Beyond ‘Flash Boys’: Matt Levine Interviews IEX’s Brad Katsuyama

The hero of Flash Boys helped make market structure mainstream. Yet Katsuyama really only crossed the starting line in September, when his stock exchange, IEX, finally went live—and nothing gets him talking like the nitty-gritty of Wall Street.

Stock exchanges are part of the plumbing of Wall Street, and the details of how they’re run have never exactly captured the public imagination. That changed with Brad Katsuyama, 38, the co-founder and chief executive officer of IEX, who brought equity market structure to the mainstream as the hero of Michael Lewis’s book Flash Boys.

Katsuyama was working as a trader at the Royal Bank of Canada, helping big investors buy stock, when he noticed it was getting increasingly difficult to do so without moving the price. As he investigated, he came to the conclusion that stock exchanges weren’t always looking out for investors’ interests and the market favored high-frequency traders at the expense of long-term investors. (In Lewis’s words, the market was “rigged.”) This led Katsuyama to start IEX, an exchange with a “speed bump” designed to slow down high-frequency traders on behalf of longer-term investors.

After a grueling application process full of fierce resistance, IEX’s Investors Exchange gained approval from the U.S. Securities and Exchange Commission in June. “We just wanted a chance to compete,” says Katsuyama, who spoke with Matt Levine of Bloomberg View about the nuances of market structure shortly after the exchange went live in September.

This interview appears in the November issue of Bloomberg Markets.

This interview appears in the November issue of Bloomberg Markets.

Cover artwork: Zachary Walsh

MATT LEVINE: There’s an origin story 1  of IEX in which you were a trader, you realized it was hard to buy all the stock that seemed to be available in the market, and you set out to fix that. You created IEX with a “speed bump” but also with a router that skipped the speed bump, so that high-speed traders couldn’t see a trade on IEX and race ahead of your clients to buy up all the shares on other exchanges. But the reality of IEX doesn’t quite match that story anymore: As part of the SEC process, you had to give up the advantage your router had. Do you miss it?

BRAD KATSUYAMA: We originally started IEX under the premise that we would be this router for clients. Any broker could send a routable order to IEX, and we could get that order out to the market. Thank goodness it evolved. 2  The way it did was asking the question, “Why can someone get a signal from one place and get to the next?” It really came down to co‑location and this idea that people could pay to put their servers as close as possible to the exchanges to get information fast; now, with microwave connectivity, it’s even faster. So when we thought about it, RBC, as a broker, could only solve a very narrow part of the problem. And the problem was broader—not only for orders that were routed, but also for orders that are resting, or pegged, on the exchanges. How are those orders treated when there’s this huge kind of disparity between who knows what, when? And that evolved into things such as discretionary peg—D-peg—and midpoint peg, 3  and basically the advancement of the speed bump. If we had stayed only focused on the router, I don’t know if IEX would have made it. The router became secondary to our operation as a market.

ML: You don’t miss the former structure of the router?

BK: The benefit, and I guess the slight change now, is that the router trades a little bit less on IEX. The router doesn’t know there are certain orders that are displayed and others that are not displayed. The non-displayed orders aren’t being broadcast to the router because that would create an unfair advantage, even though that advantage exists for certain exchanges today. From our standpoint, it trades slightly less on IEX. I guess that’s a small change, but the customer experience doesn’t change. That, for us, is most important. 4

ML: The customers on IEX, the people you talk to, what’s important to them? What do they think is your advantage, vs. other exchanges?

BK: I think, most important, is they feel like we represent their interest.

ML: What does that mean?

BK: It means that the market had evolved in a way where they were the last to know. Clients can’t be members of exchanges; a broker is a member. A client has to route an order through a broker-member to get to an exchange. Because of that relationship, they were on the outside looking in. As a result of a lot of these things that evolved over the last decade, we earned a lot of credibility when we were at RBC meeting with customers and saying, “Here’s an explanation for this experience you’re having when you can’t buy or sell what you see.” That level of trust parlays into us as an exchange. Most of them benefit from things like discretionary peg and our speed bump because it means a client on the West Coast can send an order, and once that order makes it to IEX, it’s IEX’s responsibility to price that order if we receive instruction to peg it. 5

ML: Why was the discretionary peg less a part of the Flash Boys story?

