Marty Chavez
Investor, Entrepreneur, & Technologist
My mom would say that this is the only way to get things done, that you’ve got to be highly organized and highly disciplined… in retrospect, I know I get exhausted just thinking about that level of organization. But it was my mom and dad that really created that priority.
Summary
On this episode of “Leadership Matters,” Marty Chavez joins Alan for a rich discussion about his pioneering career on Wall Street and in Silicon Valley. Formerly the CIO and CFO of Goldman Sachs, the current Sixth Street Partner and Vice Chair had a wealth of experience to share with Alan.
Marty dives into his early life and explores his formative years in Albuquerque, New Mexico, discussing life around the house and the values instilled in Marty by his parents. Alan and Marty have a rich conversation about the latter’s education at Harvard and Stanford and his initial ambitions of becoming a doctor. They discuss his journey towards Silicon Valley and the beginnings of his career as a computer scientist, before speaking about the chance recruitment offer which brought Marty to Wall Street.
In their conversation, Marty and Alan discuss many of the critical moments from throughout Marty’s early life which led to his later success at Goldman Sachs and Sixth Street.
Mentions & Resources in this Episode
Guest Bio
R. Martin (“Marty”) Chavez, Ph.D. is widely renowned as a trailblazer and leader who turned the Wall Street trading business into a software business, revolutionizing the way that capital moves and works. He is a partner and vice chairman of Sixth Street Partners, where he works on research and development; diversity, equity, and inclusion; the sourcing engine; and the More Than Capital business, driving deeper engagement with the portfolio companies. Before joining Sixth Street, Marty served in a variety of senior roles at Goldman Sachs, including Chief Information Officer, where he oversaw the firm’s 9,000 engineers; Chief Financial Officer; and global co-head of the firm’s Securities (now Global Markets) Division. Marty was also a partner and member of the Goldman Sachs management committee.
Marty has achieved singular acclaim in the financial-services industry for his work on SecDB, an early platform that transformed the trading business into a software business. He is also known for bringing the front and back offices together. By training, he is a computer scientist who successfully advocated for the elevation of engineers. Yet Marty is also a highly regarded investor in his own right, leveraging his background in machine learning to push the industry forward.
Far from the stereotype of a banker, Marty is a disrupter at heart. He was among the most senior Latinos on Wall Street, as well as the most senior openly gay executive at Goldman Sachs. In 2016, a New York Times profile described Marty as “a departure in sensibility from the buttoned-down partners of Goldman lore.”
Adept in both Wall Street and Silicon Valley, Marty co-founded San Francisco start-up, Quorum Software Systems. Later, he was CEO of Kiodex, a New York risk management systems company that SunGard Data Systems acquired in 2004.
Beyond finance, Marty has long held a passion for converging the life sciences and software, and has an eye for new applications of AI and technology that will transform industries. Since retiring from Goldman Sachs, Marty serves as an advisor and board member to multiple startups and projects that are accelerating breakthroughs in their fields.
Marty serves on the Board of Directors of Alphabet Inc. (NASDAQ:GOOG). In addition, he chairs the Board of Directors of Recursion Pharmaceuticals Inc (NASDAQ:RXRX), a digital biology company industrializing drug discovery by harnessing the power of cloud-based machine learning models. He advises numerous other companies, including Abacus.AI, an AI startup developing new approaches to deep learning; Cambrian Biopharma, a distributed drug discovery company developing medicines to extend healthy lifespans; Earli, which delivers new technologies for identifying and localizing early-stage cancers; and Ketch, which delivers responsive infrastructure for compliance and data security.
Marty serves on the Board of Directors of the Broad Institute of MIT and Harvard, the Stanford Medicine Board of Fellows, and the Board of Directors of the Los Angeles Philharmonic. Previously, he served on the Board of Overseers of Harvard University (and as President for the 2020-2021 academic year). He also served on the board of directors of Grupo Santander; the Institute for Advanced Study, the International Swaps and Derivatives Association (ISDA), Paige.AI, PNM Resources, Inc., and Sema4. A passionate patron of the arts, he has served on the boards of the Friends of the High Line, amFAR (the Foundation for AIDS Research), and the Santa Fe Opera. Marty resides in the Berkshires, and is the proud father of two.
He holds an A.B. (1985) magna cum laude in Biochemical Sciences and an S.M. (1985) in Computer Science from Harvard, and a Ph.D. (1990) in Medical Information Sciences from Stanford (Architectures and Algorithms for Probabilistic Expert Systems).
Clips from this Episode
Episode Transcription
Alan Fleischmann
You’re listening to “Leadership Matters” on SiriusXM and leadershipmattersshow.com. I’m your host, Alan Fleischmann. I’m here today with an inspirational investor, entrepreneur and technologist.
