Getting to the Data: An Interview with Antony Jenkins
Plus: GameStop, GameStop, GameStop
Happy Friday and welcome to another issue of Net Interest, my newsletter on financial sector themes.
Bit of a different format this week. It’s not often a finance story captures the public imagination like GameStop and I can’t not talk about it. But I also want to publish an interview I did this week with Antony Jenkins, former CEO of Barclays and now active in the world of financial technology. So let’s do this: we’ll start with the interview, and then I’ll turn to GameStop. It’s a long one, so you may want to click through on the title and read on the site.
And, as always, if you like what you read, please share with friends, family and co-workers. Thanks.
Getting to the Data: An Interview with Antony Jenkins
We spend a lot of time at Net Interest thinking about the impact of technology on the financial services industry. That means looking at the challenges incumbent banks face, like in Banks vs Fintech: A Coronastory and in The End of Banking. It also means looking at emerging models of financial services, like in Buy Now Pay Later and Online Lending: the Good, the Bad and the Ugly.
To discuss some of these themes, we’re delighted to host an interview with Antony Jenkins. Antony is uniquely placed at the intersection of financial services and technology. Formerly the CEO of Barclays, he went on to found a financial technology company, 10x Future Technologies, which builds technology platforms for banks. Antony is also a director of Blockchain, a main board member of Fannie Mae and later this year will become an external member of the UK’s top financial regulatory committee. Few people have as broad a range of experience across banking, technology and regulation as Antony so it’s great to speak to him.
You spent much of your early career managing credit cards businesses, at Citigroup and then at Barclays. We’ve written about Buy Now Pay Later a fair bit in Net Interest, as a model looking to disrupt credit cards. How vulnerable is the credit card industry?
AJ: I think, in some ways, credit cards were one of the first virtual financial products. I always remember when I was running Barclaycard, I used to get a kick out of standing in the supermarket queue and somebody would pull their Barclaycard out and pay; that was one of my customers! But I had 10 million of them and I never met any of them. Unlike branch banking, where people used to come to the branches all the time. So it was, in many ways, one of the first digital and one of the first virtual products.
Of course, a lot of what goes on across credit cards is just straight payments. But on the credit component itself, my own view is that the credit scoring models that banks use are still very imperfect, and the ability to collect data more quickly, and make better decisions is limited by how data is collected and curated inside the banks themselves, because they do have all this data – it’s just very hard for them to put it together. And it’s possible to imagine a world where you have a much finer, much more granular credit decision making, which would allow, not only the quantum of credit but its pricing to be much more tailored to the individual, and therefore, frankly, better outcomes for the individual. And I think that is a more intriguing and transformational way of thinking about things, rather than some of these other models like Klarna which is a great business, but it’s basically point of sale financing and that’s existed for 100 years. So I’m not sure that’s an existential threat as much as the ability to use data, better.
This idea about using data better—can you expand on it?
AJ: Sometimes you get to hear bank CEOs at Davos say, oh we’re a technology company now. Banks are not technology companies, they are absolutely not technology companies; they are data companies. They use technology to handle data, often not very effectively but they need to think of themselves as data companies because data is effectively your raw material and you need to handle it much more effectively.
Banks talk a good game about big data. But actually getting to the data is really difficult, because it’s so fragmented across many different systems. But if you could really get to the data and use it real-time, then your ability to make better lending decisions, handle operational matters more quickly, deal with fraud more quickly, target products better, would be much greater.
And so it’s a combination of opportunities to serve the customer better, opportunity to squeeze out defectory work and cost that way. And frankly, just the cost of maintaining these legacy systems is vast; huge mainframe systems sitting in data centres, with full redundancy and all of that old software that doesn’t work very well. Also very hard to secure from a cyber point of view, because it wasn’t really designed for that. Some of this is no doubt the fact that the banks business – the way they serve customers – hasn’t fully transitioned to digital yet. But most of it is because of the sort of millstones of technology and operations that they have to deal with.
After you left Barclays you founded a company, 10x Future Technologies, to help banks deal with some of these technological issues. What was the thinking behind that?
