Embedded Files
The main technology featured in the movie was high frequency trading. There are two major components in high frequency trading, which are the high frequency trading algorithm and the method of receiving and sending market orders. The method of receiving and sending market orders that Vincent uses is a fiber optic cable. Later in the movie, Eva manages to introduce microwave towers as a new method instead of fiber optic cables.
High Frequency Trading (HFT) Algorithms
A HFT algorithm in its simplest form is a program that processes multiple market prices in the stock market and decides on whether to place orders depending on the information that it collects. These algorithms were designed to be fast, and they compute in a matter of microseconds. HFT algorithms use their speed to find disparities in stock prices, buying them at the lowest prices and selling them almost instantly at normal prices. These disparities are not necessarily large, and most traders gain less than a cent of profit after selling the stocks. However, these algorithms are fast and allow traders to make thousands of transactions in a short amount of time. Hence, speed becomes an important factor in gaining an advantage over other traders. Whoever has the faster algorithm can process transactions faster and thus gain the largest possible disparity in stock prices.
Anton was able to strip 100 microseconds from the existing high frequency trading algorithm
In the movie, Anton mentions that the algorithm that he created for Eva's company could process transactions in 500 microseconds. After stripping away unnecessary code, he managed to reduce the time to 400 microseconds. Given the nature of high frequency trading, a 100 microsecond advantage is huge as it allows Anton to place an order faster than other traders. As of 2011, the processing speed of most algorithms are under 1 millisecond or 1000 microseconds [5], which makes Anton's algorithm realistic given the time period. However, nowadays, algorithms can process transactions in 64 microseconds [4].
There are a few implications surrounding HFT algorithms, and a quote from the movie's opening scene appropriately summarizes it: "The speed of transactions has become such an important variable in the whole market economy, that faster computer algorithms can make millions in a matter of milliseconds or can crash entire markets by triggering mass defensive responses from automated computer systems." While HFT algorithms could make the traders profitable, it increases market volatility [8]. Even though the disparities in prices that the algorithms take advantage of are small, the transactions happen so often that it can cause constant changes to stock prices. Because of how fast these algorithms operate, any major bugs that occur as a result of a faulty algorithm will incur heavy damages to the stock market before it can be stopped [8]. One example of this is the 2010 flash crash, which will be explained more in the Computer Reliability section.
Fiber Optic Cables
Anton and Vincent want to make a straighter fiber optic cable line to reduce the latency. They managed to reduce the latency to 15.33ms (without processing time).
The method of network communications that Anton and Vincent chose to use was a fiber optic cable. In a fiber optic cable, data is sent in the form of light particles and travels by bouncing off the insides of the cable [6]. A primary factor in the efficiency of a fiber optic cable is its speed. Hence, the distance that the light particles have to travel must be minimized [2]. Anton and Vincent's goal was to make the line as straight as possible to reduce latency, which is the time it takes for information to get from one place to the another. At the start of the movie, the time to beat was 17.22 milliseconds, and Anton's initial algorithm could process transactions in 0.5 milliseconds. This means that the latency of the cable line at the start of the movie was 16.72 milliseconds for a two-way journey, which is 2000 miles in total. After building a straighter line, the time was 15.73 milliseconds, and it took the final algorithm 0.4 microseconds for processing, which led to a latency of 15.33 milliseconds.
The line also included 154 optical regenerators to revive the light particles as they transfer data, and each regenerator has a latency of 0.02 milliseconds. This means that 3.08 milliseconds are being used to travel through the regenerators, leaving a travel time of 12.25 milliseconds in the cables themselves. If that time is divided by 2000 miles, it takes light to travel in a fiber optic cable 6.13 microseconds every mile. In 2011 and up to the present, the latency of fiber optic cables is around 8.05 microseconds per mile [2], and the movie might have exaggerated the speed of these fiber optic cables. The movie also exaggerated the latency of an optical regenerator as its latency in 2011 is around 100 microseconds [3].
Microwave Towers
Eva uses microwave towers to beat Anton and Vincent, managing to process and send orders in 11 milliseconds.
To race against Anton and Vincent's line, Eva used microwave towers as method of data transfer. Eva hired an NYU student who developed a pulse width modulation algorithm that enabled high speed data transfer using the towers. She managed to bring down the latency to 11 milliseconds with a processing time of 0.5 milliseconds. Over a stretch of 2000 miles, the time it takes for microwaves to travel a mile is 5.3 microseconds, which is consistent with today's microwave speed of 5.4 microseconds per mile [1]. Realistically, high frequency trading that incorporates microwave towers for data transfer can be accomplished. In 2013, one of the earliest projects that utilize microwave towers for high frequency trading is a series of microwave tower dishes from London to Frankfurt, which reduced the latency of existing fiber optic cable lines by 40% [1].
