A panel of financial services investors and executives looks at the risks faced by the new generation of fin tech companies, including lending startups, robo-advisors, and more.

Investors in financial technology are identifying some problems with robo-advisors’ business models.

In fact, Alan Gallo of American Express painted the business of robo-advising as even more fundamentally challenging than that of online lending marketplaces, which have been under fire lately for mixed results and governance issues.

Robo-advising is tough for one simple reason.

“You can always find people to borrow money,” said  Gallo, speaking at CB Insights’ Future of Fintech Conference earlier this month. “Getting people to invest tends to be a more serious question.”

Gallo took a relatively bearish view on this category, arguing that for young people robo-advisors may add very little relative value versus traditional savings or investment products, while wealth management at the high end of the market is so complicated that humans are still necessary.

“Particularly for young people, it’s much more important that you invest, rather than exactly what you do,” said Gallo, who is executive vice president of strategic initiatives. “The human intervention is to convince someone to do it, it’s not really [about choosing] product A or product B.”

Flash Webinar: Alternative Data
CB Insights senior analyst Matthew Wong hosts a flash webinar (only 10 minutes) on how startups are enabling the rise of alternative data in institutional investing and more.

Other panelists included Deven Sharma, formerly of Standard and Poor’s and a fellow at the MIT Media Lab, and Liron Gitig of growth equity firm FTV Capital. The moderator was Ainsley O’Connell of Fast Company.

Robo-advisors might be useful for long-term financial planning, said Gitig of FTV Capital. As for robo-advisors that help in stock picking, Gitig rejected in strong terms. “Alpha is dead,” he said.

In fact, he argued that one of the main reasons human advisors are important is that they help  individual investors make emotional decisions, which is something a robot cannot do.

“I think the most helpful role for human advisors is that they help you walk away from the cliff, when you’re about to buy at the top or sell at the bottom.”

Online lending, which was covered by another panel, also came up. The panelists pointed out that the past few years have been favorable for online lending startups — with low interest rates, an expanding economy, and big banks distracted by regulatory challenges.

Lenders’ new underwriting models and uses of alternative data won’t be truly tested until we see a downturn, and that’s on the horizon, said

“For all you ‘Game of Thrones’ fans, winter is coming,” said Gitig of FTV Capital. He added that cyclical downturns should be expected and he is still “very optimistic long-term” on lending startups.

The panelists described larger opportunities in currency exchange and payments, especially business-to-business payments and healthcare payments.

“Businesses that should know better are still cutting checks in the back room, sending each other checks,” said Gallo. “It’s a totally ridiculous way to do business in 2016.”


Ainsley O’Connell, Business & Tech Reporter, Fast Company: Hi. My name is Ainsley O’Connell. I’m excited to be here today on stage with our three distinguished panelists. I’m gonna let each of them introduce themselves and tell you a little bit about what they do and also how that informs their point of view on what’s happening in the FinTech space at the moment. Do you want to start, Deven?

Deven Sharma, Former President, Standard & Poor’s: Yeah. So I look at things from three lenses, one as a private investor, two, advising private equity and corporates, around broadly information and analytical tech space, and third, as it says, I just get to hang out at Media Lab for a bit. They think they’re getting advice from me, but I think I’m learning from them.

I sort of really look at these opportunities from really four fundamental lenses. One is I focus on what role anybody’s trying to cover in the finance function, because the roles are intact. They’re not gonna change, and they’ve been intact for centuries, and they’ll stay intact. It’s connecting the borrower and the savers of money, intermediating between short and long-term duration of timing, and then allocation of capital and risk and return. That remains intact. The question is, how is one addressing this? How deeply they’re addressing, is what I look at.

The second is the fact is risk infrastructure makes up a substantial cost of the banking system, and somewhat of the asset managers and then some of in the insurance sector. That will never change. It will in fact only go up, because the risks are going up. So the question I look at there is, how is somebody addressing that risk infrastructure? How, either as an opportunity or as a way to manage their own marketplace business?

The third is the customer expectations are intact also, but how you deliver that is changing. You and I want safety of our money. You and I want predictability of when I want the money, I want the money back, in the financial structure. Thirdly is I want to get… I don’t want to pay arm and leg to get any services. I want to do it as cheaply as possible.

Now the convenience, where you do it, how you do it, is absolutely under complete disruption. The last thing I look at is scalability. Unless I see some opportunity of path of scalability, I stay away. The scalability is in the origination. It’s in the funding. It’s in the processing. So that’s sort of the way I look at three or four things.

AO, Fast Company: And you, Alan?

