Introduction

Patrick: [00:01:53] Hello, and welcome to Business Breakdowns. I'm your host, Patrick
O'Shaughnessy. Today we'll be diving into Visa. Started in 1958 as a bank
BankAmericard credit card program in Fresno, California. Visa then became a
nonprofit consortium of banks that operated the Visa network. Over the first few
decades of its existence, Visa became the protocol layer that allowed essentially all
banks in the world to communicate with one another. In 2007, Visa completed a
corporate restructuring that took it public and now boasts a larger market cap than all
of the banks that previously owned it as part of the consortium. In this breakdown, we
set the stage with the role that Visa plays in the card transaction, describe the
lifeblood of Visa revenue interchange, and then dive into its unique history as a
consortium turned multi-hundred-billion-dollar public business.

We then explore Visa's unique moat and network effect, how Visa makes money today,
and look at the potential threats from other businesses and macro-economic forces.
Visa is a fascinating business and I recommend you check out our website at
joincolossus.com, where we provide additional articles, books, and podcasts for those
that want to keep unpacking the Visa's story. To help me break down Visa, I'm joined by
Alex Rampell, a general partner at Andreessen Horowitz, where he focuses on
investing in financial services. Prior to joining Andreesen, Alex co-founded multiple
companies, including Affirm and TrialPay, which was acquired by Visa in 2015. Please
enjoy this Business Breakdown of Visa.

An Overview of Visa's Business

Alex, Visa is such a fascinating business because, well, I think literally everyone
listening will know the name or have used the product. Probably some tiny percentage
will have any idea of what is going on when they swipe a credit card with Visa
emblazed on it. Can you walk us through just to set the scene, what is happening and
who the key players are in a single credit card transaction? Because I think that
backdrop is really important to everything else we're going to talk about with Visa and
its business model.

Alex: [00:03:47] There are actually five parties in a transaction. I'll name the parties, and
then I'll try to chronologically show how this works. So imagine that I, Alex, I'm buying a
Chipotle burrito. So it's obvious, I need to give money to Chipotle, but the way that works
is I'm buying it with a Capital One credit card. That's called the issuer. It's called an
issuing bank. Capital One's a bank, they gave me a card, the card has a credit line
attached to it. Chipotle has a merchant bank, it's also known as an acquiring bank. It's
called an acquiring bank because they acquired the merchant. Let's just say that
Chipotle uses Bank of America merchant services to go sign up, or First Data, one of
those.

And then there's a network that connects the acquiring bank to the issuing bank. So if I
buy my $10 burrito at Chipotle, what happens is, Chipotle swipes that card, it goes to the
acquiring bank, so call it Bank of America merchant services. Bank of America
merchant services sees, "Oh, this card starts with a four, that means it's a Visa, we got
to send that over to Visa, if it starts with a five, we have to send to MasterCard, but if it's
a four, send it over to Visa." Visa sees, "Oh, it's 414720, That's the BIN, Bank
Identification Number. The first six of your card actually just identifies the bank. So
414720, if I remember correctly, is a Chase card. So I guess I just broke my metaphor
here with Capital One, but that means that Visa sends it to the issuing bank, Chase. And
then Chase says, "Do we approve this transaction? Yes or no?" Because the risk is
actually being borne, assuming that the customer actually made the transaction, I might
not pay Chase back. Maybe I do pay Chase back, that's a credit limit or a credit decision
that Chase has to make. So there are really five parties involved. And the reason why
most people don't really understand the card networks like MasterCard and Visa well is
that neither merchants nor consumers actually deal with them. They're simply a router
that connect the issuing banks to the acquiring banks. The issuing banks are the ones
that own the consumers, the acquiring banks are the ones that own the merchants.

Source: Fin, Published by Plaid

Patrick: [00:05:49] Let's say there's $100 transaction that happens, and $3 of that 100
gets sliced up amongst those three players. So I'm paying 100 as a consumer, I lose
100. As a merchant, I'm getting 97. There's $3 left over for the acquiring, the issuing
and the network. How, traditionally, is that $3 split up?

Alex: [00:06:09] So the vast majority goes to the issuer. And this is actually what's really
interesting about, if you think about what is a product in the Visa world or the
MasterCard world, a product is effectively an interchange rate. It's very complicated to
answer your question because it depends on the type of card. So if I pay with the debit
card, it's not anywhere near 3%, unless it's a debit card issued by a small community
bank with under $10 billion of assets. But what if I pay with a Visa Signature card? So
what actually happens, it's a little tangent, but it's relevant to your question, Visa has
something called a Visa Signature card, that comes with a higher interchange rate. And
the higher interchange rate is actually used to incent banks to issue that card because
then they get to keep more money. The first answer to your question is, the vast majority
of that goes to the issuing bank. A little bit goes to the acquiring bank, not that much.
And then the network, I think on average, like Visa, makes about seven and a half cents
per transaction, but the interchange, which is the inter-banking rate between the issuing
and the acquiring bank, and that's set by the network, the issuers don't like it when Visa
or MasterCard changes that rate.

There is a rate card that you can actually download. If you go to Visa's website, if you go
to MasterCard's website, there's a rate card, it's a PDF, it's probably 25 pages long to
show you exactly how many rates there are. And depending on like, is it a corporate
charge card? What if it's an international transaction? What if it's a gas card? And then
what kind of merchant is it? Oh, it's a utility bill. Well, that's different, it's card not present,
meaning that you're not actually presenting a piece of plastic in person, it's a
card-not-present transaction, which is entered on the internet or over phone. So there
are so many different rates, but the reality is that the issuer gets most of the economics,
the issuer uses most of those economics to then, I don't want to say bribe, that's the
wrong word, but reward the consumer. And this is why Visa and MasterCard has just
such powerful moats. If it's a double-sided moat. It's not just a network effect in that
everybody uses one of these cards because everybody else accepts that card. I use
WhatsApp, my friends use WhatsApp, classic network effect. It also has this very bizarre
set of economic incentives where the money that the issuing bank is getting paid
rewards the customer. So again, go back to the Chipotle example, but maybe not as
relevant for $10 burrito, I'm buying like 1,000 burritos now, so it's going to be a $10,000
bill.

