Incident: Flash Crash in India's Stock Market due to Trading Glitch

Published Date: 2012-10-05

Postmortem Analysis
Timeline 1. The software failure incident of the flash crash in India's main share index happened on Friday [Article 15172]. 2. Published on 2012-10-05. 3. The software failure incident occurred on October 5, 2012.
System 1. Trading technology system at India's National Stock Exchange (NSE) [Article 15172] 2. Automation process system at US broker Knight Capital [Article 15172]
Responsible Organization 1. A brokerage placed 59 wrong orders, triggering a sell-off that led to the flash crash in India's National Stock Exchange (NSE) [Article 15172]. 2. High-frequency trading, which involves using software to post orders for microseconds at a time, was mentioned as a factor that could have exacerbated the error leading to the flash crash [Article 15172].
Impacted Organization 1. India's National Stock Exchange (NSE) [Article 15172] 2. Investors in global financial markets [Article 15172]
Software Causes 1. High-frequency trading using software to post orders for microseconds at a time, exploiting tiny differences in share prices [15172] 2. Bug in the automation process at US broker Knight Capital that generated thousands of mistaken orders for stocks, resulting in a $440m loss in 45 minutes [15172]
Non-software Causes 1. Human error: The failure incident was blamed on human error by the NSE, although experts like Joseph Saluzzi doubted this explanation [15172]. 2. High-frequency trading: The use of high-frequency trading, which involves software posting orders for microseconds at a time to exploit small differences in share prices, was identified as a factor that could have exacerbated the error [15172].
Impacts 1. The software failure incident led to a 16% plunge in India's main share index within minutes, resulting in a "flash crash" and causing a sell-off that wiped nearly $60 billion off the value of the country's biggest companies [Article 15172]. 2. The National Stock Exchange (NSE) had to halt trading for 15 minutes due to the incident [Article 15172]. 3. The incident damaged the credibility of exchanges and rocked investors' faith in financial markets, as seen in previous trading errors that have occurred [Article 15172]. 4. The glitch caused by the software failure led to a $440 million loss in 45 minutes for US broker Knight Capital after thousands of mistaken orders for stocks were generated due to a bug in the automation process [Article 15172]. 5. The incident raised concerns about the impact of trading technology on the stability of markets and highlighted the risks associated with high-frequency trading, which can exacerbate errors and market instability [Article 15172].
Preventions 1. Implementing stricter controls and checks on trading algorithms to prevent erroneous orders from being placed [15172]. 2. Conducting thorough testing and quality assurance of trading software to identify and rectify any bugs or faults before deployment [15172]. 3. Enhancing monitoring and surveillance systems to quickly detect abnormal trading patterns or glitches and take corrective action promptly [15172]. 4. Providing adequate training and education to traders and brokers on the proper use of trading technology to minimize the risk of human errors leading to market disruptions [15172].
Fixes 1. Implementing stricter regulations and oversight on high-frequency trading practices to prevent market manipulation and excessive volatility [15172]. 2. Enhancing software testing and quality assurance processes to detect and prevent bugs or glitches that could lead to erroneous trading orders [15172]. 3. Developing fail-safe mechanisms within trading systems to automatically halt trading or cancel orders in case of abnormal market behavior triggered by software failures [15172].
References 1. India's National Stock Exchange (NSE) [Article 15172] 2. Joseph Saluzzi, co-founder of US agency broker Themis Trading [Article 15172]