Alan Gallo, Executive Vice President, Enterprise Strategic Initiatives, American Express: Hi, everybody. I’m a finance person by trade. I’ve really spent my entire career with American Express, and recently I was appointed to lead a group that’s really focused on strategy and reengineering. So I come at this question from the standpoint of a big company, and from the standpoint of how changes in the industry will affect our strategy, and how it affects our approach to risk. I’m not directly involved in the company’s fund where we make investments in FinTech. You may have heard from my colleague, Rohit, earlier today, who is directly involved in that. I’m certainly indirectly involved, and again, look at it from the standpoint of its impact on our strategy, and what we can learn from companies in this space, and if there are commercial agreements that can come out of that, that could help our company, help our customers.

We’re clearly strategic investors, not in it from a financial standpoint. So hopefully I’ll be able to enlighten folks on a big company’s approach to that.

Brad Bernstein, Managing Partner, FTV Capital: Sure. Hi. I’m Brad Bernstein, and our firm, FTV Capital, has $2 billion under management. We specialize in growth equity investing in financial services, enterprise technology related to financial services, and transaction processing and payments. Our firm was founded on the idea of a strategic network that informs the way we come up with themes, the way we diligence those companies, and then the way we provide business development support to those companies, so that they can get to the right people in these strategic partners, partner with them for channels, as customers, variety of different ways, as lenders, and we’ve been doing this since 1999, made over 90 investments, are typically looking for companies that have achieved $20 million in revenue plus, and are growing at least 20% a year.

Our network looks like 350 executives that are a lot like Alan in a variety of different enterprises that are dealing with these challenges every day, where we can access their insight and gain perspective on what’s happening in the financial services landscape.

AO, Fast Company: Wonderful. Well, I feel like I have to start with Lending Club. Hard to have a conversation about FinTech today without Lending Club coming up in some way. Has the turmoil there changed your perspective at all on both lending specifically, perhaps, but also the FinTech space overall?

AG, American Express: You can start. That’s always the question. There should be a pecking order. I’ll start.

AO, Fast Company: Yeah. Yeah, feel free to dive in. Yeah.

AG, American Express: So it hasn’t really changed my perspective. So as sort of an incumbent in the lending industry, we always look at these things as they come up, and there’s always this tendency to say, “Oh, they’ll get their comeuppance,” whether it’s from regulatory scrutiny, or some sort of control or compliance failure, which you sort of see in this case, or from the next recession. Right? Which is gonna affect everybody in this space, including people who are new to this.

But I don’t think it fundamentally changes the landscape, from the standpoint of lending has a lot to do with underwriting. Underwriting has a lot to do with data. Customer experience is a matter of how intuitive the customer interface is. Matching borrowers and lenders is about matching. Those are all things that the digital world does really well. So I think there’ll be ups and downs in the next recession, which is when, not if. Right? There’s been a recession every 10 years in the U.S. for the last 200 years. And yet, we’re always surprised every time one comes up, and they always take sort of different shapes. So there’s gonna be ups and downs. I think it’s here to stay.

Question really is when you think about how big it could be, and you think about the sort of pie of lending by consumers in the U.S. and most other parts of the world, you have secured lending for mortgages and autos. You have lending for education. You have student loans, and then you have people who borrow to buy stuff. Most people do that on their credit cards, and most people have access to credit cards. So when you think about how big that part of the industry could be, you have to answer the question, “What are people borrowing for?” if it’s a consumer need.

In the small business space, I think there’s a lot more kind of niche opportunities, because small businesses have more tailored needs, and many more specific use cases, like the nature of supply chain financing in a particular industry, for example. So I think it’s an exciting development in the industry. I think it’s here to stay. It’s gonna get a lot of bad press, and there’s gonna be a lot of pain, particularly when the next recession comes.

BB, FTV Capital: Yeah, I think, from my perspective, when we… the biggest danger of being a growth equity investor is the false positive. No business sort of has more potential for the false positive than lending, because it’s really easy to grow when you’re giving money away. So when we look at a business that… a new emerging lending model… we do believe the technology can empower underwriting and cost of acquisition and make the experience better. But until there’s been a downturn, nobody really knows whether all the analytics and all the underwriting and all the pricing is really working the right way.

So I think, very much to that point, we’re eight years into an economic recovery. It’s been, in many ways, the perfect environment to build a next-generation lending business. Banks have been incredibly distracted by regulators and trying to heal themselves from the financial crisis. Cost of capital is at zero. And yet, you can still charge the maximum legal rate. So spreads are kind of as wide as they could possibly be.

So when you think about what’s gonna evolve over the near term, as banks get healthier, as the regulatory environment changes, as interest rates go up, or as the economy goes into a downturn, all these things are gonna create some very interesting challenges that I think, I think that while I’m still optimistic long-term, that these types of online lenders have long-term potential, for you “Game of Thrones” fans, winter is coming, and it’s gonna be interesting to see how that plays out.