If I put this on my Citi Double Cash Back MasterCard, where I get 2% cash back, I'm
going to get $200. And Chipotle doesn't want to spend the 3%, as you mentioned, they
would love it if interchange went away because they would make more money, but how
do they induce me to sign up for a new payment mechanism or go pay them with cash?
I don't want to go to the bank and withdraw $100 bills and pay them with cash, I'm going
to pay them with the Citi Double Cash Back MasterCard, I get $200. I now have an
incentive to use that card wherever possible. Taking another step back, Visa is able to
convince a bank like—Fidelity has a double cash back card, 2% cash back on the Fidelity
card is one of their top cards. Fidelity is trying to decide, "Do I issue a MasterCard or a
Visa card?" Because this is what Visa and MasterCard really do, they compete for
issuing banks because acquirers have to acquire all cards. They just get the merchant
and they want to monetize all the merchants' traffic so they can't just say, "We're only
going to process Visa cards. We're only going to process MasterCard." They have to
process all cards.

So, what happens is, Visa and MasterCard compete over issued contracts. So if they
have a higher interchange rate that they're charging to merchants, which merchants
hate, of course they hate paying a higher interchange rate, the issuing banks love it
because then the issuing banks are able to go into the market and say, "Hey, guess what,
consumer, instead of getting 2% cash back, you get 2.5% cash back." And that's not
coming out of negative unit economics, the 2.5% cash back is coming because we have
this great product that Visa invented to charge merchants 3%. So it's one of these weird
things where normally you would expect competition, and there is competition, Visa and
MasterCard, American Express, like it's an oligopoly, but it's not just a monopoly where
there's only one player. You would expect, at least in that kind of construct for fees to go
down, but outside of regions like Europe or Australia where regulatory issues have
forced down fees, fees go up because that is more compelling for the issuing banks to
go get their cards into market and for consumers to use those cards, because it's kind
of like an intelligence test, like, "Okay, you're buying the $10,000 of burritos at Chipotle,
do you want $200 cash back or $10 cash back?" Well, I'd rather get $200 cash back.
Well, in order for that to work, interchange has to be above 2%, and therefore, if I make
interchange 5%, Chipotle will be even more angry. That might be the breaking point for
them. But now, a new bank can go get me to sign up for that 5% cash back card, and I'm
going to use that one at Chipotle as opposed to the 2% cash back card.

How Visa Started

Patrick: [00:10:54] I'm thinking about this visual where there's these two nodes on the
bottom, the consumer and the merchant. Above each as a bank, they each have their
own bank. And then the way those banks communicate is through this centralized
network, primarily Visa and MasterCard, and Amex is a whole different beast that we
won't tackle in this conversation. So really, this central network is really not a finance
company so much as it's an information routing system. It's literally just the
communication node through which banks communicate with each other. And it begs
the obvious question, which necessarily brings us to Visa's very unique origin story of,
with this 3% thing hanging out there, why wouldn't the banks just talk directly to one
another. So maybe we could rewind all the way back to the beginning of the origin
story of Visa, which is among major corporations, certainly one of the most unique,
and describe why there's this central player and the banks don't just talk to each other.
Alex: [00:11:47] The first day of Visa, well, it wasn't called Visa at the time, but it was
September 18th, 1958. And what happened was, Bank of America, which at the time,
was a California only bank. Even though we have crazy banking regulations today, it was
much crazier than the 1950s. What happened was that there was this mid-level
manager at Bank of America, Joe Williams, who came up with this idea for the
BankAmericard. And the idea was, "Okay, let's just go mail 60,000 cards to pretty much
everybody in Fresno." I think Fresno had like 250,000 people at the time, 45% of them or
something like that were banked by Bank of America. The way that you're going to solve
a chicken and egg problem is just give everybody a card, they don't even have to apply
for it. There aren't that many merchants in town, it's not like there's an internet where you
go have to sign up 100,000 merchants in order for this to be relevant. So I think the first
merchant that they signed up was Florsheim Shoes, which is still around today. So you
got all the merchants signed up. They gave everybody a card. There wasn't an
interchange, but they said, "Okay, it's Bank of America." Bank of America is talking to
both sides. I think the rate was something like 6%. But it turned out, because they just
gave this to everybody, there was no application, they were giving people credit, fraud
was out of control. Joe Williams basically got fired and had to leave. But the idea
worked because the problem with most of these network effect businesses is it's like,
okay, you either get the chicken or you get the egg. And because Bank of America
already had half the customers in the town, I mean, 250,000 is a pretty large sized town,
and it wasn't that hard to just go to mainstream and get all the merchants, you were able
to get this off the ground. And then other banks took note of this as well.

So if I'm, whatever, Chemical Bank or Chase Manhattan Bank or the banks of the 1950s,
I don't want to go issue a BankAmericaCard, that doesn't make any sense. So basically
what happened, fast forward a few years, but you had all these different banks that were
doing these cards drops, card drops were eventually declared illegal, you can't just like
give people credit and not tell them what they're in for, and then prevent them from
getting a mortgage in the future because they didn't pay you back for something that
they didn't understand. So the laws clarified a bunch of these things. But what used to
happen in the '50s and '60s and '70s is that you might have like 10 cards in your wallet,
only certain cards would work in certain places. And basically what happened was,
there were two consortiums that developed, there was Master Charge, which was one.
And then the BankAmericard became Visa. So there was a set of banks that operated
on this Visa network, there was a set of banks that operated on this Master Charge
network. Master Charge obviously became MasterCard. And then both of them were
nonprofits, and they're nonprofits not like The Red Cross is a nonprofit, they were
nonprofits the NFL is a nonprofit kind of thing, which is owned by the constituents.
And that's how they existed for dozens and dozens of years where effectively, it was a
communication protocol, exactly as you said, it was a router and a technology platform
for banks to talk to banks, and they owned it. Visa wasn't taking any economics. I mean,
they would take economics in terms of like they had to pay for software development
and whatnot, but it was this nonprofit entity that was controlled by banks so it was
totally fine for this thing to exist. And then what happened, I think it's 13 years ago, is
they both went public.