Marty Chavez is a Wall Street veteran who revolutionized the role of computer science and engineering in finance. Marty emerged as an industry leader during his two decades at Goldman Sachs, where he held a variety of key roles including Chief Information Officer, Chief Financial Officer, and co-head of the firm’s securities division. Marty is also an esteemed Silicon Valley founder and executive. He launched Quorum Software Systems, and later served as the CEO of Kiodex, a New York-based risk management company. A longtime private investor, he now serves as a partner and vice chairman at Sixth Street, a leading global investment firm.
Despite his stature in the financial community and influence as one of the most senior Latino and gay executives on Wall Street. Marty remains humble to his core. He is a dear friend, and he is truly the renaissance man that I know. I’m thrilled to have him on the show today to discuss his amazing life and career. And also to get some of his wonderful advice for our great audience.
Marty, welcome to “Leadership Matters.” It is such a joy to have you on with us today.
Marty Chavez
Thank you, Alan, I certainly don’t merit even half of that. And I just have to note that, in leading the revolution of engineering, software, and math on Wall Street, I was one of many leaders and definitely showed up at the right place, at the right time, and with the right preparation and built on the work of many people. I really worked with thousands of people to make that happen.
Alan Fleischmann
Well, you’re credited with being a great leader and a great team player, as well as an inspirational force that makes other people want to join you. So, the fact that you had a movement behind you to make that happen does not surprise me — that humility again. It shows me that it takes a village, but you were able to build that village, which is impressive, and we’re going to get into that, too.
Let’s start with your early life. You grew up in Albuquerque, New Mexico, and have four siblings. You’re the oldest. You have an amazing story with your family and your parents, and I’d love to get a little bit of a sense here: could you talk a bit about what life was like around the house?
Marty Chavez
It was intense. I think my brother had a great description of growing up in the Chavez household. He said it was a startup. And that’s probably the best way to think of it. My mom and dad had a crazy vision. And we were all just living into that vision. It was highly organized by backward induction. And by that, I mean that it starts with the family legends and maybe a vision as well.
My mom, the story goes, at 10, made a vow that she was either going to be a nun or have 10 children and send them all to Harvard. And my mom, being a tough grader, would give herself a 50% because she ended up having five children and sent them all off. So I don’t know when Harvard started showing up in our mental images, but it was pretty early on. And so if you’re growing up and barely hanging on to the middle class, in Albuquerque, New Mexico, where no one’s ever been east of Dallas, let alone all the way to the heart of the Yankee universe that’s Cambridge, Massachusetts, this is a really crazy idea. One where there’s a lot of work that needs to happen. So it’s a highly improbable outcome. But that was it.
I remember there was a lot of love. There was a lot of education and music. There was also a lot of religion. I could have skipped that part. But for my mom, that all went together, in no particular order. If there’s a central image that I want to leave you with, it’s going to school. In the morning, my mom would line up the boys in the bathroom, and in one second she would comb each boy’s hair the way it needed to be combed so we looked the part. And I didn’t necessarily want my hair combed that way, so that objection was noted. It seemed really intense at the time, but I appreciate it.
My mom would say that this is the only way to get things done, that you’ve got to be highly organized and highly disciplined. After all, criminals started by spitting on the sidewalk. So there was very little tolerance for anything other than what would lead to the successful outcome. And in retrospect, I know I get exhausted just thinking about that level of organization. But it was my mom and dad that really created that priority.
Alan Fleischmann
Where’d they’d get that from, Marty? To say, “I’m going to raise five children and they’re all going to go to Harvard.” It’s stunning. And then to realize it. I think you’re the only family who has ever done that, and Harvard recognizes it. What gave them that kind of grit and determination? That vision, and also the confidence to say it?
Marty Chavez
I think causality is hard to demonstrate. And genuinely, we’re just making up stories after the fact that explain what happens. So here’s my story. It’s as good a unifying story as any, which is, Albuquerque is a very interesting place. I realize that now. At the time, I thought it was the most boring place in the universe. But I now really appreciate its oddity and its eccentricity.
So, what are some of the strange things about it? It’s a liberal, mostly Democratic oasis in a part of the country that tends to be much more conservative, more Republican, surrounded by Colorado, Arizona, and Texas. So that’s the strange thing about it. And another strange thing about it is that it is a microcosm of the Hispanic world: it is heavily Hispanic — something like 75% — but, even inside Albuquerque, there are the kinds of things that regrettably happen in the Hispanic world: the tyranny of small differences. So we’re all Hispanic, but there are some Hispanics who’ve been here for a long time — since the 1500s, allegedly — and others who have more recently arrived. So there’s a kind of discrimination based on how long your family has been there.
There was also the reality that we were definitively second-class citizens in a state where we were technically the majority. And so everywhere I looked, if I looked at the power structure — at the doctors or lawyers, the engineers at the National Laboratories — they were all Anglo-Saxon. So that was something that you just couldn’t miss growing up. So we had an intense underdog mentality, and my mom instilled in us that desire, for better or worse, to belong to institutions where it wasn’t clear that we were all that welcome.
