Finance disrupted

Speech by Mr Benoît Cœuré, Head of the BIS Innovation Hub, at the 23rd Geneva Conference on the World Economy, Geneva, 7 October 2021.

Good evening.

It is a great pleasure for me to be back in Geneva and see you all. This is only my second in-person speech since I joined the BIS in early 2020. I would like to thank the ICMB for inviting me.

Let me begin by stating the obvious: we live in an age of disruption. We hear every day about businesses, industries, and governments being disrupted. And, of course, our private lives have been disrupted by the pandemic. But tonight, I would like to talk about a specific type of disruption – disruption arising from technological innovation in the financial sector.

Two years ago, "Banking disrupted" was the title of the 22^nd Geneva Report. The report presciently raised competition issues arising from big techs and fintechs. Tonight, I would like to make the point that disruption goes well beyond banking.

Think of cryptocurrencies¹, the rapid rise of decentralised finance, or DeFi, and digital ID systems using biometric data. Think of the explosive growth of data and how firms – particularly big techs – exploit these data. Think of the massive hacks that regularly compromise the personal data of millions of individuals.

What do these stories tell us? They tell us that technological innovation and associated disruptions can be good or bad. New technologies can foster greater efficiency, financial stability and inclusion. But they can also do the opposite, spawning financial instability, loss of privacy, and financial exclusion.

The Great Financial Crisis didn't stem from technological change but from opacity and greed, but it taught us a useful lesson: when finance doesn't work, it takes a heavy toll on society. It took a decade to clean up the mess and reform the financial system.

New technologies are disrupting every corner of the financial sector. Should we let disruption run its course, whatever the consequences? Or do we want to harness the power of innovation in a way that preserves the best elements upon which the financial system is built?

I think the answer is clear. And this is where central banks must step in.

Tonight, I will focus on the role that central banks are playing in ensuring that technological innovation is a force for good, and in developing innovative technological solutions themselves – both domestically and, increasingly, internationally. I will focus on the work we are carrying out at the BIS Innovation Hub, and on the technological disruptions we see in payments and money, banking supervision, and financial markets. At the end, I will share some tentative thoughts about the consequences of innovation for monetary policy implementation.

Digitalisation disrupting payments and money

Payments and money first jump to mind. This is an area of rapid and unprecedented change. Cash is declining while digital payments are on the rise. Covid-19 has just given another jolt to this transformation.

Change started with how customers make payments. We can use our smart phones and watches. Contactless and mobile payments have become part of daily life in many countries including emerging and developing markets.  Four in five Kenyans are using a mobile money service like M-Pesa². Alipay and WeChat Pay accounted for 94% of all retail payments in China last year. Gloves with payment functions are being prepared for the Beijing Winter Olympics.

Most of innovation has been on the "front end" but in recent years, it has moved to the "back-end," the part of the payments system that consumers don't see, involving money flows, and clearing and settlements between financial institutions – the part that not long ago used to be called "plumbing".³

Fast payment systems are a great example, although consumers don't see the new plumbing that is needed to ship money in real time between banks. Services such as the UK's Faster Payment Scheme or the ECB's TIPS allow real-time payments 24/7 and deliver new benefits to consumers.

But other changes come with risks as well as opportunities. Think of global stablecoins, especially if issued by big techs. They are promoted as a way to provide faster and cheaper cross-border payments and deeper financial inclusion. And they do. But they also pose significant risks: they can create closed ecosystems or "walled gardens" that fragment the monetary system, by potentially taking large volumes of payments outside the system that has central banks at its centre.

Stablecoins may also pose risks for financial stability. As clarified yesterday by the Committee on Payments and Market Infrastructures and the International Organisation of Securities Commissions, stablecoin arrangements should observe international standards for payment, clearing and settlement systems to safeguard financial stability, if they perform a payment function and are found to be systemically important.⁴

Walled gardens also have serious implications for competition. They augment the already significant market power of big techs. They also risk threatening consumer privacy and challenge existing regulatory practices.⁵

The history of private money initiatives is not a happy read. Whenever faced with the conflict of interest between making their money stable no matter what and making a profit, private issuers have always chosen profits.

This is where central banks come in.

Money is ultimately a public good whose stability and use needs to be protected by the public sector. This is why so many central banks around the world are working on central bank digital currency, or CBDC – essentially, to ensure that the next generation of money continues to serve the public interest.

If well-designed, CBDC could be a safe and neutral means of payment and settlement asset, serving as a common platform around which a new payments ecosystem can develop. It could enable an open finance architecture that welcomes competition and innovation; and preserve democratic control of the currency.⁶

The BIS Innovation Hub (BISIH) is helping to foster the international development of CBDC. Our centres in Hong Kong SAR, Singapore and Switzerland are building six proofs of concept, or prototypes, with ten central banks in Asia, Europe, the Middle East and Africa, looking at different types and uses of CBDC.

We are looking at wholesale CBDC, which may be used only by central banks and large financial institutions, to facilitate cross-border payments and avoid the use of the correspondent banking system that we all agree is slow, opaque and expensive. And we are investigating the digital equivalent of cash – general purpose (or retail) CBDC.⁷ With the opening of new BISIH centres and partnerships, there will be more projects to come. 