And they went public, not because it's like, "Wow, let's seize this opportunity because it's
great to be a public company," but it was largely to avoid antitrust scrutiny because once
a year, it's a little more than that, but there was this pricing council it actually still exists
at Visa today, it's like, "Okay, how much should we charge merchants?" Imagine that all
the airlines... There were two groups of airlines. There was Star Alliance and whatever
the other evil airline alliance is, and they got together and they said, "Okay, Star Alliance
has decided we will charge this much for the London to San Francisco flight." Well, that's
called price fixing, you can't do that. So Visa owned by the banks and MasterCard owned
by the banks, this is what they would do. They would come up with like, here's pricing. It
always somehow finds a way of going up every year. The merchants hated it. The
government wasn't too fond of it. So the IPO's of both MasterCard and Visa, and these
were very profitable entities once they reorganized such that they could actually control
and keep their profits, it was primarily secondary. It was very, very little primary. It was
the banks really selling their primary stakes in the company. And now, what's really
interesting is that Visa has a higher market cap, I believe, than any bank in the world.

And it's one of these weird cases of, it's almost like patricide, where Visa
was birthed by Bank of America and now has a much, much higher market
cap than Bank of America, has a much better market position than Bank of
America because they are the central player. They are the central router
where all commerce goes through whereas Bank of America is one of end
participants.

The Protocol Effect

Patrick: [00:16:25] So it's almost as if both of these network examples got to very
quietly, really without competition, mature their network effects over decades. And
that ultimately when they went public, you have these two really unique entities in the
entire corporate sphere. Are there any other businesses that you can think of that have
what I'll call a protocol effect like this, where they're almost just a pure information
router that due to fragmentation around them, there's tons of banks on both sides, it
doesn't make sense to build the technology, to connect... every bank to connect to
every single other bank? Are there any analogs like this in the world?

Alex: [00:17:03] If you remember early on in the days of Twitter, and I know you use
Twitter, I use Twitter, Twitter was more protocol and there were a bunch of Twitter
clients, dozens of them. There was TweetDeck, that was one that I used. And eventually
Twitter realized, "Wait a minute, we can't serve as effectively. If we want to make this
work, not only are we the centralized router, but we want to control the client experience
on both sides." Really one side, because I'm writing tweets, I'm reading tweets, I'm using
the same product to really do that. But I would even say Twitter does have other clients
where they probably charge you usage fees or something. TweetDeck still is around, or
there are products that help big merchants that have unhappy customers on Twitter,
they're not just using the Twitter app. When United Airlines cancels a flight and 200
people on the flight go complain, there's not some random person in the call center,
that's just logging into Twitter, they're charging for access to the protocol. I agree, it's not
as common, because normally what you see is you don't have this like weird joint
venture of like dozens of banks and you go for this structure, you say, "If I'm building a
network effect, classically like eBay network effect, OpenTable network effect, it's
always the centralized player that controls the experience and controls the economics.

Actually, this was one of the things that I was getting to around pricing, Visa has a
pricing committee. Somebody like Amazon will say, "Hey, wait a minute, it's not fair that
we pay card-not-present transaction rates because our fraud rates are lower," and yada,
yada. And Visa might be sympathetic to their arguments. But if Visa says, "Oh, it's a
good point. We'll cut interchange for ecommerce transactions for trusted consumers to
1.7%." That's going to really off the issuing banks, because the issuing banks' like, "What
are you talking about? We just got a double cash back card to pay out 2%. Now, you cut
our rate to 1.7%? You can't do that. We're going to switch to MasterCard."

So it actually makes improving the product and improving the experience so much
harder because they don't control the end-to-end platform in the same way that you
mentioned like American Express. American Express is what's known as a closed loop
network where they control... They acquire the merchants and they acquire the
customers directly. It's one entity. You're just dealing with American Express through
and through. Whereas Visa and MasterCard are called open-loop networks. It's a very
rare business model which you'll only see that happen if you get the constituents from
day one. Maybe some of the markets are like this. Like if you look at, obviously
everybody knows NASDAQ, but if you looked at BATS, the Better Alternative Trading
System, that was a company where they got started because they were able to get more
high frequency flow from various traders. Those were equity owners in the actual
exchange.

I've seen this happen from time to time, but it's just a very, very hard
playbook to operate on because normally network effects, they start a
small, like Facebook starts small, but they control the client, they control
the end-to-end experience.

The Card Network Business and Innovation

Patrick: [00:19:46] It's almost as if the model for both of these companies is like if you
were an evil genius designing like a perfect competitive moat, you might design
something like this, just deeply embedded, deeply hard to rip out. But nonetheless, it
begs the question of like, "Okay, maybe we can understand its moat. I think as you
described to me, by describing the ways in which it may be vulnerable. There's a
couple different ways we could do this. Before we get to that, I'd love to understand,
what is the business then in terms of new technology or existing technology being
built? Is there a lot of new spend to improve or maintain this network?Because at its
core, it's been working the same way, it seems like, for a long, long time. What does the
business itself look like? Revenues and expenses, like just at a basic level, What does
the internals of the business look like?

Alex: [00:20:33] Yeah. I worked at Visa. Visa bought my company. I wanted to answer
this question every day. What do all these people do? It's a very, very simple business to
operate, and I mean that is a compliment and not an insult. They need probably 10
people to keep the lights on. And if the lights are kept on... And it's actually kind of
remarkable, when was the last time they had an outage? Whereas all of these
newfangled tech companies with amazing engineers, they seem to go down all the time.
Twitter has their Fail Whale and PayPal goes down and eBay goes down and Amazon
goes down and then Netflix goes down, but Visa and MasterCard do not go down. It's
pretty remarkable that they've been able to pull that off.

So it goes back to, there's not that much that they can really build or do. Product for Visa
and MasterCard needs a new branded logo or a new interchange table where like Visa
Signature cards, what do you get from the Visa Signature card? Well, if you go to Napa
Valley pre-COVID, you got a free wine tasting at these two wineries, you get insurance in
case you buy an expensive item that breaks, or you get, if the price drops when you
bought something at Macy's, we'll reimburse you. It's a lot of these kinds of breakage
benefits, they get included for consumers, but it's almost like classic enterprise sales
where I'm a bank, let's say I'm Barclays, and actually I remember this. When I first joined
Visa as part of the acquisition, we had to do a little bit of a dog and pony show because
Barclays was trying to determine, we've got this portfolio, Barclays is behind the LL Bean
card. And I know that because my wife has an LL Bean card, but they have a bunch of
these affinity cards like LL Bean. And Barclays is like, they're up for renewal, it's a
three-year or five-year renewal cycle. Do they go switch everybody to a MasterCard or do
they keep on Visa? And maybe they might even split their portfolios so it's half and half.
And we have to say like, "Here are all the great things that we do at Visa. We have
chargeback technology, we have fraud technology. We have the wine tasting at Napa.
We bought this amazing company that I started called TrialPay that provides offers to
consumers so you can say, 'Hey, get $5 off at a non-competitive... Like you get $5 off at
Fandango and Fandango will pay for that,' or something." And then MasterCard goes
with their own dog and pony show, and it's this matrix where it's like, you've probably
seen this if you've looked at marketing collateral, when one company is... like SAP is
selling versus Workday or something, it's like, "Hey, do they have on-premise cloud? Blah,
blah, blah. I was like, 'Yes.'" You always have a checkbox and like other guy always has
an X. Amazing how that works. That's a lot of the product development, is stuff like that
because it's just really hard to change the protocol itself often to the detriment of the
market.

