We don’t know the future of blockchain, but we do know that organizations across the financial sector are deeply invested in discovering its capabilities.

Nasdaq and Citi are partnering on a new blockchain-based payments solution. JP Morgan filed a patent for a blockchain-powered network for interbank settlements. Malaysia’s central bank is exploring ways to integrate blockchain into the banking system. And on and on.

This push by the financial sector to experiment with blockchain makes it a good bet that various asset types could be moving on blockchain networks. That has major implications.

Blockchain’s first major use was in support of Bitcoin, which had a stated goal to eliminate the need for intermediaries in transactions. But the technology has evolved, and that initial aim isn’t practical or desirable for many of blockchain’s uses and potential uses. Now, third-party intermediaries are often the ones pushing blockchain for networks for critical asset flows. Also, auditors, payments network rule-enforcers, and data reporting entities wouldn’t be able to do their jobs without a view into the network traffic. And regulators might gain from a real-time view into those transactions, to understand what’s happening and better prevent problems with the organizations and industries they monitor.

That’s why it’s time for thoughtful discussion about blockchain “supervisory nodes,” which I see as a key part of the blockchain’s future.

In our recent white paper, “Beyond Theory: Getting Practical With Blockchain,” the Boston Fed introduced and detailed the concept of a supervisory node within a distributed, decentralized network. We aren’t presently experimenting with this concept at the Fed, and we know it raises a host of complex technical issues, practical concerns, and risks that must be considered. But the sooner we start looking for answers, the better, because they won’t be easy to find.

Not the “Wild West”

Blockchain for Bitcoin was introduced after the 2008 financial crisis because of mistrust of financial intermediaries. Blockchain is at the heart of many cryptocurrencies, which generally operate without regulators in an industry with a Wild West reputation. But an unmonitored future is not likely for blockchain if and when it becomes a major conduit for global financial assets. We can’t alter the underlying fabric on which critical assets move without watching it for risks to the system or to individual banks related to technical problems, market weaknesses, liquidity problems, etc.

In addition, these assets are largely controlled by financial institutions which aren’t about to move them onto a free and open, anybody-can-do-anything network. That’s where supervisory nodes come in.

In our white paper, we’re clear that we see “supervisory node” as a generic term with potential business functions well beyond the regulatory supervisor role. But the Fed would obviously have a specific interest in the regulatory capabilities of these nodes. I believe they could be critical to making blockchain a transformative technology for regulators, if combined with machine learning and artificial intelligence to offer real-time views and insights into institutions and broader markets. J. Christopher Giancarlo, the chair of the Commodity Futures Trading Commission, is convinced the response to the 2008 financial crisis would have been faster and better-informed if blockchain was around.

“What a difference it would have made on the eve of the financial crisis in 2008 if regulators had access to the real-time trading ledgers of large Wall Street banks, rather than trying to assemble piecemeal data to recreate complex, individual trading portfolios,” Giancarlo told the Senate Banking Committee in 2018.

We have questions …

To become more than a concept, developers of a blockchain supervisory node will need good answers to a host of questions, some technical, some philosophical, some both. Here are a few of them:

• How can we ensure supervisory nodes see only what they should see? Regulators, for instance, may be restricted to viewing only certain types of transactions, but a blockchain network may have lots of other things going on inside it. How do we design the node to protect an organization’s privacy and stop those operating the node from seeing things they shouldn’t?
• Do supervisory nodes create a moral hazard just by being on the network? When people know they’re being watched, that’s often accompanied by some sense that they’re also being protected. Will organizations respond to supervisory nodes by unintentionally changing their behavior to become more lax and put themselves at greater risk?
• How can third-parties insert themselves onto blockchain networks without disrupting them? If you’re in the network, even as an observer, you’re part of it, and that has implications. For instance, what if a supervisory node’s software goes haywire? How will that affect the rest of the network? More broadly, how can a supervisory node be inserted onto a network without negatively altering it?

The Boston Fed’s blockchain paper lists more of these important questions, and has additional details about our thinking on supervisory nodes.

Answers to questions like these will drive the technical design of supervisory nodes. It’s time to start asking them.