AJ: Yeah, I sort of triangulate it into 10x from three perspectives, really. The first was this question of technology transformation. And I’ve been a big believer in that for many, many years, I’ve done a lot of work in financial services around innovation. I keep going back to the mid 90s, when we were trying to build, when I worked in the US, a receivables processing system. And the key there was, could you scan invoices and process them for large corporates. The gating factor was the cost of a T1 line at the time. So we built the business case on the fact that we could just about afford one T1 line, then we realised we needed another T1 line for redundancy. So technology continues to move on, increases in capacity capability. And that means that complex verticals like financial services become capable of transformation when perhaps they weren’t in the past.
But I think it’s also allied to a couple of other things. One is that banking is the least customer-focused business in the world. And the industry has traditionally made a living out of pumping product, most of which the customer needed, some of which they didn’t. But the idea is can we become more focused on truly identifying and meeting the needs of the customers by deploying technology.
But the third area – which will be of interest to you given your background as a bank investor – is that the industry really has a lamentable return on equity. And that’s partly a function of the amount of capital the banks have to hold post the crisis. And, you know, I’m a big supporter of banks having strong balance sheets, but the economic model really hasn’t adjusted. And so when you get a situation where banks are delivering, you know, 5, 6, 7 percent ROE, when their cost of capital is at least 10 percent, and you see this reflected in their their price to book, you know, it’s a real problem, because that’s effectively an economic transfer away from the ultimate owners of the business, the shareholders.
For me, these three issues are issues that I’ve been grappling with virtually all my career: how do you create transformation using technology, how do you serve customers better and thirdly, particularly in my time as CEO of Barclays, how did you deliver a sustainable return on equity?
But really, the answer to that question is the bank’s cost base is just too high, is too high relative to revenue and too high relative to capital. And if you really unpack that, and that’s to do a lot with the fact that technologies that banks are working with are incredibly inefficient, very expensive and highly fragmented. So all of these things, just as we were approaching a time where there’d be a market for a radically different solution. And that was, frankly, something that I was looking for when I was doing various jobs at Barclays and including being the group chief executive and I couldn’t find anything in the market. So I decided to build it myself.
What makes it so difficult for banks to get technology right?
AJ: It’s a combination of technology, but also sort of leadership, cultural and organisational issues inside banks. And like all organisations, there are different fiefdoms and territories and banks tend to be organised around – I’ve got my retail bank, I’ve got my mortgage division and my credit card division and my savings division and my lending division, and the systems have tended to follow those organisational constructs. And if you think about data, data is only really useful when it’s attached to an individual. It might be interesting to look at, you know, I’ve got a million mortgages in the UK and look at their characteristics in aggregate. But I can’t really do something until I know it’s about your mortgage, or my mortgage.
But it’s very difficult for banks to form the view of the customer because these systems are architected around products. And additionally, there are many multiple sort of subsystems inside the bank, there might be a core mortgage processing system, there will be a customer service system, a fraud system, a payment system, a collection system, all of these systems have a different view of the customer. So it’s very hard to create that aggregate picture of the customer and the customer’s business with the institution. It’s effectively as if the customer appears thousands of times in the records of the bank.
And so, this information should be incredibly valuable. But the difficulty is getting at it, and banks will talk to you about we’ve got this great customer master file – we take all the data out of the product systems and put it into a data lake. But that data then has to be cleaned up, it has to be made sure the Antony Jenkins in the mortgage system is the same as the AP Jenkins in the credit card system is the same as the A Jenkins and the current account system is connected to the AM Jenkins in the savings system, who happens to be my wife. So cleaning that data up and operating with it is really tricky for banks.
And one of the things that is hard for them as they compete with a Google on the one hand, or a FinTech on the other, is they’re just much better – they’re much better at the user experience, of course they are – but they’re just much better at handling data.