Nobody knows what SKUs are being bought. That's crazy. Why do I have to get a paper
receipt when I go buy something at Walmart? Why can't that just be added to my credit
card bill? Well, the reason why is because there's literally technologically no way for
Walmart to send back to First Data, to send back to Visa, to send back to Capital One, to
put in my app, what items I bought at Walmart. And a lot of this is just based on how old
this technology is. But also because for protocols, the hard thing is not actually
improving the technology. I was talking to somebody about this for like, "Why do we
have T+2 settlement for equities? And why isn't it T+1 or real time?" And it's not a hard
tech problem, it's just you got to get everybody to agree. If five big parties don't agree,
then nothing happens for 10 years. The longest that a merchant name can be in the Visa
protocol is 20 characters. It's crazy. It's like, well, I don't know, my MacBook Pro that I'm
calling you from has a terabyte of space that can fit every merchant name in the world,
even if the merchant name was 100 characters long, but that wasn't the case in 1960,
when a megabyte probably cost a million dollars.

So you have a lot of stuff like that that's very hard for them to change, again, not
because of incompetence, but just because as the protocol layer, they would have to get
all of the banks to agree to go change this protocol. And then as a product layer, there's
not that much that they can do because it's like, "Oh, we want to do these five things that
make the consumer experience dramatically better. But wait a minute, number one, we
don't want to compete with the issuing banks because the issuing banks are how we get
our business. If we pissed off all the issuing banks, we would go bankrupt. They
wouldn't issue our cards, they'd issue our competitors' cards." So that's a bad idea. So
that's innovator's dilemma number one. But number two is, there's no dilemma, it's just
like, "Well, we don't control the customer, that's the issuing bank. So we can make tools
for issuing banks that maybe they roll out to customers, but it's kind of a maybe."

Patrick: [00:24:47] So if you think about just the business, is the vast majority of their
revenues simply this, call it seven cents per transaction, and therefore the way to think
about Visa's top line is just that they're writing on, not only transactions, but the move
from cash to credit as the primary means of changing value or paying for things? Is
that where almost all their revenue comes from?

Alex: [00:25:06] Yeah. I would amend that to say cash to digital. Debit is a huge part of
Visa's business. There are a lot of threats to credit, like buy now pay later is a threat to
credit, it's not a threat to the eradication of cash. The pandemic has supercharged that.
The number of places that I go to say cash is not allowed. Well, guess what? That's
good for Visa. There are certain areas where they disproportionately benefit like cross
border transactions are a huge, huge part. That's their most profitable segment. That's
where if you look at the recent financials where people don't travel as much and
therefore they don't do cross border transactions, that's bad for Visa and MasterCard.
But all of this, anything that's ever bad that's ever happened to them has always been
overshadowed in a good way by just the eradication of cash. It is this bridging
transaction.

And then when you have regulatory stuff that pops up, an interchange gets squeezed,
normally Visa's fee doesn't get squeezed as much because they're not taking any credit
risk. It might get squeezed a little bit just because there's less pie to cut up. Mostly, it's
taking away from the issuers because they're the ones that get the majority of the
economics.

Patrick: [00:26:10] So on the one hand, I would presume Visa has exceptionally high
margins. When the tech doesn't have to be updated that often, you have this incredible
network effect that you sit on top of, you can have very high margins. How do you think
about the thing that might fuel its growth though? So unlike a Walmart or a Costco
where you can just keep opening new stores, they don't seem to have this reinvestment
moat that's possible and seem instead to be really subject to, not out of their control,
but just like broader forces in the economy. Is that fair that we think of Visa simply as
like a tax on the growth of commerce, generally speaking?
Alex: [00:26:44] You could look at it as a tax, you can look at it as an enabler. That's how
Visa would say, which is, "Hey, look, we enable all these transactions to happen."
Therefore, like the reason to justify card not present being more expensive than card
present is just it's classic value based pricing. How much value is the internet in the
economy getting versus how much are we taking? I think your point is valid, which is,
there's not that much that they can reinvest in besides buying back their own stock or
something. I don't think it actually weakens the moat though. They have done things
that are potentially well-intentioned, but didn't really make sense in retrospect, which is,
should Visa have a wallet? PayPal is a wallet, we should have a wallet too. Should
MasterCard have a wallet? Should American Express have a wallet?And I feel like
companies often do things that make strategic sense, but where there's no market need.
The great example that I like to use of this is, should Google have a social network?
Facebook's grown, it's like, of course they should. It's a no brainer. Social networks are
going to be big. Wait a minute, Facebook solved this problem in like 2007, the now
shuttered Google Plus launched thereafter. So you can't just build something because
you should strategically have it. You should build something because, A, you should
strategically have it, sure, only if it's key. The market actually desires it.

Visa had this ill-fated project called Visa Checkout, where they spent all this money
getting merchants to go accept, basically you could pay with a Visa Checkout
apparatus. You'd go to a website just like you pay with PayPal. You go click there, you go
pay with Visa checkout. But again, it just didn't solve a fundamental consumer need, and
like PayPal solved that problem 20 years ago so it was hard for Visa to compete. So that
been said, does it make sense for them to do it? Sure. Companies try and fail things all
the time. I'm in the business of investing in companies that seem to fail all the time. But
the ones that work, work really, really big. There are a lot of investments like that that
Visa can do too, but it's not as natural, to your point, like, let's go build 1,000 new stores.

