Interview: “China is creating more capital market-enabled financial systems through technology platforms”

Emmanuel Daniel, The Banking Conversation (ED): Lufax is one of the pioneering and perhaps largest and most successful peer-to-peer (P2P) lending platform on every front—both for institutional investors and the retail borrower. Is Lufax just a platform or is it a wholly owned asset-holding business that mirrors banking?

Gregory Gibb (GG): Lufax has changed over time from the day we started in 2010, to what we are today. If we base on where we are today, it really is a platform. But it’s a platform that needs three key pieces to make it work. First, we have to be able to source good products. In China, that means on the individual borrower’s side. Lufax does its own sourcing. We do own the offline and online components and are bringing in borrowers and screening them. So we’re very much control in in a full way.

Second, we also have a lot of people working with banks, trust companies, leasing companies, and finance companies to take their assets off their balance sheet and create or structure fixed-income products. So the whole asset sourcing is key because that’s where the beginning of quality control begins. If you don’t have a view of who should be coming in or not into your platform, it can create a lot of problems.

The third part, which we started to get into about 9 months ago, is all of what we call standard products in China. These are all of the mutual funds, insurance products, and investment-based insurance products or protection-based insurance products. For example, today we have about 2,500 mutual funds in the platform.

ED: What is the size and composition of your asset base today? Do you buy the assets?

GG: The total volume that went through our platform last year was RMB1.7 trillion. Of that, RMB600 billion was accounted for by retail customers. The big part—at one point, RMB1 trillion—were institutional investors. In the latter, majority came from banks—assets that have been on balance sheets of banks for a short period of time, or have come through the trust companies, which have come through security companies. We make a selection based on the data and the performance of the institution.

Basically, we identify assets that will be placed in the customer base. So the customer base buys something that has just been freshly created a second ago, and then it goes on the platform and is sold. If you break down the assets, all of the consumer-financed, or what we call peer-to-peer (P2P), is where the borrower is an individual and the investor is an individual. Out of the RMB600 billion done last year, P2P is probably about RMB50 billion.

Peer-to-peer is where we started Lufax five years ago. But other businesses have grown more quickly, hence the second part I was describing earlier as coming from the banks, trust companies, and security companies. The reason it’s gone this way—from P2P to fixed income, and now to equity-linked or standard product-linked offer—is that as a platform, if you’re too narrow, you’re only meeting part of the needs. If you’re positioned to be into wealth management for individuals, or providing a platform for wealth management for individuals, if you just do fixed income and the stock market booms like we saw in 2015, you’re only meeting part of the needs. You can say we can offer you a return of 6%–8% a year annualized. But people are thinking they can make 7% in a day on the stock exchange. So you’ve got to be broad enough.

And that’s what’s driven us, that whenever we move into an asset category, it’s often very not quickly enough to meet demands of the investor base. And so going broader while trying to have control has been key. Asset sourcing across P2P, nonstandard assets, fixed income, and then into the standard part has been absolutely the foundation.

ED: Did a lot of that also originate from your parent organisation on the product front?

GG: The second category, yes, probably over time. Originally, if you go back over time, it might have been half; now it’s probably about a third.

ED: So you’re not white label, in a sense. Are you a product-neutral sort platform?

GG: Yes, we are product-neutral in that sense. We now work with probably about 250 institutions in China. And because we have various securities licenses or trust licenses to create the products, processing is also an open platform. So we probably work with about 20 packagers in the process.

ED: Is the sales process completely online? Do you see the need for manual intervention to make the numbers work?

GG: Once we’ve gathered the product, there’s a middle part, risk control, which is saying who gets to go under what terms, and then we get to the sales process. The sales process for institutions is still an offline component because the purchasing amount they’re making could be quite large. It could be a couple hundred million; it could be a couple billion, in which case they still want to see a person before they make the transaction to make sure all the contact terms are right.

For individuals, sales is 100% online. The entire account opening is 100% online. But what we found about a year ago is the referral process has generated enormous volumes of customers who are older and also have a much higher income. So for example, when we source the customer online, the average or first investment they make with us is probably about $5,000 to $6,000. When it comes from an offline connection, or what we call the online-to-offline or O2O model, that number goes to probably about $30,000.