Software Taxonomy of Faults

Category Option Rationale
Recurring one_organization, multiple_organization (a) The software failure incident having happened again at one_organization: - The article mentions that earlier in the year, US broker Knight Capital suffered a $440m loss in 45 minutes due to a bug in the automation process that generated thousands of mistaken orders for stocks [15172]. - Additionally, the article refers to the botched Facebook flotation where Nasdaq's system crashed under the weight of demand for the stock, indicating a previous software failure incident within the same organization [15172]. (b) The software failure incident having happened again at multiple_organization: - The article highlights that the flash crash of May 2010, which saw the Dow Jones industrial average plunge by 998 points in 20 minutes, was fueled by high-frequency trading, raising fears of a worldwide stock market collapse [15172]. - It also mentions that European legislators have been moving to limit high-frequency trading, with the European Parliament voting to force trading venues to slow down the speed of orders, indicating concerns and incidents related to high-frequency trading across multiple organizations [15172].
Phase (Design/Operation) design (a) The software failure incident mentioned in the articles is related to the design phase. The incident at India's National Stock Exchange (NSE) was caused by a brokerage placing 59 wrong orders, triggering a sell-off that led to a flash crash and a significant drop in the market index. This incident highlights the impact of trading technology on market stability and the credibility of exchanges, indicating a failure due to contributing factors introduced by system development or updates [15172]. (b) The articles do not provide specific information about the software failure incident being related to the operation phase.
Boundary (Internal/External) within_system (a) within_system: The software failure incident related to the flash crash in India's National Stock Exchange was attributed to a brokerage placing 59 wrong orders, triggering a sell-off that led to a 16% plunge in the main share index. This incident highlighted the impact of trading technology on market stability and raised concerns about high-frequency trading exacerbating errors [15172]. (b) outside_system: The article does not explicitly mention any contributing factors originating from outside the system that led to the software failure incident.
Nature (Human/Non-human) non-human_actions, human_actions (a) The software failure incident occurring due to non-human actions: - The article mentions instances of trading errors and glitches that have impacted financial markets, such as the flash crash in India's National Stock Exchange and the market glitch causing a significant increase in Kraft's share price on Nasdaq [15172]. - High-frequency trading, which involves using software to post orders for microseconds at a time to exploit tiny differences in share prices, is highlighted as a contributing factor to market instability and flash crashes [15172]. (b) The software failure incident occurring due to human actions: - The NSE blamed the flash crash on human error, specifically mentioning that a brokerage placed 59 wrong orders, triggering a sell-off that led to a significant drop in the share index [15172]. - However, experts like Joseph Saluzzi expressed skepticism about attributing the crash solely to human error, suggesting that high-frequency trading and automation processes could have exacerbated any errors made [15172].
Dimension (Hardware/Software) software (a) The software failure incident related to hardware: - The article mentions that a brokerage placed 59 wrong orders, triggering a sell-off that wiped nearly $60bn off the value of India's biggest companies. This incident was attributed to human error by the NSE, but some experts like Joseph Saluzzi doubted this explanation, stating that there is no human being in the world that can take down the stock market by 16% [15172]. (b) The software failure incident related to software: - The article highlights the role of high-frequency trading in exacerbating errors in trading. High-frequency trading involves using software to post orders for microseconds at a time to exploit tiny differences in share prices. It is mentioned that high-frequency trading was widely accepted to have fueled the flash crash of May 2010, where the Dow Jones industrial average plunged by 998 points in 20 minutes [15172].
Objective (Malicious/Non-malicious) non-malicious (a) The articles mention incidents that point towards non-malicious software failures: 1. The flash crash in India's main share index was attributed to a brokerage placing 59 wrong orders, triggering a sell-off and causing a significant drop in the market value. This was described as a trading error and blamed on human error [15172]. 2. The article also discusses other instances of trading errors, such as Nasdaq having to cancel trading in Kraft due to a market glitch and Knight Capital incurring a $440m loss in 45 minutes after a bug in the automation process generated thousands of mistaken orders for stocks. These incidents were not attributed to malicious intent but rather to technical glitches and errors in the trading systems [15172].
Intent (Poor/Accidental Decisions) poor_decisions (a) The software failure incident mentioned in the articles was primarily attributed to poor decisions rather than accidental decisions. The incident at India's National Stock Exchange (NSE) was caused by a brokerage placing 59 wrong orders, triggering a sell-off that led to a 16% plunge in the main share index. This was described as a "flash crash" and resulted in nearly $60 billion being wiped off the value of the country's biggest companies [15172]. Additionally, the article highlighted other instances of trading errors and glitches in the financial markets, such as Nasdaq having to cancel trading in Kraft due to a market glitch, Knight Capital incurring a $440 million loss in 45 minutes due to a bug in the automation process, and the flash crash of May 2010 that raised fears of a worldwide stock market collapse. These incidents point towards a pattern of poor decisions and errors in trading technology that have shaken confidence in financial markets [15172].
Capability (Incompetence/Accidental) development_incompetence, accidental (a) The articles mention incidents that can be attributed to development incompetence. For example, the article discusses how a brokerage placed 59 wrong orders, triggering a sell-off that wiped nearly $60 billion off the value of India's biggest companies. This incident highlights the impact of human error on market stability [15172]. (b) The articles also touch upon accidental failures. One instance is the case of US broker Knight Capital, which incurred a $440 million loss in 45 minutes due to a bug in the automation process that generated thousands of mistaken orders for stocks. This incident demonstrates how accidental software glitches can have significant financial repercussions [15172].
Duration temporary (a) The articles do not mention any software failure incident that resulted in a permanent failure where contributing factors were introduced by all circumstances. (b) The articles discuss temporary software failure incidents caused by specific circumstances. For example, the flash crash in India's National Stock Exchange (NSE) was triggered by a brokerage placing 59 wrong orders, leading to a sell-off and a 16% plunge in the main share index. Similarly, incidents like the market glitch causing a 30% surge in Kraft's share price on Nasdaq and the $440m loss by US broker Knight Capital due to a bug in the automation process highlight temporary software failures caused by specific factors [15172].
Behaviour omission, value, other (a) The articles mention incidents of crashes in the financial markets due to software failures. For example, the flash crash in India's main share index led to a plunge of 16% within minutes, triggering a sell-off and wiping off nearly $60bn in value [Article 15172]. (b) The incident involving the brokerage placing 59 wrong orders that triggered the sell-off in the Indian stock market can be seen as a case of omission where the system failed to perform its intended functions correctly at that instance [Article 15172]. (c) The timing of the software failure incident is evident in the fact that the National Stock Exchange (NSE) had to halt trading for 15 minutes to address the glitch caused by the wrong orders, which affected the market [Article 15172]. (d) The software failure incident resulted in incorrect performance of the system, as seen in the $440m loss incurred by US broker Knight Capital in 45 minutes due to a bug in the automation process generating thousands of mistaken orders for stocks [Article 15172]. (e) The behavior of the software failure incident can also be categorized as byzantine, as high-frequency trading fueled the flash crash of May 2010, raising fears of a worldwide stock market collapse due to erroneous behavior and inconsistent responses in the market [Article 15172]. (f) In addition to the mentioned behaviors, the software failure incident can be classified as a glitch, where the market experienced a series of trading errors that damaged the credibility of exchanges and rocked investors' faith in financial markets [Article 15172].