I went to a fantastic prep school in Albuquerque, a school with some 500 students. There was a tiny number of Hispanic kids — you could count them on the fingers of one hand. So we started off thinking that we were way, way, way behind in this race. Because it is a race, and we were going to win. But while we didn’t have any money and we didn’t have any power, we could get an education. And so the value of education was just drilled into us from very early on. My mom made sure that we could all read and write and do addition and subtraction and multiplication up to 12 before we were even in kindergarten. I was aware, even in kindergarten, that I knew things that the other kids around me didn’t know.
It really accelerated for me from there. I was good at math from a very tender age. And schools recognized that, and they accelerated and pushed me further; by the time I was 11, I was taking math at the University of New Mexico. As my mom would say, she’d say that she didn’t necessarily start off with this vision, but that she had a little bit of it. I was good at math, and that led to a self-reinforcing virtuous cycle. Some teacher early on said, “your kid is a really, really great student, he could go to Harvard.” And that became the mission.
Alan Fleischmann
Tell me about that fact — you were a 12-year-old at a university, in essence. I think you’ve said that the computer lab was your first love?
Marty Chavez
So I was 11 and I was good at math. My school, Albuquerque Academy, had skipped me from fifth to seventh grade. Then in seventh grade, I remember the math teacher saying you really don’t need to be here. He didn’t think I really needed to be there for eighth through twelfth-grade math, and that I should go take calculus at the University of New Mexico. So I remember taking the bus downtown and sitting in the very back row, always. Look, I was an 11-year-old kid with mostly 18-year-olds. At that age, seven years is a great gulf. So I didn’t really make friends, just went and did the lectures to do the problem sets.
I remember the first day of class — I can still picture a computer printout that I was handed, which had a little sticker on it. The sticker had a username and password for the IBM 360 time-sharing system at the UNM Computing Center. Some professor, Clete Miller — who’s a legend in his own right and was one of the co-authors of MATLAB — had this idea: let’s give all the kids a Computing Center account. That was not an obvious thing to do. And most of the students didn’t do anything with it. But I thought, “I’m going to go check this out.”
So this is roughly 1975. I went there and there were punch cards. Right at that transition point, there were punch cards for just a year or two, and then all the punch card machines got taken away and replaced by what we now see now as terribly old-fashioned teletype writers. I taught myself to write computer programs and submit them in batch mode on this time sharing mainframe, and I taught myself some of the early programming languages: PL360, which I think nobody remembers, PL1, Algol, Algol-W. Then eventually Pascal.
I eventually persuaded my parents that this was a fine activity and my parents really embraced it. I remember my dad telling me at this very tender age, putting an arm around my shoulders, just like our version of that famous scene from the movie, The Graduate. And he said, “Martin, computers are going to be huge in the future.” This is 1975, so it was a non-obvious statement. “And you’re going to be really good at them.” In the extraordinary discipline and structure that my parents created for us, the one area where they let me run wild was in the walled garden of the UNM Computing Center. My father would take me there all day, and night. I’d pack a sandwich and would just go down to the Computing Center. As long as they knew I was there, then they were fine. And so for me, writing software equals freedom. And I still feel that way about it.
Alan Fleischmann
I love that actually, I think that’s so cool, that that’s where you actually found your first love.
Marty Chavez
Yes!
Alan Fleischmann
But also, part of that first love was the intellectual journey that you launched, that renaissance man who I described. Also the freedom part, I can’t imagine being there at 11-years-old. I can’t imagine how Intro to Computer Science could be my first love, but the fact that it was for you speaks volumes about how you’ve managed to take the most technical and complex matter and make it simple for others to understand.
Marty Chavez
It was an amazing time — I just have to emphasize being in the right place, at the right time, with the right preparation. 1975 was the early days of a piece of software coming out of Bell Labs with the odd and geeky sounding name of Unix. I remember some UNM students, maybe in their early 20s, who were the early hackers at the University of New Mexico. They would tell me about this Unix operating system, and they’d say: that’s the future, not this IBM 360 thing that we’re all using. But Unix is the future, and we would figure out ways to get access to Unix machines.
That’s when I first heard that there was this thing called the ARPANET, and that the Defense Advanced Research Projects Agency was building a network of computers that would resist a nuclear war by being highly distributed through being packet switched. So there was no circuit to communicate, to connect one computer to another, but they were still all connected. Any packet could travel any route there over that network, and they’d be reassembled in the right order at the other end. There was a DoD weapons lab, the Air Force weapons lab, and it supposedly had a node on this nascent network. I remember doing whatever it would take to get that phone number and they would keep changing it.
My dad would bring a computer home from Sandia laboratories where he worked, and it was a TI Silent 700 with an acoustic coupler in the back. I’d tie up the only phone in the household for hours and would have the Silent 700 on the kitchen counter. Whenever the dishwasher would go on, there would be noise that would go through the coupler, and it would mess up the program I was typing.
