Dr. Indre Viskontas:
Hello and welcome to Inside the Mind of a Trader, the show where we explore how brilliant traders are wired for success. I'm Dr. Indre Viskontas, neuroscientist and Professor of Psychology at the University of San Francisco. On this episode, I'll be talking with Matt Andresen. Matt is a pioneer in the fields of electronic and quantitative trading. He was CEO of Island, the first electronic stock market which merged with NASDAQ and is still used by NASDAQ today. Now as founder and CEO of Headlands Technologies, he and his teams have programmed computers to do millions of trades a day. Hi, Matt.
Matt Andresen:
How are you doing?
Dr. Indre Viskontas:
Good. How are you exploiting human weaknesses?
Matt Andresen:
Well, human beings have wonderful strengths, they also have certain weaknesses. Our ability to process large amounts of information quickly is never going to be as good as machines that we program to do those tasks for us. We all have outsourced our memories to our phones. I know a lot of random errata. It used to be great fun at cocktail parties because I could remember sports scores and historical dates. Now no one cares. They're like "I know that too I have my phone." Just as in other industries where you have technology replacing rote and repetitive tasks, trading is the same. And for the last 25 years I've been working in that transition using computers to do ever more of the processing work of trading.
Dr. Indre Viskontas:
I'm sure you're still a lot of fun at cocktail parties. I wonder if you can tell us what are the ways in which these programs actually make life better, even for traders that maybe aren't using them as much?
Matt Andresen:
I wish people asked me that question more often. I think to the ideas that people have that well, it's machines versus humans. It's not, it's humans harnessing machines to do things. If you look at the ... What it costs to trade now versus when I came to Wall Street nearly 30 years ago. Back then a trade commission on a trade was about $250. The spread you would pay, which is the difference in the market between what someone was willing to buy or sell it for, which is at an implicit cost of trading. I mean, Intel had a half-point spread, that's a half-a-dollar spread, which anybody tuning in now under the age of 40 would find incomprehensible.
We're in a world where brokers charge no commission to trade, and where the execution price they're getting is virtually the midpoint meaning, there's no slippage in the quality of the execution that the end customer gets. And that's not because there was no law that said that had to happen, it's what markets did. And as people got better with computers, got better with harnessing quantitative science you ended up with all that competition driving prices down in terms of explicit costs and then narrowing spreads to the benefit of everyone.
Dr. Indre Viskontas:
So if a person is just starting out in the trading world, what advice would you give them in terms of the skills that they need to develop to become brilliant at it today?
Matt Andresen:
I think it depends on what type of trading you want to do. To get into my industry I would tell someone that ... Go hit the computer science textbooks, take your linear algebra, and your other advanced math classes and those will provide a grounding to then apply technology to the markets. There's lots of decisions made every day by human beings in the marketplace. And so my advice to people who want to be the decision maker about given trades is to make sure you understand your time horizon for the trade. If you're trying to trade ... Put on a trade that's going to realize in four seconds, I'd recommend that you reconsider. At those timescales, you're competing with machines. What moves prices at those very small timescales is very infinitesimal fleeting information that a human being can't possibly ingest and translate into a trade.
But you go out for three months and your guess is as good as any machines about whether AMD's earnings will be good or bad in that quarter. As you go further out on the time horizon you get the benefit of all the machines competing for these very short-term trades in the form of very efficient prices. But then you are competing really against other people and against rudimentary machines for how to price out further out.
Dr. Indre Viskontas:
Yeah, that's a really interesting point that the human skill now is not going to be in these little tiny decisions that ... On the millisecond or even the second range, but rather understanding the larger patterns of human behavior or market forces moving out. When do you think that might not be true anymore?
Matt Andresen:
I don't know. I'm sure everyone read up on the OpenAI drama at their board. One of the things that was sort of out there as a conspiracy theory was that they ... Perhaps the OpenAI team had invented Skynet and the board was trying to stop it from launching the nukes and burning the sky. I'm skeptical of that. Clearly, given the pace of technological advancement, people will have to remember that technology does not ... It's not a linear process it's geometric. You think that in 1903 the Wright brothers flew. They were on the moon by the time I was born, right? You're talking about 60-odd years. You can go from biplanes in World War I to jet aircraft at the end of World War II. Each individual leap forward builds on all of the prior leaps and so you get this very big steepening. I'm reticent to make a grand pronouncement about when we face Skynet. Clearly, you have to be prepared for things to change much quicker than they did when you were a kid.
Dr. Indre Viskontas:
So one of the things I've been wondering is whether you need to actually be a brilliant trader in order to program a computer to do it.