ED: And for that, you need an actual sales force going out?

GG: We don’t manage any offline sales forces. We basically focus on the product, on the platform. It would be too expensive to really have your own sales force. But if there is an installed sales force who’s already got a customer base, then you’re just adding a service to them and providing a small commission. For example, under the Ping An Group and Lufax, the majority shareholder is Ping An. If we work with the life insurance sales force and put a Lufax-sourced product onto their app, then the life insurance sales force says, by the way, as I'm selling insurance, have a look at our app and you can find Lufax products in there.

And from there everything else is closed on the platform. As a platform, we’re both in secondary and primary markets. The primary market is where we first put up the product. Everything that we put on the platform, we try and make it tradable. So that anyone who buys anything and thereafter wants to sell it can set his or her own price and sell it to someone else on the platform.

ED: Financial institutions have learned what sells and what doesn’t sell, online and offline, over time. The supermarket concept of selling mutual funds has never worked. So something needed to happen along the process to come to where you are. On the consumer side at least, what customer behaviour are you selling into, or what makes it possible? Is it price? Is it convenience? Is it choice?

GG: When we started, we asked, What are we trying to do? What are we trying to give the customer? And we realized that there’s just a shortage of investment options domestically in China. And there were really no private fixed-income markets that had liquidity. So what we’re offering to the customer is a return that is better than their bank deposit, which is as transparent as can be in terms of what is the underlying asset and risk in there. And it’s something that they can do online with a relatively high degree of convenience.

ED: How would you distinguish yourself from Yu’e Bao, for example, which is a mutual fund but scales so quickly?

GG: Basically, if you look at the Ollie Group, Ant Financial Group, or Tencent Group, what they’re doing is they have huge online customer flows already. And what they’re doing is taking a standard product like a money market fund, and saying if you’ve got extra cash sitting in the account before you buy your next item on Tao Bao, why not earn some interest on that?

ED: You’re not in that game.

GG: We don’t have the luxury of having 300 million, 500 million online customers. So the way Lufax works is really using our ability to select products and targeting the middle and middle-upper-income customer. What we have found is that people under 30 in China really don’t have much money to invest. But at about the age of 30 up to about 50, this is the sweet spot. And people who can invest about $10,000 up to about $1 million are not necessarily getting great service offline today, but they’re probably not going to go through a payment account to do their investing. And so we’re providing internet-enabled financial DNA, which says we’ve got probably the best product selection you’ve got in the market at a decent price at a high convenience. That’s basically it.

ED: That’s on the retail side. At which point does the institutional investor come into the equation? I take it that most of that is O2O.

GG: Yes. For the institutional side up to now, it is really O2O. I would imagine over the next 2–4 years as we build up more flow, as we’re able to move more of it online, as we take nonstandard stuff and standardise it more, we can move it more online. Since in China everything is separately regulated, there are very few nationally created platforms that cross those lines. Generally the participants are asked to stay within those lines.

If you look at the United States, just to give an example, for the OTC markets, the biggest provider of assets would still be the banks, investment banks. And the biggest buyers of assets are going to be insurance companies, mutual funds, and retirement funds. In China, that line crossing hasn’t really happened. So we’re providing a framework. I wouldn’t say it’s highly technology-enabled yet but we’re providing a framework for that crossing to start now.

ED: On the institutional side, what are you learning in terms of what institutional investors look for? Is this a sellers’ market at the moment?

GG: Right now it is a sellers’ market because particularly in the last 6–12 months, so much cash has been released into the economy that all of these institutions are standing around, looking for yield. The asset supply in the market, particularly over the past six months, has started to contract with the macroeconomy. So you have bonds from real estate companies go out at 3%; and securitised mortgages with 5-, 10-year tails on them go out at less than 3%. So there’s a huge amount of money chasing yield right now.

ED: At the same time there are about 2,000 P2P platforms like this. So why have you been successful and been able to create scale and the others haven’t?

GG: Obviously, we have a big advantage in brand from where we started with our shareholders, having the Ping An Group to back us. Second, we actually started very early. Five years ago, we didn’t know it was called P2P. We didn’t know that Lending Club existed. These are things that just by forces of regulation and market demand we entered into. If you look at our team of over 1,000 people, our core team, it’s probably about 40% people with financial background and 60% people with internet background.

