Wolf hunters from Wall Street. Part 4 (and last)

Adaptation from the book by Michael Lewis "Quick Boys"

[ This is the last part of a series of adaptations of the recently published book by Michael Lewis. The first three parts can be read here: 1 , 2 , 3 - approx. transl.]



In May 2011, a small team that Katsuyama formed — Ronan Ryan, Rob Park, and a couple of other employees — were sitting in the office at the same table surrounded by the award-winning applications of the Technology Innovation Award from The Wall Street Journal. The RBC marketing department informed them about the competition the day before the application was completed and offered to participate - so now they had to figure out which category to act in and how to make Thor look like a favorite. “There were papers everywhere,” says Park. “Everyone was talking somehow unnatural, as if instead of us there were people who had recently recovered from a deadly disease.” “It was stupid,” says Katsuyama, “there was not even a suitable category for us. I thought that in the end we would apply in the category Other. ”



The sense of futility of this occupation was in the air when Park declared: “I had a crazy idea.” His idea was to give one of the exchanges a license to use the technology. The border between Wall Street brokers and stock exchanges was blurred at the time. For several years now, Wall Street brokers have run their own private exchanges. The stock exchanges, in turn, attempted (inevitably ending in failure) to become brokers. The largest exchanges offered services that allow brokers to transfer their market orders to them, which brokers then redirected - first of all, of course, to their own exchanges, and then to everyone else. The service was mainly used by small regional brokerage offices, which did not have their own routers, but this “brokerage” service, by its very nature, opened up - at least in the Park's view - new opportunities. If only one exchange had an instrument protecting investors from market predators, small brokers from all over the country would flock there, it could take the lead of all other exchanges.

')

Forget it, said Katsuyama. "Let's just organize our own stock exchange."



“We sat motionless for several minutes,” says Park, “staring at each other. Create your own exchange. What does that even mean? ”



A few weeks later, Katsuyama flew to Canada and tried to “sell” to his bosses the idea of ​​creating a stock exchange under the auspices of RBC. Then, in the fall of 2011, he discussed this idea with a couple of the biggest financial managers in the world (from Capital Group, T. Rowe Price, BlackRock, Wellington, Southeastern Asset Management) and several influential hedge funds (led by David Einhorn, Bill Akman , Daniel Lowbeb). They all responded equally. They liked the idea of ​​creating a stock exchange that would protect them from predators from Wall Street. They also believed that in order to assert with confidence that the new exchange is independent of Wall Street, a Wall Street bank should not participate in its creation. Even as honest as RBC. If Katsuyama wanted to create a benchmark for the other exchanges, he had to quit his job and start his own business.



The difficulties were obvious. He needed money. He needed to convince a lot of highly paid people to leave his job on Wall Street in order to earn a small share of their current salaries - and, perhaps, even to invest money in the company themselves. “I wondered: Will I be able to get the people I need? How much can we survive without earnings? Will our influential environment allow us to complete the task? ”. Allowed, and the Katsuyama team followed him in this venture.



But he also needed to understand whether the big banks on Wall Street, who controlled almost 70% of all investors' orders, would like to send these orders to a truly safe stock exchange. It would be much more difficult to create a “fair” exchange if the banks that controlled the vast majority of customer orders were not interested in honesty themselves.



Back in 2008 , when Brad Katsuyama realized that the stock market had turned into a black box, whose internal activities eluded the understanding of the average person, and began to look for gifted technology experts who could help him open this box and understand its contents. He started with Rob Park and Ronan Ryan, then others joined them. One of them was a 20-year-old Stanford junior student named Dan Aisen, whose resume Katsuyama had found among the heaps of others in RBC. The line that caught his attention was: "Winner of Microsoft's Puzzle Contest Competition." Every year, Microsoft sponsored this one-day 10-hour national brainwashing marathon. He attracted more than a thousand young mathematicians and programmers. Eisen and his three friends took part in the competition in 2007 and won all the prizes. “It's a bit of a mixture of cryptography, cipher and sudoku,” says Eisen. To be able to solve such puzzles, you need to have not only a mathematical mindset, but also an exceptional ability to recognize patterns. “This is partly mechanical work, and partly insight,” says Eisen. Katsuyama gave Eisen a job and a nickname, the Puzzle Lord, soon shortened by traders in the RBC to the name Paz (Puz). Paz was one of those who helped create the Thor program.



Paz's outstanding ability to solve puzzles turned out to be most welcome. Creating a new exchange is a bit like creating a casino: its author needs to be sure that customers cannot use it. Or, in the worst case, he needs to know exactly how to beat his system in order to be able to monitor its use - just as the casino watches those who count the cards while playing blackjack. “You design the system,” says Paz, “and you don’t want to beat it.” The complexity of the stock market - with all of its private and public stock exchanges - was that they could just be beaten - and beaten up with ease - first, the smart guys from small firms, then the "prop shops" that turned into large banks with Wall Street. This, Paz thought, was the problem. From the point of view of the most sophisticated traders, the stock market was not a mechanism that redirected capital flows to productive enterprises — it was a puzzle that had to be solved. “Investing is not about how to beat the system,” he says. "It's about something completely different."