Matt Andresen:
That is a great question. So in our company there is a dividing line between two broad job functions, one we call software developer and one we call quantitative researcher. The difference in that they both have to be able to code at a world-class level. So if you looked at their resumes they look extremely similar. The difference is sometimes maybe someone has just a little bit better intuition for how to computer ... To program a computer more efficiently, to use less memory, to be ... Operate a little faster. The quantitative side is where they have what we call trading intuition. So even in a place where every single trade is happening at the command of a computer, to program effectively the actual strategy and ... You really have to have an intuition for the trading. So it's that same trader intuition that Paul Tudor Jones or Ken Griffin or the other trading greats have. It still has to be there in the minds of the people programming.
Dr. Indre Viskontas:
And when you get this huge increase in speed and efficiency, does that come with a greater risk, not just for your own trades but for the market in general?
Matt Andresen:
It certainly can. I mean, I'm a believer in man's progress and technology being largely a force for good. Electronic trading has been around for 30-plus years so we've had lots of chances to make mistakes and to learn as an industry. Have you ever heard of a fat finger trade?
Dr. Indre Viskontas:
No.
Matt Andresen:
And they still happen very occasionally where ... Let's say you're my customer and you say, "Buy me 100 shares of Intel." I go, "Got it." And I type 100, and then I take a sip of my coffee and hold my finger down and it adds 20 zeros to it.
Dr. Indre Viskontas:
Oh, that's bad.
Matt Andresen:
And then I go and I hit send. And then the market has to react to someone trying to buy a trillion shares of Intel and chaos ensues. In the industry these are known as fat finger errors. The explanation would be the guys or gals fingers were too fat to fit on the one key and they did. Not a super great HR environment back in the day apparently on the trading floors. You can still have those in electronic trading. The benefit is that you have all these computers that you can program, and test, and test, and program such that ... Machines don't operate outside of their lines. So if you tell them, "Don't ever send a trade away." Whatever else I say, if I ever tell you "Try to buy a trillion shares of Intel don't listen to me." And that's something you can actually put at different parts in your system as a check and balance.
Dr. Indre Viskontas:
So do you do that?
Matt Andresen:
Oh, yes. That's a critical part of our risk management is not just designing the strategy to do what we'd like it to do and to be profitable but then to not send orders that we don't want to send. Don't send an order outside of this size or don't send the same order multiple times. That's an advantage I think we have over humans who are more error-prone. But the flip side of that is that ... The worst thing a human can do probably is send one erroneous order. Whereas if you had a poorly programmed system where they weren't following the rules, they weren't testing it thoroughly and they do something bad it can get out of hand very quickly. There was a trading firm called Knight, K-N-I-G-H-T, and they had an incident that is known colloquially as the Knightmare. See?
Dr. Indre Viskontas:
Oh, yeah, very clever.
Matt Andresen:
They are a large complicated trading firm. They were trading lots of markets in lots of different places with lots of different trading teams. But they had started out as a human-driven trading shop and then evolved to be more electronic. Anyone who's tried to build technology know that if the foundation is less secure there can be little gotchas in there that you don't realize that can come up and bite you. And they were hooking up to a new market. The NYSE had a new auction order type that they were ... That finally went live. And it was just this very small little thing that they had to do and it caused a series of unexpected things to happen in their system whereby they bought the proverbial trillion dollars worth of Intel.
Dr. Indre Viskontas:
Wow.
Matt Andresen:
And that actually caused them to be insolvent at the end of the day.
Dr. Indre Viskontas:
Well, that's a tough day. Well, let's take a quick break and we'll be back with Matt Andresen after this.
James Kostulias:
A lot of what we do and a lot of what we talk about at Charles Schwab and with our whole offering of Schwab Trading powered by Ameritrade is the educational component.
Lorraine Gavican-Kerr:
We really like to teach our clients about risk management, about using a platform. I think that we really want to set people up for success for whatever they want to do. Traders have different aims, different goals, they trade different types of products,
Chesley Spencer:
Enabling them to see what they are able to do, particularly with things like derivatives or options where it's not nearly so much a buy-in area of hey, you think this is good, buy it and it'll go up. You think this is bad, sell it and it'll go down. All of those have very particular actions one can take based on them that it's a lot more nuanced than just a simple buy or a simple sale.
Lorraine Gavican-Kerr:
The people who work behind the product and the experience are traders. They're super engaged, they love what we do.
Chesley Spencer:
Advice is very much a part of the overall Schwab business. But in the trading space it's much more about empowerment.
James Kostulias:
Whether you're somebody who's more fundamentally based or more technically based there's so much to learn across how to trade different instruments, across how to trade different markets. I think from my perspective what I would say is, continue to learn and continue to educate yourself. There's no sort of one playbook to trading.
