My wife Yinh’s podcast Growth From Failure is in its 5th year. Her guests are extremely wide-ranging and rarely found on “the circuit”. Many of you know her and would agree — her multi-standard deviation superpower is a genuine interest in people’s individual stories. It’s a power that is not simply recognized by people she meets — it’s felt. You feel like she’s listening to you because she listens with her whole self. Some may wish for an invisibility cloak or a crystal ball. I’d take her ability to listen (and for every keto dude working on his abs — I’d be willing to bet that this listening skill is the OP aphrodisiac — the problem is it’s rarer than a six-pack.)
She doesn’t need me promoting her pod but yea you should check it out if you want to break out of your normal pod routine especially if you get the awkward sense that you’ve slotted too many quarters into the same dudes-who-use-business-idioms-like-open-the-kimono merry-go-round.
That said, if I promoted every episode I’d dilute my recommendations. I restrain unless the interview is:
☑️if they intersect with quant finance. Past in this category include interviews with my former colleagues Kelly Brennan, partner at CitSec, and options trader Tina Lindstrom or power trader Noha Sidhom.
Today, I’m boosting her latest episode because it checks both boxes.
🎙️ALLISON BISHOP – COMPUTER SCIENTIST, WRITER, COMEDIAN. PRESIDENT AND CO-FOUNDER, PROOF TRADING (Growth From Failure)
This is the story of Allison Bishop, president and co-founder of Proof Trading, an institutional equities execution platform. In this episode, we cover Alison’s journey from aspirations in creative writing, which had a rejection and led her down a path that was much more quantitatively oriented than she thought. That led her to Masters and PhDs, in math, computer science, and cryptology.
I learned so much from her like the branch of math called combinatorics, which I’ve never heard before. And also the practical applications of cryptology and computer science and for the first time, finally, understanding how those were actually used.
But more importantly, I learned about controlling how you set people’s expectations and reclaiming that power, either emotionally or physically. Allison literally changed the way she moved in the world.
On her transition into math after being rejected from the hallowed Princeton Creative Writing path, her entire reason for choosing Princeton in the first place:
I was devastated, it was the entire point that I was here. And at the time, I was also supposed to be just getting rid of my general distribution requirements. One of those was to take a math course. And I refused to take calc, which was what most people are taking. I’d taken calc AB in high school, and I find it kind of boring. And so they’re like, you could just take Calc II.
Meh, what else you got?
I ended up taking number theory because I thought it was the “physics for poets” of math. It is not. It’s a foundational course in cryptography. And it was a course math majors were taking and it set me on this very different path of learning that I liked math and that math could be creative. But it happened because of this simultaneous confluence of events that I got rejected from Creative Writing, and I refused to take calculus.
I also did take the “physics for poets” version at Princeton. I thought it was great. It was really philosophical and insightful, and I quite enjoyed it. But yeah, the math class was not at all that. It was a serious math class and I also had no conception of what Princeton math is. What the Department was or what that meant. And I’m really glad I didn’t know because I think I would have found it very intimidating. I got pretty deep into this process as a math major before people started, like telling me, you do know, this is one of the best math programs in the world. Oh, no, I didn’t know that. You know, the guy teaching linear algebra is a Fields medalist. What’s the Fields Medal? And it was just very perplexing to the people around me how I got there because there are usually about 60 to 100 people every year applying to Princeton for math and in my year, only 15 graduated with that major. I ended up being like this wild card person who did not know what was going on, but was in for the ride.
What she found creative in math
It was the first time I actually got to write proofs. This concept of using math not as a vehicle to just calculate something, but as a way of reasoning about the world. That was new to me because I feel we lose so many opportunities in the way we teach math at the elementary, middle school and high school level. Here’s the procedure and you do it. If we’re lucky — “here’s what that means”.
But there’s not “here’s how people came up with it. Here’s the history of it, or here’s the question from first principles, and have you come up with the procedure?” So the first time having these open-ended questions be asked, and then having to reason about it myself, assemble proofs out of the building blocks that we’ve seen in the class.
The one that I think fascinated me at first, which is actually a relatively simple one that you can teach at the middle school level, is a proof that there are infinitely many prime numbers. The fact that you could prove there are infinitely many of something as complicated as prime numbers. What’s fascinating to me, and the creativity that goes into this, okay, how do I reason about this? How do I create a procedure to always produce a new prime outside of the set of primes that I produced so far? That constructive building process was very appealing to me.
But when you’re writing a proof and realize how creative it is, but then you’re going through it again and it’s still the same answer does it then feel black and white, and then QED?
Not really. And I think this is something that mathematicians are at varying degrees of denial about, because there is this process of peer review that happens when we prove new theorems and people put them out. And it’s highly nontrivial to go through and check someone else’s proof and make sure everything makes sense. And in some sense, when we can reduce each little step to this logic axioms, then you can check that we mostly gain confidence. But the process of putting those things together really shapes the theorems that you’re proving. So it’s not really that there’s this a universe of facts. And then there’s these proofs that get attached to the facts. And then as long as we have a proof attached to the fact that we can check we can move on. There’s so many different things that we could reason about, or relationships between objects that we could try to prove theorems about so much more of the interesting stuff goes into the human process of what do we find worth studying? What questions do we ask?…I think mathematicians are a little bit reticent to admit how subjective the process of curating what’s important and what things we study is, because the process of checking a proof is already nontrivial, but is solvable in some sense. But the process of deciding which proofs are worth doing, and which questions are worth investigating is where most of the fun stuff happens.