DS, Standard & Poor’s: So I think separating out Lending Club and broader… I agree with most of the points that are made here. Financial sector is a very interesting space, and both of you know this, and the panelists here. It’s a game of trust and confidence. Once the trust and confidence goes, that’s where bank runs happen. Not because the bank cannot give the money away. People feel fearful that you may not get the money, and then you start withdrawing, and the crisis in most of the financial sector happens because of liquidity. Liquidity is the biggest risk financial sector faces.

In this situation, actually, Hans Morris sort of made the right statement. He’s the exec chair of Lending Club, that “we want to have full confidence and trust of all the investors, the borrowers, the shareholders, the regulators.” He understands that. Without that trust and confidence of people, it can’t survive.

So whether Lending Club survives or not, I don’t know. The challenge there is this is a business of volume. If you have volume, you’re getting a fee, you keep your costs low, and then you can originate more volume. But if your volume drops, your cost goes up. Then you have two choices. Either you charge more. In that case, the origination goes down, or you charge the same, and you don’t make money. In that case, investors start questioning whether I should put the money here or not. So it’s a bit of a cycle. How they come out of that cycle it’s to be seen. Have banks who have faced bank runs come out successfully? Absolutely. IndyMac is such a successful bank now.

So it’s a question of how they handle themselves, but the point I want to emphasize is everybody complains about regulation and compliance. Yes, the rules are onerous. But if you really step back and look at it, most of the regulations and the rules that are required, really the intent is exactly spot-on, that you and I, as a customer, would want that from a provider. Except when the rules get written, unfortunately, sometimes written by people who have no industry experience, it gets very onerous. If you don’t follow those things, because you think we don’t need to, ultimately, it catches up, because this is a game of trust and confidence.

In terms of broader sector, I think I agree with everything, but in the marketplace lending, the only thing I would sort of continue to sort of highlight is it’s a game, scalability is out of access to capital. This is the same situation that I observe that happened in the mortgage sector. The yields were phenomenal. Everybody chased the yields. Everybody pushed the capital. The moment the yields came down, people ran away. Liquidity wiped out, and that sort of stopped the musical chairs. I see a little bit of that happening now, because the investors are getting worried about the losses going up, as both of them said.

Credit is a cyclical business. Anybody who says it’s not, they’re dreaming about it. It’s a cyclical business. It’s gonna happen, and as I have poked around a number of these lending platforms, few of them, many of, they all manage their transactions very well with underwriting analytics, but I’m not sure they manage their portfolio well, and the fact is the portfolio losses will go up. As that goes up, the question is, will their access to capital go up?

AG, American Express: And when the mortgage industry sort of went down, mortgage brokers went with them too. So a lot of times, we talk to companies like this, and they say, “Well, we don’t take the balance sheet. We’re just the matchmaker, and we sell the balance sheet to stupid banks like you.” We don’t do this, but all the banks buy these loans. But when it goes down, flow dries up. So even if you’re just a platform, you’re not immune to downturns in the industry.

AO, Fast Company: One of the lending companies that seems to be managing its balance sheet at a pretty remarkable scale is SoFi, and they are now moving aggressively into areas like wealth management and sort of trying to think about ways they could cross-sell. I know Brad, for example, you’ve looked at the robo wealth management space. I’m curious where you see business models there taking off, and what’s sort of exciting to you there. Do you think companies can make that transition to doing more than just sort of one job with the customer in the startup side?

BB, FTV Capital: There are a couple different questions.

AO, Fast Company: Yeah. A lot baked into that.

BB, FTV Capital: Yeah. I think the whole “try to be all things to all people” is really, really hard. So that part, I’m pretty skeptical about. I do think there’s some interesting things that SoFi is doing, but I don’t know how that is gonna relate to robo advisers. I think, on the question of robo advisers, that is a model that obviously has been getting a lot of press, and there’s a lot happening in wealth management. I think that there’s clearly a huge role for technology and the digitization of the wealth management experience. The question is, how is that gonna play out?

To some extent, there’s going to be a commoditization of that digital experience. Every adviser is gonna have some digital robo solution available. The question is, where is that gonna fit in the landscape? How is that going to impact the core of what wealth advisers do, which is serving people with over $1 million of worth, versus serving younger folks who have $50,000 in a savings account, who, frankly, advisers don’t want to speak to now anyway. So you really have to segment the market, look at different market consumer needs, and where they are and what they’re gonna benefit from these different solutions.

AO, Fast Company: I’m curious. What’s the opinion of the panel? Are we gonna have wealth advisers in 10 to 15 years?

