Max Levchin On Being ‘Employee Zero’ At PayPal
Affirm CEO Max Levchin recounts working alongside Elon Musk and Peter Thiel at PayPal, shares the best (and worst) leadership advice he ever received, and addresses “phantom debt” tied to the buy-now, pay-later industry. This interview was recorded on May 9, 2024.
Category
🤖
TechTranscript
00:00The reason United States consumers have $1.1 trillion in current outstanding credit card debt
00:05is because there's no real motivation on the part of credit card issuers to tell you,
00:09hey, you got to pay this thing off. In fact, that balance is sitting there revolving,
00:13which means that the interest accrued every day, whatever the accrual period is,
00:18goes right into the principal and just spins up and spins up on itself. I have infinite conviction
00:23that if U.S. consumers would just switch entirely to a firm, and not just us, although I'm
00:29biased, but I think we're the best, if consumers would switch to using buy-now-pay-later products
00:35instead of credit cards, we would be in a healthier financial position as a nation.
00:42The Chronicles of Max, the very abridged version. I'm not as young as I used to be,
00:47so there's a lot to cover in a brief period of time. I was born in Kiev, Ukraine,
00:51back when it was part of Soviet Union, in a secular Jewish family, which did not prevent
00:57it from being on the receiving end of a bunch of anti-Semitism, which unfortunately has not
01:02gone away in today's day and age, but at the time was maybe even more of a thing there.
01:08And so in 1991, my family immigrated to the United States as refugees to the wonderful city of
01:16Chicago, which I consider home. Probably when people ask me, where are you really from? I say
01:21Chicago and then reluctantly admit to having not been born there, but I consider Chicago.
01:26The roots and the formative period of where I really am. Went to the University of Illinois
01:30at Urbana-Champaign, which was the optimal intersection of state and therefore least
01:35expensive school and the best computer science education immigrant money could buy,
01:39which was extraordinary luck because I matriculated in 1993, which is the year
01:45Mosaic Browser, which became Netscape, which gave birth to this whole web thing we're all in it now,
01:52was built on my campus. So I was there overlapping with the likes of Marc Andreessen and sort of the
01:57entire Netscape gang. And my path towards PhD in computer science, which is certainly what my
02:03very education-minded family wanted, was derailed and for a while terrified them, but it all worked
02:10out. I did not drop out of college, even though I really wanted to. And my family intervened and
02:16said that my sainted grandmother would die if I did this. And I decided to finish my undergraduate,
02:23which makes me the least educated person in my family. But I did get a bachelor's in computer
02:27science and promptly moved to Palo Alto to start a company, which turned out to be PayPal. Co-founded
02:31that with Peter Thiel, merged with another company called X, which you can guess by the name, was
02:37founded by Elon Musk, which makes us the three amigas of the PayPal saga. Was there until the
02:43acquisition by eBay. Did not last very long in corporate captivity, but it was a good deal for
02:51all involved. And PayPal obviously ended up being culturally significant and economically important.
02:56Went on to start a bunch more companies outside of payment space, because I thought that how can I
03:00possibly beat my freshman album with any sort of financial services company after. So I wanted to go
03:07and do anything but financial services. And they all worked out. Some worked out better than others.
03:12Sold the company to Google, spent a year working at Google, which was a fascinating experience.
03:16But ultimately just felt like my calling was actually back in that memory of getting screwed
03:22by a credit card on campus at University of Illinois. And so somewhat reluctantly decided I'll
03:28incubate a financial service company, but maybe I'll let someone else run it.
03:31And then six months in, I was running a firm and that was 13 years ago.
03:36I was in Palo Alto in an on-air conditioned room that I was bumming a very small part of from one of my
03:43college friends who had moved to Palo Alto a year before. And it was very hot in the summer of 98.
03:48So I went to listen to a lecture at Stanford during summer school, which was open to the public.
03:54And it was given by this guy, Peter Thiel. Afterwards, I chatted him up and just introduced
03:59myself and said, we have a common friend. In fact, two common friends. And he said,
04:03we have a common friend. In fact, two common friends. Oh, that's awesome. What are you working
04:07on? I'm not working on anything, but I'm definitely going to start a company. That's amazing. You
04:12should start a company as soon as possible. This is what this whole place in time is all about.
