Core Banking Modernization: Why Legacy Ledgers Can't Handle Crypto

Most banks still run on core banking systems built decades ago. These platforms were designed for a world where markets close at 5 PM, settlement takes days, and every transaction fits into a neat double-entry ledger denominated in a single fiat currency. That world is disappearing. As stablecoin payment volumes climb toward trillions of dollars annually and regulated banks begin issuing their own digital tokens under frameworks like the GENIUS Act, the gap between what legacy cores can do and what the market demands is becoming a strategic liability.

Core banking modernization is no longer optional for institutions that want to participate in digital asset markets. The question is no longer whether to modernize, but how: full replacement, middleware integration, or a hybrid approach that layers new capabilities on top of existing infrastructure.

What Legacy Core Banking Systems Were Built For

The architecture of most incumbent core banking platforms dates to the 1970s and 1980s. Systems like FIS Profile, Fiserv DNA, Jack Henry SilverLake, and Temenos T24 were designed around batch processing: transactions accumulate during business hours, then settle in bulk during overnight processing windows. This design made sense when the fastest payment rail was a wire transfer that closed at 6 PM ET, and the primary asset types were deposits, loans, and certificates of deposit.

These systems are deeply embedded. Roughly 43% of banking systems worldwide still rely on COBOL, a programming language introduced in 1959. An estimated 95% of ATM transactions in the United States run on COBOL-based mainframes. The average COBOL developer is over 55 years old, and the talent pipeline is thin.

None of this means these systems are bad at what they were built for. Batch-based core banking handles traditional deposit operations, lending, and regulatory reporting with proven reliability. The problem is that crypto and stablecoin operations violate nearly every assumption these systems were built on.

Why Legacy Cores Fail at Crypto Operations

The mismatch between legacy cores and digital asset operations spans multiple dimensions: time, speed, asset types, and settlement models.

No support for 24/7/365 operations

Traditional banking operates on business-day cycles. Core systems have end-of-day processing windows, weekend cutoffs, and holiday calendars baked into their architecture. Bitcoin and stablecoin networks never close. A stablecoin payment arriving at 2 AM on a Sunday needs to be booked, reconciled, and reflected in customer balances immediately. Legacy cores cannot do this without significant re-engineering because the batch cycle is not a feature: it is a fundamental architectural constraint.

Batch settlement versus real-time finality

In traditional banking, a payment instruction and its settlement are separate events. ACH transfers batch and net across banks once or twice a day. SWIFT messages initiate flows that settle through correspondent banking chains over one to five business days. Even RTGS systems like Fedwire operate only during specific hours.

On-chain settlement is atomic: the message is the settlement. When a stablecoin transfer confirms, the value has moved. There is no separate clearing step, no netting window, no T+1 or T+2 delay. Core systems built around deferred settlement struggle to represent this: they need to book the transaction, but the reconciliation process that confirms it may not run until hours later.

Single-asset ledger design

Legacy cores model the world as fiat-denominated accounts. Adding a new currency is already a significant project, typically requiring months of configuration, testing, and regulatory approval. Now consider adding not just a new currency but an entirely new asset class with different settlement properties, different decimal precision (Bitcoin uses eight decimal places, not two), and different custody models. Most cores lack the schema flexibility to represent digital assets natively, forcing banks to track them in separate systems and reconcile manually.

No multi-chain awareness

A single stablecoin like USDC exists on more than a dozen blockchains, each with different confirmation times, fee structures, and finality guarantees. A bank handling stablecoin deposits needs to track which chain a payment arrived on, verify its finality, and book it accurately. Legacy cores have no concept of chain IDs, block confirmations, or transaction hashes.

The Three Modernization Approaches

Banks evaluating how to bridge this gap generally choose from three strategies, each with different cost profiles, risk levels, and timelines.

Full core replacement

The most ambitious approach: rip out the legacy core entirely and replace it with a cloud-native platform built for modern requirements.

Vendors like Thought Machine (Vault Core), Mambu, 10x Banking, and Finastra offer platforms designed from the ground up for API-first, event-driven architecture. Thought Machine reported a 57% year-over-year revenue increase in 2025, with recurring revenue reaching $100 million by Q2 2026 and 68 bank clients across 30+ countries. Lloyds Banking Group and Standard Chartered are among the institutions migrating to its platform.

The challenge is cost and risk. Commonwealth Bank of Australia spent over $1 billion across five years on its core replacement. Most large banks estimate multi-year timelines and nine-figure budgets. A 2026 survey of 150 UK banking leaders found that core system maintenance already consumes about 24% of IT budgets, roughly £3.3 billion per year across UK banks alone.