DS, Standard & Poor’s: You have to think about the whole wealth management in maybe three buckets. There’s a distribution bucket, and really advisers are asset gatherers. They are not advisers, per se. I have an adviser. It’s a fact. I have an adviser, but then they have product specialists. So I know what I use the adviser for. I use that adviser for access, and they’re asset gatherers. So will the robo adviser play a great function in being able to be more efficient gatherers of assets? Absolutely. Would this 75 basis points and 50 basis points… Will I give anybody that? Do I gather my assets? Then they’re gonna give me a product that’s based on somebody else’s product, and I have to pay expenses on that? It’s a different question.

So I think, to Brad’s point, on the front end, everybody is gonna adopt it, because it’s an efficient way to engage, to share information. The question is then how good an access they have to manufacturing the products of investments. So if I can go and get Vanguard the same product with no extra basis points, why would I pay that? So that’s where I think this model will evolve, because there’s no differentiation at this stage of the game, just because the front-end interface is nice. It’s a very interesting situation, but I don’t know how this will evolve.

I’ll tell you one quick story. Some of you may remember. I know you two will remember. In early 2000, Spitzer passed a legislation, basically a ruling, that said, “All banks have to do equity research and advising independent of investment banking.” I was at S&P leading it, and we benefited from it. So we sell it. It was required that all investors should be given free research. Out of 100 emails, how many people do you think opened 1% of emails that were open?

BB, FTV Capital: Not many.

DS, Standard & Poor’s: One percent. So it’s a… I don’t know how this will evolve.

AG, American Express: So Brad said it before. You could always find people to borrow money. Right? People like money. People want to buy stuff. They’re always gonna borrow money. To get people to invest tends to be more of a serious question. So I think this part of the industry will grow, but there’s sort of two things that, in my mind, require human intervention. So I think there will be wealth advisers forever, because at the high end, things get pretty sophisticated, and your choices get very complicated, and you want human advice from people who have experience and that sort of thing.

So I’m a financial professional, and sometimes I don’t even know what they’re talking about. It’s very complicated, in terms of estate planning and how the law works around this stuff. At the lower end, there’s sort of a different issue in my mind. Once somebody’s in the system at the lower end, I think the robo adviser thing works well. But for most people on the planet, under the age 30, it’s really much more about if they are investing than how they’re investing.

So when I talk to some of the young people in my life, they say, “Uncle Alan, what should I invest in?” I’m like, when I hear that, I give them a hug, and I said, “You’re 90% of the way there, because you’ve already decided to put some money aside and invest.” Particularly for young people, it’s much more important that you do it than exactly what you do.

Now, don’t take that to the extreme. Don’t give your money to cousin Vinnie who wants to start his yoga studio or anything like that. But basically, within certain parameters, it’s really sort of the binary, are you doing it or not doing it? Insurance is a similar space, where the human intervention, again, to convince somebody to do it… it’s not really, “Do I buy insurance product A or B?” It’s just convincing them to do it. So the huge thing that young people, young people who don’t have children, totally miss, is that the most important form of insurance for people like that is disability insurance.

So people, life insurance is the top of mind for people. When I talk to somebody who’s under 30 and doesn’t have children, I say, “Listen. No offense. But if you’re dead, you’re dead. You don’t have to worry about it. Right? The worst thing that can happen to you is that somehow you’re disabled, and you have to support yourself, and you’re then much higher medical bills for the rest of your life.” So when a young person comes to me and says, “What kind of disability insurance do I need,” I give them a hug and said, “You’re done. You already know you need to buy disability insurance. Exactly what kind of disability insurance you need? Far less important.” So human intervention at the lower end is really about the binary choice of are people in the system or not.

BB, FTV Capital: So if I could just chime in here.

AO, Fast Company: Yeah, of course.

BB, FTV Capital: I actually disagree that the wealth adviser is just an asset gatherer, but I do think that the role and the value-add has been evolving over time. I think that’s an interesting topic. So if you go 20 years ago, if you were a wealth adviser, you were basically a stock picker. You were telling people, “Hey. You should buy IBM.” Then 15 years ago, you were talking about mutual funds. Over the last 10 years, you’ve been talking about ETFs. More recently, you’re talking about asset allocation.

When you think about where you are today, we live in a world where, effectively, at least in the liquid markets, alpha is dead. There is virtually no alpha. So what’s happened is all people really care about is the cost of investing, and getting their allocation out. But I do think that, as you talk about the higher end, and I agree about the more wealth you have, the more complicated decisions, what you’re seeing is the role in where we believe we’re going is really around financial planning, really being that planner, that CFO, that longer-term thinker that helps people who aren’t that financially sophisticated think about, “How am I going to save for my retirement? How am I going to plan for my healthcare costs? Will I be able to afford my daughter’s wedding? My second home?” Whatever it might be.

