I buy my books later and cheaper. Waiting a few years means I score pristine hardcovers for a buck at a Friends of the Library book sale, which I savor and then donate back in a virtuous cycle of frugality and reading. This is how I came across Flash Boys by Michael Lewis, published 6 years ago and devoured on my sofa over the past few days.
Lewis is a talented writer, and he brings finance alive through vivid characters facing difficult moral questions despite tremendous financial temptations. Having worked at Salomon Brothers in the 80s, he has an insider's understanding of the flawed personalities and false veneer of stability on which the equities market rests. He specializes (and delights) in unveiling how average investors are exploited by shady actors incentivized to exploit loopholes that stay one step ahead of regulation.
What follows is a highly simplified abstraction intended to give a newbie a basic understanding of how high frequency traders (HFTs) operate. If this piques your interest, I strongly encourage you to read the book.
The central tenet of Flash Boys is the law of unintended consequences: SEC regulation intended to fix one vulnerability in the market unintentionally creates a new vulnerability which enables rapacious behavior an order of magnitude greater than the one it intended to outlaw.
It was the 1980s: thin ties, cocaine, and Patrick Nagel art were all the rage. The vulnerability to be solved was that investors depended on phone calls placed to human brokers to execute the investors' desired transactions in a timely manner.
During the 1987 stock market crash, brokers on Wall Street stopped answering phones to avoid having to buy stock as the crash commenced, which was counter to the brokers' financial interests. Legislation enacted by the SEC moved equity transactions online to reduce the potential for human interference with investing activity.
The unintended consequence was that the new computer-based market technology created opportunities for savvy technologists to learn market moves in advance of typical investors, allowing them to manipulate markets for personal profit at the expense of average investors.
Market exchanges where stocks are bought and sold are distinct entities located in geographically distant areas (i.e., New York, Chicago). The electronic signals used to execute buy and sell orders travel through systems that take different amounts of time to process data, with older and slower computer systems lagging newer and more sophisticated systems. Additionally, the quality and length of fiber optic cable through which the signals travel affect the speed with which signal transmission occurs.
High frequency traders (HFTs) learned that by investing heavily to outspend competitors, they could pay top programmers to design speedy new software to operate on expensive, state of the art hardware to pass information preferentially through private, higher quality fiber optic cables designed to minimize the lag time from when an order was placed to when it arrived at an exchange to be executed. These advantages were so pronounced that firms paid high fees to "co-locate" their computer systems in buildings closer to the exchanges to further reduce lag time.
The exchanges, which are privately held, for-profit entities, were only too happy to oblige the requests for co-location as a novel revenue stream. They quickly realized that selling proximity to their servers was a service that HFTs were willing to pay dearly to obtain to the exclusion of their competitors.
These technological advances, which amounted to milliseconds, created an accumulation of marginal gains in speed that, when taken as a whole, were critical in amassing billions of dollars in profits while minimizing risk to the HFTs.
Allow me to present a fishing analogy. This is intended to be an illustration of one of many ways that HFTs front-run ordinary investors as well as banking competitors. (They employ additional strategies beyond the scope of an introductory article written by a simpleton like me.)
An HFT might "bait" multiple exchanges at the same time by placing small orders to sell a stock at a particular price. When one or two purchase orders from a client struck their lures and "took the bait" by purchasing their stock at the offered price, this information was relayed more rapidly to the HFT computer systems than it was to average investors. Advanced algorithms residing in the HFTs computer systems were able to interpret the market activity and respond faster than competitors to use the information for profit.
For example, an investment banker at an office in New York sends purchase orders to both the New York Stock Exchange and the Chicago Stock Exchange at the same time as part of a large purchase of a single stock. Each order is routed through a set of complex fiber optic cables with distinctive, consistent lag times. A system designed to receive those orders (by setting "bait" above) and interpret those lag times can often recognize which bank was the source of an order and, critically, infer those cases where these smaller purchases are likely to be preludes to larger purchases yet to come.
Back to our analogy, those initial strikes of bait reveal to the HFT that a school of fish is about to make massive purchase orders. Using their technological advantages, the HFTs would "front-run" the buyer's big order by purchasing all available shares of that stock at a lower price only to immediately turn around and offer the same stock for sale at a higher price, pocketing the difference.
This rapid turnover strategy was unbelievably low risk to the HFTs, as evidenced by the fact that rarely had a year where they lost money. The trades were executed in a matter of milliseconds - and based on the low likelihood that volatile price changes would occur in the fraction of a second during which the HFT actually owned the stock - were sources of perpetual profit as long as the advantages were maintained.
The way this appeared to our investment banker, who was watching the market in real-time on a monitor on a much slower time scale than the milliseconds in which the HFTs operated, was that she placed what she thought was a good faith and presumably instantaneous buy order for 10,000 shares of Widgets Inc. when she saw a stock selling for the favorable price of $1.00, but by the time her order was executed the average price of your order had been jacked up to $1.10. The market she saw was not the market she experienced when she pushed her button.
More to come in part two...