BK: D-peg came later. So IEX launches, and brokers started analyzing their fills. We got a lot of independent reports saying, “Wow, you guys are top-tier—our No. 1 pool from a performance standpoint.” Over time they started to see that performance get slightly worse; we were still better, but the distance between us and others was starting to close. We started to research, Why is this gap closing? The whole point of the speed bump was to say, when information is disseminated, someone has it before us. D-peg was our answer—to say, “Some people are guessing, with a very high degree of accuracy, that the quote’s about to move.” 6  Do we necessarily want to take a buyer and throw them up to the midpoint if we think they’re going to get run over? It’s our most popular order type. But this all evolved, and it’s still evolving. We’re still learning things today that we didn’t know at RBC or when Flash Boys was being written. It’s this constantly evolving understanding of what’s happening.


Photographer: Celeste Sloman

ML: Would you say that a customer uses IEX for more of the specific, structural reason—like, how D-peg prevents a crumbling quote? Or would you say that it’s more a general sense of, “These are the guys we trust.” Is it about you, or is it about D-peg?

BK: We have a good group of supporters—some based on data, some based on trust. We work a lot with buy-side clients who do a huge amount of analysis and actually teach us about the market. They are using us for a particular reason. At the heart of it, these people are seeing it in their executions. Others haven’t dived that deep into market structure but say, “We trust you, and you have a sophisticated following.” I don’t think we would have survived if it was just hype. People are too smart. There’s too much scrutiny on this for it to survive based on hand-waving.

ML: Back to the origin story. You are the hero of a Michael Lewis book, which is an unusual situation for anyone in finance to find themselves. And for a generation on Wall Street, Lewis popularized this notion that equity traders are dopes, right? “Equities in Dallas” became a famous insult after Liar’s Poker. 7  Did you ever talk about that with him?

BK: When I started in the business, the first book I read was Liar’s Poker—and I was on the equities trading desk. So I remember “Equities in Dallas.” I saw that as an opportunity. But, no, I don’t recall us ever talking about that.

ML: Lewis kind of created the way that Wall Street thinks about Wall Street—and you were telling him how Wall Street works. Was that a strange experience to be, like, “No, no, Michael, you have it all wrong”?

BK: Moneyball was a hugely influential book for me. Starting at RBC back in 2002, we needed to find ways to be different without the same resources as the biggest banks. So I read Moneyball, and I felt like that had a pretty big influence on the way we thought. The discussion with Michael really started with helping him write a story about somebody else. He had stumbled upon the story of Sergey Aleynikov, who was thrown in prison for taking computer code. Michael went to a couple characters in The Big Short, whom I know. The first few months was really just giving him background. He’d ask, “What is this?” And we’d try our best to answer. That evolved into him getting to know and understand our story. In many ways, we just interacted with him as if he were a buy-side firm or broker who came in and started asking questions. I tried not to think about the fact that we were talking to Michael Lewis.

ML: Did you think you were interesting?

BK: We had been told for a while, “This is a Michael Lewis story.” Did I necessarily believe that I could be the main character in one? Probably not. It’s funny, because one of the things he told me in the middle of his research was, “You’re a fairly boring person. It’s hard to make you jump off the page.” I’ll be the first to admit that the fact he was interested was a stroke of luck. 8

ML: You had buy-side clients, you had a business going, and it seems like it would have worked without Flash Boys, at least in some commercial way. Has going so big and so pop culture made it harder?

BK: There have been upsides and downsides. But I do think the upsides far outweigh the downsides. The story had much greater reach than Ronan [Ryan, IEX’s president and one of its co-founders] and I could have ever had going door to door. That was important, not to be a midsize ATS [alternative trading system, or dark pool], but to be an exchange and try to take this to the next level, I think that was critically important. The downside was that it upset a lot of people. And we made the decision early on that we wanted brokers to be our members, and our customers, which was hard because the buy side was our strongest relationship and we were sticking the broker in between.

ML: There’s a set of facts on the ground, and then there’s this broader, emotionally charged narrative. Do people have a point when they’re upset with you?