So we see a lot of tools and technology that’s being applied to that part of the value proposition. Now that being said, I think one of the least appreciated aspects of what an adviser does is help walk you away from the cliff when you’re either about to buy at the top or sell at the bottom. If you look at the history of the markets over the last 20, 30 years, the market’s been up, about average, 6% a year. The average individual investor is up about 2%, and that is because they tend to buy at exactly the wrong time and sell at exactly the wrong time.

AG, American Express: Bad choices.

BB, FTV Capital: One of the interesting thing about robo advisers, back to our question about the market turning, is what’s gonna happen to all that money that came on over the last three or four years. When I say all that money, it’s actually not that much money, but all that money that did go on. We have enthusiastic millennials put some money into the markets, and suddenly their account’s down 20% or 30%, and they have nobody to call, no one to say to them, “Gee, you got to stay in through the cycle. Be patient.”

So it’s gonna be very interesting to see what happens to some of those firms. Some of those firms, they’ve gotten these massive valuations, which have gotten a lot of press, but most of them, in terms of assets under management, are what Bank of America would call a bad Merrill Lynch office. I mean it’s a… these are not very large amounts of assets. If you think about the economics of what they’re charging, at 30 basis points, the high end of their accounts are typically $100,000. That’s $300 a year. So you need a lot of those accounts to make the math work. I think that’s gonna be a really interesting challenge to see how that plays out.

AG, American Express: So if alpha is dead, should we tell Seeking Alpha to stop looking?

AO, Fast Company: Someone better give them the message, I think.

AG, American Express: Thank you for that softball.

DS, Standard & Poor’s: There are two things to keep an eye on in this space. This will evolve. Australia is a very interesting market, which went superannuation market many, many years ago, where basically everybody had to take ownership of their own retirement funds. The second is if you look at the… and most companies have moved to a 401(k) plan and direct contribution versus… and so that sort of trend is moving. We give you the money. You invest. You figure out what you want to do. It’s not well-managed, but it’s gonna evolve a lot faster than one thinks.

Third is I think the healthcare model is gonna influence this a lot more, at least the sense I have, which most people who now get healthcare insurance, even in companies, you have to pay your own self a little bit. When you start paying more money, and you have to manage your expenses, whether it’s health savings accounts, or this and that, then you start looking at the whole thing and say, “Okay. How do I manage all this?” So I think that thread might sort of influence this trend.

Having said that, Hearst Ventures started a business called Manila, 15 years ago, to do many of these things what the robo advisers are doing today. It didn’t work. But that was 15 years ago. So now it’s a different time with a different generation of people, with different things happening in the sector, so it may take on. But, in a different way.

AO, Fast Company: Yeah. Feels like, with all these companies competing, a lot of what they’re competing for is trust. How do you win that trust? I know, Alan, you’ve been involved in sort of thinking about what does American Express’s sort of digital brand look like. I’m curious. Are there things or challenges that you’ve solved there that sort of inform how you take a brand that has won trust sort of offline and instill that same value online?

AG, American Express: Yeah. Yeah. So we think that our brand and the capabilities that we have as a company have translated well into the digital space over the last, say, 20 years. There’s some evidence of that, including the brand attributes that you mentioned: trust, security, service. Those are… we like to say those attributes are more important in the digital world than they are in the physical world. So it’s translated well.

Having said that, we have a healthy paranoia about the way the industry might evolve, and there’s sort of things that we do to manage against that. One of them is making sure that the basics of our digital interaction with customers is as good as it is offline. It’s that good or better in the digital world. We try to remind ourselves that there’s a lot of the interesting and fancy and sexy things out there. But on any given day, the vast majority of our customers, more than 95%, just want the basics done well. Right? They want the card to work if they’re using it in a digital setting, just like they want it to work when they swipe it at a store. If they’re on our mobile app, there’s a 90% chance they’re doing something very simple, like they’re checking their balance. They want to know when their payment is due, etc.

So we just try to remind ourselves that operational excellence on the basics is key. Beyond that, what we do is we partner with companies in the e-commerce world, large and small, to make sure that our payment functionality, and in some cases, our membership rewards currency, for those of you familiar with those points, are embedded in the commerce flow, where our customers want to transact, because our customers want choice. So almost every single American Express card holder in the U.S. also carries another credit card, almost every single one of them. So they want choice. They want convenience. So we want to be where they are.

So we’ve done a lot of things to embed both payment functionality and sort of pay with MR points, into commerce flow everywhere our customers want to be, and that’s from everything from a big decision, like we immediately decided to be in the Apple Pay wallet, down to you could use your MR points to pay for your Uber.