From math to computer science
Math separates into two sub-fields. For me, at least the way I think about it, some of them have very complicated objects, and then you prove relatively simple theorems about them. So all the complication goes into the objects themselves. And so there’s things like topology, which is about shapes in the world, and things like that. And algebraic geometry. And combinatorics is more like the objects are pretty simple. Graphs, you’ve got dots, and you’ve got lines connecting the dots are things but then the theorems and the patterns that you study on them are complicated, which is much more aligned with computer science.
I ended up wandering from math to computer science. So as an early PhD student at Austin, I would go around to the different mathematicians in the department and try to learn about their research, I would always ask them, “What is the application of this in the world?” And I got a range of answers from there aren’t any applications of this in the world to tons of applications to other areas of math. And so I was pretty underwhelmed by the potential impact of the things that were being studied. And there’s a lot of fundamental research that happens. And we don’t know yet how it’s gonna get applied. And it eventually gets applied. And I think this stuff is really cool. But from the perspective of I have one life, and I like being very people-facing and doing things that impact people directly, it wasn’t really compelling enough to me. So I started wandering over to the computer science department.
It’s always a joke I make with mathematicians, “what are you guys doing over here?” It’s all the same stuff. But it’s got much better research grants in computer science, and it’s much more potentially impactful and relevant, pretty directly. So I ended up switching halfway through my PhD to be in the computer science department, which meant doing all of their requirements, again from scratch. So I basically did two PhDs and left with one of them. Totally the wrong way to do it. But I ended up in a good place sort of finishing my PhD with an emphasis in cryptography in the computer science department.
[Skipping over a discussion of the 2 major branches of cryptography which solve the same problems with different tools but remain silo’d for complex reasons as well as her human interest in the field]
While I’ll also skip how she started working with IEX before starting Proof Trading I found the transition story useful:
Cryptography and finance are similar fields — they both try to scare people away with enough acronyms. So I was somewhat comfortable at that point, being the person who just asked everyone questions all the time. And so my strategy on the trading floor was just sitting in the middle of all these people and whenever they their take their headphones off, I would ask them a question. “Hey, I’m wondering what this graph means. Can you come over here?” I was just shameless about asking questions. But I think also just having that prior experience of learning a jargon-heavy field, knowing that the fact that I didn’t know what the things were, didn’t mean I was stupid. And it didn’t mean it was going to be hard. It just meant that I had to ask, although I do think finance makes it unusually tough.
I’ve also been documenting some of my process. Proof has put out a market structure primer, which is basically my writing down the things that I didn’t know, and therefore asked and putting that into a form that hopefully helps other people. Because we do think that, unlike cryptography, where there are textbooks and there are public facing things and surveys that are helpful, in finance, you’re piecing this together from different people’s memories. I’m in the middle of the trading floor asking “What does this four-letter code mean on this trade flag?” There are no public references for so much of this. So I do think as a field, we need to do a better job of giving people friendlier entry points, but the people on the trading floor at IEX, around me were great at just having the patience to answer a million questions.
What problem is Proof Trading trying to solve?
The biggest surprise to me learning about the stock trading system was that the entire purpose of stock trading is to provide public information about prices. That’s the whole reason for being of the stock market. And yet, absolutely, every structure around that is completely opaque, which I just think is ludicrous. I will make the same argument with people about pay transparency, we need better insight into where these prices are coming from, for exactly the same reasons that we need public price discovery for stocks, but taking it specifically to the broker layers.
What we saw from our seat at IEX, was that we were trying to build tools that would ultimately help long-term investors on the buy side, but we didn’t have direct access, because that’s not who IEX is — its customers are brokers. And every time we’d ask a question, well, why are brokers doing this? Or why are brokers acting this way? There’s no insight available into that because brokers are this black box.
Thinking about this coming from cryptography, cryptography is all about keeping secrets, but they are much less secretive than this. Because fundamentally, whenever somebody comes to the cryptography community and says, I have a cool new way of encrypting data but I can’t tell you how it works, we laugh at that person, then we break their system immediately. Usually, “I can’t tell you how it works” means it’s not going to stand up to scrutiny. And the reason we believe in the encryption algorithms that we have is that they’ve been exposed to decades of people trying to break them and improving them and all these things. So it seemed crazy to me that the science of electronic trading algorithms of how to round certain ways or how to build or design a schedule, or how to optimize to certain goals for order performance, that that was all individualized inside these black boxes, instead of benefiting from an open science culture. It seemed completely handicapped to me as a scientific field and completely unnecessary.