Big data and algorithms disrupting banking supervision

It's not all about CBDC though. Far from it.

Innovation in the financial sector is usually referred to as fintech. Let me narrow the focus to regtech and suptech – the use of technology to improve financial regulatory compliance and supervision. 

Algorithms, artificial intelligence and machine learning, empowered by big data, are transforming financial services. When big techs and credit platforms provide credit, some say that they turn data into collateral⁸. Actually, what they are doing is using data to reduce the need for collateral.

These firms know a lot about us. They collect enormous amounts of data about our preferences, spending habits and payment history – and those of our peers, who may be similar to us - even before we ask for a service or apply for a loan. By using artificial intelligence and machine learning to study a treasure trove of data – typically more than 1,000 data points – they can determine how much we can borrow and repay. And they do it in part by using information that until recently did not have much financial value, like the model of smart phone someone has, or their browsing habits.

Collateral is needed when lenders don't have enough information about a borrower. Data help close the gap. This is very beneficial for the so-called "thin file" customers, those who could not get a loan because of lack of credit history and couldn't build a credit history because nobody would give them loans – a chicken-and-egg problem. There is a financial inclusion benefit from the use of non-traditional sources of data for lending decisions, but also a potential risk for privacy and the management of personal data.

Against this landscape which is changing in quick and often unpredictable ways, financial supervisors have mostly analogue tools. Their workflows are heavily manual. Data collection typically involves reports submitted by paper or email, with file size restrictions and operational and security risks.⁹ For example, performing a cross-firm review often requires going through spreadsheets and PDF files from different sources, minutes of meetings and data from different systems and in different formats. Most of these data are one or two quarters out of date, some could be from previous years.

Figuring out what is going on from these fragmented and outdated pieces of information is, to put it mildly, challenging.  Understandably, supervisors are increasingly worried about being left behind.

Technology can change this game, by giving supervisors access to a lot more data, structured, unstructured, with better quality and granularity than ever before. It can also give them effective means to extract, query and analyse data. To perform the same cross-check review that I just mentioned, a digitally native supervisor could build integrated platforms to avoid using spreadsheets and PDFs. She could use artificial intelligence tools to crunch the data and apply natural language processing and machine learning algorithms to real-time, typically unstructured data from news and market developments.

The BIS Innovation Hub is doing exactly that. The BISIH Singapore Centre is working with the Monetary Authority of Singapore, the Bank of England and the International Swaps and Derivatives Association on project Ellipse, a prototype which investigates the feasibility of an integrated regulatory data and analytics platform. Tools based on project Ellipse would enable supervisors to digitally extract, query and analyse in real time large and diverse sources of structured and unstructured data that are relevant to the residential mortgage market, and anticipate supervisory action.¹⁰ Looking ahead, we will also investigate ways to use the suptech toolbox to support the green and sustainable finance agenda.¹¹

Changes in market structures

I just mentioned the problems of having data that are a quarter too old. Let me now discuss the risks of data that is a thousandth of a second too old.

I'm talking about the intense digitalisation of financial markets. High frequency and algorithmic trading have introduced new actors and redefined market structures. High frequency trading deploys latency arbitrage, or "sniping". Simply put, if you are ultra-fast, you can jump the queue and make a great deal of money. The speed required to gain an edge - millionths or billionths of a second - renders humans and slower computer systems obsolete. A recent BIS working paper estimated that latency arbitrage accounted for 20% of FTSE stock trading volume and imposed a roughly 0.5 basis point tax on trading, increasing the cost of liquidity.¹²

This transformation has given rise to new risks. As noted by the BIS Markets Committee, FX execution algorithms have enhanced the matching process between liquidity providers and consumers in a highly fragmented market, but they shift the execution risk from dealers to users (that is, firms and investors), who may be less capable of managing these risks.¹³

The problem is not new, of course. In the past decade, algorithmic trading has contributed to the occurrence of flash crashes. In May 2010, a flash crash wiped $1 trillion of value of the Dow Jones index in about half an hour. ¹⁴ It may get even worse as computers become even faster and more powerful.

And just imagine what the future advent of quantum computing could do. If quantum computers were able to resolve in minutes the calculations that might take conventional computers months or years, imagine what they could do with the processes that normal computers can already process in nanoseconds.

Other changes in market structures stem from non-banks and fintech firms expanding their footprint in financial intermediation and challenging the role of banks. How can we trace financial contagion spreading in the non-bank financial intermediation universe, by trying to aggregate data from banks, asset managers, other non-bank institutions, clearing houses, and so on? Looking ahead, how can we identify and analyse risks arising from decentralised finance platforms, or DeFi, and the ways in which they spill over to traditional finance?

Here also, new tools will be needed, and here also, the BIS Innovation Hub is helping. The BISIH Swiss Centre is building a tool to monitor trading in fast-paced foreign exchange markets in real time. Project Rio is a central bank-specific market monitoring platform. The cloud-based stream processing platform will process real-time financial data feeds and compute relevant liquidity and market risk measures in real time. Watch this space!  