The third thing we do… I have to say this might depress you guys, but it is the least important in our minds, which is we have an investment fund. And we invest in… we make small investments in a lot of FinTech companies, basically to learn, and to get exposure to the talent in that space and to figure out is there, like I said before, is there a commercial relationship, either in the beginning or over time, with companies like this, that might benefit us as a company and might benefit our customers. So that’s our approach to it, but we wake up every morning, freaked out that something is gonna change in the wrong way. I think that’s healthy, because that sort of drives us to stay on top of it.

AO, Fast Company: You mentioned payments, something we haven’t talked about yet. What are the trends in payments that you all are most, I guess, thinking will be finally maybe transformative? I feel like payments has been kind of stuck in first gear for a little while in the digital world. What’s gonna kind of get it over the hill, I guess, to this next phase?

BB, FTV Capital: Well, there’s a lot going on in payments. I think earlier, people were talking… I think B2B payments is an area that’s really exciting and completely untapped, but just focusing on consumer and some things that I’m sure my colleagues will have some thoughts on.

One of the areas we’re very excited about is FX. So what you’re seeing… when we think about where do we want to look for more innovative solutions, we tend to look at where are there excess profits. If there’s one area I think probably everyone would agree we’re all getting screwed, it’s when you’re getting your foreign exchange conversion on your credit card bill, and there are excess profits there that I believe banks and, sorry, credit card companies are extracting.

AG, American Express: Carry a gold or platinum card, and you’ll never have to pay that.

BB, FTV Capital: There you go.

AG, American Express: Just saying.

BB, FTV Capital: So we have a company that specializes, called World First, in powering the FX for larger transactions, where we charge substantially less, but also powering merchants that are in marketplaces to disintermediate the credit card companies, the banks, and provide a far better FX solution to merchants, and that company is growing close to 100% a year in that business and really has found a very interesting segment of the market to focus on. So that would be one example that I would look at.

The whole landscape of e-commerce, fraud detection, in that payments landscape, is gonna be a critical area to watch. Again, we’ve made massive inroads electronifying consumer payments, but B2B payments still virtually 100% really being done by checks, and that’s a massive opportunity for change.

AG, American Express: So payments is my thing, and I agree with Brad. So there’s excess profits in FX, for sure. Listen. I don’t know if there’s a regulator in the room, but there’s excess profits in the credit card industry as well. The good news is it’s mostly buried in banks, whose other businesses suck so much that you don’t see it. They have a 10% return on equity. But in there is a credit card business that has a 30% return on equity, like American Express has, but there’s reasons for that.

So the payment industry has been growing faster than the economy, since it was introduced in its current form, roughly 50, 55 years ago, and is going to grow faster than the economy, not just long after I’m retired, but, believe it or not, long after all of you are retired, because it’s been at it for 50 years. To this day, over 80% of the world’s transactions take place like this. A human being walks into a physical place and hands another human being a piece of paper. More than 80% of the world’s commerce takes place in that fashion. Businesses, including businesses that are big enough to know better, cut checks in the back room and send them to each other.

Totally ridiculous way to pay for things in 2016, but it’s stubborn, and part of the reason it’s stubborn is because, because a lot of this commerce takes place in the physical world, it makes it harder to get the network effect. So Visa, Mastercard, American Express spent 50 years of laying the pipes that brings all the physical world together from a merchant network standpoint, and all the new entrants have basically just ride on those rails.

Again, we’re paranoid that that will change, but all the changes in the payments space, as it relates to e-commerce and the digital world over the last 20 years, has helped incumbents like American Express, like Visa and Mastercard, because it’s basically accelerated the trend toward electronification. If you think about buying something at a store versus buying something online, the share of payments online for credit and debit cards automatically goes up, because you can’t stuff cash into your phone or into your laptop, right?

So all the changes so far help the incumbents, and I think that is, that sort of barrier to entry for creating a network in the physical world, where there are sort of 30 million outlets all over the world where people accept payments, is part of why there’s excess profits. When the government starts to come after you, you know that’s when you have excess profits. Right? So the interchange regulation in Australia, in Europe, that’s gonna continue to affect the industry as well. I think the industry, the banking part, the issuing part of the industry, is going to have to turn more toward its traditional role of matching depositors and lenders.

DS, Standard & Poor’s: I think I agree with everything. I think B2B is a space. I mean there’s a lot of stuff happening in the consumer B2B is getting a lot of attention. I think that’s where blockchain, as it evolves more, will become a fairly good enabler of this.

BB, FTV Capital: We just announced an investment in the first quarter, in a company called VPay, which is one of the leaders in something called virtual payments. This is the use of a credit card number generated for a single-use pay. So really the first application, which we can all, I’m sure, appreciate the inefficiency of, is healthcare payments, when a healthcare insurer is trying to pay a provider for a particular claim. So we’re able to generate a single-use credit card number to the provider, and they can then use that to pay. We think that type of a solution, using the rails that have already been put in place, and then just having this sort of single-pay credit card numbers can be an interesting example of a B2B solution, in addition to blockchain and some of these other innovative things that people are working on.