A lot of it, for me, I do FinTech stuff here in venture capital land, the biggest banks of 10
years from now might be non-banks that look like little tiny tech companies. Stripe was
not a big company 10 years ago. Now, it's a reasonably big company. The way that it got
there was actually saying, "Hey, let's go sign up everybody in this little tech incubator
called Y Combinator and get them processing cards with us, even though their volume is
insignificant, and even though all of the big merchant acquirors, just ignore them. And
so it's like, why would I want to go process payments for somebody that does two
payments per year? That seems very boring. But if you get somebody like Instacart
early, you're going to process a lot of payments for them in your aid, but you got to get
them early. So I think that's actually probably the farming motion that both Visa and
MasterCard have to get used to, which is, so many companies are now monetizing via
financial services. So you wouldn't be surprised to see a Lyft or an Uber offer a card to
all of their drivers that displaces their bank. Why would they do that? Well, because they
get to retain those drivers longer. And if the driver's low on money, they can say, "Hey,
drive for us and make more money." There are all sorts of benefits to actually controlling
the financial infrastructure. But you wouldn't think of Visa saying, "Okay, the way that
we're going to get big is we're going to win the Citi deal or we're going to win the Chase
deal." That's what they've historically done. It's not like, "Hey, let's go sign up 10,000
non-banks that might offer a card in the future that currently are very small." Anybody
who works in FP&A would find this job very challenging to say like, "Well, why would I do
that? What's the size of the market?" The answer is, it doesn't really exist yet, but it might
be big. That's probably something that they can get into in a much more meaningful
way, as opposed to the zero sum game that they play around jockeying for bank
renewals, which is not zero sum in as much as the eradication of cash is floating all
boats.

But you can make it even more positive some where it's like, "Hey, let's go get 10,000
brand new neobanks to go issue cards, or 10,000 new, I don't know, SaaS platforms that
handle small business, something or other. Like there's a company called ShopMonkey
that helps body shops run their business. Why not give every one of those body shops a
card? Is it a Visa-issued card or a MasterCard-issued card? They should both be fighting
over ShopMonkey, which is emotion that they're just not as used to.

Risks and Challenges

Patrick: [00:30:47] I'd love to understand Visa through the angles of things that may
happen that affect it or its moat. The first is fragmentation. You've already talked
about how they're the beneficiary of their being, even from the earliest days, lots of
different banks on both sides, if you're the centralized communication layer,
fragmentation is good for you. Presumably concentration is then bad. So how do you
think about that as a risk factor that reveals something interesting about Visa.

Alex: [00:31:13] It's the biggest risk factor, and the reason why, going back to what I said
before, which is the only thing that really drives Visa MasterCard's business is issuer
relationships. Citi is the biggest issuer of MasterCards. If Citi says, "Screw you
MasterCard, we're going to Visa." I would not be surprised if MasterCard stock dropped
by very, very large double digit percent because of how big Citi is for MasterCard. And
you could play the same scenario for Chase and Visa. So it's very, very important for
them to hold onto those issuer relationship because again, acquires have to acquire the
merchant, and the merchant is supposed to accept all the cards and they don't want to
be caught up in some game of chicken around negotiating rates. The reason why this is
relevant is Chase renewed with Visa, but Chase is actually the biggest acquirer of online
merchants. So Amazon.com I believe uses Chase payment tech, which is their payment
acquiring division or merchant acquiring division for online payment processing. But
wait a minute, Chase is also the biggest issuer of Visa cards in the US. The
Amazon.com credit card is issued by Chase, Chase issues the Sapphire card. They have
tens of millions of customers that actually have Chase cards. Why the heck am I Chase
paying Visa to talk to myself? That doesn't really make sense. And the reason why is
because it was in my contract, but when Visa says, "All right, Chase, please renew with
us again for the next 10 years and don't go to MasterCard," Chase is like, "Wait a minute,
I have an idea. I don't want to pay you anything, if I'm talking to myself," which is a pretty
reasonable request if you think about it. This how a lot of acquiring and issuing
relationships work outside the US. This is called On Us Transaction. So it's perfectly
easy to do, it's not that hard, technically, it's just a contractual thing, and then a very easy
snippet of code to say, "Hey, 414720 go, just send it this way, the other part of our
building."

So Mexico, for example, almost no transactions in Mexico go over Visa. The acquiring
base and the issuing banks are effectively one and the same, they just talk to each
other. Now, Visa is very useful for this interoperating when you get into spaghetti land,
where it's like, "Okay, I live in Mexico, but now I went to the US and I'm shopping at a
store in Texas. Well, that's not going to be an honest transaction, that's going to go over
Visa rails, and it's an FX transactions." That's even better for Visa. One of the dangers
that Visa and MasterCard face long term is that it's not so much the concentration, it's
more of concentration where the issuing and acquiring is done by the same entity.

The reason why that's relevant is because Visa doesn't really care about the acquirers.
I'm sure they will say they care about them, but they don't care about them as much as
they care about the issuers. If you're a big issuer and you've got a big acquiring division,
you can easily push Visa and say, "Hey, I want what Chase had," which is their Chase net
thing, "I want that too. And I want to pay nothing for it." And this is actually how things
work in China. The largest payment network in China, like old-fashioned non-QR code
based payment network is called China Union Pay, CUP. And CUP actually interoperates
with Visa. So if you live in China and you have a China Union Pay card, which many
people do, and you leave China and go buy something in Hong Kong or Japan, or the
United States, that's a Visa transaction. But when you go back in the country, then it's a
China Union Pay transaction. And that's effectively one form... Again, I think Visa and
MasterCard would still do well even under this situation, but if it turns out that there are
only 10 major, major banks in the US and they all issue cards and they all acquire cards,
you could actually see Visa seeing fewer and fewer transactions if it's banks talking to
themselves, or even banks saying, "Hey, why don't I just route directly to this other banks
and not pay Visa anything?" But if it's the long tail of like, "Hey, I'm visiting Jordan from
Columbia," Those two banks, Banco de Columbia isn't going to go connect with the Bank
of Jordan directly, too much work.

Patrick: [00:34:54] You've previewed the second one, which is the role of regulation in
different geographies. So we know that interchange fees are way lower outside the US
than they are in. Maybe describe what's driving that and how international standards
relative to the US is, may play a role in the future story of Visa?

Alex: [00:35:13] If I remember correctly, one of the first countries to do this was
Australia, which regulated interchange down to a little under 50 basis points. This was
the first test of like, "Well, what happens when you do this?" It actually ended up being
bad for consumers. Now, how could lower prices be bad for consumers? What kind of
monopolist am I? Well, it turned out that it's bad for consumers because the rewards
went away and it just ended up being a transfer of wealth between the big banks and the
big merchants.