We’re probably unique in being able to attract the financial talent to the platform. And then when we designed the products, risk has always been at the core in terms of building rating tools and building the transparency around it. When we create the products, we consider the first loss component, the securitised structure—how do you do that and get that balance right for the institutional buyer who’s buying the subordinated class, the portfolio part that may go to retail? All of that is very financial DNA.

And because we started early and we’ve been able to build up data, that’s been critical for informing our credit models on the consumer finance side. It’s been critical for us to get the retail investors classified, as well. Because internet finance for us is matching the right risk to the right investor.

ED: What about your retail customer base, in terms of the DNA, as you call it, would you say that your risk modeling mirrors what the banks already do? Or are you able to do it a little differently?

GG: The interesting thing is going to the borrowers’ side of the P2P business. Most of these customers who come to us may never get a loan from a bank because they’ve never borrowed before.

ED: So you don’t have an equivalent of a credit score here? Do you create your own?

GG: We don’t have a FICO score provided by a nationwide credit bureau. We basically use application data. There is a central credit bureau that can give positive and negative data on about 300–400 million of the population, but it doesn’t give you a score. And then you’ve got third party data, negative list data—all kinds of other information that you can triangulate against. About half of customers who come to banks could get a loan with the bank but it would take them 2 weeks to a month, whereas we get that done in two days. We’re able to use third-party data to stay at a low-cost point. If we don’t have a lot of data on a customer, we give them a very small loan and let them build up their own credit history with us.

Generally, for banks in China, if the loan is under $30,000 and unsecured, it’s just too expensive. Our average loan size is $10,000–$15,000. So we’re basically operating in the space that the banks are not that keen on. And we’re doing it in ways that are lower-cost.

ED: Though they are unsecured, are they also classified as business loans, working capital loans?

GG: It’s very interesting. If you look at what’s making up P2P, at least in our case, in China, P2P has been used to describe everything you could possibly do. But when we talk about P2P, it really is a way of saying consumer finance. And for unsecured loans, which is $10,000–$15,000, 70%–80% of the users of that loan are operating some form of small business. Borrowing costs are still not very low so you have to have a cash flow behind that loan. In contrast, in the US, 80%–90% is refinancing your credit card debt.

ED: Looking at Lending Club and so on, at the end of the day they are like one step behind what the banks do. They lend to mortgagers; they still ask for 30% equity; and they look at very specific types of borrowers. You could even say that P2P in the US and the United Kingdom is still on a learning curve. Are you also still on a learning curve, super conservative? Or are you creating your own risk model?

GG: I'll give you a number to give you a sense. For our unsecured credit business, the annualized charge-off is about 7%. It’s a near-prime market. It’s dealing with a lot of first-time borrowers, small business owners using their personal name to borrow. The approval rate from end to end is probably 30%–35% from the time the guy collects all of his information, makes it through the screening process, makes it to the credit model, and gets funded.

If you compare this to Lending Club in the US where they may be taking a FICO score of 700 and above, Lufax’s approval rate is 10%. The game in the US is much more refinancing of existing debt. With us, about half of people don’t have an established credit record, and the other half who do have some credit record want access to the funds more quickly.

ED: How many such retail customers do you have, and are they localized in specific parts of the country?

GG: On the borrowers’ side, over the last five years, we've probably gone to about 3 million borrowers now. We cover 250 cities. What’s interesting is the investors across the products, including P2P, of which a relatively small portion (probably 20%–30%) come from the tier one cities. Around 70%–80% of the borrowers are coming from tier two, three, and four. And that to a certain extent, reflects people in tier one cities who have probably built up more of a credit record, who have been in white collar jobs longer. Banks are more willing to interact with them, even at a younger age. Whereas you go to tier two, tier three, tier four cities, banks really are not looking at the consumers.

ED: What about migrant workers?

GG: Our platform is still probably not touching them. We’re still looking for people who have been settled for two or three years. If they don’t have a very clear income record, at least they should have a very clear job record, and a very clear location.

ED: Do you take a view on asset-based lending?