The replacement paradox: Banks spend billions maintaining legacy cores they know are inadequate, but the cost and risk of replacing them often exceeds what any single business case can justify. According to DXC, 98% of banks planned to upgrade their core systems within three years as of 2025, yet most have not moved beyond planning.

Middleware integration layer

Rather than replacing the core, this approach adds a specialized layer between the legacy system and digital asset operations. The core continues handling traditional banking while the middleware manages crypto-specific functions: wallet management, on-chain transaction monitoring, multi-chain asset tracking, and real-time settlement.

Platforms like Fireblocks offer APIs that integrate directly with core banking systems including Temenos, Avaloq, and FIS. Banks like Germany's Bankhaus von der Heydt have deployed Fireblocks as their digital asset custody and transaction layer. Anchorage Digital, a federally chartered crypto bank, provides similar infrastructure for institutions that need regulated custody alongside their existing banking operations.

The middleware approach is faster to deploy and lower risk, but it creates architectural complexity. The bank now operates two parallel systems that must stay synchronized. Reconciliation between the legacy core and the crypto middleware becomes a continuous operational challenge, and the batch windows of the core can create temporary inconsistencies with the real-time state of the middleware layer.

Hybrid: the "side core" model

The emerging middle ground is what analysts call the "side core" or "hollow the core" strategy. The legacy system retains its role for traditional products, while a cloud-native side core handles digital assets, stablecoins, and real-time settlement. The two systems share customer data and reporting infrastructure but operate independently for transaction processing.

This is the approach most commonly recommended by BCG Platinion and McKinsey for large incumbent banks. IDC projected that 40% of global banks would run modern cores alongside legacy systems by 2026, rising to 70-80% by 2028. The Jack Henry Fintech Integration Network, for example, now includes Stablecore, enabling community banks and credit unions to offer 24/7 stablecoin transactions without replacing SilverLake or Symitar.

Banks That Have Already Modernized

Several institutions offer case studies in what modernization looks like in practice.

JPMorgan: Kinexys (formerly Onyx)

JPMorgan did not wait for vendors. It built its own blockchain-based settlement platform, originally called Onyx and rebranded to Kinexys in late 2024. The platform has processed over $1.5 trillion in transactions since its 2020 launch, averaging more than $2 billion per day. It supports intraday repos, cross-border payments, and on-chain foreign exchange settlement. Clients include Siemens, BlackRock, and Ant International.

Kinexys operates alongside JPMorgan's existing core infrastructure. The bank did not replace its core: it built a parallel settlement layer for programmable, 24/7 operations. This is the side-core model executed at the largest possible scale.

BIS Project Agora: the central bank experiment

The Bank for International Settlements launched Project Agora to test tokenized cross-border wholesale settlement. In July 2026, 28 financial institutions and central banks completed live-value trials, settling approximately $1 million (800,000 Swiss francs) across 17 transaction scenarios in six currencies. The average settlement time was roughly 80 seconds. Participants included the Bank of England, Bank of France, Bank of Japan, JPMorgan, Citi, Deutsche Bank, BNP Paribas, UBS, Standard Chartered, and MUFG.

Project Agora is still experimental, not production infrastructure. But it signals that central banks and the world's largest commercial banks see tokenized settlement as the direction of travel: not a question of if, but when.

The Regulatory Catalyst

For years, regulatory uncertainty gave banks an excuse to delay crypto integration. That excuse is disappearing. The GENIUS Act, signed into law in July 2025, created the first comprehensive federal framework for stablecoin regulation in the United States. It sets reserve requirements, establishes issuer licensing, and creates a clear path for banks to issue or handle stablecoins.

The law itself does not mandate specific technology infrastructure. But its requirements for real-time reserve attestation, continuous compliance monitoring, and 24/7 transaction processing implicitly demand capabilities that legacy cores lack. A consortium of 21 major banks is reportedly building a shared stablecoin, with a target launch in the first half of 2027.

The compliance shift: With the GENIUS Act in place, the primary barrier to bank stablecoin participation has shifted from regulatory uncertainty to technology readiness. Banks that delayed infrastructure investment waiting for regulatory clarity now face the reality that competitors are already building.

The Cost of Not Modernizing

The financial case for modernization grows more urgent as stablecoin volumes scale. Allium data shows stablecoin payment settlement reached approximately $482 billion in the first three quarters of 2026, up 45% year over year. Payment transaction counts surpassed 1 billion, a 71% increase. Business-received stablecoin volume grew 37% to $246 billion, representing roughly 64% of total payment volume.