The easiest way to design a stock exchange that you cannot beat is to hire the best of those who can do it and ask them to demonstrate their capabilities. Katsuyama did not know the other champions of the country in solving puzzles, but Paz knew. The problem was that none of them had ever worked on the stock exchange. “The Puzzle Lords needed a guide,” says Katsuyama.



And then Konstantin Sokolov stepped on the scene, who once helped Nasdaq create his own matching algorithm — a computer that “reduced” buyers and sellers. Sokolov was Russian, he was born and raised in one of the cities on the Volga. He could explain why so many of his compatriots came to high-frequency trading. The Soviet education system divided people into techies and humanities. And the state-controlled economy, terrifying and confusing, was riddled with various loopholes — such an environment allowed those who could subdue it to prepare well for the order on Wall Street of the beginning of the 21st century. “Such a system has existed in our country for 70 years,” says Sokolov. "The more actively you cultivate a class of people who know how to bypass the system, the more you get people who know how to do it perfectly."



Puzzles' masters might not have thought about it at first, but, trying to design their stock exchange so that investors coming there would not become victims of predatory traders, they would foresee various ways that high-frequency traders went to their victims. For example, these traders have helped public stock exchanges create all sorts of options for clever “order types”. On the New York Stock Exchange, for example, they created a type of order that was executed only if the opposite order turned out to be smaller — the purpose of creating this type of order was to protect high-frequency traders from buying a small number of shares from an investor who was ready to bring the market down. at the scanty price.



Exploring order types, the Puzzles Overlords created a taxonomy of predatory behavior in stock markets. Generally speaking, it seems that there are three types of activities that lead to a huge number of cases of fantastically dishonest trade. They called the first electronic forward transactions - traders saw that the investor was trying to do something in one place, and were ahead of it in another (this happened when Katsuyama was trading at RBC). They called the second type “rebate arbitration” - it was carried out using legal “kickbacks”, or rebates, as they were called in the industry. Rebates were the result of complications designed to take advantage of any bottlenecks in the system: exchanges offered rebates without actually providing them with liquidity, which supposedly should have been their main trump card. The third type of activity, in all likelihood the most common, they called arbitration in the sluggish market. It arose when the high-frequency trader got the opportunity to learn about price changes on one exchange and "broke off" trading on other exchanges before the exchanges were able to respond to it. This happened around the clock, every day, and, in all likelihood, this type of activity generated much more profit during the year than all the other types combined.



All three predatory strategies depended on speed. And the very first, still immature idea of ​​confronting them arose precisely in Katsuyama: everyone tried to get as close to the exchanges as possible - so why not try to distance the “predators” from the exchange to the maximum possible distance? Detach yourself from the rest and do not allow anyone to get close. The idea was to place stock exchange machines searching for matches between buy and sell orders at a considerable distance from the places from which traders connected to the stock exchanges (these places are also called points of presence) and require everyone who wanted to trade on exchange, connect to it from a certain point of presence. If you place each market participant at a considerable distance from the exchange, you can level almost all, if not absolutely all, the benefits that speed can provide. The company has already decided that the machines that search for matches will be located in Vihoken (where they were offered a good price for places in the data center). The only question was: where will the point of presence be located? Someone suggested: “Let it be in Nebraska,” but the company understood that it would be very difficult to get the already reluctant cooperating banks with Wall Street to connect to their market if the banks had to send their people to Omaha. Although, in general, there was no need for anyone to go to Nebraska. The delay required for their new exchange, as soon as the client’s purchase order was executed on it, had only to prevent high-frequency traders from entering the race for stocks on other exchanges — that is, to prevent electronic leading transactions. The required delay, as it turned out, was to be 320 microseconds; during such a time, they had, at worst, to transmit a signal to the most distant from them stock exchange - the New York Stock Exchange, which was located in Mahwah. Just in case, they rounded the delay time to 350 microseconds.



To create a 350-microsecond delay, they needed to place a new exchange about 38 miles from the place from which brokers were allowed to access the exchange. And that was the problem. Having closed one good deal in Vihoken, they received a new offer: to organize a point of presence in Secaucus. The two data centers were located less than 10 miles apart, and other stock exchanges and all high-frequency traders already rented space there (“We went straight to the lion's den,” says Ryan). A brilliant decision occurred to one of the new employees, James Cape, who recently joined the project, leaving the high-frequency trading company: you had to rewind cables with optical fiber. Instead of using a straight cable running from one point to another, why not wind 38 miles of optical fiber into a coil the size of a shoebox and connect them to the main cable to simulate the effect of transmitting information over a distance? So they did.