GG: On the P2P space, in addition to unsecured, we also do home-backed second mortgages, where the mortgage has been repaid and there’s free equity there. We will go up to about 50%–60% loan-to-value ratio in certain cities.

ED: But there’s documentation involved?

GG: It’s documentation and it’s full collateralization of the underlying asset. So if the person does not repay, that house is going to the investors. That model has been operating for about 2 years; it’s a very interesting space. It’s big in terms of asset size and number of customers because a lot of the small businesses in China for the last 10–20 years have actually stored value in second, third homes.

ED: Unlocking the assets.

GG: So it’s unlocking that under. That’s a bigger ticket. That’s more like a $50,000 to $100,000 loan ticket, fully collateralized.

ED: If I look at the entire supply chain of your business, where is the biggest cost base? Is it in the origination process?

GG: Basically it looks very much like a classic e-commerce platform. The bits that matter a lot in terms of risk management and product design actually make up a relatively small part of your cost. What costs the most is the acquisition on both ends. In the case of the P2P business, acquisition cost is high—where you’re dealing with someone who is a small borrower, getting that flow, and getting that name in people’s heads so when they say I need a loan, they think of going to a Lufax-enabled platform.

And then the other big part of the cost to the business is acquiring investors. Even though you do it all online, it still costs you $10, $15, $20 per active investor once you’ve actually put the ads out and you go through all the conversion rates of him becoming active on your platform. We’re now at about a 40%–50% activation rate for anyone who comes in and registers on the platform. When you’re growing a business at a rate of 5–7 times a year, that’s where your acquisition costs come.

Where the model eventually plays out is once you’ve got that installed base of, let’s say, 10–20 million borrowers and investors, then you’re at a scale where your growth probably drops down to maybe 50%–70% a year. And then it’s very easy to generate the profits because you have a low cost base.

ED: Right now, your growth at least on your origination side is amazing. You grew RMB3.3 billion in 2013 and today, you’re lending RMB14 billion?

GG: If you just take the consumer side of the business, which is really the core for us, yes, if you go back to 2013, the total placement of the platform was probably less than RMB5 billion or something like that. In 2014, it was about RMB80 billion. Last year, it was about RMB600 billion.

ED: That’s amazing scaling.

GG: The truth is, if you take us and you take these 2,000 platforms in China, and you take Alibaba, ANT Financial, TenCent, WeBank, all of these best available technology-enabled platforms and add them all together, we probably, in terms of flow or outstanding on these platforms, represent something like maybe 1%–1.5% of the total financial system in China. And that’s been over five years. In the next forward-looking 3–5 years, does that become 3%, 4%, 5% of the system? We’ll see what happens. But it’s still a very small part.

ED: On the institutional side, how is it playing out? There’s this nervousness about bond markets in China right now, and anything to do with assets. Do you actually hold the assets on your own books?

GG: We never hold anything. Basically our operating model is we identify an asset, we have it rated, we have it disclosed, and we have it placed. Now, obviously as that process is getting ready on the asset side, we’re always on the institutional side tapping into the potential buyers saying this is coming down the pipe, are you interested? And that helps us set pricing for things to clear. But it has to clear. If it doesn’t clear, it fails.

ED: But does the model involve you holding it at some stage in the transaction? So you take it on and then try and sell it?

GG: No, never hold. At the scale and volume that we operate at, if we had to hold anything, you’re talking about $200–$300 billion a month.

ED: When and where does the risk come to roost? What makes you afraid? Because it’s like a musical chairs kind of thing. At which point does the working model break down?

GG: If you look at what is driving failure—and obviously we’re seeing big failures in the platform—I think you can break it down into three things. The most pervasive driver of failure right now in the market is the business model with a construct problem. Which is, these platforms are taking in assets that are not fragmented. When we talk about the P2P business, we are talking about millions of borrowers and each one is $10,000. If your assets are very, very lumpy, like financing for that 20-story building and that 20 story building doesn’t work out okay, it’s not a fragmented risk; it’s a zero one outcome. That’s the first problem with the business model; it’s not actually looking at fragmented risk, which is the law of large numbers.

