Thank you Center for Global Interdependence for inviting me to speak today. My remarks touch on crypto assets, but focus more broadly on innovation and the future of finance.1
Innovation is defined as “the introduction of novelty; the alteration of what has been established by the introduction of new elements or forms”.2 This definition does not take a position on whether innovation is good or bad. To our modern ears, innovation generally has a positive connotation. But we must also remember that we have a long tradition of viewing innovation with suspicion. This will be the basic point. Innovation is a double-edged sword, with costs and benefits that affect different groups of people in different ways. This is a lesson for us financial regulators to take to heart. Former Fed Chairman Paul Volcker is a well-known critic of financial innovation, and while credit default swaps and collateralized debt obligations may have helped some to hedge investment risks, the world He famously praised cash machines as a beneficial innovation for banks, but was skeptical of most other financial innovations.3 When thinking about cryptoassets and other forms of innovation, we need to think carefully about which cutting edge we’re looking at. Will innovation create new efficiencies, reduce risk, and promote financial inclusion? Or does it create new risks or exacerbate existing ones?
You recently spoke about crypto assets,Four While we are discussing how individuals can make their own choices about whether to invest in speculative assets, banks and other intermediaries who choose to engage in crypto-related activities should ensure that they are safe and sound. At the same time, the world is changing rapidly and we need to be open to the potential benefits of innovation. From that perspective, he will focus on two areas of innovation that could bring real benefits: tokenization and artificial intelligence (AI).
tokenization
Let’s start with tokenization. As we have explained in our previous remarks on crypto assets, we believe that the crypto ecosystem consists of several parts, including the database management protocol used to record transactions, commonly referred to as blockchain. Blockchain is ultimately a type of decentralized database that can be used to record data, including ownership of assets and transactions within assets (i.e. changes in ownership of assets).Five To date, blockchain has primarily served as a ledger for crypto assets, but it may be well suited to play a similar role for traditional assets such as securities and derivatives.
Before blockchain can be used to facilitate the trading of traditional assets, the assets must first be “tokenized”. In other words, it must be represented on the blockchain so that it becomes a ledger of records for assets. At that point, parties can engage in transactions with tokenized assets by updating their records on the blockchain.
Why would financial institutions undertake this process to tokenize their assets? What are the advantages of blockchain over traditional approaches to conducting transactions? I don’t intend to provide a complete answer to your question, but I would like to highlight a few areas.
First, blockchain can provide fast or near real-time money transfers on a 24/7 basis. This allows parties to precisely control settlement times, potentially increasing efficiency and reducing liquidity risk. Note, of course, that these benefits are not unique to blockchain. The Federal Reserve’s FedNow service, which is set to go live in July, does not rely on blockchain. And we provide secure and efficient instant payment services 24/7 in real time.
Another potential advantage of tokenized assets is that they are “programmable” and have “smart contract” capabilities. A smart contract is a computer program stored on the blockchain that can be programmed to perform predefined actions when certain conditions are met.Once assets are tokenized, they can be built using smart contracts and Execute transactions related to assets. Once the smart contract is activated, transactions will proceed automatically as long as the specified conditions are met. This is what smart means the contract is smart. It does not depend on the parties to the transaction to implement it. Instead, it will implement itself, subject to terms specified by both parties.
Smart contracts may enable so-called “atomic payments”. Rather than relying on each party to execute a leg of the transaction separately, the smart contract combines his two or more legs of the transaction into a single unified “atomic” act executed by the smart contract. can be effectively combined. This is an additional robust method of achieving delivery-to-payment (“DVP”) and payment-to-payment (“PVP”) functionality, where only if one leg of a transaction is settled, so is the other. It is settled only when it is settled. Atomic payments are useful because they reduce the credit risk of the settlement and the counterparty. Conversely, if the buyer does not pay, the seller will not deliver.
In fact, private organizations are testing use cases to better understand the benefits and risks of this technology. Firms are using blockchain technology and smart contracts to execute forex trading to improve efficiency. Apart from this, financial institutions are using blockchain to facilitate intraday repo transactions. The parties to these trades may have more flexibility as to when the trades are settled, which may result in additional capital and liquidity efficiencies. Blockchain atomic payment capabilities could also serve as another way to achieve important risk mitigation. Using a repo agreement as an example, the repo “seller” can have confidence that it will receive a specified loan amount in exchange for the collateral it conveys. The “buyer” of the repo knows to receive the specified collateral.