But I was on the ARPANET, which as you all know, is the predecessor of the internet. And I connected to a group of people who call themselves hackers at MIT, which sounded like the coolest place in the universe, infinitely far away from me. But at the age of 11, I was part of this community of people who were just writing software for fun. And I encountered this wild guy named Richard Stallman, who had this insane idea that software should be free. He’s the progenitor of the open software movement. So I got very early exposure to this stuff at the tender age of 11.
Alan Fleischmann
That’s amazing. And it sounds like you had some great mentors there as well.
Marty Chavez
I did. I’ve been fortunate my whole life — perhaps there was something in me that they saw. But mostly, I think of it as them being generous, and I was just there. But there were undergraduate students and then there were postdoc students at MIT. I was just some kid — you can send pictures then and have no idea how old I was. We were just typing in two versions of chat boxes. And yet, we became friends. And I remember, when I actually showed up as a freshman at Harvard, I went down the road to MIT to meet them in person. What an amazing, wonderful experience that was. Then early on…I think this shows how life has these pivots that you can’t predict or really set up. I remember arriving at Harvard freshman year, but had taken sophomore standing. I was very young, I was 16. And as a sophomore — I was just arriving but arriving as a sophomore — the first thing you have to do in the first week of school is choose a major, which Harvard calls a concentration.
Alan Fleischmann
You were how old, 16?
Marty Chavez
I was 16. I started as a sophomore, and I’d never been anywhere before. I hadn’t visited Harvard, I only went to Harvard — as opposed to MIT, or Caltech, or Stanford, where I thought I was going — because I had my mother, of course. I had an English teacher who said, “You don’t like to hear this, but your mom is right. You should go to Harvard. Because you’re already really good at math and science. We know that, but Harvard will be a chance to explore the liberal arts.” And so I turned up at Harvard and started as a sophomore. Then reality hit and I had to declare a major.
I thought, “Wow, I guess I could declare a math major.” But I’ve already taken all the math classes in the sequence at the University of New Mexico, so that didn’t sound that interesting to me. And computer science is a lot more humble now. But back then I thought, I already know computer science. I knew how to hack programs together, but I didn’t really know computer science. So I walk into the Science Center, and all the professors are recruiting undergraduates for their majors. There’s Professor Stephen Harrison, he’s a legend, sitting behind a table. And I remember he accosted me, and says, “What are you?” This was like a Harry Potter moment, but Harry Potter didn’t exist yet. And I said, “I’m a computer scientist.” And he said, “The future of the life sciences is computational.” And I thought, wow, what does that even mean?
This is 1981, right? Nobody said things like that in 1981. He said, “If you sign up as a biochem major, we will construct the major where you promise me to take the intro biochem. You take one wet lab class, that’ll be all. Everything else can be any science course you want and we’ll package it up into a biochemistry and molecular biology major.” So I did some work in his lab and I was absolutely terrible in the wet lab. I’m not sure I ever got a single wet lab experiment to make any sense. I was all thumbs with the pipettes. But he put me on a project, which I remember very well. He said, “Let’s crystallize some proteins in the capsules of some plant viruses. And let’s choose them because they can be crystallized and the names of the viruses are so remarkable.”
I remember them to this day. One was the tobacco mosaic virus, and the other was the tomato bushy stunt virus. We crystallized the proteins and used X-ray diffraction to figure out how the proteins folded. I remember at the time, Professor Harrison saying, “The combinatorics on this problem are brutal. Even with things getting faster and faster, I don’t know: we might never solve this one, it may take 1000 years or not, nobody knows. But we are going to start crystallizing proteins one at a time. And we are going to document the way they fold in three dimensions” — their so-called tertiary structure — “and maybe someday if we build enough of these as a training set” — he didn’t use that term, that came later — “maybe we’ll figure something out.” Well, roll forward to last year, and the folks at Deep Mind have a piece of software called AlphaFold. Building on those laboriously assembled datasets for 100,000-plus protein structures, they bootstrapped it to generate some predicted sequences, which they fed back in as a training set, and something magical happened. Then they published expected structures for all 200 million known proteins. So there you have it: his prediction of life sciences becoming computational is coming true right now. It’s 40 years later, but it’s happening.
Alan Fleischmann
Was that hard for you to be that young — you were so young, already a sophomore, and, I would argue, on one of the most difficult tracks as well. You were doing what you love, so it sounds like you were happy to be where you were. But I’m just curious, was it also hard?
Marty Chavez
There was maybe a bit of Catholic-Jewish masochism going on there. Let’s choose biochem, because I don’t know anything about it and it’s known to be one of those very difficult majors, not something that I thought was going to come easy. And in fact, it did not come as easily to me, like organic chemistry. I found that this was not a good match for the way my brain works. Yes, it was daunting, but also it was exciting. And the reason it was exciting is that very early in life, even growing up,
I didn’t mention this, but my first summer job was at the Air Force weapons lab. One thing my mother insisted on was that we were going to work and the proceeds would go into the family pot. And we were going to work using our brains, not our hands. This is a Hispanic underdog concept, right? We are not going to work in the farms, or work at the carwash, or mow lawns. But if you can get a job at the Air Force weapons lab, that’s fine. So I got myself a job at the Air Force weapons lab. GS1, step one: literally the lowest, lowest rung in the civil service at minimum wage. But what I was doing was interesting: the government had decided that blowing up neutron bombs in the Nevada desert was bad all around; it made Nevadans unhappy, it was dangerous. So someone had this idea, “What if instead of blowing up the bombs, we could simulate their explosion in software? Maybe let’s pay up and buy one of those crazy supercomputers.”