Second, any borrowing needed is usually going to be 1–3 years for any business. But these platforms in China have been consistently selling that asset in intervals of 1, 2, 3, 6 months; or even gotten down to being T+1 liquidity. So you’re running duration mismatch. And the only way you can run a duration mismatch on an asset that’s 1 year and you’re selling a product that’s 1 month is a new investor has to come in to take out the old investor at day 30.

If that new money doesn’t come in, that old investor doesn’t get his money back. Liquidity breaks down.

ED: How do you offer this liquidity to both sides?

GG: For the market as a whole, for the investor side, 85% of bank wealth management products in China are sold at a duration of 6 months or less. And obviously the assets that are held by the banking system are going to be 1–3 years. So the banks run a huge duration of mismatch in China. But they’ve got the central bank behind them; they’ve got 17% reserves with the capital. The internet platform by definition should be capital-light. If it’s not capital-light, it’s not creating new efficiencies in the system.

And because it is capital light, it can create secondary markets to create liquidity for products on the platform. However, it cannot be running duration mismatch. That’s the biggest problem right now in China. Of these 1,000-plus platforms that have failed, problem number 1 is they’re running duration mismatch. So if there’s any loss in confidence or there’s any change in money supply or competitive dynamics that doesn’t allow new money to come into that platform, it just unravels.

ED: Some of the platforms that have failed, they’ve got this Ponzi element to it.

GG: Fraud is the stuff that makes the headlines. For Ezubao, apparently 90% of it was just a money game and there were no real assets. Those grab the headlines. There are 1,500 platforms that have disappeared over the last 4 years. That’s maybe 5%–10% of the problem. 80% is people who thought here’s an interesting space, and they started to set it up and they didn’t get the DNA right.

ED: How is it that an industry mushrooms overnight like this and the learning is all over the shop? It hasn’t been formalized in any way. It’s like there’s no common language by which some of these started.

GG: If you go to the deeper, underlying root cause as to why this has panned out this way it’s that there’s just enormous demand for investment needs and for financing. There’s enormous need to make use of existing assets on balance sheets and free them up. And there’s enormous need for investors to have better yield and liquidity. So these are just fundamental things. And I think the regulators at the highest levels, over the last 5–10 years have been saying to the banking system: we have to serve the real economy; we have to serve these needs. Like many parts of the world, that just doesn’t happen that easily. So I think the attitude was someone’s got to do it; why don't we kind of let it run a bit?

ED: What are the profiles of these institutional investors on the buyer side? Are they state-owned enterprises? As you say, they’re looking for yield but who are they?

GG: The ones that we deal with and I would say that we’re a little bit unique because most P2P platforms don’t have any real interface with institutional customers. Because institutional customers are not going to go and deal with someone who is 1/1000th or 1/10,000th of their size; it’s just not worth their trouble. We’re able to do it because of our backing and our scale and our first mover advantage. And who are they? They’re going to be, generally speaking, mid-sized banks. They’re not going to be the top 10 or 20 banks, not going the top five insurance companies. They’re going to be mid-sized insurance companies. So if you take insurance, banking, trust, securities, these are guys that have assets or have cash who are looking to optimise or diversify geographically or across lines. And they don’t necessarily have a trusted agent to show them this is what it is, this is the documentation, we’re a safe pair of hands that will help you process the transaction. That’s really what we’re offering.

ED: Is that temptation to create more debt into that business, derivatives?

GG: There’s always temptation but up to now, none have been followed because it’s a very long path. And one of the things where we stop is, are we building something here that actually, over time, is going to take more offline resource to do well? If the answer to that question is yes, then we stop. Because there are a lot of banks in town that will do that. We have to design our product set if we’re being online in orientation. And now 80% is on the retail side, and all through mobile. So to your point earlier, no one who has done the supermarket has ever worked; I totally agree. Because the more you put on a small screen, the less people are going to know what to buy.

So it has to be very much more a tailored model of the matching. And that has to be kept to a certain level of simplicity. If you start to get into very, very complex products, for example, if our RBI customers go over $100,000 with us, we give them the option to say yes, I would like to have contact with an online representative. So if I have questions on the products or questions on the services, there’s someone I can call. We do that for the top 3%–44% of the customer base who do make up half of what goes on in the platform in terms of volumes. But if you’re getting to a situation where you have to be sitting down with somebody to explain to them what it is, then we stop.