Concluding remarks

Let me conclude. The central banking community is playing a critical role to ensure that technological innovation and its associated disruptions are a force for good. With this goal in mind, central banks are pursuing innovation and actively developing technological public goods. All of this is meant to have the global financial system deliver greater benefits for citizens. 

Could technological innovation also disrupt the conduct of monetary policy? This could be the next battle line. There are so many dimensions to this question, and I certainly don't have all the answers. Tonight, I will simply raise two questions for our discussion.

First on CBDC.

A great deal is being written about the effect that a retail CBDC could have on monetary policy. Some academics¹⁵ have suggested that it could provide an effective way to implement deeply negative interest rate policies and overcome the "effective lower bound". But if CBDC offers lower interest rates than cash or commercial bank deposits, would you still want to use it? The answer is: up to a point – but we don't know what that point is.

I believe in fact that CBDC could have a greater impact on fiscal policy. Think of the extraordinary support that some governments provided to the population during the pandemic. Some countries showed great ingenuity in using digital technology to reach those most in need¹⁶. Others mailed cheques to people while bank branches were closed because of lockdowns and people were told to stay home. Imagine how much easier it would have been to transfer digital money to people's e-wallets in real time.

My second question is about technological innovation in market structures.

Financial intermediaries are key nodes for monetary policy transmission. Could changes in financial structures affect how they work? It all depends. Disintermediation in forms that by-pass these intermediaries may make this channel less effective – unless central banks consider broadening access to their balance sheets or introducing new instruments.

As a thought experiment, let´s imagine a world where treasuries are exchanged as tokens on decentralised platforms. How would this affect market concentration and the role currently played by large primary dealers and custodian banks in liquidity provision and price discovery? Would current frictions be reduced? How would shocks to the supply and demand of Treasuries be transmitted through the market, and ultimately affect financial conditions? We may never go there, but if we do, consequences for financial stability and monetary policy effectiveness will deserve careful scrutiny.

These are all complex questions that will engage policy makers.  What is clear is that we are in the midst of an age of disruption – for the financial system and the world. The road ahead is exciting abut we don't know where it will take us three, five or 10 years from now. However, what we know is that the direction we take will be defined by our own choices as policy makers and as market participants. I look forward to joining you all for the journey.

Thank you.

¹       See International Monetary Fund, "The crypto ecosystem and financial stability challenges", Global Financial Stability Report, October 2021, Chapter 2.

²       See Central Bank of Kenya, Kenya National Bureau of Statistics and fsd Kenya, "2019 Finaccess Household Survey", April 2019.

³       On central bankers as plumbers, see B Cœuré, "Farewell remarks as CPMI chair", Buenos Aires, 2 October 2019.

⁴       See CPMI and IOSCO, Application of the Principles for Financial Market Infrastructures to stablecoin arrangements, October 2021.

⁵       See A Carstens, "Regulating bigtechs in the public interest", speech at the BIS conference "Regulating big tech: between financial regulation, antitrust and data privacy", 6–7 October 2021. See also S Claessens, F Restoy and H S Shin, "Regulating big techs in finance", BIS Bulletin, no 45, August 2021.

6       See B Cœuré, "Central bank digital currency: the future starts today", speech at the Eurofi Financial Forum, Ljubljana, 10 September 2021, and the reports by a group of seven central banks together with the BIS, "Central bank digital currencies", September 2021.

⁷       For an overview of BIS Innovation Hub projects, see www.bisih.org.

⁸       See L Gambacorta, Y. Huang, Z. Li, H Qiu and S Chen, "Data vs collateral", BIS Working Paper, no 881, September 2020.

⁹       See B Cœuré, "Leveraging technology to support supervision: challenges and collaborative solutions", speech at the Peterson Institute for International Finance Financial Statement event series, 19 August 2020.

¹⁰      See BIS Innovation Hub, "Ellipse: regulatory reporting and data analytics platform", 5 October 2021.

¹¹      See B Cœuré, "Digital rails for green transformation", speech at the Salzburg Global Finance Forum, 22 June 2021.

¹²      See M Aquilina, E Budish and P O'Neill, "Quantifying the high-frequency trading "arms race"", BIS Working Paper, no 955, August 2021. The authors define latency arbitrage as "an arbitrage opportunity that is sufficiently mechanical and obvious that capturing it is primarily a contest in speed". Data was collected over a 9-week period in the autumn of 2015.

¹³      See Markets Committee, "FX execution algorithms and market functioning", October 2020.

¹⁴      See A Haldane, "The Race to Zero", speech at the International Economic Association Sixteenth World Congress, Beijing, 8 July 2011.

¹⁵      See M Bordo and R Levine, "Central bank digital currency and the future of monetary policy", Hoover Institution Economics Working Papers, 2017.

¹⁶      See A Gelb and A Mukherjee, "Digital Technology in Social Assistance Transfers for COVID-19 Relief: Lessons from Selected Cases", CGD Policy Paper, no 181, September 2020.