AG, American Express: So that’s, by coincidence, a great example. We bought a company, I don’t know, 10 years ago, that has that capability, and that’s a big product offering in our corporate card space, and our biggest vertical for that product is the healthcare industry.

BB, FTV Capital: There you go.

AO, Fast Company: Let’s briefly talk about risk, and then I think we’ll take some questions from the audience. There are lots of companies out there doing crazy things with LinkedIn data and social data and other ways of sort of thinking about risk and turning it away from some of the standard measures. I’m curious what you all think about that, and where you see sort of the gaps in the data that we currently have in making decisions about underwriting and that kind of thing.

BB, FTV Capital: We think that there is a lot of very interesting data. Big data is a real thing. I like to distinguish between big data, which is just lots of data, and big information, which is where it’s something that’s actually useful, and we have to find out which one we actually have. But we think that there’s enormous potential, in terms of analytics being applied to the data that’s being generated. Again, we have to go through a downturn to see how good it is in helping with underwriting, but clearly in things like underwriting insurance risk, we’re seeing the benefits of that. And as an investment theme, it’s an area we spend a lot of time and believe that virtually every single financial services company is looking at that.

Now, some of the stuff you’ve seen, some of the over-hyped online lenders, talking about using Facebook data to determine whether you’re a good credit, I think we’re a little more skeptical of that. But if they come out of the other side of a recession with 1% loss rate, God bless them. They’ve figured something out that nobody else did. So it’ll be interesting to see how it plays out.

AO, Fast Company: Yeah.

DS, Standard & Poor’s: I think as a backdrop… it’s very interesting. The data science is at a much different level with the technology even evolving. So you’ll see a lot of use of lots of new data sets and testing and evaluation. Brad’s right. You have to wait to see how the results are to see which worked and which didn’t work.

I think sort of… it takes me back to… if you look at, among the different contributors to the crisis, there was, one of the key areas of crisis was the data and the modeling. That’s what sort of risk I sort of look at looming out there. If you look at… there was a lot of junk in and junk out. People are using all kinds of mortgage data here that was fraudulent, that was biased. By the way, now I see a lot of people doing underwriting based on social data, and there’s enough research now done that shows, for example, product reviews and customer reviews are very biased, both positive and negative.

So the question is how do you decipher that to really make it a meaningful data intelligence? There are people who are working on that. I’ve just been looking at something in that direction. The second is modeling. There is a lot of attention being paid by FTC, and I think you’ll see this become a big issue. There are folks working on different angles around this, at four or five different universities, that sort of show basic, how do you test the biasedness and discrimination in these models? That’s gonna… that’s looming, and it’s gonna come up, and people have to figure out what’s a solution to that, and don’t know yet, but I know I’ve been working with a couple people who are looking at that.

The third is really, that drove the part of the crisis, which is a lesson learned for me, is that when you start using predictability, you become over-confident about it. Then conditions changed, or the deals changed, but people say, “Oh, this fits. Let’s try it. It’ll work.” It doesn’t work. Those are the three things I think, in the use of alternative data, that I think there’s a risk that may be looming, based on past experiences, based on what I observe now and some of the rush to use all that.

AG, American Express: I think that’s why we’re still gonna need humans, at least for the time being here. We use a lot of… we’ve evolved our use of big data to really help us predict fraud. So far, it’s been very successful. Actually we have the lowest fraud rates in the industry. If you carry an American Express card, you may notice, over the last couple years, you get more text messages like, “Was this your transaction,” because we have all these sort of predictive models that help to spot anomalies and out-of-pattern trends.

Underwriting is tougher. Both of you guys said it. The difference between data and insight and people being over-reliant on models. There’s an old expression, “Risk is never where you think it is.” Right? So the smartest people on the entire planet worked in the risk department of big banks in 2005, 2006, and 2007. Where did that get you? Right? So recognizing the limitations of data and even the limitations of intelligence, and recognizing that there’s prediction, which you always have to be skeptical about, and then there’s preparation, which you always need. Recessions are going to come. They’re probably going to look different from the last recession, and you need to have sort of a plan to respond to that. You need to have enough capital, obviously, but you can’t fall in love with your own models, your own data.

DS, Standard & Poor’s: The last thing I’ll sort of say, one of the big observations, there’s a lot of noise. The key in all of this analytics and modeling is how to sort out between the noise and the weak signal, because the weak signal is the powerful thing. How do you know the weak signal won’t take you off the cliff? Or is this a really a weak signal that’s gonna take you somewhere? Hindsight, everybody will say, “I knew the weak signal.” You know? But the question is… there’s a lot of work going on there in lots of places, in academics and other companies and some fascinating stuff. I think that’s an area of development that will really be very fruitful.