And one of the core principles that underpins 10x the platform is, it’s built around the customer. So that once the customer is on the platform, we can expose a set of functionalities to the customer that looks like a product, but isn’t delivered through a product system. And the great thing about that is you only have to onboard the customer once, so when the customer signs up for a current account, and the next day, they want a savings account, you don’t have to create an entirely new customer record around savings. And equally for a customer service person, they could look at the customer’s entire business in one place. The key design principle of designing around the customer as opposed to around the product is one of the things that distinguishes us.
It sounds like organisational structure is a major issue at the big banks as well as their ability to adopt new technologies?
AJ: I mean, it’s the long history of incumbents isn’t it? Why do incumbents usually fail to make the transition? And a lot of it has to do with leadership and cultural issues in the organisation. I always loved the fact that Kodak were the people that invented digital technology. There’s lots of case studies about Kodak, but they were so dominant in the film sector, their profits were immense, both in absolute terms and in margin terms. If you ran the film business in Kodak, you were basically King of the World, and nobody could challenge you. And so even though this group of bright folks over here had invented digital photography, nobody ever thought it was going to be anything much, because it could never compete with the big behemoth film business. And it’s pretty much the same if you look at Blockbuster, you look at Nokia, you look at all of these companies that often have huge market positions, but they lost them, they ceded them, because technology created a better way of doing things.
Now, I think, for various different reasons the banking industry is less susceptible to those trends, but it’s not completely defended from them. Clearly regulation does play a huge role in our industry. Brand matters. Incumbents have millions of customers, lots of money, so their position is strong, but it’s not invincible. And I think there’s a realisation increasingly, not so much that banks are going to be totally replaced by FinTech or Big Tech, but that what will happen is the pieces of the business will be sliced away, leaving banks with progressively less and less profitable parts of the business, and more and more legacy costs. And I know in this country, it’s very controversial because people don’t like branches closing in their neighbourhoods. And also there’s a lot of people out there who are advocating, we should continue to have cash and we should continue to have cheques and these things, and they are incredibly expensive for banks to deliver. So I think all of those forces are sort of coming together.
It’s interesting isn’t it, how banks remain so vertically integrated, in contrast with other industries. Does technology represent a threat to that?
AJ: Yeah, if you go back, say 50 years, people who consume financial services – and let’s be honest, it was mostly middle-class people who had access to the banking system – consumed it through one provider. And then you got the proliferation of provision of services through non-banks, so you’ve got supermarkets out there, doing financial services, you had online credit card companies. And then to make sense of all of that, you had the comparison websites pop up, which allow customers to begin to assess competition based on price. But none of that really created…
The reason why I called the company 10x is because I believe we only adopt technology when it makes our lives a lot better; we don’t adopt technology when it makes it a little bit better. So you know, Netflix was kind of interesting, when they would mail you a DVD. But it didn’t change people’s lives until you could stream a vast amount of content. So part of what I think hasn’t happened yet, is we haven’t had that sort of 10x moment, where the experience suddenly becomes 10 times better and banks haven’t created that themselves yet. But it is possible, of course, for Big Tech companies to do that, particularly when you put that together with Open Banking. So, if you’re a Facebook or a Google who knows a lot about you, as a person, or me as a person, it’s not a strange thing, why not just integrate all the banking activities in there so that you can take out the cognitive load of people having free to run their finances.
And for everybody, even people who work in the industry, managing your money is just a painful chore, you know, you always feel like you don’t know enough, you always feel like you might be making the wrong decision, you know, you have to address things nobody wants to confront, like, what happens if I die, you know, can I provide for my dependents and so on. And so if there’s a way to sort of reduce that cognitive load, and allow people to engage with the system, in a very kind of light touch but highly effective way, then that’s when I think you start to get a shift in behaviour.
And you can begin to see a capability that allows people to manage their financial lives much more effectively on the one hand, with the banks themselves, certainly getting pushed back down the value chain, to being managed as a balance sheet, you know, highly regulated capital intensive balance sheet businesses. And you can probably make reasonable money out of that. But you have to be very good at handling your business efficiently and effectively. And you have to deploy technology efficiently and effectively. And if you look at what’s happened in China, for example, people get a lot of their stuff through things like WeChat. And the banks, the banks have to compete for your business. There’s 20 banks in the panel, and they have to compete for your business. And it’s all automated and click of a button and you’re done. So, you know, there are some models that I think can suggest where things are likely to go.