There are things that need to be protected and confidential, like, what stock are you trading and when are you trading it. But if I’m claiming to you that I have a way of taking big trades and breaking them up into pieces in the market, so that they blend in with the noise, if they really blend in, then I should be able to tell you my mechanism, and it should still work. And if I’m forced to expose that mechanism, it should open up all of the surface for public science and collaboration that should help us get much better outcomes. So I’m a big believer in making science public, whenever there’s not a great reason to do otherwise, rather than secrecy being the default.
So one of the main things we wanted to solve for as a broker was to make as much of our decision-making process public as we could. So to make us more accountable, to open that up as a surface for collaboration with our clients that if we tell them what we’re doing, they can say, Oh, well, that’s not aligned with our goals in this case. And that’s a much more rich surface for communication and improvement than just saying, Here’s the box, you can use it or not. So it’s kind of our hope that bringing a scientific process into the open in this area would really jumpstart progress toward better solutions.
[Skipping a balanced discussion of the challenges they face as a remote-first company]
Can you give me an example of what Proof Trading does for a client? (I highlighted the refreshingly honest part of the answer)
Once a client has decided they want to buy a large amount of stock or sell a large amount of stock, our job is to figure out how to split it over time and space into pieces, so that the market doesn’t disproportionately react. How do we make their total activity blend in with the random activity of the market in such a way that we can get done what they want to get done without the price going up, because they’re buying or going down, because they’re selling so it’s reducing impact relative to getting the trade that you want? That is the value that we’re looking to add.
One thing that’s very tricky about this process is that the market is very noisy. So there’s a lot of noise in the data performance. And it’s very hard to compare two algorithms, apples to apples, because they get different kinds of orders and different market conditions, huge sample sizes. So it’s a very tough sales position. I can’t point to a single order. “Look, this is the amount of money we saved you.” And people seem to expect that, which seems super weird. When we go into the sales call it’s “explain to me how many basis points your algorithm is going to save me versus my competitors.” One, your competitors won’t tell me anything about their performance. So how would I know that? And two, I can’t even measure my own performance cleanly, because it’s a function of all this noise. And so I get a lot of befuddled expressions on the other side, it’s like, well, then what are you selling?
We think we’re better. We just also think this is really hard to show and I’m not going to show you some half-baked thing.
Addressing her recovery from a deeply personal trauma:
As a growth moment, it was pretty significant because I was having symptoms of depression and PTSD as a young faculty member. And what was strange to me about it was that my ability to function and do my job was the last thing to go. I was not sleeping, I was not eating well, everything was off. And yet, I was still publishing papers and teaching classes on paper, everything looked fine. And I had to convince myself that it was bad enough to heal, just give myself permission to heal.
And I think one of the things that surprised me about that process was it was less of the growth — it was more, almost shedding a limb, just completely resetting. And that was actually the time a lot of people have asked me about. I started dyeing my hair, blue and purple, and I started getting tattoos that were meaningful to me and changing my appearance. I felt like I was walking through the world looking like this person that I wasn’t able to be anymore. And having people react to her the same way they used to react to her was stopping me from healing, people would expect things in me and I would do them because that’s what I’d always done. And so I needed to find a way to change what people expected of me. And so I think one of the biggest things was reclaiming my physical presentation and space. That was also when I started learning, boxing, and just changing the way I move in the world.
So for me, I think the biggest growth moment was realizing that sometimes you need to reset, and you need to reclaim some of the things that are in the way of that progress and giving myself permission — also to quit, as a faculty member and not say it is a failure. I think a lot of academics think of leaving academia is always intrinsically a failure. It was a continuation of my journey, the willingness to be seen as a failure. And that sense, I think was really important.
And I loved this part:
Realize that we have more control over setting people’s expectations for us. And I think it takes a certain amount of separation from people’s reaction to you just realize that that lever is there to some extent. And for me, it really came out of a place of necessity. And from feeling it in myself, I had to wake up in the morning and look in the mirror and see this person that I didn’t feel existed anymore. And that was this weird, disconnecting feeling.
This was always the biggest thing I tell people thinking about public speaking or comedy, is that if you make yourself laugh, it’s contagious. If you’re getting on stage, and you want to convey an emotion, or you’re pitching your startup or whatever, if you feel it, the audience will feel it. And so it’s getting to the point where you feel good about what you’re doing, or you feel excited about what you’re doing is 90% of it. And the rest is just icing on the cake.
Check out the full episode to hear about how Allison defines success, what she learned from her blunder as an 18-year-old intern working on the Mars Rover with the JPL (Jet Propulsion Lab), her boxing and comedy outlets and how her professor Jordan Ellenberg (my notes on his book) inspired her to embrace her varied interests and her recent acceptance to a creative writing MFA program!
Allison also founded a conference called CFAIL: Failed Approaches and Insightful Losses in Cryptology.
It’s an annual event where we give people a platform to talk about research that didn’t ultimately succeed in whatever they were trying to do. So it’s whether they tried to prove something and they couldn’t, or they tried to build something, and it was broken. And that has been, I think, just one of the funnest and most energizing research conferences that I’ve been a part of people embracing talking about failure.
I’ll leave you with this skit Allison created:
“The MMWF Award” — a fictional award for Men Who Support Men Who Support Women in Finance.