BK: It all depends on perspective. When people view the world through their own lens, I’m sure they have reasons to be upset. Some people from the high-frequency-trading community—and again, I think this is a valid argument—say, “These inefficiencies are there, someone is going to capitalize on them, and that’s what we do.” I think that’s a legitimate claim, which is really why we started an exchange. We believe that the exchange needs to be held to a higher standard than just anybody. And the fact that these exchanges are creating these inefficiencies, that’s the heart of it. Because if we really felt like high-speed trading was solely responsible for all of this, it probably would have been more about regulation or lobbying or something like that. And we would have banned high-speed trading from our market, which we didn’t do. We basically said, “We’re going to set the rules, and anyone can show up.”

I think automation and high-speed trading can provide a lot of value. But the exchanges changed—in how they built their market, who they catered to, and how they were influenced. From our standpoint, for the long-term investor, knowing that the stock market exists for them, they’ve been massively underserved. As a result, they’ve been taken advantage of with these little nuances, order types, machinations. They always seem to be on the other end of that. Maybe it’s time to try and change the tide, and not to outlaw anything that exists, but just to give people an alternative—or to give brokers who trade on their behalf an alternative.

ML: So you’re now an exchange, and you’re, like, 1.5 percent of volume. There’s one way this plays out, with different exchanges catering to different investor types. And the other way is you need to kill everyone: You have the right model, and everyone else has the wrong model. What does success look like to you?

BK: It’s hard for us to set a specific market-share goal. A lot of the decisions that we make mean we trade less volume. D-peg is about preventing trades from happening; the speed bump deters some types of activities. Not paying rebates as an exchange is a huge differentiator, and we think it’s a benefit to IEX as well as the investors and brokers we’re catering to. But not paying someone to send you an order when a bunch of other exchanges are doing that, that’s also a disincentive. I think if we were paying rebates, we’d be a lot more of the volume. 9  We try to focus mostly on, “Are the brokers happy with their execution?” It’s an educational process. The rebate is such a huge short-term incentive. And, for us, you have to have a very strong sales pitch to get people to change their behavior and come to a market that won’t pay you to come. It’s a longer sales cycle. If I were to drop the hammer and say, “We have to get to X percent by this time,” people internally would probably start making different decisions. That’s something we don’t want to do. We don’t have a revenue target; we don’t have a market-share target. We care mostly about the experience on our market.

ML: On the one hand, it sounds like you don’t have any near-term goal to be 100 percent of the market. You’re happy that there’s an ecosystem of different competitors. But on the other hand, you count it as a win when your competitors move toward your approach. Is there only one right way?

BK: Well, it depends on whose perspective you view  is the right way. The exchanges clearly have set their trading models up to sell latency, to provide certain advantages to people who want to pay for them. That caters to a certain class of the market. We’re trying to do the opposite, where we create a universal speed bump. There’s one way in and one way out. Do we think we’re serving the most critical player in the market? We do, because at the end of the day, the stock market was built to serve long-term investors and help them invest in companies. But does that mean that everyone should have to serve that one constituent? No, because that’s what makes for competition.

ML: Can we talk about speed bumps? The one I think is the most interesting is Chicago Stock Exchange’s. They’ve basically created a speed bump that applies only to incoming liquidity-taking orders, which means, effectively, that Chicago market makers have a “last look.” They can put up an order, and when a new order comes in to take their liquidity, they get a few microseconds or whatever to decide whether they want to pull their order. That’s not exactly right, but that’s basically right—and sort of the opposite of the story told in Flash Boys. What do you think about that?


Photographer: Celeste Sloman

BK: What they’re trying to do is similar to what Nasdaq, I think, proposed a few years ago with a 5-millisecond speed bump as opposed to 350 microseconds. So it will be interesting to see how the SEC handles it, because it’s an asymmetrical application of latency, which means that slowing down certain parties and not others will, I think, create a challenge. 10

ML: Speaking of your competitors, how much sympathy do you have for them?