Banks that cannot process these flows risk losing them entirely to fintech competitors and banking-as-a-service providers that operate on modern infrastructure. The risk is not just lost transaction revenue: it is disintermediation from the payment flows that define banking relevance.

• Cross-border B2B payments are migrating to stablecoin rails that settle in seconds instead of days through correspondent banking chains
• Payroll and contractor disbursements in emerging markets increasingly use stablecoin rails, bypassing banks entirely
• Treasury management teams at multinationals are exploring stablecoin positions for 24/7 liquidity, creating demand that banks with batch-only cores cannot serve
• Institutional custody clients expect a single view of fiat and digital asset positions, which requires integrated infrastructure

The competitive pressure is not hypothetical. As detailed in our analysis of embedded finance and banking-as-a-service, non-bank providers are already capturing payment flows that traditionally belonged to banks. When those flows move to stablecoin payment rails, banks without modern infrastructure cannot even participate in the competition.

The Buy vs. Build Decision

For most banks, building proprietary digital asset infrastructure from scratch is impractical. JPMorgan can afford to build Kinexys because it has one of the largest technology budgets in financial services. Regional and mid-size banks need a different calculus.

When to buy

Banks that want stablecoin and crypto capabilities within 12 months should look at integration partners. Middleware providers like Fireblocks and custody platforms like Anchorage offer pre-built infrastructure that connects to existing cores through APIs. BaaS providers and fintech integration networks, like Jack Henry's partnership with Stablecore, package these capabilities specifically for community banks that lack large technology teams.

When to build (or co-build)

Banks with strategic ambitions in digital assets: issuing their own stablecoin, running a tokenized deposit platform, or becoming a crypto custody provider: need deeper control over the infrastructure. This typically means deploying a cloud-native side core (Thought Machine, 10x Banking) and building custom integration layers on top. The timeline is longer, but the result is a differentiated platform rather than a commoditized feature set.

The SDK approach

A third path is emerging for banks that want to support Bitcoin and stablecoin transfers without building or buying full digital asset infrastructure. Protocol-level SDKs provide clean APIs for specific operations: sending, receiving, and settling digital assets: without requiring the bank to operate blockchain nodes, manage channel liquidity, or build its own wallet infrastructure.

Spark's SDK, for example, enables instant settlement of Bitcoin and stablecoins through a single integration. Banks and fintechs using the Spark SDK do not need to run Lightning nodes, manage payment channels, or handle the operational complexity of on-chain transaction monitoring. The SDK abstracts the protocol layer, exposing simple send and receive operations that a core banking middleware can call like any other payment rail. This is the same infrastructure that powers wallets like General Bread, which offers users instant Bitcoin and stablecoin transfers built entirely on Spark.

What a Modern Core Banking Architecture Looks Like

Whether banks replace, wrap, or augment their legacy cores, the end state converges on a common set of capabilities. A core banking architecture ready for digital assets must support the following.

• Event-driven transaction processing that eliminates batch dependencies and supports continuous booking
• Multi-asset ledger design with configurable decimal precision and native support for tokenized assets alongside fiat
• Real-time reconciliation engines that continuously match on-chain state with internal ledger state
• Chain-abstracted settlement interfaces that normalize differences between blockchains, Layer 2 networks, and traditional payment rails
• API-first architecture enabling integration with custody providers, compliance tools, and protocol SDKs
• Programmable compliance that enforces KYC/AML rules, travel rule requirements, and jurisdiction-specific regulations at the transaction level

This architecture maps directly to the composable banking API model that forward-looking fintechs are already deploying. The difference is that incumbent banks must achieve it while maintaining continuity for millions of existing customers and trillions of dollars in existing deposit balances.

The Road Ahead

Core banking modernization has been discussed for decades. What makes this moment different is that the market is no longer waiting. Stablecoin payment volumes are growing faster than any new payment rail in history. Central banks are actively testing tokenized settlement. And for the first time, there is a clear regulatory framework that tells banks what compliance looks like for digital assets.

Banks that modernize their infrastructure now will be positioned to capture always-on settlement flows, serve institutional clients with integrated fiat and digital asset views, and compete with fintech challengers on speed and cost. Banks that delay will find that the batch-based, business-hours-only model they built their operations on is increasingly irrelevant in a market that never sleeps.

For a deeper look at how stablecoin rails compare to traditional payment infrastructure, see our analysis of stablecoin payment rails versus traditional systems. Developers exploring how to integrate Bitcoin and stablecoin capabilities into banking applications can start with the Spark SDK documentation.

This article is for educational purposes only. It does not constitute financial or investment advice. Bitcoin and Layer 2 protocols involve technical and financial risk. Always do your own research and understand the tradeoffs before using any protocol.