04:17Let's have breakfast tomorrow. Tell me about your company ideas. So I met him the next morning for
04:21breakfast at Hobie's. He had red, white, and blue smoothie. I had an omelet. And I pitched him on two
04:26different ideas. One is lost in the oceans of time, and the other one was PayPal. And it was
04:33not PayPal as we see it today. It was not even remotely similar, but it was the thread on which
04:37the two of us pulled. And initially, on the spot, Peter said, I would like to write you a $300,000
04:43check so you can go develop this idea. And I walked out feeling like this could have never
04:49happened to me in central Illinois, where I went to school. I'm so glad I moved to Palo Alto.
04:53And then maybe 12 hours in, I have no idea how to run a company. I should get that guy to run it,
04:59and I can write the code, which is how he became the chief executive, and I'm the chief technology
05:03officer. The very original idea for PayPal. So I was really, really into security. In fact, a lot
05:09of my various projects throughout my life, both professional and personal, are all around things
05:15like personal privacy, cryptography, security, data security, underwriting. The math that goes
05:23into underwriting, which is what we obviously have to do at a firm really well, is very similar. A lot
05:29of linear algebra, a lot of number theory, machine learning, even what these days people call AI,
05:35all of that is kind of in the same bucket of math that I love studying and working on. And
05:40cryptography is one of the fundamental applications. So at the time, handheld devices,
05:46handheld computing was just sort of taking shape. Newton had just failed, and PalmPilot was just
05:52rising. And so I pulled out my PalmPilot, and I told Peter, I think this thing is going to be
05:57profoundly important for all kinds of tasks. I don't know exactly what they're going to look
06:01like, but you're going to have your corporate documents in this thing, and you're going to have
06:07important letters that are going to be very secret, and no one's thinking about security.
06:11I want to encrypt it all. We should have a cryptographically secure PalmPilot. And that
06:17seems really visionary. I don't know if you're right, but that seems right. That seems like a
06:22good idea. So that was the very beginning. And then within six months of sort of trying to convince
06:27anybody else to believe it, sort of realized, all right, this is probably not, it may be well ahead
06:31of its time. In fact, maybe ahead by, turned out to be two decades. But in the process, he and I,
06:38and a bunch of other co-conspirators that we had pulled in, he from Stanford and me from University
06:43of Illinois, had sort of become quite infatuated with the idea of security inside handheld devices
06:51for some purpose. What can we use this class of ideas for? And not me, someone came up with the
06:58idea of like, well, what about IOUs? What about writing down on a piece of paper, I owe you $10?
07:04It's insecure. Like I can forge your signature, but what if you had cryptography and security in
07:08that? And so we built a secure IOU between PalmPilot. And that was fascinating for like
07:16five people in the world, but we really loved building it. And it was really cool.
07:20On the side, I built a demo that showed how this would work just in a webpage,
07:24because not everybody had a PalmPilot. I think PalmPilot peaked with 4 million devices sold
07:28or something along those lines. And there were already more than 4 million people on the web.
07:31So I thought, well, I should have something that looks like this whole exchange of IOUs,
07:36but just in a webpage, just for a demo. And so I left it be, and then I looked at weblogs,
07:41and it was getting used by a lot of people. And first thing that I wanted to do is shut it off
07:47and make sure that it's not, because it's insecure. I didn't really think deeply about
07:51security for that product. But when I shut it off, I got lots of angry emails from people saying,
07:56well, the hell are you doing? I'm using this to sell stuff on eBay. And that's when we
08:00hit the product market fit, where even though it was not on PalmPilot, it was on a webpage.
08:04It was, to me at the time, painfully boring. I tried very hard to convince Peter and a bunch
08:09of others that this is just the stupidest thing. People on the web using this to buy
08:13tchotchkes on eBay, what are we, like, this is not a noble pursuit. But between the time I started
08:19convincing him and the time we actually decided, hey, we should do this, it grew like 10X.
08:22And so it was very clearly, the market pool was strong. That's what PayPal became.
08:28Affirm is a payments network. We help people buy things and pay over time without gimmicks and
08:35late fees and all the other junk that the financing industry is famous for.
08:39I thought there was a better way to do payments and lending going back more than 30 years,
08:46when I moved to the US as an immigrant and got my first credit card on campus and promptly got
08:51into a bunch of debt and paid a lot of late fees and had my credit score go down and all the bad
08:56things and sort of stayed with me where I thought I was super undereducated, didn't really understand
09:04personal finance at all and certainly meant no harm. And once I got that cleared up, I can't
09:09think I've been late for any kind of bill ever since. And my credit score wouldn't recover for
09:14a decade. And borrowing was really difficult. I tried to buy a car after I took my first company
09:21public and was told that I had to pay cash because my credit score was too low. So I decided
09:25that some decades later, I should go and build a better payments platform that had accommodations
09:33for people that didn't understand how lending worked and just optimized in favor of treating
09:38them right and making sure they don't make mistakes. And if they do, wouldn't take advantage of them.