Imagine that I'm buying a Coke for a dollar, if I now buy that Coke for 99.50 cents, I don't
really care about 0.50 cents. But you know who really cares about 0.5 cents is Walmart
because they sell like a billion of those every hour. So it's very, very hard to fix these
problems of concentrated benefit and diffuse harm, or figure out how you rejigger the
shares, if you will. So what really happened was that rewards went away in Australia.
That was very, very clear, there's no money to fund the rewards. Merchants basically
kept more money, but it wasn't a huge amount of money, huge, at least on a per
transaction basis. If you're buying something for $10,000, sure. Maybe now as a
consumer, it's like, well, 1.5% times 10,000 is real money. So I want to save money and
I'm getting a lower price of factors held constant. Whereas for like the $10 transaction,
the $20 transaction, it's basically just rewards went away, but still, it was popular, but go
Google like Australian rewards credit card, you won't find that many, or at least not that
many without fees. In fact, annual fees used to be a mainstay of credit cards in the US.
For you to get a credit card without an annual fee, that was a rare thing. I forgot who
was the first one to introduce the no-annual fee credit card, but now none of them in the
US, very, very few of them have annual fees. Australia, a lot of them have annual fees
now.

And then Europe did the same thing, but even lower, half the rate of Australia
interchanges, although not for all cards, like a small business cards, somehow that has
exempt from regulatory regime, so they can charge higher on a change there, but a lot of
countries and regions like EU being a whole region said, "Yeah, interchange is too high."
And look, they are partially right, merchants, they hate this stuff, it is a tax on their
commerce. In many cases, it's not enabling people to buy high-priced items that they
otherwise couldn't afford, that they didn't have cash for. It's like, "I'm just buying a tube of
toothpaste CVS for $4." And then like 10%...actually, this was one of the things that
backfired in the US around regulation, so banks with over $10 billion of assets are
subject to this Durbin Amendment, which basically allows the Federal Reserve to set
interchange for debit cards because they're like, "Look, this is a joke. You shouldn't be
able to charge 2 or 3% because you're not extending credit, you're not even forking over
money. There's no float. You're just taking the money from the person's checking
account and giving it to the merchant." So it's going to be five basis points, plus 21
cents, because there's a fixed fee. But wait a minute, if I buy a cup of coffee at Dunkin''
Donuts for $2, this is a real issue, the QSR, Quick Serve Restaurants, they hate this
Durbin thing because most of their customer base uses debit cards, not credit cards.

A $2 transaction means like 10% plus goes away in the form of interchange, and that is
the regulated form. So ironically, the non-regulated form is cheaper for them. Long
winded way of saying, there are a lot of countries that are experimenting with different
regulatory regimes under the idea that it's better for merchants, which is largely true, but
it ends up being a little bit worse for consumers, which I think was unexpected. You
would not expect again that lowering prices would be an issue.

But there's a second thing which I would call geopolitical risks, which I actually think is
much, much bigger because interchange will recap, but the Visa will always figure out a
way to make money, issuing banks will make money, like fine. But the geopolitical thing I
think is actually much bigger. I'll explain this by means of a story. So when I joined Visa,
this was right after Vladimir Putin had a next Crimea, which was once Ukraine, or maybe
I'll get in trouble if I say once, still Ukraine, or depending on which map you're looking at
is, either Ukraine or Russia. All these sanctions were passed against Russia, Russians,
but in particular, around this one geographically distinct area of Crimea. And if you were
a Ukrainian consumer and you had a Sberbank, that's the biggest bank in Russia. You
went to go use that card at a merchant in Crimea, it now wouldn't work because of US
sanctions that were implemented targeting this part of Ukraine/Russia, but there's a
company called Visa based in San Francisco, they could just change that routing and
say, "Nope, doesn't happen." It gets rejected. And this is really dangerous. And I'm not
making a normative judgment here, I'm just saying like, if you're a country, you're an
amoral country, and you're like, "Okay, I got to make sure that some other country
doesn't get me in trouble." Think about OPEC, where I'm the US, I don't create a lot of oil
or pump a lot of oil back then, it didn't have an oil surplus. And then OPEC doesn't like
me, and then cuts off oil shipments to my country, and now people have to line up, not
for vaccines, but for gas. And only odd days if your license plate starts with this, or even
days that... this was the '70s in the US, so what do we do? We build these strategic
petroleum reserve so that the government could help solve some of these geopolitical
risks. The economy wouldn't shut down if OPEC decided to cut off oil, but this is like
what we did to Russia. Again, I'm not making a right or wrong judgment here, but this is
what happened when Crimea gets annexed and the United States says, "No more
commerce for you in that region." And a lot of people don't use cash, they use cards. So
if I'm Russia, I'm like, "That's not a good idea. I got to go build my own network." But even
if I'm France, I'm like, "Wait a minute. I don't know if I like that plan. I got to go build my
own payment network as well." UK, "I'm good allies with the US but I got to go build my
own payment network."

And it just shows to me that networks, they are the new strategic petroleum reserve. It's
like China has Didi and the US had Uber, and Uber was competing in China. Imagine that
everybody in China was using Uber and only Uber, there was no Didi? Again, network
effect business, Uber wins. And then nobody owns a car anymore in China. And then the
President of the United States says, "Hey, Uber, we're going to shut off China." That
would be a disaster for China. So again, it's the same thing, but this is commerce.

This is what's so interesting for me is, countries never used to have to fear
this because it's they are the sovereign, they print the money, and the
money is the only legal tender. But the legal tender now happens over a
network. And the network is controlled by a private entity that is potentially
mediated by a government that's outside of your control.

So I think that's actually a big thing for, again, there's nothing to do about this, except
they might have to create wholly separate instances of the network that is under the
purview and control of the country.

The Merchant Struggle

Patrick: [00:41:50] Networks as a national security asset is incredibly interesting, we

Alex: [00:42:29] Well, yes and no. This goes back to companies always see what
strategically makes sense, but which there is no market demands. And the merchants
think it sucks. And they're correct, it does. I think I told you this. I remember looking at
the net income for Target and comparing it to their interchange fees. And basically, if
they were able to eliminate interchange, their after tax profits would double. It's crazy
how important this is. If you have 90% margins, like whatever, you don't really care that
much. If you have 2% net margins, this is a huge amount of money. So therefore, you put
a huge amount of resources into two things. One is suing them whenever possible,
these MasterCard are sued so many times.