BK: Having talked to people who have worked at some of these firms, I don’t think people set out to end up where they are. I think you put one foot in front of the other, and eventually you go, “Where am I and how did we get here?” I don’t think it was this massive chess match, where they saw every move coming and played it the way it is. Part of it was a reaction. I think Bats [Global Markets] had a tremendous influence on NYSE and Nasdaq. I think that, competitively, they started to chase different types of revenue streams and ended up where they ended up. Do I have sympathy for them? I don’t. But do I think it’s part of this massive plan that’s unfolded for the last decade exactly as they thought? Probably not.

ML: How are you going to avoid ending up in 10 years somewhere very different from where you want to be?

BK: I think a big check on that is really how IEX was formed and whom we’ve been supported by. Our No. 1 asset is the relationship and trust we have with the buy side. If we lose that, we lose a lot of credibility and it becomes harder to do what we want to do. We’re driven and motivated in a different way. It took us nine and a half months to raise the money we needed—and those were pretty tough moments. Deep in our hearts we wanted to be neutral. We wanted the buy side to own us, and we wanted the sell side to be our members, and we wanted to try and align them in a way that was unique. Our relationship with the buy side is just something that we’re not willing to sacrifice for some money.

ML: Many of your employees have worked for other exchanges, high-frequency-trading firms, and even the SEC. Do you ever have philosophical discussions internally about what’s right?

BK: None that come off the top of my head. Well, I’ll give you one interesting one. As the application process escalated, we were doing our best to address the questions through the comment period. And then the exchanges took it to the media with pretty sensational language. We felt like we needed to respond in an equally strong way. I think that created a bit of a push internally to say, “Why don’t we just keep going? We’ll win on merit.” And others said, “No, we’re losing—we’re playing one game, and they’re playing another.” We have debates internally a lot. I had written, like, six op-eds over the application process.

ML: Wasn’t there one on Bloomberg View?

BK: There was. And I’m thankful to my team for not letting the others get published. I was trained as a trader: Something happens, and I react. I’m learning to be patient. Once you get in this habit of responding to everything, you almost have to keep that up. Whereas if you’re smarter and more strategic about how you react and how you engage, it actually lets us focus on the most important thing, which is running the company.

ML: So now that you’re a stock exchange, what’s next from a business perspective? One obvious thing is listings. What’s going on there?

BK: One of the benefits of Flash Boys was that it got the market-structure story to publicly traded companies, who had no idea that their stock wasn’t trading 90 percent of its volume on NYSE, or that other markets could even trade their stock. Listing is something that we’re definitely interested in, but we’re still in the process of working through how we can be different.

ML: It feels like listings are very central to being a U.S. stock exchange.

BK: There’s 10 that don’t and two that do.

ML: But it seems like a very obvious business opportunity, especially with all the interest you’ve gotten. Do you see an opportunity to innovate in listing standards, or would you copy what’s already out there?

BK: We think that listing standards is a hard thing to disrupt out of the gate. There will be a lot of regulatory pushback if we try and dramatically change the standards for a company to be publicly traded. We’ve just been trying to have market-structure discussions with corporates in the same way that we’ve had them with the buy side. I think the exchanges becoming so overtly against IEX and the buy side being so supportive made a lot of people scratch their heads and think, This fight seems a bit odd to us.

ML: There’s this discussion about your helping long‑term investors in a very specific market‑structure way. There’s another big discussion in U.S. equity markets among people saying shareholders or companies have a too-short-term perspective. It seems to me that there’s an overlap there. Do you think about that?

BK: It is a bit further afield for us. The Long-Term Stock Exchange—that’s an interesting idea, and I like Eric [Ries] and what they’re doing. Philosophically, we’re aligned. But to dramatically change the listing standards, it’s not our core expertise to arm up with lawyers and dot the i’s and cross the t’s a different way. I think what they’re trying to do is extremely interesting. We do see short-termism as a problem, but solving those types of regulatory rule-based standards, I don’t think that’s high on our list. We give them a lot of credit—what they're doing has equal merit—and think we have a good relationship with them.

ML: Do CFOs come to you and say, “My stock turns over a lot, it’s really volatile, can you fix that?” Is that the tenor of the conversation?

BK: No, because what we say is, No. 1, “We can’t fix it instantly, because we’re one market of 13. Your stock can trade anywhere. Here are some explanations as to what’s happening and ways that we’re trying to stop it on our market. But make no mistake, we can’t help what happens on other markets.” So sometimes just knowing what’s happening is an equally good relief to hearing, “I’m going to fix all of your problems.”