I think increasingly you have to make a decision. Are you going to be in that business? Or are you going to be in the balance sheet business? You might be able to be in both, but you’re going to have to be very good at both if you want to be in those businesses.
It’s slightly unfortunate for the big banks that this massive shift towards digital technology occurred in the decade that they were otherwise distracted cleaning up after the crisis.
AJ: Yeah, and I think you’ll know this from your time as an investor. I suppose when you’re talking to a lot of CEOs, in the 2012/13/14 period, there was a view a) Actually these regulations, people aren’t serious, they’ll back off in a while. We’ve got to take our medicine but actually, regulators will back off. That didn’t happen. Secondly, interest rates are low, but they’re going to go back up, they’re going to revert to the mean, and that’s massively beneficial for our economic model. That didn’t happen. And then the third thing was this tech stuff, you know, who’s going to put their money with a FinTech nobody’s ever heard of, we don’t have to worry about that; our customers like doing things face-to-face. So the leadership of the banks in the first part of the last decade were wrong on all of those points.
As well as your role in 10x, you’re also a Director of Blockchain. Where do you see the potential for blockchain?
AJ: Yes. I got involved in blockchain because I was interested to see what was going on as much for the technology as for cryptocurrency. With cryptocurrency, you can have various views on it, whether you think it’s just a scam or whether it’s digital gold, or you can have a view on it. Interestingly, many of the sceptics on crypto seem to think that fiat money is in some way superior. But, actually fiat money is not backed by anything either in today’s world, so it is literally worth the paper it is printed on.
But putting all that to one side, I was really intrigued by the technology. And one of the things that I think has happened if you go back to just how we transfer value, for example, in the old days a pound note was actually a receipt for gold held at the Bank of England but rather than me go and get my gold and then give it to you, I just gave you the receipt. It was actually backed by assets. Now, it’s not backed by assets today, as I said, but as the world’s become more complicated because I interact with 100 people and I can’t do bilateral exchanges.
What happened was we developed a series of counterparties. So if I want to make a payment to you today, I instruct my bank. My bank then makes payment through some central clearing mechanism - faster payments or chaps or whatever. It goes to your bank and then it goes to you. Now, in that system, there’s a lot of friction. And there’s a lot of risk if things break at any point along the way. Theoretically speaking in a distributed ledger system that could happen instantaneously between you and me, just as I would send you an email.
And when you think about that, when you think about the friction that’s involved in international trade, for example, when you think about the sort of rent-seeking that goes on in international currency markets, you could imagine a huge economic boon to using some of these technologies to transfer value.
The same is true of real estate. When I bought my first apartment in the US, and this admittedly was around 1999, 2000, but I’m told things haven’t changed very much, about eight people show up at your house and you have to write them all cheques, because they’re all there to provide something, the title insurance person - what do you mean I have to insure the title? Anything that can be represented digitally, whether it’s real estate, fine art, diamonds any of those things, you could imagine a world in which ownership could transfer instantaneously. And also, you could then link it to financial products. So if, for example, we put the land registry on a distributed ledger capability, I could then connect a mortgage to that.
But more importantly, if I wanted then to sell that mortgage on, I could do it again at the click of a button. So when you buy mortgage backed securities, you’re basically buying a basket of securities that has been selected for you with certain characteristics. In this world, you can literally pick them by the ones. Now you wouldn’t do that manually, you’d have some sort of algorithm doing it for you. So you can imagine almost every aspect of the financial system being reinvented in a distributed way as opposed to an intermediated way. That I think is a fascinating set of possibilities.
Thank you, Antony. It’s been great to catch up.
