This post is about entrepreneurs and secrets.
In Zero to One, Peter Thiel spoke about how secrets still exist and how entrepreneurs must and can find them. Michael Lewis’s Flash Boys can be seen as a story about persons who discovered secrets about the US stock market and eventually became entrepreneurs.
The book, published in 2014, explains the story of Brad Katsuyama, the Royal Bank of Canada equities trader who discovered how stock markets had changed, how trade orders were really executed and how market fragmentation had created huge advantages for the fastest traders; this led him eventually to start IEX, a new stock trading venue where orders were slowed down to mitigate the advantages of high frequency traders (HFTs).
In 2007 Brad Katsuyama began having problems when trying to execute stock market orders; he could see bid-ask quotes on his trading system, but when he tried to trade the market seemed to disappear. He could not find answers at first; googling was no good; and it took him a long time to figure out what was happening; as one investor he spoke to much later said “there’s no book you can read, it’s just calling up people and talking to them”.
While investigating, he discovered various things: he was not the only one, other investors at firms like SAC Capital and T Rowe Price also had difficulties trading; new exchanges and trading venues had appeared over the last years; different trading venues applied different fee structures and these changed over time; some exchanges and venues paid a fee to market makers while others charged them a fee; some brokers were sending their clients’ order flow to the trading venues that paid them a fee for doing so; HFT firms were paying brokers to be able to access and trade against their order flow; exchanges were offering colocation: for a fee, traders can set up their computers inside the exchange and gain faster access to information.
In the early 2000s most equity trading took place on the NYSE and Nasdaq exchanges; shares traded on either one or the other, but not both. Then, regulation was introduced to increase competition, and this led to the creation of many new trading venues and to the same share being traded in all venues simultaneously. Also, the matching of trading orders became electronic.
The 2000s also saw the growth of dark pools: private trading venues, often managed by large banks and broker firms, with more limited information requirements than the public exchanges. The advantage of dark pools for large investors is they allow them to transact anonymously without moving the market. Brokers were also allowing HFT firms to access their dark pools.
The fragmentation of equity markets created opportunities for HFT firms, which were enhanced by the services offered to them by the exchanges (colocation, flash quotes) and by new types of trading orders, created by the exchanges, tailored to HFT needs.
At first Brad Katsuyama could find no one who could explain what was happening to his trading orders and how orders were executed. He learned that HFT firms accounted for a lot of trading volume and they were making a lot of money in this new environment, but no one from these firms would talk to him. He searched among people who had worked in or near bank’s technology departments and started to build a team.
He and his team performed trading experiments and they found out some surprising facts: if a trading order was sent only to one exchange, then the order was executed in full; if the order was sent to various exchanges, only a fraction was executed. They also learned that if they timed their order to arrive at the same time at all trading venues, then the order was executed completely. This led them to eventually learn their orders were being front run from one exchange to another.
HFT firms were sending multiple small buy and sell orders to multiple exchanges. If some investor took the other side of one of these small trades and they guessed this investor was interested in buying or selling more, they would front run him to another exchange. This was enabled by trading venues (both exchanges and dark pools), who took fees for letting HFTs act as market makers.
Another person featured in the book is Dan Spivey who in 2008 realised there was a big difference between the trading speed available at the time between Chicago and New York and what was theoretically possible. There was latent demand from traders for faster communication speed between these two cities and the telecom carriers were not catering to it. This led Dan Spivey to think of building a fibre optic line much faster than what was available at the time.
Dan Spivey had to study maps and the laws and rules on rights of way and on laying fibre optic along roads; he hired a person who worked in construction to drive with him for two days along the path he planned for his line; he later founded Spread Networks, raised money from investors and hired a construction engineer and construction crews. Speed Networks had to speak to and reach agreements with hundreds of local authorities and property owners. The line had to be as straight as possible.
The surprising thing about Spread Networks is they started building the line before speaking to customers about their plans. It was not until March 2010, three months before the expected line completion, when they started reaching out to investors via personal contacts.
Spread Networks’ line was valuable if access was scarce. They settled on providing access to 200 firms and pricing this access at 300k a month. Clients could opt to pay 10,6M upfront for a 5-year lease.
IEX was founded in 2012 and went live on October 2013. As of writing, it holds a market share of around 3% and handles more than 300M shares daily.
In November 2017, Spread Networks was acquired by Zayo Group Holdings for $127M.