IoT System Layer

Layer Option Rationale
Perception None None
Communication None None
Application None None

Other Details

Category Option Rationale
Consequence property, non-human, theoretical_consequence, other (a) death: There is no mention of any deaths resulting from the software failure incident in the provided article [15172]. (b) harm: The article does not mention any physical harm caused to individuals due to the software failure incident [15172]. (c) basic: The incident did not impact people's access to food or shelter [15172]. (d) property: The software failure incident resulted in a significant financial impact as nearly $60 billion was wiped off the value of the country's biggest companies due to the flash crash triggered by wrong orders placed by a brokerage [15172]. (e) delay: There is no mention of any activities being postponed due to the software failure incident in the article [15172]. (f) non-human: The software failure incident impacted the stock market, leading to a plunge in India's main share index and causing a sell-off that affected the value of companies [15172]. (g) no_consequence: The software failure incident had real consequences, such as the plunge in the share index and the financial impact on companies [15172]. (h) theoretical_consequence: The article discusses potential consequences of trading errors and glitches on the stability of financial markets, as well as the concerns raised by regulators about the impact of trading technology [15172]. (i) other: The article mentions the loss of $440 million by US broker Knight Capital in 45 minutes due to a bug in the automation process, highlighting another financial consequence of a software failure incident [15172].
Domain finance (a) The failed system was related to the finance industry as it caused a flash crash in India's main share index, leading to a significant drop in value and trading disruptions on the National Stock Exchange [Article 15172]. (h) The incident was specifically related to the finance industry, as it involved trading errors, glitches, and high-frequency trading practices that impacted the stability and credibility of financial markets [Article 15172].

Sources

Back to List