They settled one of their last lawsuits $10 billion and the stock went up. You know
you're doing something right when you settle for only 10 billion for a company. So
Walmart was the biggest rah cheerleader, anti these MasterCards, again, as they should
be because they pay more in interchange than anybody. And they tried introducing
something called MCX, but it basically was meant to be this competing system. But the
problem is, again, how do you get the consumers to use it? Does every merchant want
to pay nothing? Yes, 100% hands or rates. Does any consumer want to link their bank
accounts to some new single thing that they don't really trust? And then why are they
doing it?

Well, they're doing it to get no rewards because you can get no rewards if the
interchange is zero. So Target has something called the Target REDcard, and you get 5%
cash back if you use the Target REDcard. And Target is very proud of this because
they're not paying the 2% of using a MasterCard, instead, they're paying 5%. Wait, wait,
what? That's not good. If every one of their consumers had to be bribed with 5% cash
back, that's not good. MCX was a complete disaster. PayPal try doing this offline as
well. And just none of these things either solve the consumer pain point. Consumer pain
point is one way of looking at it, but nor did they appeal to consumer grief. If you don't
appeal to solving pain or satiating greed, it's just really hard to say that you have a
solution. You have a solution to one side of the network, but just not the other side.

New Technologies and Visa

Patrick: [00:44:32] How do you think about how some of the newest technologies and
technology-based companies matter to Visa. I'm thinking here of Plaid, Stripe,
cryptocurrency, this bucket of new technology challengers that are building
businesses around the transfer of value, to what extent are these companies relevant
and interesting in the Visa's future story?

Alex: [00:44:53] I would ask you to start with a different group of companies there. I
think it's a good question, but I would start with Apple and Google as the most relevant
once there, because once upon a time, people would carry their credit cards, their ID,
their social security card in there, what I would call dead cow wallet. I got mine at my
bar mitzvah. I still have it today. I got my eight cards in there, I got my SSN, I got
everything. But all of that is clearly moving to the phone. I don't think anybody would
argue with that. It's just a question of when not if. And it moves to the phone, then some
weird stuff happens, which is, I don't know if you remember the Samuel L. Jackson
commercials for Capital One, like what's in your wallet?

Why would they spend tens of billions of dollars doing that? Well, because it wasn't even
trying to get new people on Capital One, it was like subconsciously, when I go to
Chipotle and I have five cards in my dead cow wallet, why do I pull up my Capital One
card? Which one is first? What is the dominant card? It's the power of the default. That
goes away because now it's software. I think Apple just alphabetizes it. So American
Express is going to do great. So it just changes the default paradigm that everybody in
this space has to play by. And also you can imagine, instead of getting junk mail to go
open up a new credit card, just there's an app store for apps that use my phone, what do
apps do? They rely on a system of permissions that are granted by the operating
system. So when you download a new app, it's like, "Hey, can app XYZ access your
contacts? Can it access your location? Can it access your photos? All photos or only
some photos? This is all about permissioning. That's going to happen, I believe with
financial services as well, so, "Hey, I want to get a new card, what cards should I get? I
don't know, I'm going to get Capital One, add it to my wallet. Done right then and there,
I'll make that the default." That's just a very, very different world. And then, because
Apple has a way of communicating, it can communicate with the payment terminal from
Verifone, there actually is a way of potentially bypassing these networks and you can
also have an interoperate much more easily because you don't have to carry around like
15 different cards.

This is the funny thing, is that the reason why you have Visa and MasterCard, is because
people are carrying around 20 different cards. One, MasterCard actually was originally,
it's like all of the retailers of the 1960s. I was like, "Hey, restaurants would have like a
little separate tab, why not have one cards unite them all?" Well, now you can actually go
back to a world where you have like 50 different cards, it's so easy to sign up, like, "Hey,
do I want to sign up for the Walmart card? Well, heck no, I get nothing out of it, and it
takes too long. Wait a minute, I just put an app on my phone and then it's there. And
then yeah, I get 1% cash back right away, or I get $5 off this item immediately right now,
but never get money off again. And that's my way of failing the marshmallow test, but
Walmart winning it. Totally, I'll do that because it's easy to permission me into the
system."

So I think those two are actually by far, they're going to have the biggest
impact, potentially changing things, because there's an app store right now
for apps that you play games and take pictures and do stuff like that, but I
think there's a whole other set of apps and permissioning that really aren't
even apps, they're more like financial products that you're going to discover,
utilize and sign up for on your phones. That's the group that I think is going
to have the biggest impact. But then to your question, I think Stripe is
extraordinarily interesting because they actually do issue cards as well.

They are one of the biggest acquires of merchants online in the world. If they become
one of the biggest issuers of cards, they could just self-clear and self-settle, and they
don't have to pass anything through Visa, and Visa will like them for it because it's like,
"Hey, don't switch to MasterCard, keep issuing your cards through our thing because
those cards might get used to Chipotle that actually does pass through our network." So
I think a lot of these that are able to become issuer and acquirer will have a much, much
better shot of pulling off what Chase somewhat failed to pull off, because Chase
honestly, should have run the table on online, acquiring merchants because they were
like, "Hey, wait a minute, we're 15, 20% of all Visa cards in the entire country, goes switch
from Stripe or Adyen, or Braintree or any of these things that your first data or whatever,
switch to us, and we're going to give you a 20% lower rate, which we can do because we
will charge you nothing for a Chase transaction." They would have gotten all the major
merchants, I don't know why they didn't do that, but they just didn't. Whereas Stripe, I
have a lot of faith that they could.

Patrick: [00:48:58] In the case of something like Plaid, which in my mind is this
fascinating business, which maybe is the closest to this protocol effect that we talked
about earlier, where it is the centralized communication hub between financial services
companies of all different kinds including banks, is a company like that potentially
important in this ecosystem as well even though it's very different from Stripe?