AG, American Express: So I love to use the specific example in the case of Amex before the financial crisis. One of the factors in our models was that if we saw that you owned more than one home, you had more than one mortgage, that was a positive indicator, because we figured well you were an affluent person if you had more than one home. You obviously… you’re gonna pay your bills. That turned out to be exactly the opposite of what happened in the financial crisis. A lot of people who owned a second or third home, that shouldn’t have been owning a second or third home, shouldn’t have been borrowing for a second or third home.

So great book, “Signal and the Noise.” I don’t know if anybody’s ever heard about it, by Nate Silver, the guy that does the 538 blog. I love that kind of stuff.

DS, Standard & Poor’s: Just so… the crisis really happened because consumers historically always paid their mortgages…

AG, American Express: Mortgages first.

DS, Standard & Poor’s: … and if they went default, they defaulted on credit cards. So if you look at the mortgage defaults, very low. Credit card defaults, higher. Somewhere along the line, that curve shifted. Then the question became, “Gee, is the historical data that’s predicting these models right or wrong?” That’s one of the things. It was very hard to see the behavioral shift.

AG, American Express: We always use five-year historical data in our models. I was the CFO of the consumer business, right as the financial crisis was beginning, in a stunning stroke of great timing. When I looked at it, I said, “Well, you use five-year backward-looking historical information.” There’s a recession in the U.S. every eight years, for the last 200 years. So by definition, our models are gonna move us to lean into a recessionary period. So it wasn’t like we didn’t have smart people working on it. I think the smartest people in the world were working on this stuff, but you sort of… human beings stepping back and having context, not so that they can predict what’s gonna happen because nobody can, but so that you could recognize the limitations, and prepare for it really.

AO, Fast Company: Great. Well, let’s take just a couple of questions from the back.

Kerry Wu, Tech Industry Analyst, CB Insights: Yeah, so we just have two questions. The first is for Alan, and that is, “When your venture arm comes to you with an investment opportunity that will directly impact American Express, what are the biggest things that you evaluate, from a risk standpoint?”

AG, American Express: So as I said before is that we make relatively small investments. So the entire investment fund is a tiny little nit on the American Express balance sheet. So we don’t really assess it from the risk of loss. We assess it from the sort of opportunity cost risk, like, “How does this compare to the other things on your list?” So we really try to look for the commercial connection. Is there something unique about this technology or about the approach this company is taking, that we think can teach us something? Is there something unique that we can bring to our customers? We definitely favor businesses that we think that could be a commercial relationship with over time. The financial risk is actually a minor thing for us, because like I said, the entire fund is not big in the context of American Express.

AO, Fast Company: Great. I guess one last question.

KW, CB Insights: Right. Our second question is, “When a startup comes to you for investment, or for guidance, or to sell you their software or hardware, what is the biggest business model or technology risk that you find they often underestimate?”

BB, FTV Capital: Wow. I would… I mean look, I think that if you are an entrepreneur, if you are a founder of a business, by definition, you’re not thinking a lot about risk, but you, because you have to be pretty crazy just to start your own business to begin with. So we deal with very optimistic people, who are typically building the next Google or Amazon in their minds, but I think the reality is that one of the key fundamental issues is just, I think it was referred to earlier, is scale. There are just challenges of scaling a business, even if you’ve got a great product, that really requires a lot of thought about how you build an organization, how you build the right infrastructure. People tend to be very focused in startups around growth in a vacuum and not backfilling and building the support and the foundation to make sure that that doesn’t go off the tracks. So that’s really where we spend a lot of time, making sure that companies have those key pieces to get to their destination.

AG, American Express: I’ll give you the obvious sort of regulatory scrutiny issue. So a lot of small companies come to us, and we talk about, “Have you thought about this regulation? Or with how the CFPB is gonna react to this,” etc. They say, “Well, you don’t need to worry about us, because we’re so small that we’re gonna fly below the radar.” We say, “Yeah, but if you’re successful, you’re gonna be big. If you’re gonna do business with American Express, all of a sudden, they’ll notice.” You really have to think about, “Are you gonna build that muscle over time, so that you can scale and operate in the financial industry?” Also, do you recognize that when you’re in the financial industry, the regulators don’t let you do other things? So we had to get out of the publishing business, because when we became a bank holding company in 2008, because once you’re… if you’re a regulated bank, that’s it. You can’t sell shoes.

AO, Fast Company: And we’re not sure you would want to either. Great. Well, thank you all so much. I really appreciate your taking the time.

AG, American Express: Thank you.