ML: Do people come with expectations that listing on IEX could change how their stock trades?

BK: If they do come in with that notion, I’m strict with the people internally to say, “It’s a step in a direction that can help build a market that better suits their interest.” We’re clear about that.

ML: The IEX model is intended to solve very specific issues in the U.S. equities market structure—with D-peg and the shoebox—that don’t necessarily exist in currencies, or bonds, or even equities in India. Obviously those markets all have their own issues. There’s an opportunity for someone with your symbolic clout or narrative to say, “I’m going to come in and take market share by fixing those problems.” But the technical solution would be totally different.

BK: We’ve looked at other asset classes and talked about other geographies. Some people at IEX have expertise in different areas, but right now we’re 74 people universally focused on U.S. equities. That’s the only experience I have; I’m not shy about saying that. As we’ve looked at other opportunities, it’s definitely in partnership. And finding the right partner has been tough. Although we might be interested, it’s important to find the right people.


Photographer: Celeste Sloman

ML: You now have an affiliate, TradeWind, that trades gold on a blockchain. 11

BK: That’s the first alternative business line that IEX is formally involved in. Part of that is the fascination internally with the concept of a blockchain and us making sure we’re positioning ourselves for where we think the puck is going. We’ve felt like there’s a way for this type of technology to evolve that will have implications in our business. It should be no surprise that Nasdaq is extremely interested in blockchain because they don’t have a vertically integrated exchange like CME or ICE does. Finding a way to disrupt the clearing-and-settlement process makes a lot of sense for someone like Nasdaq. It also makes a lot of sense for someone like IEX.

ML: When you think about blockchain and vertically integrated exchanges, do you have some advantage because you’re a disrupter who’s starting fresh and isn’t burdened with a few decades of baggage? Or is it that you’re relatively small and new, and you’re focused on building out that idea rather than getting distracted by other stuff?

BK: We’re small. We can’t get distracted. We can only focus on a few things. We have to find projects that we think are meaningful, where we can ensure that we’re efficient in how we spend our time. On the flip side, I think we are a disrupter. We have built technology that is three years old, and it’s ours. And there are huge benefits to that.

ML: So it seems like you have one big idea, and you’re working with full focus to implement and sell that one big idea, and you’re not looking to branch out into other big ideas at this point?

BK: Not yet, we’re in only the second or third inning. We’re well-capitalized, the business is profitable, and we know exactly what we need to do. If the business starts to mature, could we invest and look at other areas? Absolutely, but we’re just days in.

ML: Let’s talk about the SEC process. What was your impression of how they dealt with the pressure from both sides?

BK: The people who worked directly on IEX’s application were extremely thorough and asked a million questions, and we did our best to work with them. It was a fairly rigorous process. It’s one of those things where some people are upset when they get patted down in security lines at airports. I don’t mind. At least they’re in there trying to vet things out. They did a thorough job. The pressure—and you’re right, it was coming from both sides—showed what was at stake and probably caused more deliberations internally to make sure they were making the right decision for the right reasons. At the end of the day, we’re just thankful that process is behind us and we have the approval. We felt that we met all the regulatory standards. We just wanted a chance to compete.

ML: A lot of the comment letters were from retail investors. Why is starting a stock exchange the new way to advocate for the little guy?

BK: Clearly part of it is just the time period in which we live. People feel underserved. And I think people are inspired by those on Wall Street trying to do the right thing for the broader markets. For us, I think it was equally gratifying to get letters from Goldman Sachs or Texas Teachers or Capital Group—very sophisticated people who understand the nuances of market structure. We weren’t going to win their votes with magic and tap dancing. They have dug into our model at the deepest possible level. They’ve used us, and traded on us, and then got out there and supported us. Our support was a broad spectrum of sophistication, and that was very gratifying. You don’t have to be a market-structure expert to believe, or at least trust, that we’re trying to do the right thing for them, but at least people who do understand markets and are very sophisticated were supporting us as well.

ML: What’s your strategy for retail?