High Frequency Trading Competition: Milliseconds Matter
This section has emphasized the need for speed in the world of high frequency trading as it could provide a competitive advantage over other high frequency traders [4]. With a faster network connection and more efficient algorithm, traders can gain access to market orders quicker than their competitors. However, this competition benefits these traders at the expense of other parties involved in the stock market, particularly the investors [7].
There are two types of high frequency traders: market-makers and speculative traders. Market-makers are responsible for putting out the bids onto the stock market. They use trading algorithms to match a seller's bid price with a buyer's ask or offer price very quickly. These buyers are normally investors. The market maker's profit comes from the difference between the ask price and the bid price, which is called the bid-ask spread. Speculative traders target market-makers, and they use high frequency trading algorithms to predict whether the market-maker's bids would increase in price. After calculating the disparity in the price, they will purchase orders from the market-makers and resell that order quickly at a higher price. In the market, speculative traders try to compete with other speculative traders on who can take advantage of the market makers quicker using faster algorithms and faster network speeds. Additionally, market-makers compete with these speculative traders by increasing their bid-ask spread to compensate for speculative traders purchasing and increasing the price of the offers intended for investors. [7].
A study conducted by the European Central Bank details the effect of competition in high frequency trading based on stock data from Nasdaq Stockholm [7]. It measured the volume of high frequency trades and market illiquidity when another high frequency trader enters the market (graphs below). The darker line represent a control group, which only has one high frequency trader. The lighter line initially has one high frequency trader, but a second trader was introduced at the vertical dashed line.
The study found that the introduction of the second trader increases competition and it increased the volume of high frequency trades by 12%. It rose from 29% to 41%.
With this increase in speculative high frequency trades, the illiquidity of the market also increases. Illiquidity is a phenomenon that occurs when a stock cannot be quickly sold because investors are not willing to buy it. When an increase in high frequency trades due to competition occur, the study showed that the illiquidity of the market increased by 18% using the Amihud measure of illiquidity.
When speculative traders buy stocks from the market-makers, they increase the price of the stocks that were intended for investors. This means that market-makers are forced to compensate for that illiquidity by having to increase the bid-ask spread, which often meant that the ask price was increased. Ultimately, the investors are the ones that end up paying more as that ask price increases. Hence, high frequency traders profit at the expense of the investors. [7]
References:
[1] Eric Onstad, Lasers, microwave deployed in high-speed trading arms race, (Reuters, 2013) https://www.reuters.com/article/us-highfrequency-microwave/lasers-microwave-deployed-in-high-speed-trading-arms-race-idUSBRE9400L920130501 (accessed September 28, 2021)
[2] Low Latency - How Low Can You Go. (Infinera, 2020). https://www.infinera.com/wp-content/uploads/Low-Latency-How-Low-Can-You-Go-0188-WP-RevB-0920.pdf (accessed September 28, 2021)
[3] David Mazzarese, Minimizing latency in long-haul networks, (Light Wave Online, 2011), https://www.lightwaveonline.com/network-design/article/16649529/minimizing-latency-in-longhaul-networks (accessed September 28, 2021)
[4] Eric Reed, What Is High-Frequency Trading?, (Smart Asset, 2021), https://smartasset.com/investing/high-frequency-trading (accessed September 28, 2021)
[5] Markets Committee, High Frequency Trading in the Foreign Exchange Market, (Bank For International Settlements, 2011) https://www.bis.org/publ/mktc05.pdf (accessed September 28, 2021) [5]
[6] Craig Freudenrich, How Fiber Optics Work, (HowStuffWorks, 2001), https://computer.howstuffworks.com/fiber-optic1.htm (accessed September 28, 2021)
[7] Breckenfelder, Johannes. Competition among high-frequency traders, and market quality. (European Central Bank, 2019). https://www.ecb.europa.eu/pub/pdf/scpwps/ecb.wp2290~b5fec3a181.en.pdf (accessed October 5, 2021)
[8] 2010 Flash Crash. (Corporate Finance Institute, 2015) https://corporatefinanceinstitute.com/resources/knowledge/trading-investing/2010-flash-crash/ (accessed October 1, 2021)
Page updated
Google Sites
Report abuse