So I joined a project where we were writing Fortran programs that imagined simulated the scattering of electrons — Compton electrons, they were called — from neutron bombs. We solved Maxwell’s equations using finite difference methods, and we simulated the motion of individual electrons probabilistically scattering and maybe generating an electromagnetic pulse EMP. As a kid at the age of 14, I got this deep immersion into this notion that we can write programs, using math and software, that simulate reality to such a high fidelity that we can now perform experiments in software simulation and do things that would be expensive and dangerous if we did them in real life — like blowing up a neutron bomb. But if we’re just simulating the explosion, then it can be as destructive as we want it to be. It was all just happening in an imaginary world, right?
So this, this idea of simulating a bomb, simulating proteins and the way they fold, simulating financial markets and the way they fold and unfold: that is something I encountered early on. And I’ve been doing it in one version or another my entire life.
Alan Fleischmann
Wow, that’s amazing. When you were that young and you were figuring and thinking about what you wanted to do, was that what you wanted to be when you were 16-years-old? You’re 17-years-old and you’re young, you’re in college doing what you were doing. Did you imagine yourself as a researcher, as an academic, as a scientist?
Marty Chavez
So, it’s funny that you ask that. I have, all my life, had this idea that I wanted to be or maybe should be a university professor. And at various times, I’ve done job talks, received a couple of different offers to be a professor, but never quite did that for most of my career. And as I went on, and did other things in industry, almost always in business, I always had this idea: did I miss my calling?
Should I have been a professor somewhere along the line? I had a wonderful life coach who said, “Look, don’t chuck your Wall Street career and quit and go become a professor. Before you do that, why don’t you work up a single lecture and offer to give that one lecture? They’ll probably say, yes, you have some credibility, working on Wall Street, and see if you like it.” And so I did that: I did a single lecture and I gave it at Georgetown, Harvard, Stanford, Oxford, and Cambridge. And that was an amazing experience. But it didn’t really seem like something I wanted to do full time. One good lecture, I’m good.
After I eventually retired from Goldman, I did teach a whole class. It’s one of the hardest things I’ve ever done at Stanford GSB, and was an amazing experience. I’m so glad I did it. It also clarified for me that being a full-time professor, like many of the role models I had… I admired them, and am so grateful for them. But it was not my calling. My calling is actually using math and software to build digital twins, and to discover new scientific and industrial realities. I love sharing that with junior people. I love mentoring and sponsoring them.
I don’t know who my role model would really be? My hero is Alan Turing, the gay father of computer science, and he’s my hero. I wish I could invent a whole new field of inquiry the way he did. I don’t know that that’s happening or is going to happen for me, but I do often find myself asking, “How would Alan have approached this? What would he do? What would he say?” as a guiding principle.
Alan Fleischmann
Wow. You received this incredible degree in biochemistry at Harvard. And you graduated when most people would be going into college?
Marty Chavez
Eh, I was a bit older. I did a bachelor’s in biochemistry and in the same year did a master’s in computer science, theoretical computer science. So yeah, I was pretty young to have a master’s degree for sure.
Alan Fleischmann
And then, you went off to get a PhD from Stanford as well. But what made you switch away from medicine? Because if I’m not mistaken, there was an MD track that you had, which your family had wanted for you.
Marty Chavez
Yeah, that was very much part of it. So I’m finishing up at Harvard and I discovered theoretical computer science, which I absolutely loved. In all my hacking at the University of New Mexico and elsewhere, I completely missed out on computational complexity theory, automata theory, theoretical computer science, and fell in love with it. I had an amazing chance to learn from, really, the professors who created the field, many of whom were at Harvard and just down the street at MIT. And then I was going to get a PhD in theoretical computer science from Harvard, that was the plan.
But once again, my mom arrived at reality. She said that she had a dream of a doctor in the family. I negotiated with her and said, “I will apply to one medical school, as long as the same university has a fantastic theoretical computer science program. And I’ll apply to that PhD program as well. If I get into med school, as well as the PhD program, then I’ll then I’ll do both.” And my mom’s side of the bargain was that she would type up the application — I’d write it, of course, and then she would type it up. So we did that.
