This article originally appeared on Forbes.com.

The systems the world uses for transferring money—vast, sprawling, complex and costly—are ripe for change. And change is coming, in the form of distributed ledgers, digital signatures, and virtual currencies. Banks, the traditional intermediaries in the movement of money, must now move quickly to figure out how to deal with this technological threat to their business.

Some $300 trillion in transactions flow through international transfer networks annually, generating $150 billion to $200 billion in revenues for the banks, Bain & Company estimates. It’s not hard to see why banks would be slow to alter a system that’s been such a steady source of fees.

Given the enormity of what’s at stake and the spotty track record of digital currencies, it’s understandable that banks would want to proceed with caution. In early August, for example, Bitfinex, a Hong Kong-based exchange where bitcoin is traded, was hacked, resulting in the theft of about $65 million. Bitcoin is underpinned by the blockchain, the best known distributed ledger.

However, the trend toward digitization of transfers is inexorable, spurred both by a stream of innovations from fintech companies and by a demand for better service from corporate customers. Recent surveys by the Association of Financial Professionals and Temenos show that about 60% of corporate treasurers are dissatisfied with the payment services provided by their banks and have been reviewing their options.

The adoption of distributed ledgers likely will happen in two waves. First is the development of specific systems focused on international payments, a trend that is already well underway. Second will be a broad disruption of domestic payment networks triggered by digital fiat currencies supported by central banks. While the second wave will occur further in the future, it may have more profound consequences for banks, potentially destroying the entire market for checking accounts and credit cards.

It’s worth reflecting on the radical nature of distributed ledgers. Traditional payment transfers rely on a central authority to mediate transactions. When money passes between participants, the central party records and supervises the transfer to prevent cheating. Participants then reconcile the transaction with their own systems, a process that protects both parties but has remained costly, time-consuming, and bureaucratic.

Distributed ledgers, by contrast, operate as secure, shared databases, where each participant has its own copy of the stored data. When a payment gets made, it’s validated by all participants collectively, and updates across the network almost immediately. Transactions can be initiated only by certified parties using encrypted digital signatures. Rather than relying on a central authority to authenticate a transaction, the design of the system itself guarantees one version of the truth. Distributed ledgers cut out the middlemen.

In international correspondent banking, innovators are proliferating. For example, Ripple, a San Francisco-based start-up, has built a functioning payment system around a customized protocol and currency. It’s working with about 30 banks to run pilot tests of its software.

How banks choose to respond to these developments in international transfers depends on their size and location. Payment systems, regardless of their technological foundation, need a critical mass of participants to be effective. That means banks will want to build alliances.

Super-regional banks that already operate in multiple countries may want to consider forging relationships with a few natural allies whose geographic coverage complements their own.

Global powerhouse banks, which handle the biggest volumes of international transfers, face the often daunting prospect of disrupting their own business. They can do this by developing distributed ledger pilot projects internally, or by partnering with the most promising third-party platforms. These megabanks have to meet stringent anti-money-laundering and know-your-customer rules—requirements they can turn to their advantage when marketing the security of their transfer systems to their customers.

While the impact of digital ledgers on international correspondent banking will be significant, the new technology’s consequences for domestic payments systems could be seismic. Already, the People’s Bank of China and the Bank of England are exploring the idea of issuing a national digital currency based on distributed ledgers. If such systems become available to the general public, consumers and business could make electronic payments without using a credit card or a checking account. That could severely constrict the flow of deposits to banks, a critical source of their funding. Sensing this threat, banks are accelerating their investments in digital wallets and payment apps.

Nobody knows how rapidly these two waves of digital innovation—international and domestic—will come crashing onto the shores of global banking. But those banks that begin preparing now will be in the best position to adapt. Executives can start by asking themselves which customers, peers, and third parties they want to work with. Whether a bank chooses to be a cutting-edge innovator or a fast follower, it can map out a plan to win the battle for customer wallets.

How banks respond to the challenge from distributed ledgers will help determine the shape of the global financial system. For banks, coming to terms with digital payment networks is no longer a strategic choice, but a necessity.

Glen Williams and David Gunn are partners with Bain & Company.