On the asset side, we’ll go deep because you have to get the quality right on the asset side. And that’s leveragable because with 300 to 400 people, we can cover the country on the institution side to source assets. For the P2P business, we do have 40,000 people across those 250 cities, 750 locations doing the validation of IDs and screening of stuff. You’ve got to do that. But on the investor side where we think the platform really comes into its own and is scalable, it’s online. It has to be.

ED: So all things working well, at which point does profitability kick in given that you’re making a fee on every transaction, on every front?

GG: Profitability basically kicks in when your growth rate drops below about 150% a year. We’re growing at 500%–700% a year. So you’ve got all of the acquisition costs that are coming in. Up front you’ve got to build all your systems. You’re hiring all your people and all your developers. You’re building brand. The model for us is basically this: it comes down to across all the product flows that we serve. When you add in the primary market, the secondary market, and everything, it’s about 100–120 basis points in revenue. Your steady state service model of the existing assets and the existing investors probably comes down to about 40 basis points; maybe a little bit less if you get more scale but consider it a 40–60 basis points range. So you have 50 points of long-term margin that you can realize if you’re growing the business with a growth rate of 150% or less a year.

ED: This is what you think.

GG: So far, it’s what the last 2–3 years of data have shown us. This year for example, if we say we’re not going to grow three or four times, we’re just going to grow 50%, you become immediately profitable.

ED: And you can’t regulate that?

GG: You certainly could. The interesting dynamics of e-commerce around the world is the winners in e-commerce have ended up representing 10%–50% of their vertical. And so if you look at an Amazon or Talbow, the investment to scale up quickly because of the long-term advantage it gives you is a very strong incentive.

Now, if you want to be cash flow positive in year five; I don’t want to take 5–7 years to get to profitability, you can do so much earlier but you do so at a much smaller scale. And the risk of that, particularly in China, is that someone then just blows past you. And you say I was making money earlier; he’s still burning cash. But he’s now three or five times as big as you and when it really push comes to shove, distribution wins the game at the end of the day and then you’re a far second.

It’s a very tricky dynamic and very hard for traditional financial institutions to respond to because they have to make two big tradeoffs. Firstly, if they want to replicate this sort of thing, they have to give more to depositors in terms of return. So they give up on their own margin. And secondly, if they boost this thing up, it is going to affect their return on equity. Would their shareholders support them when they’re going heavily online, when their profits are going to report a downtick of 20% for the next years?

That is one of the elements of capital markets: people say you’re building a new model that doesn’t have the burden of the old, and they’ll support you as long as it plays through.

ED: What’s your funding base like right now? Ping An is your main investor. How much money do you have to burn?

GG: Ping An represents about 47%. Last year for the platform as a whole, we burned probably about $150 million.

ED: That’s fair, right? It’s not as expensive as it would be in another part of the world.

GG: Yes, it’s given us an active investor base of, to date, about 5 million investors and $50 billion in money that’s invested across the products. The question we always ask ourselves is if we had to spend $150 million to build a small- to medium-sized bank in China, and we could do it within three years, would we do it? The answer is yes, we would do it.

ED: Half the capital is stuck in capital costs as a bank.

GG: Exactly. Here, most of it is going across your customer base.

ED: On the markets front as a markets player, are you an OTC?

GG: Yes, in effect we are an OTC.

ED: Isn’t there a little temptation there, as well, that you would want to build that side of the business and be more of a capital market player, eventually?

GG: The thing is that you have to keep going back to where the value is in the business.

ED: So you stick to your origination process.

GG: There are three, if you will, constituents in this game: the asset providers or the product providers, the investors, and the liquidity and trading that happens in between. And you kind of have to make a choice, that when push comes to shove, which one are you going to stand for? For us, we’ve made the choice that we're going to stand for the consumer in making sure we’re giving them the best choices in the best efficiencies.

ED: I have a note that says Lufax inclusive financial department. Do you want to go into the index business?

GG: The inclusive finance is basically our sourcing division for the consumer finance loans. The 250 cities’ coverage that’s sourcing those borrowers of $10,000 to $15,000—that’s the inclusive part of it. It’s basically providing funding for people who don’t get funded by the bank system.