I got into Stanford, medical school, MD, PhD. So there, I kept up my end of the bargain. And I didn’t go to Harvard, which was for a PhD, and which was heartbreaking for me and maybe a couple of my mentors at Harvard. But I was actually very excited and got to go to the West Coast. I thought, “I’ve been cold for the last four years. I’m a Westerner to begin with, and wouldn’t it be great to go out to Stanford?” Stanford had a program, even in those days, called the artificial intelligence in medicine, or AIM Program. My PhD sponsor was the progenitor of that program, Professor Ted Shurtleff, who’s one of the legends in the early days of artificial intelligence. He created a program called Meissen to diagnose blood bacteremia. So you give the computer program symptoms and findings, and it tells you what is infecting your patient and recommends a course of treatment. And it was why they’d be successful.
He had created this joint degree program. I thought, “Okay, well, this is perfect for me.” So I have the dubious distinction of being the person at Stanford med school who got closest to MD graduation — 10 months away — and dropped out. I had a big life change and it happened all at once: I realized that I was gay. I realized that I didn’t really want to be a physician. So I did a very sharp pivot. I dropped my remaining clerkships, took a leave of absence, and went back to Boston. And did some consulting for a company that I’d helped co-found back when I was in undergrad.
One thing led to another, and I never went back to med school. When I got back to Silicon Valley, I thought that I just had to move up to San Francisco — I couldn’t live in the Valley for one more day. And so I moved into the gay neighborhood of San Francisco, the Castro, and started a software company, Quorum Software Systems. My co-founder and I started a company in a really difficult time in Silicon Valley with a crazy idea: could we trick programs into thinking they were running on a Macintosh computer, but they were actually they were running on a Windows computer, or on a silicon graphics computer or on a Sun SPARC station, but they wouldn’t know the difference. They would call all the Macintosh toolbox functions, and create windows and talk to the file system. They would happily run thinking they were on a Mac, but they wouldn’t be. This is not an emulation, but it’s an early version of an API. Which is, if the program is calling an API to create a window, the program doesn’t really know what’s under the hood, it shouldn’t need to know what’s under the hood. So we can completely replace what’s under the hood.
This idea was controversial, but it actually worked and I met many of the legends of Silicon Valley in the process. There was a moment when it seemed that Apple was going to license their software, though it didn’t end up happening that way. Apple went in a very different direction. 10 years later, of course, it ended up acquiring NXT computer and putting everything on a UNIX foundation. Of course, I would say they could have done that 10 years earlier. But that movement, that change, that plan not working out, almost in a cliche, led to the closing of one door and the opening of another in the middle of that time, which was a nuclear winter in Silicon Valley.
I received a letter in my Stanford mailbox from a headhunter and didn’t think I was too special, because several other grad students had gotten what was clearly an identical FedEx package. I opened it up, and I still remember the headhunters name, George Marino. And he said, “I’ve been instructed by Goldman Sachs to make a list of entrepreneurs in Silicon Valley with PhDs in computer science from Stanford and ship them in for an interview. Would you come?” Now at that time, we had this dream of feeding a computer an X-ray or a CAT scan and it would tell you the patient’s illness. That was one of our ideas, and the dream wasn’t working out so well. We were getting PhDs, but we were solving toy problems because it turned out that the computing power was just way too small in the late ’80s to get anywhere close to what we were envisioning.
So when I look at that little program of AI in medicine, or the section on medical informatics as it was called, it was a real rogues gallery. One of the people in that group is now the Chief Science Officer of Microsoft, Eric Horvitz. One is Amazon’s distinguished scientist in machine learning, David Hackerman. One is me. And the fourth is by far the most successful. He had this idea of using machine learning and AI to solve an easier, but important and much more solvable problem, which was, what movie are we going to watch tonight? Do you know who that fourth person is?
Alan Fleischmann
Yes, the gentleman who started Netflix.
Marty Chavez
Yes, Reed Hastings.
Alan Fleischmann
What an amazing group there, my friend!
Marty Chavez
Okay, so maybe the universe wasn’t ready for real artificial intelligence in medicine. But as a side effect of that, I went to Wall Street and learned a few things about finance. And Reed gave us Netflix, and Eric just co-authored the US government’s report on artificial intelligence, together with a few other friends. Everybody’s gone on to do some good stuff.
Alan Fleischmann
That’s amazing. Is that around the time that you got the offer to go to Goldman?
Marty Chavez
It all happened around the same time. I got this letter from Goldman in 1993. You have to imagine how it looked from my perspective. I dropped my last few med school clerkships and wasn’t going to finish med school, so I had to pay back my National Institutes of Health grant, my MD/Ph. D. Grant. That was a big bill, a big debt that I owed, and my dreams for AI were clearly not working out that great. My software company, which maybe was going to sign a deal with Apple, didn’t end up doing so. Things were looking pretty dark and I was pretty broke. I remember I had gone to the bank and borrowed a bunch of money. Basically, they loaned me the money just because I’ve gone to these good schools and for no other for no other reason. I was basically living on that credit line and credit cards.