While these efforts are still in their early stages, we expect to see more participation and growth as their capabilities are expanded with more currencies, eligible securities, and new products.
This is not to say that there are no risks associated with tokenization and the use of smart contracts. Smart contracts can have bugs and potential cyber vulnerabilities. Instant settlement poses a set of risks of its own. But there is considerable potential, and we look forward to seeing what private sector participants come up with to potentially enhance the way traditional trading takes place.
artificial intelligence
The next area to be covered is artificial intelligence. Can you go anywhere without hearing about AI? As you know, AI is currently of increasing interest thanks to so-called generative language models. These types of models are conversational sentences that are very close to the famous “Turing test” in artificial intelligence, and can provide complex responses to user requests. You can even ask him to write his 10 page story about the fox. A fairly sophisticated short story can be ready in a matter of seconds. You can also create presentations, summarize documents, do rudimentary coding, and perform many other functions, all at superhuman speed.
To date, these generative language models remain error-prone, and it is well documented that the technology still has a way to go. That response is often imprecise, but even in such cases, it is often sound As if it knew what it was talking about. As with any source, it is important for anyone using these tools to look at the output with a properly critical eye and not take it at face value.
But progress continues, and one can only speculate as to what these models are capable of in the years to come. So what does all this mean for banks? are testing or using AI in Banks have started using AI models to generate personalized product suggestions for their customers. AI can even generate and send customized marketing emails to that customer based on the model’s product recommendations. Banks are also looking to AI for a variety of customer service applications, such as chatbots that can help reset passwords, locate branches and ATMs, and check account balances without the need for human intervention. AI has also proven useful in fraud surveillance. For example, it helps banks uncover potentially fraudulent credit card transactions by identifying new spending patterns that indicate fraud.
Banks are also beginning to explore the potential of using AI to improve credit underwriting processes and analytics, potentially speeding up underwriting decisions and lowering loan prices.
Like many innovations, AI comes with new risks, or at least new variations of old risks. AI models are only as good as the data they are trained on. This can pose problems when AI relies on large amounts of different types of data, and can complicate the task of detecting problems and biases in datasets. Another important consideration is the “black box” problem. This is because some AI models can have difficulty explaining how to arrive at an output given a set of inputs (this is often referred to as lack of explainability). In some cases, even the AI developers themselves may not know exactly how their AI approach works.
As already mentioned, we are already seeing banks using AI in a variety of ways, and we are having regular discussions with them about understanding and managing the associated risks.6 It remains to be seen if and how they will utilize generative language models. This technology could bring new efficiencies to banks’ software development processes. Or have an application in customer service. Or it could be useful in some way I haven’t foreseen yet. Similar to tokenized assets, we are curious to see how banks can do more with AI. We also want to make sure that we do it responsibly.
Conclusion
Tokenization and AI are just a few of the innovations that will eventually play an important role in banking and, by extension, the economy as a whole. We had a quick talk today about Web3 and quantum computing. All of these innovations have champions who make a case for how they change the world. And I think it’s important to look at such claims critically. But it is equally important to challenge the skeptics who claim that these innovations will do nothing and will be a disaster. The world is changing, and we need to encourage innovations that show the potential to benefit society, including in the financial services sector.
1. The views expressed herein are my own and do not necessarily represent those of my colleagues on the Federal Reserve Board.back to main text
2. “Innovation,” Oxford English Dictionary, accessed 18 April 2023, https://www.oed.com/view/Entry/96311?redirectedFrom=innovation#eid.back to main text
3. Paul Volcker, “Paul Volcker: Thinking Bolder” wall street journal14 December 2009, https://www.wsj.com/articles/SB10001424052748704825504574586330960597134. Return to main text
4. Christopher J. Waller, “Thoughts on the Crypto Ecosystem,” (Speech at Global Interdependence Center Conference: Digital Money, Decentralized Finance, and the Puzzle of Crypto, La Jolla, Calif., Feb. 10, 2023).back to main text
5. The rest of this speech refers to tokenized assets represented on the blockchain. However, assets can also be tokenized in other forms of distributed ledger technology.back to main text
6. Request for Information and Comments on Use of Artificial Intelligence, Including Machine Learning, by Financial Institutions, 86 Fed. Registration 16837 (31 March 2021).back to main text