Dr. Indre Viskontas:
Welcome back. We've been talking about what might be the remaining features of the human mind that have not been bested by machines. Instead of talking about this necessarily as this dichotomy one or the other, I wonder if we could talk a little bit more about how humans can use machines in ways that make the most of their abilities. Is there something in the trading world where you're programming your machines to do something and they do something or they find something that you think never would've been possible?
Matt Andresen:
Oh, without question. I think you've hit on the right concept which is that it's nothing to be feared. We train machines to do stuff for us, and we offload the things that we don't need anymore. But the technology, our feelings as people in society is that change is always disruptive. I don't know what the age is, let's say it's 35 years old. We basically love all technology advances until we're 35 years old. And then after that, no.
I remember my grandparents were ... When they were still alive they were like ... When the answering machine came out and they were resistant like "Who" ... "No one needs that." My dad, who's an older guy now, getting him to move off of a flip phone is a ... It's a crucible. Because he's like "Well, why would you need that," right? Once you have it you don't understand how you got along without it. I mean, I'm sure the candle makers went to DC and lobbied for jobs, jobs, jobs after the light bulb but it hasn't made our lives darker. There is plenty of opportunity out there to harness the new technologies. And you're always going to need a human to direct that next layer of innovation.
Dr. Indre Viskontas:
So what does it mean to you to trade brilliantly? Where do you use technology? What skills do you ... Would you develop? If you were building from scratch a brilliant trader, that was human, what would you put into it?
Matt Andresen:
Well, I think the best traders I've seen have insatiable intellectual curiosity. So Ken Griffin, who I worked for for many years, the most notable thing about him wasn't his intellect which is amazing, it wasn't his work ethic which is great, it was his childlike intellectual curiosity. I mean, he was just like "Tell me more." And any new thing his eyes get wide and he's like "Tell me more." Paul Tudor Jones is someone I've gotten to know very well. Another great trader, exactly the same. Any new thing that comes along they don't fear it. They don't say, "Cryptos new I don't want to do it." They would not have been the guys complaining about the light bulb, they'd be the guys that couldn't wait to figure out a way to harness it and use it. So I think that intellectual curiosity is number one. Because otherwise, certain parts of your skills will become outmoded by competition or by technology. The only way to stay up with that is to be curious enough to motivate yourself to keep up.
Dr. Indre Viskontas:
One of the other things that neuroscientists are learning about the brain is that sometimes struggling with a particular decision or a set of information is the way in which you define your own values. You're voting in the booth, and you could write an algorithm that uses all kinds of points that gives you the ultimate choice of who you should vote for depending on ... You plug in your values and then who's the best candidate. But the actual work of looking through the candidates, figuring out where their stances are seems to be how we actually get those values to begin with. So in a sense, you have to do that work in order to understand yourself. And so I wonder if there's an equivalent of that. As you've been programming these what have you learned? And what do you think gets lost if we outsource all of those decisions to a machine?
Matt Andresen:
Oh, that's a great question. Well, I'm going to make another pop culture reference from a dystopian future, The Matrix. And if you remember when Neo, who's the character, he wakes up and finds out that his life is a lie, it was a simulation. And now he's awake, and he's never used his muscles before, and they're ... And he's having to learn everything from scratch. And then one of the other teammates sits him down and they start uploading skills. And it was like everyone's fantasy. You're like I want to speak Mandarin. They're like duke and he's like "I speak Mandarin." And then he goes "I want to know kung fu." And I was like wow, that's amazing, wouldn't' we all like that.
But I always was a little disturbed by it because I do think ... If you didn't put in the hours and exert the discipline to acquire the knowledge you don't value it and you don't understand it the same way. When you've been on the grind and been trying to figure out these problems, I do think that ... It strikes me as correct that that's how we actually learn and how we develop values, how we develop insights. And if you didn't do that how would you know where to start?
Dr. Indre Viskontas:
That is one of my fears with outsourcing a lot of these tasks that ... Even when computers become better at writing essays or doing this work where we're going to lose some of those critical thinking skills that are only developed with that struggle. Do you have a sense that there are specific skills that we shouldn't outsource just because we need to be able to flex those muscles and learn to use them?
Matt Andresen:
That's a good question. We talked about memory earlier and how ... Memory used to be an essential part of education, you had to learn how to memorize. And the concept of a memory palace and other techniques where you try to turn ideas into a visual abstraction that you can then file away and sort of deep storage in your brain and be able to recall. When you read books or see a movie involving people from the Victorian age and people just walk around and ... Off the top of their head they could quote a whole act of Shakespeare. Nobody could do that now at all. We've allowed that skill to atrophy. Sometimes I think, because I have a really good memory, that sucks because you've outmoded my ... A key advantage that I have. It's good because now you can go do something else with those brain cells. It's just an opportunity for us to evolve as well. The machines are going to help us evolve. We don't all have to grow our own food anymore, look what we do instead of that. Even 200 years ago we probably would all be growing food.