ED: The thing that keeps you honest is bank interest rates. If it went up any higher, your business would look very different. If it went down any lower, your business would look very different.

GG: Yes. It’s very interesting. What we have done, in effect, is we have started to create the secondary markets in interest rates. Because all of the product that’s gone on the platform over the last couple of years has gone out the door at a fixed interest rate. And as interest rates have come down, it provided incentive for existing product holders to actually sell to the next investor. If you look at the volume on the platform each month, about 70%–80% of it comes from primary placement of new products. But about 20%–30% of the volume is coming from secondary markets.

So people who bought a product of 12-month duration that was yielding at that time 7%, may now be a new product going out the door maybe only yielding 6% or 5.5%. So people are able to trade and make capital gains on the stuff that was sold a year ago. Now, we’re in a declining interest rate environment so that is actually a very good impetus for secondary market development. If it was a rising interest rate environment, you’d probably see less activity in the secondary market. But it does create a pricing effect.

One of the interesting problems in China is you always hear the senior leadership in China complaining about the cost of funding for small and medium enteprises. And the reason for that is because there is no liquidity in the system. In the US, if you buy a 7-year bond or a 30-year bond, the yield doesn’t change that much because there’s liquidity. So the more that you can create liquidity around different durations, eventually what you’re solving for is a funding price efficiency issue. And for us, we price our products or we work with the providers to price the products, so that they’re yielding slightly better than a fixed-term interest rate or as a bank structure product.

That’s the motivation for someone to change their behaviour. And we track that. So as the banks go down, we go down. If the banks start to move up, then we would move up. There are boundaries in each track.

ED: When you say you move up, you provide transparency.

GG: We provide indications to the providers to say if you price here, you have a 95% chance of clearing, whether it’s a primary or secondary market. If you price here, best of luck.

ED: And so far it works, right?

GG: So far it works. We can see it. Where it gets tougher is when you get spikes in the system. It happened about a year and a half ago; all of a sudden the Copenhagen Inter-bank Offered Rate shoots up to 18% overnight. At that rate on that day, there were no new assets coming in. It doesn’t work with extremes. If interest rates get low to a certain point, then we can see certain parts of our customer base start to drop off. The higher end of the customer base will say at these returns, I'll buy a house and bet on that. If rates go to a certain level, then you start to attract different segments in.

ED: When you reach a certain scale in liquidity, what’s tempting for you? What are things that are left to be done that can be done? Investment banking?

GG: I think the temptation is dynamic pricing. One of the things we debate on, and we haven’t figured this out yet, is the people who do pricing in the most sophisticated way are not financial institutions. You could argue they’re airlines. Right?

ED: Supply and demand.

GG: Supply and demand. So technically as a platform, you could reset pricing by the minute. And that price could be built around supply and demand. It can be built around competitive situations. It can be built around reverse auctions. As you get to a certain amount of participation and data and behaviour, on the one hand you’re going to increase the efficiency of the matching; hence, risk-weighted assets, customer risk-weighted, as well as in those matching engines. And then how do you price that supply and demand across those different flows? You could be much, much more dynamic online.

I think the story of internet finance more broadly is today, the game is mostly about taking advantage of channels—the channel advantages of internet finance. Lower cost, higher efficiency, higher customer convenience. I think eventually, 3 years from now we probably won’t be spending much time talking about the channel advantage.

ED: I agree.

GG: It’s going to be—and I don’t want to say big data because it’s so overused and abused—but I think it is how you take advantage of scale.

ED: Big data is only the starting point because we don’t know whether the market is going to go fragmented or consolidated. So if you take the airline analogy, there are a few dominant places that the dynamic pricing will work differently from a highly fragmented market.

GG: Right. The future of internet finance in terms of why does internet finance have competitive advantage or differentiation over traditional finance, comes down to two things. One is that you’re able to use the data well on the asset and the investor side to do the matching better. And a platform is better than the average teller or a relationship manager in driving advice. A platform doesn’t say you’re perfect, but you’re better. And that’s because of data and dynamic pricing—a better understanding of the investor because of their behavioural information—so that you have the ability to drive the right product to the right person at the right time. That I think is the Holy Grail. I think that’s where it ends up. All that debate about customer convenience and pretty mobile stuff I think will not last.