So yeah, sure, I thought I’ll go to New York: I’ll scam this bank for a free trip, probably nothing will come of it. Then they made me an offer. The offer was for multiples of what my income was, in those days. And the project they wanted me to work on, I honestly thought was completely insane. The project was, let’s build a piece of software, into which we will put all the trades, all the time series, all the positions, all the market data, all the reports of the trading business. And then, whenever we wonder where we put something, well, we’ll know the answer, because we always put it into that one place. And now we’ll have a software artifact, the high fidelity digital twin of our trading business. And we can ask “what if” questions: well, what if this happens? And what if that happens?
That is the essence of risk management. Asking not just “what if the thing I want happens?” but “what if all the things I don’t want happen?” What will ensue? And how do I mitigate or prevent them? That’s risk management in the almost purest example. I know of the value of building a digital twin, because you can have all these unpleasant outcomes happen in a software simulation. And what I thought, Alan, was particularly insane about the project is they wanted to write a database from scratch and said the databases that exist are not fit for purpose. We’re going to build this completely different kind of database, it’s going to be in memory. I thought, okay, that’s crazy. And it’s going to be object oriented, but that is extra-special crazy. And then also, I remember thinking, why me? I didn’t take the database classes in college. They were really not my thing, and I managed to evade them and take everything else.
Look, I don’t know if it’s a good or bad thing, but I think it’s because I skipped the database classes that I didn’t know how hard it was. And I had an amazing team. So it was three guys who were building this piece of software called SecDB. They’re the author’s progenitors, and I was the fourth person to join to take some of what they were building into production. The program was successful. It was so successful in the Goldman foreign exchange trading business, run by a partner named Lloyd Blankfein… Then, he was head of ForEx, he wasn’t he wasn’t the legendary Lloyd Blankfein back in those days. But it was so successful that this database was about to blow out of the address space on the IBM machines they were using. So they needed an instantaneous port to Sun SPARCstations. I seemed like I knew something about Sun SPARCstations, so I got the job. Again, it was an example of being in the right place, at the right time, with the right preparation.
It did not look promising. I remember people telling me, “You’re joining commodities trading at Goldman Sachs? That is the dusty backwater of Goldman Sachs. The main business of Goldman is mergers and acquisitions. If they were serious about you and you were serious about them, that’s where you’d be working.” What did any of these people know? Goldman actually saw that, in its foreign exchange trading business, they had the need, the desire, and the ability to pay for somebody to create a digital twin of that business. I just showed up at the right time. What I could not have known is that this piece of software would be so successful that it would help Lloyd and other traders, as they would say, see around corners and see bad things happen before they actually occurred, when you could still do something to prevent them from occurring. This conferred a kind of superpower.
Lloyd would always say, “we’re not predictors of the future, we’re excellent predictors of the present.” And what he meant by that is that we have modeled up the present, and we’re slicing and dicing it in a million different ways. We’re seeing the seeds as early possibilities of good and bad things that might happen, and we’re adjusting ourselves accordingly. That was so successful that it propelled his success in foreign exchange and the revenues became much more stable. So then, the desire was to extend it to commodities. I worked with several others on that particular project that contributed to the transformation of the commodities business, run by another partner, Gary Cohn, who was my boss. And then Lloyd and Gary just continued, saying “we’re in” whenever they inherited a new business. Usually, because that business had a problem with its risk management, they would say, “I need SecDB, and I need those quant people, those strategists.”
I think I was strat number 12. The group, over time, grew to 2,000 people. I eventually became co-head of that group. By that time, it had permeated all of the trading businesses of Goldman Sachs. So I would say that I was just the most junior of that whole group of people who embraced not only the software, but the really radical notion, for those days, of hiring people like me and putting them on the trading desk. Of course, a quant shows up on the trading desk and the quants are a little bit of a different group of people compared to the traders and the salespeople.
The fellow who hired me and who is the progenitor of the strats group at Goldman, Armen Avanessians, likes to tell the story of the partner — whom I won’t name, but is very well known in the financial industry — who first saw how much the firm proposed to pay Armen for his services as a quant doing math and software in the trading business. This partner is said to have remarked, “You pay that kind of money to those kinds of guys?” That, Armen says, is going to be the title of his autobiography, if he ever writes it: “you pay that kind of money to those kinds of guys?”
I was just one of the early people in a firm that recognized that it was worth it to pay that kind of money. Because you gave a leverage, or a force multiplier, to your traders and your salespeople. You allow them to make much more money and lose less money. Importantly, that virtuous cycle just was part of what propelled everyone’s career success.
Alan Fleischmann
And that’s when you really put the engineers first, during that time, too?
Marty Chavez
When I joined in the ’90s, we realized very early on that strats, or strategists, could change the calibrations in their models and create or destroy revenues on that basis. Right? Because the mark to market is based on modeling assumptions, and not all of the model parameters are readily observable. And if you move them around, you change the mark to market. This is a very impressive power that you want to be very careful about.