Dr. Indre Viskontas:
So you've been successful in your life, and I'm sure there are moments where you were under a lot of stress. I wonder if you have some advice for people who are trading and are feeling the stress, and are working to manage risk. What are some of the things that you've learned over the course of your career?
Matt Andresen:
The big lesson I've taken in business is just not being emotional. You're going to be emotional. But my rule is, anything ... A new thing that happens you're not allowed to be emotional in the moment, right? So if you give me super bad news and you ... I'm not going to react. And I'm really not going to react. I'm going to be like "Okay, that's really bad news, I'm going to file that away and I'm going to deal with that later, the emotional side of it.
Dr. Indre Viskontas:
So you don't even feel your heart racing? Okay.
Matt Andresen:
And so that to me has been very helpful one, in managing the stress but two, not doing something dumb because you're being driven by emotion instead of that. No one likes being wrong.
Dr. Indre Viskontas:
No, I totally get that. That's a theme I think among trading is that you have to be able to separate your emotional response. But there's also this myth out there that we have a reptilian brain that is emotional, and then we have this evolved human brain that is all rational. The truth is that emotions are constructed moment to moment. There is an evolution in how we respond to them. I don't know that it's necessarily true that you just tamp down the emotion. There's an element of understanding when the emotion has value and when it doesn't. I'm sure you get emotional, I hope when your child does something that makes you weak in the knees.
Matt Andresen:
Sure.
Dr. Indre Viskontas:
You probably have very deep feelings still when there's something in that realm, but yet you have been able to not get carried away into, as you mentioned, some of these bigger emotions.
Matt Andresen:
I went to Duke, I sent two kids there. A big Duke basketball fan, I can't watch it.
Dr. Indre Viskontas:
Really?
Matt Andresen:
Literally, I can't watch it. My sons will text me like "Oh, did you see that comeback or whatever?" I'm like "I'm watching Netflix with mom." I can't deal with it it makes me crazy. And I can't control my emotions in it so I just take a powder and I'm like not ... Let me know how it went. I remember losses that Duke had and in the moment I'm just crushed. And I look back on it I literally do not care at all right now looking back on it. So you just have to remember that there'll be a time where this is not as traumatic, and emotional, and destabilizing as it is in this moment. And so to say, look, let's table that. I'm going to come back to that and grieve later for that loss.
Dr. Indre Viskontas:
And I think there are two strategies to handle this if you're just starting out. One is understanding that experience will give you that. As we all know, the teenage years are fraught with big emotions in part because it's the first time you're experiencing rejection or falling in love or all of these things. And then once you've experienced it a few times you've lived through it, you ... It doesn't seem to affect you as much anymore.
There's another element of not putting your self-worth into an outcome that you can't control. I work a lot with musicians and helping them deal with performance anxiety. And the times where people get totally crushed by some criticism or hitting the wrong note is when they equate their value with someone else's opinion or with a perfect performance. But that isn't where their value is, right? It's in all of the stuff leading up to it, it's in all of the other people that may or may not have been touched on it or the thing that they've expressed. And so I wonder if that's part of the strategy is, if you don't put a value on yourself as a trader, or as a person, on the outcome of a particular trade then it's not as bad.
Matt Andresen:
I wish I could do that, that sounds like fabulous advice. There's clearly strategies to be able to be functional while you're doing that. And I think one thing that I try to draw on is to look back at other times in your life when you've had adversity and said, "Geez, that was really bad. It felt like the world had ended but it didn't. And I persevered and things turned out okay."
When I was an athlete, I remember losses, right? I had the 12 steps of losing I always thought about. Every time I lost a tournament I would come home and I'd be like "I quit, I'm never doing that again," right? And then I'd be in denial and all this. And then three days later I'd be like all right, I just got to train harder that's the answer. But I had to go through that sort of emotional ... Because you have so much tied up in it. And I had my self-esteem. Your self-image is tied up in this endeavor, and when it doesn't go the way you dreamed you have to recalibrate. I try to draw on that and look back at times in my life where I've ... Hey, look, I've been beat down so many times, and look, I always managed to get up. And the answer is always work a little harder.
Dr. Indre Viskontas:
Or program a computer to do it for you.
Matt Andresen:
Or that.
Dr. Indre Viskontas:
Matt, thank you so much for coming on the show.
Matt Andresen:
It's been a pleasure.
Dr. Indre Viskontas:
Join us next time as we explore the minds of brilliant traders.
On-screen text: [Schwab logo] Own your tomorrow®