ED: So if that’s the Holy Grail, giving customers what they want, where they want it, and the liquidity, where do you think we are headed to? Bank interest rates today can be topsy-turvy. And it’s not just data. When you think about things like robo advisors, for example, does that plug the advice component better?

GG: That’s a great question. I think the answer is it seems to be working in the US. But when I've asked people in the US why has robo investing become popular in the last 2–3 years, the story they tell me is first of all, you’ve had the development of the exchange traded fund (ETF) industry. And the ETF industry has allowed you to create subtracking of any part of the index you want with reasonable efficiency. And it’s been able to do so at a cost point that is much lower than a managed fund. And then you match that with American investors, who for better or for worse, believe that long-term investing is a good idea. And that if you do a portfolio base, you’re better off than stock picking.

So if you pick a series of ETFs through your robo investing, which is setting up a portfolio for the next 5–10 years and you go with this advice, you’re better off than punting. If you come back to Asia, to China specifically, I'm not convinced that we’re there. We don’t have developed ETF markets. There isn’t that cost efficiency. There isn’t that transparency. There may not be that rationality yet in the markets.

ED: The customer doesn’t even hold those assets yet.

GG: The breadth in standardisation that has to go into place in a capital market environment for robo investing to work is a question mark for developing markets. But for our equation, we’re saying if you have someone who comes in, I know from which part of the country they’re from. I know what their risk appetite is based on the device they’re using, based on the email address they’re using, based on how much they invest with me, from the first to first 12 months, as I expose them to different products. As I look at third party databases I also know more about them.

So with ability to use that data to say you bought one or two products with me, if I can do a better job than the teller at the bank counter or securities counter of calculating what should be the third product, then I should be able to do that at a more efficient cost, and do a better job for the customer.

ED: That’s on the investor side. What about the institutional side, the sourcing of institutional assets?

GG: I think that as to the serving of institutional investors on that level of sophistication, we cannot beat the institutional customers themselves.

ED: And you’re not looking at ETF.

GG: That sort of thing, where it becomes deep, deep, single asset structure, de-specialisation on a big deal, that’s going to be a bilateral market. It doesn’t really lend itself to a platform.

ED: And you need someone to manually package it and pass it on.

GG: Yes, exactly.

ED: What are you learning in terms of the intermediation business now? China is unique, unlike the US because banks are almost stuck; their hands are tied with directed lending. I get the sense that in the US and Europe in general, the P2Ps have to sort of attract the banks. Whereas here, you seem to be running independent of it. You are actually catering to a market that is waiting to be served.

GG: Certainly I think, if you look at China, we have much bigger opportunities here than in the US or Europe because there’s a lot more unmet need. But we’re also realizing that to be scaled, you’ve got to work with a lot of institutions in different areas for custodial services, for payment services, and you have to get that right, too. But the thing that I always ask myself is why is it growing faster in China, and how big can it be in China? And then if I dial back and look at the US, what is it that doesn’t make it possible in the US, or how has the US dealt with this?

My feeling so far is if you look at the US, the truth is traditional financial players have been very active in retail for 40 or 50 years. And your average 40-year-old in the US has a Citibank account or he’s been dealing with Charles Schwab. This is already set. He’s not going to go change his behaviour. There’s a lot more flexibility in China than there is in the US. Also, if you look at the US, 30, 40 years ago the markets at that point were commercial bank-dominated. And then over the last 30, 40 years we’ve moved to heavily securitized markets and OTC markets, and investment banks have played a very heavy disintermediation role. You had all of these sub-OTC markets that created clearing mechanisms, which eventually then flows into the fund market, which then flows to retail.

So I think what’s happening in China is that you are having this disintermediation occur and the commercial banks are, if you will, reduced in their role in the overall financial market. But in a world where technology is very different.

If you went back to the US 30, 40 years ago and said if you can have all these tools, how would your market develop? It may look more like what’s happening in China today. So I think China is going through the same process of disintermediation, of creating more direct flows, of creating more capital market-enabled systems to finance, but it’s going through technology platforms that are different.