The very prudent decision in those days was that we’re going to have a strict separation between the people who commit code — strats, like me — and the people who commit capital, the traders. And that made so much sense to us, you have to have the governance, that tight boundary. You can’t have a trader committing code, because the trader could create profits and losses based on changing the code. For the exact same reasons, you couldn’t have a strat committing capital, so this made sense. At the same time, we were noticing more and more markets going electronic. It was subtle at first, Alan, but what we started to notice was that we went from building tools that helped the traders decide which trades to do and which trades not to do, to then, what do you do once you’ve got a whole book of trades, and how do they behave as a portfolio of trades? We moved to the models recommending trades, and then to the models entering trades into the exchange. It accelerated.
Then we got to a point where we looked around and said, “Wait a minute, the code is now committing capital.” So it’s not quite like the strats are committing capital; they’re committing code that commits capital. This led to a series of very intense and thoughtful discussions with the regulators, and with legal, compliance, strategists, traders, salespeople, and leaders all across Goldman. We said, “Well, let’s recognize this new reality: We are creating a new category of traders who code, and that category of traders who code has become an important category in markets that are heavily electronic, like the equities market. So we have traders, and we have strats. But we mostly have traders who code.”
Alan Fleischmann
Which is amazing. I know that for many, even those who weren’t at Goldman, it was the empowerment and the recognition of a whole generation of engineers and encoders. That tie to the frontlines, if you will, had never been done before.
Marty Chavez
I had the honor of being a part of that. I remember getting a phone call in 2011 from the co-head of our securities division, who had grown up in the equities business. When they were merged, he looked after the equities sub-group of of the trading division. He calls me up and he says, “Marty, I want you to be among the first to know that I’m retiring.” And I said, “well, thank you for letting me know.” “And we are asking you to become co-head of the equities business.” When he said that, I have to say that I didn’t even understand what he was saying. I should remind you that I really don’t know anything about equity. I grew up in fixed income and currency and commodities. I knew he knew that, so I just said nothing. He said, “We think there’s some value in having someone who grew up in a very different business to learn from the equities business, and maybe, take that learning into some of our other businesses.” Their equities had gone electronic earlier than the other businesses, and the other businesses had some complexities that were quite different.
This is something Goldman did very often, running experiments. So Goldman was doing two experiments simultaneously with me and my career. We like to call it a high vol trade, high volatility. There’s a lot of potential for upside, and there’s a lot of potential for downside. And the two experiments were: let’s see what a commodity derivatives person will see in the equities business. That’s different, just by looking at it from a diverse perspective. Let’s simultaneously see what happens if we finally make a strat, or strategist, a leader of the business. That did not happen before. The usual thinking of Goldman is, it’s unlikely to destroy our business. It’s unlikely to generate an interesting result. But there’s some possibility that it generates a really interesting result, and we will never know unless we run the experiment. So again, and again, I’ve offered myself up to be the guinea pig for these kinds of experiments. And it’s worked out pretty well.
Alan Fleischmann
Marty, this has been amazing. I need to ask you, would you be willing to do a part two? Because we have covered so much of the show, but it is not nearly all that we want to cover. I want to get into the period when you became a CEO after leaving Goldman for a bit, the reinvention and the personal journey of what you represent on Wall Street, and how you transitioned from being on Wall Street to being at Sixth Street. So there’s a need for another hour.
Marty Chavez
If you think so, Alan!
Alan Fleischmann
Oh, I know so! People are really going to want to hear about your thoughts on leadership, diversity. You represent so many different constituencies when you’re out there, whether it’s Hispanic, or being a top gay executive, or, frankly, being a scientist and engineer and bringing that into the Wall Street world. You’ve never let go of anything you’ve done, actually, you’ve just added to it. So it would be great to have another hour with you.
Marty Chavez
Well, Alan, I’d really love to get into some of those topics as well.
Alan Fleischmann
We’ll make it a really great one.
Marty Chavez
I think getting into diversity, why not? It’s so much of the moment. I’ve been in it and I’ve been a part of it. I’ve seen some things work really well, and I’ve seen some things work not so well. I’d love to get authentic with you and talk about that.
Alan Fleischmann
Also, the role of being a mentor. You really have this incredible constellation of people that you consider mentors. You have been a role model and mentor for so many as well, many you don’t even know. That’s the power of the world we live in now; so many people follow you and are inspired by you because of your journey. And they’re going to be even more inspired by you by hearing some of the things that you’ve done, and also the early part of your life, which frankly, people need to hear.
Marty Chavez
Once upon a time, I would have been skeptical of what you just said. But I do get a very steady stream of people saying, “You don’t know me, but I read about you and your family and my dad said we could do that. And we did.” So thank you for acknowledging my microscopic contribution there, but if it was something that was a catalyst then I’m glad.
Alan Fleischmann
I think we need to live in a world where we have heroes and have people to aspire towards. You’ve done so many incredible things and I know you well, I can tell you’ll continue to do so many wonderful things. So we’re going to need another hour to dive into the world you envision.
Marty Chavez
Let’s do it, can’t wait to pick it up.