BK: Our best way to affect the livelihood of the end user is focusing on the large institutions. Meeting with and working with Capital Group, which manages a trillion dollars in assets, is a more effective use of my time than trying to go out and rally individual retail. I may have met two or three retail brokers in four years; I met hundreds and hundreds of institutional buy-side firms. If someone wants to buy 100 shares of Microsoft, the structure of the market works fine. But if they’re invested in a mutual fund that wants to buy 10 million shares of Microsoft, yeah, we think there’s a more effective way to do that, and we can help. That’s how we think we deliver the greatest utility to the most people.

ML: What’s your advice for the next you—someone who works on Wall Street, has a big idea, and wants to change the world?

BK: It was a very hard decision to leave RBC, because there’s a certain level of comfort working at a large firm and being part of the system. I do think we’re shifting to a more transparent society. What that does is cast a brighter light on how people make money. We don’t have any issue with people making money. That’s just capitalism. We’re capitalists at heart, we’re a for-profit entity. But I think how people make money is going to be a greater focus. If people are thinking about doing something different with their lives, just think about the incentives that are there for you to do what you do. And if you don’t believe in that incentive structure, finding a different thing to do with your life isn’t as risky now as it was in the past. I think the world is more receptive to people trying to do things differently. —Interview has been condensed and edited.

Levine is a columnist for Bloomberg View in New York. Subscribe to his daily Money Stuff newsletter here.

  1. The story goes that, in 2007, Katsuyama was a trader at RBC, trading on behalf of institutional clients. Sometimes he would try to buy a big block of shares, say 100,000, but could get only 25,000. He was puzzled and initially thought it was a software problem. Eventually he figured out it was an issue with the deep structure of the U.S. stock exchanges. So his team built a router, called Thor, that would help his clients avoid the problem and get the full 100,000 shares by buying stock on every exchange simultaneously. Then he left to found IEX, a trading venue, to do something similar on a larger scale. IEX’s innovation involved coiling wire in a “magic shoebox” to create a speed bump, delaying communications to and from IEX by 350 microseconds. When people sent IEX an order for 100,000 shares and IEX had 25,000, IEX would give them the 25,000, and then before IEX told anyone else about the trade, its router would send the remainder of the order to the other exchanges, get the other 75,000, and give the client the full 100,000 shares before high-speed traders could jump ahead of them. Perhaps you read about this in Flash Boys.

  2. “We’re routing 40 to 50 million shares a day and matching 10 million on our own market,” says Katsuyama, “which means if that was our only premise, we would be doing a very small amount of volume on IEX and sending a very large amount of volume elsewhere.”

  3. IEX, like other stock-trading venues, allows customers to submit “peg” orders, typically non-displayed orders that are automatically priced at the midpoint of the national best bid and offer for a stock. So if there are bids to buy a stock for $9.99, and offers to sell it for $10.01, and you send a midpoint peg order to buy it, then your order will sit hidden on the order book at $10. If someone else sends an order to sell at $10 or less, you will trade; if that doesn’t happen, and meanwhile the bid and offer move to $10 and $10.02, your order will reprice at $10.01. It’s a way for big investors to buy or sell stock without displaying their intentions and without “crossing the spread” to buy at the offer or sell at the bid. One of IEX’s innovations was to introduce the “discretionary peg,” or “D-peg,” which looks a lot like a midpoint peg order. The difference is that IEX uses a formula to determine if bids and offers on other stock exchanges reflect a “crumbling quote”—essentially, it tries to anticipate if prices are about to change and the midpoint is about to become stale. If it decides that that’s the case, it will prevent the D-peg order from trading at the midpoint until the quote is stable again. Katsuyama says: “The thing about D-peg is about what it doesn’t do. It just stays on the bid. It doesn’t move up to the midpoint in those particular windows. The amazing part is, when you add it up, this scenario unfolds in some stocks thousands of times a day. But the amount of time that that window exists combined is, like, 60 seconds, or 100 seconds. We’re talking about very narrow increments of time where someone can see something that other people can’t. So in a 390-minute trading day, for two minutes there’s this disparity, and D-peg is trying to address that.”