Creating conditions for fair play turned out to be surprisingly simple. They did not sell to any traders or investors the right to move their computers closer to the stock exchange and did not provide anyone with dedicated access to stock information. They did not pay rebates to brokers or banks that sent orders; instead, they charged equal fees on both sides of each transaction in the amount of nine hundredths of a cent per share (or nine mil) [ mil - abbr. from mille - "thousandth part" - approx. trans.]. They provided only three types of orders: a market order, a limit order and a Mid-Point Peg order - placing such an order means that the investor order remained between the current purchase price and the sale price of any stock. If Procter & Gamble shares were quoted on the wide market at 80–80.02 (shares could be bought for $ 80.02 and sold for $ 80), the Mid-Point Peg order would be traded at $ 80.01. “It's a bit of an opportunity to set a fair price,” says Katsuyama.



Finally, in order to be sure that their own motives remain in full compliance with the needs of investors, the new exchange prohibited anyone who could trade directly in it, own part of the exchange: all its owners were ordinary investors, all their orders transfer brokers.



But the large Wall Street banks, which controlled most of the investors' orders in the stock markets, played a much more complicated role in this picture than online brokers like TD Ameritrade. Wall Street banks controlled not only the orders and their informational value, but also the dark pools on which these orders could be executed. Banks used different approaches in order to extract the maximum benefit from the orders of their customers. All of them sought to send warrants primarily to their dark pools of liquidity, and only then send them to the wide market. Inside the dark pool, the bank could trade against these same orders or sell the selected access to the dark pool to high-frequency traders. In any case, the value of user orders was monetized - by large banks from Wall Street to its own advantage. If the bank could not execute the order in its own dark pool, it would redirect the order to the exchange that paid the highest rebate.



If the Puzzles Masters were right, and the new exchange was indeed designed to level the speed advantage, it could reduce the information value of market orders of investors to zero. If the orders could not be used on this new exchange, and the information they contained about investors' trading intentions became useless, then who would pay for the right to execute them in such conditions? Large banks from Wall Street and online brokers who sent market orders of investors to the new exchange, thus, would refuse billions in profits. And this, as everyone involved in the process understood, could not have happened without a struggle.



Their new exchange needed a name. They called it Investors Exchange (Investor Exchange) - the name was then shortened to the abbreviation IEX. Prior to the opening of the exchange, on October 25, 2013, 32 of its employees suggested that they would trade in the first day and the first week of work in what amounts of shares. The average number of assumptions was to be 159,500 shares on the first day and 2.75 million shares in the first week. Below all, only one employee, who never participated in creating the stock market from scratch, appreciated the exchange's capabilities: he assumed that trading volumes would be 2,500 stocks on the first day and 100,000 stocks in the first week. Of the nearly 100 banks and brokerage companies that were at different stages of signing an agreement to connect to IEX (most of them were small enterprises), only 15 were ready to work together on the day of the opening of the exchange. Katsuyama suggested, or perhaps hoped, that the exchange would trade 40–50 million shares a day by the end of the first year of operation — such trading volumes were needed to cover the costs of the exchange. If this were not possible, the question would be how much the exchange could survive. Katsuyama thought that their intention to create an example of an honest financial market - and, perhaps, change the culture of Wall Street - could take more than a year. , : « , ».



IEX 568 524 . Wall Street, – RBC Sanford C. Bernstein. 12 . , , , IEX 50 . , 18 , 11 827 232 . . Goldman Sachs, , , , , .



«» Goldman IEX 19 2013 15 9 42 662 361 406 . , IEX, , , - . , , – , , . . .



«15 !», – - 10- . 331 – – 14 .



« !»



« !»



« AMEX», – (John Schwall), CFO, . « AMEX ».



« 120 », – , . - 300- . , $100, , - IEX . .



- : « JP Morgan, , . , , , - ».



- . « Goldman. , -. , ».



JP Morgan, , IEX, Goldman , .



« !»



51 , Goldman Sachs Wall Street, . , . , . « , -, , : « », – . « , Goldman Sachs . . ». : « , , ».



– . Goldman Sachs , , . IEX, , .



IEX : , - . , . , – , . « », – , – « , , ».



, IEX, 14 – – . $2,6 20% . , , , , IEX . « , », – . « , . , ».



, . Goldman, , , IEX. , , ; . : « , , IEX, : « ? !». IEX Wall Street , IEX – . , IEX, « 10% ». , IEX, $300 .



-, . , Credit Suisse , IEX – , RBC – IEX . , Wall Street IEX: 350 , IEX, , .



IEX , , , . Wall Street, IEX , . IEX, , , , , – , , , . Wall Street IEX, – .



, .



- : « [ ] ?»



« , , », – . « . , , , . , , , – . – , . ».



« ?», – .



«», – (IEX 94 ). RBC, Bernstein , , , . « », – . « Morgan Stanley, JP Morgan Goldman Sachs».



: « ?».



« , , […], , , – », – . , – , , Wall Street.



. , , . ( - , (Eric Schneiderman), , . Goldman Sachs, (Gary Cohn), The Wall Street Journal, , Goldman Sachs , ). , IEX , . . , Wall Street .



Wall Street . . , . , 2000-, Wall Street , , . , - , , , , , . , , . , , .