Alex: [00:49:18] It is. I'm an investor in Plaid, I certainly hope so. They don't really
compete with Visa or MasterCard at all. They're much more aggregation of data. From
Robinhood, I want to have you move money over from Bank of America, how do I know
how much money you have in your Bank of America account? Or maybe I'm Bank of
America and I want to hold on to you as a customer. I want to show you here are all of
your other assets everywhere, how do I suck that in? The metaphor holds very well in
that it's a network of other financial players, like Plaid sits in the middle and then you've
got Robinhood on both sides. You've got Robinhood as a reader and as a writer of data,
but it's not in the transactional space at all. It's not something where it's like, "I want to
go pay for something, I know I'm going to log in and authenticate with my banks." Maybe
that makes sense strategically, but it doesn't really solve a consumer value proposition,
going back to the thing that I mentioned before, which is, why would I pay with my banks
for my $2 coffee at Dunkin' donuts? Dunkin' donuts might want me to do that, and if they
launched their own payment mechanism, then Plaid might hook up there, but it doesn't
really solve a problem for me as a consumer. I want my rewards, I like to swipe, or I've
got this little app on my phone that I use, or I use my Apple Wallet to just go double click
and then pay. I think of them more as analogous to Visa but for the non-commerce
space, just for like banking information very large.

Patrick: [00:50:38] Final question before a couple of closing thoughts, which is the
obvious, crypto question. What makes this so interesting is that it's a protocol
business that's privately held, private for-profit corporation. And crypto is by definition,
a set of protocols. If you're rebuilding this from the ground up, would it look like an
open protocol like crypto seems to offer? How do you think about crypto protocols and
their role in this landscape going forward?

Alex: [00:51:05] I have a little tweet from that I wrote on this a while ago where it's like, I
think if you were to develop Visa from scratch today, and Dee Hock is actually still
around and he's like 94 and he's a crypto fan, partially, again, not based on some crazy
speculation around this will go up, and this will go down, it's because Visa was meant to
be this chaordic thing that nobody could control, but a lawyer figured out how to rejigger
that by turning it from a non-profit to a for-profit. And it turns out that protocol's trump
lawyers, and what does this mean? It means that you don't want to have the information
centrally stored anywhere.

The problem was that all the information was centrally stored in Foster City, California.
And that's where the router was, that's where Visa net was located. So if you just take
control of that, you can say, "Oh, okay, now we're a for-profit we're going to charge
everybody 7 cents, and you just got to go pay it." Versus if this were done from scratch,
you would say, "Okay, all the banks are nodes that hosts the ledger. Nobody can control
it, no lawyer can figure out how to undo that. That's just how this would work." That's a
very, very different line of conversation than like crypto will somehow supplant Visa,
which again, I think could it? I don't know. I think it would have to solve a real consumer
problem.
And right now, the fact that most crypto goes up in value and therefore is deflationary, at
least today, it goes up and down in value all the time, it's very, very hard for people to
want to spend a deflationary currency. There's a whole set of issues like that around
whether crypto is a threat to Visa, but if you were to rewind back to 1958 and say, "Okay,
how do we turn this amazing network effect and have the thing in the middle not be
worth more than the things that created it, namely the banks, the way to do that would
be to have no centralization and have it be a purely decentralized protocol. And now we
have the tools for that, but again, it's 60 years too late for Visa.

Takeaways

Patrick: [00:52:53] My closing question to you is really one lesson to take away, one as
an investor, and one as a business builder, you've been both an entrepreneur and
investor. And in Visa, we have one of the most unique business stories ever, one of the
most unique, sustainable competitive advantages ever. So this may be a hard question,
but maybe we'll start with the business hat on. As an entrepreneur studying Visa and
having worked there yourself, what one business lesson do you take away from Visa?

Alex: [00:53:19] I think it's the value of group participation, a lot of times businesses that
get to a network effect, they start off with some comically small example. Facebook
started off with the comically small example of Harvard students. But once you really,
really get going there and you've got clear product market fit, then you concentrically,
concentrically, concentrically expand. Visa did the same thing in a very non-traditional
way, but it's like, "Okay, Bank of America, we got Fresno, we got Florsheim Shoes and
everybody else, we're dominating commerce there, then go to Fremont, then go out, go
out, go out, go out. I think it's the story of honestly, every network effect business, but it's
also starting off with the right constituents, because if you start too broad, you never get
that network effect.

And almost all of the new thing, or the attempts at building a payment network, it's
number one, it didn't solve a problem. That's bad. And then number two, it went too
broad. It's like, let's go boil the ocean, like Google launched Google Checkout, Google is
smart, they're going to dominate the world. They launched Google checkout in 2006,
complete disaster. It had no purpose, didn't really solve a problem. Really smart people
that worked on it, and Google was smart to want to build a wallet and a checkout
process, but it just was too broad and didn't have the narrow-down scope. It's just not
fun if you're a big company to do a narrow-down scope, why would you ever do that? So
I think it's narrow your scope and expand concentrically is the lesson for me.
Patrick: [00:54:43] And how about as an investor? And maybe it's the same version of
the same lesson, but as an investor thinking about what Visa has taught you, anything
different of note there that you would close with?

Alex: [00:54:52] It's like a corollary to the Warren Buffett-ism of like, when a
management team known for brilliant meets a market known for disaster, it's the
reputation in the market that remains intact. It's the opposite here. It's not saying that
the management of Visa are bad, they're good. It's more of like, it is such a powerful
network effect that it's just very, very hard to disrupt. So it's more helpful to think about
the long-term issues of, to me it's geopolitical risk, it's fragmentation versus
concentration, as opposed to the ankle biters or like the Oh, PayPal launched offline, isn't
that the death of Visa? It's like, no, what is their plan to maintain this in the next 10
years? Is probably the more relevant set of questions to ask, as opposed to who shifts
the press release last week saying that they now have a potent challenger to Visa.

Patrick: [00:55:35] If there was one resource you could point people to if they're
interested in studying Visa more, is there anyone that pops to mind?

Alex: [00:55:41] Yeah, there's a great book called A Piece of the Action. It's like 30 years
old, but it's super, super relevant. It's by Joe Nocera. So that's the best one. Dee Hock
actually wrote a book called One From Many, kind of makes sense. Basically is the
summary of this podcast, One From Many. It's like you have one amazing company that
came from many, and that's also really, really good. Those two would be the best.

Patrick: [00:56:02] Well, Alex, this has been an absolute masterclass in such an
interesting business. Thank you so much for breaking down Visa with us.

Alex: [00:56:08] Absolutely.

Patrick: [00:56:08] I hope you enjoyed this breakdown of Visa with Alex Rampell. The
point that Alex made that resonated with me is that once you build a standard way for
lots of companies to communicate information with one another, a protocol, it's
incredibly hard to change or disrupt. There are not many sustainable competitive
advantages like a protocol effect.