  4. “The router change that we made was, in a way, to address a technical issue,” says Katsuyama. “People said, ‘Your router has an unfair advantage.’ So really what we did is we rearranged how we accept routable orders. We take it into the router first, and then that router is basically simultaneously blasting other markets and IEX. If it sends an order to IEX, it goes over a speed bump. When we look at the fill rates before we made the router change, they were in the high 90s, 97 or 98 percent. And when we look at the fill rates for clients after we made the change, it’s still 97 or 98 percent.”

  5. “Pegging is important,” says Katsuyama, “because it helps level the playing field for a long-term investor on the West Coast who’s not going to buy servers in the New York Stock Exchange data center—nor can they, because they’re not a member. It gives them the ability to say, I’m now entrusting the exchange to price my order according to benchmarks that are set in the broader market. And that allows us, from a latency perspective, to represent their interests in the fairest possible way.”

  6. Katsuyama adds: “We started to see the quote was changing, but it didn’t always change in the same way. Sometimes it would happen in a fairly uniform way. But sometimes the bids would crumble, which means 1 in 3 exchanges would drop, and then a millisecond later another would drop. And it happened in this rolling time period that was sometimes several milliseconds. One option was to say, ‘Do we make our speed bump several milliseconds?’ And we didn’t really want to do that, because this was a specific type of scenario.”

  7. Liar’s Poker is a tale of bond trading at Salomon Brothers in the 1980s, and bond traders and salespeople at Salomon looked down on stock traders. Lewis wrote: “We could not imagine anything less successful in our small world than an equity salesman in Dallas; the equity department was powerless in our firm, and Dallas was, well, a long way from New York. Thus, ‘Equities in Dallas’ became training program shorthand for ‘Just bury that lowest form of human scum where it will never be seen again.’ ” Katsuyama was formerly head of U.S. cash equity trading at RBC.

  8. “There was a decision point where we had to decide whether we were going to talk about us or not,” says Katsuyama. “Giving him background to write a story about someone else is very different than talking to him about you, your life, and your decisions. At the time, there was a bit of a conflict. The personal side of me was a little more reserved, and I had to really kind of talk it through with my wife and say, ‘Is this something that we really want to do?’ But as the CEO of IEX, which hadn’t launched, it was a no-brainer. And that ultimately is what won.”

  9. Katsuyama adds: “We think rebates is one of the biggest issues in the marketplace, which is why we don’t pay them. Eliminating them should be an administrative change, and it would have deep ripple effects. If you look at this huge proliferation of order types, a lot of them are designed around harvesting rebates. About $2 billion a year is being paid to brokers, who at times are acting as agents to the client. I have yet to meet a client who is collecting rebates from brokers and distributing them back to their funds. It’s a huge conflict. And I don’t think the structural implementation of changing rebates changes the market structure. I think it actually simplifies the market in a lot of ways. People will say, ‘Who’s going to show up if you don’t pay them?’ Well, the markets still functioned before the invention of rebates.”

  10. When asked if, as critics sometimes say, IEX’s discretionary peg order is itself a form of “last look,” Katsuyama responds: “No, last look gives optionality to someone. D-peg moves according to our formula, and it’s public. It’s, ‘You’ll be treated in the way that we describe.’ Not, ‘You’ll have an option.’ Last look is not even close to what we’re doing. The key part about D-peg is the fact it’s a non-displayed order. It’s not protected. People aren’t seeing something, trying to get it and it moving. People have asked us why not make D-peg a displayed order. That becomes very challenging. D-peg solves a different kind of problem for people who want their order to not be displayed. They want to send larger size to a market. The orders sent through D-peg are huge. If you display those orders, it would completely shock the market. D-peg was very specific, but within the context of displayed liquidity—seeing something and having it fade on you—we’ve been very cognizant that’s what started this whole journey for us. We don’t want to contribute to that.”

  11. Earlier this year, IEX spun off a firm called TradeWind Markets, which has plans to launch a gold exchange using blockchain technology for trading and settlement. IEX is an investor, and Katsuyama is on TradeWind’s board of directors. “I think if we were more mature,” says Katsuyama, “TradeWind would have been something we would have wanted to do inside of IEX, and that gets us back to the point about not wanting to be distracted. We’re an owner, we’re involved, but it doesn’t sit on the floor; it’s its own company.”

Before it's here, it's on the Bloomberg Terminal.