The distributed architecture of the Internet has revolutionized the way consumers shop, listen to music, read books, get around town and find a mate. So perhaps it's inevitable that the same forces will eventually upend the U.S. payment system, too.
Still, it was a bit surprising last month when the Federal Reserve System said it was actively exploring such a disruptive possibility. The Fed is a central bank, after all, and the Internet is all about decentralization.
In a highly anticipated paper about the future of the U.S. payment system, the Fed raised the idea of using the public infrastructure of the Internet to facilitate the "direct clearing" of transactions between financial institutions.
The Fed has long stood between banks that are sending and receiving checks and other forms of money. That role would likely be diminished under the scenario the Fed laid out.
The basic concept that banks will use Internet protocols to connect directly with each other, with all of the transactions logged in to a central ledger reflects the rise of technology developed in connection with Bitcoin and other digital protocols. The Fed concluded that this kind of decentralized system has the potential to cost less than today's mainstream payment system.
There are lots of caveats, including the fact that decentralization is only one of the Fed's four options for building a faster, modernized payment system.
But the Fed's report served as validation for innovative payment companies outside the financial mainstream. In particular, San Francisco-based Ripple Labs cheered the Fed's vision as consistent with its own distributed-ledger technology.
Top executives at Ripple Labs said in an interview that they have worked to educate Fed officials about the possibilities enabled by their technology.
"Like how Internet protocols work for communication, that's how Ripple works for value exchange," Ripple Labs Chief Executive Chris Larsen said.
Today U.S. banks that are sending money submit payment instructions to the Fed, which then releases the instructions in batches, five times per day, to the receiving banks.
Ripple Labs sees a scenario in which banks could use its technology to send money to each other cheaply and in real time. In this hypothetical future, "banks can then move these balances across the ledger bilaterally without having to rely on the central bank's server to reconcile and record the payment transfer," said Patrick Griffin, head of business development at Ripple Labs. The central bank would later reconcile its books with the transactions that had occurred on the real-time payment system.
The Fed raised two concerns about using the Internet as the backbone of a modernized national payment system. First is security, a paramount consideration for the Fed at a time when massive data breaches have become routine.
Bank industry officials worry that a payment system built on top of the Internet will be a sitting duck for hackers, the Fed's paper suggested. "There is an instinctual aversion to this option because of security concerns," the report stated.
The security defenses built into this kind of technology, whether from Ripple Labs or another company, have not really been tested adequately, according to Andy Schmidt, research director at CEB TowerGroup.
He said that such technology has not been the target of "industrial-level attacks that using this as a clearing infrastructure for a first-world nation would naturally attract."
The Fed's second concern involves, ironically, the decentralized nature of the U.S. banking system. This country has more than 13,000 banks and credit unions, all of which would need to interact with each other directly in this scenario, rather than using the Fed as an intermediary.
Canada already has an electronic payment system in which banks communicate directly, but that nation also has far fewer banks than the U.S. does.
"Most of the international models that have end-to-end processing have a dramatically smaller universe of financial institutions," said Cary Whaley, vice president of payments and technology policy for the Independent Community Bankers of America. "When hubs are eliminated, many bridges will have to be built."
When it comes to endorsing the disruptive potential of technology, the Fed's paper has its limits.
The central bank made clear that it is not interested in building a modern payment system around Bitcoin or any other digital currency, the supplies of which it does not control. In other words, this was not the equivalent of a major record label endorsing Napster.
"There may be technological innovations within digital currencies that become important for the broader system, as opposed to the currencies themselves," is how Fed Gov. Jerome Powell articulated the central bank's point of view in a recent webcast.
"It's not so much the currency that's interesting, but the technology," agreed Steve Ledford, a senior vice president at The Clearing House, the trade group for the biggest U.S. banks, which is sure to play a large role in the debate over the design of a faster U.S. payment system. "Everybody in the industry is looking to see what we can learn."
There are lots of reasons to be skeptical that the Fed will ever implement the concept of a decentralized payment system. Its recent paper is conceptual and vague, and a long way from becoming reality. Clearly the U.S. payment system is a dinosaur, but the way forward may turn out to be bottom-up innovations rather than a top-down plan.
Still, there is no denying the power of the Internet's distributed architecture, which has already reshaped large swaths of the global economy. In addition, technological breakthroughs have a way of evolving for uses their creators never envisioned.
The Internet itself grew out of research by the Defense Department. Technology that was developed in an effort to make torpedoes undetectable during World War II is now used in wireless networks.
So it is not unimaginable that the Fed might eventually embrace technology that was designed as part of a challenge to the central control of currency.
"I do think there are potential advantages to using a new system like Ripple, where you have distributed Internet processing of the payment," said Elizabeth McQuerry, a payments expert who is a partner at Glenbrook Partners. "It's a very big endeavor to move the entire payment system to that sort of environment," she added.