09:43We had a great fiscal Q3. So we grew another 36% of GMV, which is just the overall spending done
09:57with a firm year on year. Our revenue grew 51% year on year. So we're doing really well. Rumors
10:05of American consumers' inability to pay their bills or lack of interest in shopping are greatly
10:10exaggerated. So we're still growing really well. People are paying our bills well on time. My
10:16favorite stat, which is not a stat at all, but it's an important thing to repeat because people
10:21still don't believe us 13 years in, the total number of late fees or compounding interest that
10:26we charged last quarter was zero. And it's been zero since inception. We've never charged a penny
10:30of late fees and have no intention to, or any other hidden or gimmicky fees. And that is a
10:37profoundly important thing, not just because it's a good thing for consumers, obviously. You don't
10:41want to take advantage of people when they're struggling to make their bills. But what it does
10:46at a more fundamental level, it aligns us with the customer success. If someone borrows money and they
10:51can pay us back on time, that's great. Good for us, good for them. They get the value of the service.
10:57We are able to book revenue. When they're behind, there's nothing there for us. That means that
11:03whenever we made an underwriting decision, we had to have been more thoughtful. We should have
11:08seen something about their inability to stomach this kind of bill. And so that's an error for us.
11:12And they're not in a good place, but neither are we. And so we've learned from that and we stay
11:18aligned with their financial success. A firm is better for the consumer than using a credit card
11:23for a bunch of actually really good reasons. Buy Now, Pay Later is the fancy four-letter acronym
11:29for the industry. I don't particularly like it, but that's what the world seems to have settled on.
11:33Credit cards are Buy Now, Pay Forever. When you use a credit card, you just load up your balance
11:39and the credit card companies, banks that use your credit cards, tell you, hey, here's the minimum you
11:45need to pay and the rest is up to you. You figure out. So it seems very flexible. It's a bit of a
11:49power tool. It's a power tool with safety off. If you're not careful, you're literally going to stay
11:54in debt forever and ever and ever because minimum payments aren't designed to get you out of whatever
11:59balance you've created. They're in fact designed to kind of keep you sitting there and compounding
12:04interest. A firm has a prescribed schedule for every transaction we underwrite. So whenever
12:11you use a firm, you see visually a schedule of payments that you will have to agree to. And
12:18there's no minimum or maximum payment. You can always prepay and there are no penalties. In fact,
12:22we reward you by subtracting the interest you would have paid, estimating exponential functions,
12:26which is what interest accrual that compounds really is. And so we wanted to build a product
12:32that is not hard to understand, where the schedules are simple, where you know you're in sense of
12:36control. If you're late, we will remind you. We will be a little annoying telling you like, hey,
12:41you're behind. You really got to pay your bills, please. But we won't charge you a penny extra.
12:45And that alignment is really important for us from a business point of view, but it's also
12:49just fundamentally healthier for the consumer. The phantom debt, which is catchy, I have to admit,
12:55but pretty erroneous, is an interesting new meme, but it's just inaccurate in the following sense.
13:02So for a firm, and we're a little bit different than others, but we do quite a lot of volume in
13:08longer terms. So BNPL typically, at least to me, means paying for, which is pay over six weeks
13:14in four installments. And so all of those transactions, everything that is paid on a
13:19monthly cadence, is already reported to the credit bureaus and has been since the beginning
13:25of a firm's time. And that helps your credit score when you are paying it off on time and well,
13:31and it certainly does not help it when you are behind. And we try very hard to work with
13:36consumers to make sure that they understand it and to make sure when we possibly can, we help them
13:42navigate personal hardships and things like that. That's a well-understood part of our business,
13:46and it's the majority of our business. Paying for is typically by most of the industry, except for
13:53Apple, which with Apple Pay later just led the way and showed the industry that they can report it
13:58to Experian, is typically not yet reported. There is industry-wide support to make paying for a
14:07reportable type of transaction. The reason for it is the credit bureaus are not currently able
14:13to ingest that information in a way that would correctly impact your score with good and bad
14:20behavior. There are real challenges for them understanding exactly how to work that into
14:24the various credit scores that they generate. That is most certainly on its way to getting
14:29fully aligned and fixed, and everyone that I speak to anyway in the industry is very supportive of
14:34the idea. But just for context, the scale of paying for specifically, last year it was about
14:41$50 billion of total volume across all competitors in the US. The average lifespan,
14:47weighted lifespan of such a loan is three weeks. So the six-week total, roughly half life is three
14:53weeks. That means $50 billion worth of total originations is on the order of $3 billion of
15:01actual outstanding balances. Put that into context of $1.1 trillion of US credit card debt on a
15:08consumer side. So that's just under 30 basis points. So it's so phantom, it's less than one
15:15half of 1%. If you actually do the math precisely, you'll see that it's even less than that. It's more
15:19like 25 basis points. And so it is not a significant component, not a significant weight on the economy.
