The asset is productive and liquid at the same time. Legacy margin posting meant zero-yielding cash or slow securities pledging. Tokenized collateral eliminates that drag.

In December 2025, DTCC received SEC no-action approval for tokenization services covering Russell 1000 stocks, Treasury bills, and ETFs. Swift announced a blockchain-based shared ledger for its infrastructure stack, working with over 30 global institutions.

The goal is cross-margining: collateral that moves between venues in minutes rather than days. The unlock is straightforward. Less trapped collateral across venues means lower funding costs, tighter spreads, and higher leverage capacity with the same capital base.

A necessary caveat: tokenized collateral doesn't remove risk. It changes the risk profile. Operational risk shifts from settlement timing to smart contract execution and ledger reliability. Counterparty risk remains. What improves is capital efficiency and collateral velocity. The underlying credit and market risks don't disappear.

Stablecoins Compress Time (But Don't Eliminate Complexity)

If repo is the killer app for settlement, stablecoins are the killer app for payments. The connection to the plumbing thesis is direct: stablecoins reduce prefunding requirements and compress time-in-transit, freeing working capital that was previously trapped in correspondent banking chains.

Stablecoin transaction volume hit $33 trillion in 2025, up 72% year-over-year. Combined volumes now exceed Visa and Mastercard's annual throughput.

Visa launched production USDC settlement in December, enabling issuers and acquirers to settle VisaNet obligations on Solana. Stripe's $1.1 billion acquisition of Bridge, the largest in company history, signaled strategic commitment. JPMorgan deployed its deposit token on Coinbase's Base in November, the first bank-issued deposit token on a public blockchain.

The value proposition is specific: stablecoins compress time and reduce prefunding. A corporate treasury can move funds on Sunday and see them reflect immediately, rather than waiting for Monday's wire. For cross-border payments, this means less capital trapped in correspondent bank accounts globally.

But stablecoins don't eliminate FX conversion, compliance screening, sanctions checks, or local banking dependencies. They compress one part of the value chain. The full payment flow still requires on-ramps, off-ramps, and regulatory compliance at both ends.

There are losers in this trade. Correspondent banks face margin compression as prefunding requirements shrink. Legacy remittance rails see volume pressure. Some FX desk economics erode as time-in-transit decreases. The efficiency gains for corporates and fintechs come partially at the expense of incumbent intermediaries.

ETFs as the Liquidity Reservoir

At first glance, ETFs seem disconnected from a plumbing thesis. They're distribution wrappers, not infrastructure. But the ETF complex matters for integration because it created something the institutional market lacked: a regulated liquidity reservoir and standardized position container.

US Bitcoin ETF assets reached approximately $113-170 billion by late 2025, with BlackRock's IBIT holding over 773,000 BTC. Average trade size of $47,000 (compared to $2,400 on retail exchanges) confirms these are institutional flows.

The structural significance isn't retail access. It's hedgeability and collateral eligibility. ETF shares trade on regulated exchanges with standardized settlement. ETF options (IBIT options reached $65 billion in open interest by January 2026) provide institutional hedging tools. This standardization enables inventory management and risk transfer that spot crypto markets couldn't support.

The downstream implications connect to everything above. Institutional custody concentration (BlackRock, Fidelity, and a handful of others hold most ETF Bitcoin) creates operational simplicity. Standardized liquidity improves price discovery and reduces execution risk. Hedgeability via options makes balance sheets more comfortable treating BTC exposure as financeable.

Morgan Stanley expanded crypto ETF access to all clients in October, removing prior $1.5 million asset thresholds. Bank of America enabled 15,000+ advisors to recommend crypto ETFs starting January 2026. These moves reflect gatekeeper risk removal and distribution normalization, not sophistication regression. The wirehouses aren't endorsing crypto speculation. They're acknowledging that regulated wrappers exist and client demand is real.

What Is Not Happening

A necessary boundary: banks are integrating blockchain rails for specific operational functions. They are not adopting permissionless DeFi risk wholesale.

No major bank is providing liquidity to decentralized exchanges. No GSIB is exposing client assets to smart contract risk on public protocols without extensive controls. The integration is selective: private or permissioned ledgers for settlement, regulated stablecoins for payments, tokenized securities with familiar legal structures.

This selectivity reflects a broader pattern. The industry appears to be settling on a functional division: private ledgers for settlement, privacy, and institutional control; public ledgers for distribution, composability, and certain payment rails; with interoperability layers bridging the two.

The "public vs. private blockchain" debate that consumed earlier cycles is giving way to pragmatic coexistence. Different ledgers serve different functions. The competitive question shifts to interoperability: which networks can communicate, and who controls the bridges.

What This Suggests

Three patterns emerge from this integration.

First, partnership models dominate the current phase. Traditional institutions acquire crypto capabilities through technology vendors (Fireblocks, Taurus, NYDIG, Coinbase) rather than building internally. This is expedient but likely transitional. Margin pressure and vendor risk will push banks toward internalization over time.

Second, tokenization serves as the institutional on-ramp. Money market funds and Treasuries provide familiar risk profiles on novel infrastructure. Institutions aren't buying Bitcoin for speculation. They're using blockchain rails to move collateral faster and compress settlement cycles.

Third, regulatory clarity functioned as the catalyst. The GENIUS Act and SAB 122 rescission converted years of pilot programs into production deployments. Banks that delayed during the 2022-2024 uncertainty now face catch-up pressure.

This infrastructure buildout doesn't guarantee crypto will transform finance. Execution risk, security vulnerabilities, regulatory reversals, and interoperability failures remain possible. But the integration has reached a scale that makes it difficult to dismiss as experimental.

The rails are being tested. Whether traffic follows depends on what gets built on top.

In the coming weeks: Traditional finance and crypto markets operate on fundamentally different assumptions: trading hours, custody models, transparency norms, and the role of intermediaries. As these worlds converge on shared infrastructure, those structural differences create both opportunities and friction. We'll examine how TradFi and crypto market structures actually differ, and what that means for portfolios navigating both.

Drop me a line:

What do you think? Do you like this? Do you not like this? I would love to hear your thoughts, so please reach me at akram@span.blog

Enjoyed this? Subscribe for the weekly deep dive, or drop me a line at akram@span.blog.