More Net Interest
GameStop I: The Short Sellers
Short sellers have been getting a bad rap this week. A Reuters story reports that some have been warned to “watch out”; elsewhere they have been described as “enemies of the people”. (h/t NoonSixCap). It’s a shame, because many of them do good work. We’ve pointed out before that the potential to make lots of money incentivises short sellers to ferret out fraud and keep the markets in check. The payoff can be direct if the short book performs well but the threshold to making great returns on shorts is limited by an upside that’s capped (stocks can’t go any lower than zero) and a high cost of carry. So, more often than not, the payoff comes via a larger portfolio of long positions, which the short book helps to fund.
“But isn’t it the regulator’s job to keep the markets in check and ferret out fraud?”
Perhaps. Unfortunately the incentives aren’t quite the same. A recent paper analyses the incentives in place at the US Securities and Exchange Commission. The paper finds that promotion opportunities represent the main way SEC attorneys are incentivised to engage in enforcement. In some instances, organisational design depletes those opportunities and enforcement suffers. But even when promotion opportunities are present, the paper finds that they translate into an average monetary incentive among SEC enforcement staff of 23%. This compares with an average incentive among CEOs of 349%, which is somewhat asymmetrical. I have no idea what the average incentive is for short sellers, and its variance is likely greater, but it goes some way to address that asymmetry.
We are currently witnessing exactly the kinds of market conditions that breed fraud. As JK Galbraith observed, “In good times people are relaxed, trusting, and money is plentiful. But even though money is plentiful, there are always many people who need more. Under these circumstances the rate of embezzlement grows, the rate of discovery falls off, and the bezzle increases rapidly.” Short-sellers are needed to monitor the bezzle.
GameStop II: The Brokers
I’ll be honest, I’ve got no idea how the hot water system in my home works. I set a timer, push a few buttons and it’s just there, every morning, just when I want it. Except when it isn’t of course, and I have to call out an engineer. He comes round and shows me pipes and valves that I never knew I had. Plumbing’s like that: unnoticeable when it works and painful when it doesn’t. But at least I know it’s there. I may not know what each and every pipe and valve does, but I know there’s a system working under the floorboards to deliver my hot water every morning.
On Wednesday of this week, a record 120,000 people downloaded the Robinhood app to their phones. Daily average users of the site topped 2.6 million. Many were attracted by the gains (and excitement) available from trading stocks. But until later in the week, the plumbing remained hidden. In fact, Robinhood’s whole interface is about hiding the plumbing—removing the friction, making it easy. So not only did users not know about the valves and the pipes, few may have even known they were there.
Which is why when the plumbing did back up on Thursday, an initial response was one of conspiracy.
But Robinhood is more than a fun app; it’s a financial institution. And as a financial institution, it runs risk. That risk is reflected in its balance sheet. These numbers are already past their sell-by date, but when they were last disclosed, as at end June, Robinhood had $494 million of risk capital and $235 million deposited with clearing institutions. The deposit acts as collateral in the event that the clearing institution is unable to satisfy settlement across Robinhood’s trades. As the settlement risk increases, the deposit needs to increase. And a combination of the concentration of exposure around a few stocks, notably GameStop, and the volatility of those stocks causes the settlement risk to increase. By a lot. The founder of competing broker, Interactive Brokers, said that on Thursday evening there was roughly $10-15 billion of losses on one side of the GameStop trade with $10-15 billion of gains on the other.
So, that evening Robinhood was forced to raise more equity capital and draw down some funding. They were fortunate; other brokers confronted with market volatility in the past have been less so. In January 2015, a surge in volatility around the Swiss France led to the collapse of foreign exchange brokers FXCM and Alpari UK.
In the belly of the curve Robinhood is an app but at the tails it is a financial company. And in finance the tails should never be underestimated.
GameStop III: The Market Makers
If you’re looking for a beneficiary of all this, it’s probably the market makers. One of the biggest is Virtu Financial. Retail brokers execute their clients orders via firms like Virtu, who earn a small spread between bid and ask prices. According to the company’s Co-President, “Our business benefits from both episodic and sustained increases of market volumes and volatility.” Well, that’s been happening.
The company’s clients include Robinhood, ETrade, Fidelity and others. It does around 30% of US retail order flow. The company was already benefiting from increased retail activity and will only be benefitting more in the current environment.