15:25That said, through the consumer credit building lens, it's really important that it becomes a
15:31reportable type of transaction, and it will certainly happen. And in fact, with some of
15:36the industry players already. Leadership advice, unfortunately, is the good stuff is really obvious.
15:52It's a lot less about here. No leadership advice worth following will make you say,
15:58oh wow, that's such a novel idea. Vast majority of them are painfully obvious. The hard part is
16:03actually following them consistently. So back to sort of the PayPal story. When I was younger and
16:10significantly more obnoxious, I asked Peter, what are you good at? And he was sort of startled and
16:14said, you know, I'm a really good writer. I said, that's the weirdest thing. Like, why does that
16:18matter? He said, I think it really matters. You have to communicate your ideas. It's really important
16:21to be a good writer. He's a really good writer. He's always been an exceptional writer, but he's
16:25an extraordinary communicator. And when I watched him run PayPal, I thought somehow it takes him
16:32one-tenth the words to convey the ideas that it takes me. And people walk out with a very clear
16:38plan of action, and they're not tired. And when I try to deliver the same speech, it turns out to be
16:45this really detailed thing that really lays out the plan, but people don't seem to remember the
16:48beginning because it takes a day long. And so the best advice, the best observed thing I have to
16:55offer on leadership is just over invest in being a crisp communicator. And if you're not,
17:00then practice and practice and practice and practice. And it's really difficult, especially if
17:04your natural MO is to sort of be really thorough. So the old Fermat's line about, I could have made
17:13this letter shorter, but I didn't have enough time. Editing yourself is the hardest thing in the world
17:18because you love every word you put on paper or about to say. These are your thoughts, your words,
17:24but actually nobody cares about your words. They care about your ideas. So drilling them down,
17:28distilling them into a pithy set of, we are going to do this, and this is why it's important.
17:32And here's what it's going to take. Let's go is a much better way than, well, you know,
17:38first the earth cooled, which I certainly fall back into that pattern way too often.
17:44The worst piece of advice I've ever received.
17:50You should hire a professional chief executive officer. That is the worst advice I've ever
17:55received. I have received it over and over and over. I think a lot of investors in the early days
18:01of young entrepreneurs will tell them, hey, you're a grade zero to one guy or girl,
18:08got to get it to this level. And then, you know, it becomes all about operating and about just
18:14allocating capital and dealing with human resources. And it's really not what you're
18:19strong at. And you should just make room for a pro. And pros are not exactly born 45 with a
18:26nest in their chest. Like they are forged through years and years of trying and failing and getting
18:31bad advice and figuring out good advice. And some people are really good at it. And some people are
18:35not really good at it. But as a founder, the moment when you sort of ask yourself, well,
18:40am I still good enough? Is my team outpacing me? Is my group of people, group of co-conspirators
18:47now in need of a better leader? It's an important question to ask yourself, but the answer isn't
18:51always, yeah, like they're now in need of a pro and I need to recede into the background and just
18:56be an individual contributor, you know, write some code every once in a while. Sometimes the answer
19:01may be, okay, yes, it's time to hand off the reins to someone else. Sometimes the answer is like, no,
19:06I'm going to push myself to grow with the team and be a worthy leader for the cause and for this
19:12group of people. So at some point or another, I decided that as much as I love writing code,
19:17as much as I love thinking about big mathematical puzzles and things like that, I love leading teams
19:24towards their goals. I made the choice to become a professional CEO and here I am now. We'll find
19:29out at the end of this journey, whether I am better or worse than the pros that I've been
19:32to hire over the years, but I think taking that advice as a received wisdom is a terrible mistake.