Incident: Trading Glitch at Goldman Sachs Causes Erroneous Equity Options Orders

Published Date: 2013-08-21

Postmortem Analysis
Timeline 1. The software failure incident at Goldman Sachs occurred on Tuesday morning [21144]. 2. Published on 2013-08-21. 3. The incident likely occurred on Tuesday morning in August 2013.
System 1. Internal computer system at Goldman Sachs [Article 21144] 2. Computer algorithm used by an unnamed investor [Article 21144] 3. Software at Knight Capital that affected trading activity in nearly 150 NYSE-listed stocks [Article 21144]
Responsible Organization 1. An internal computer system glitch at Goldman Sachs caused the firm to issue incorrect equity options orders [21144]. 2. An unnamed investor using a computer algorithm triggered the "flash crash" in May 2010 [21144]. 3. Knight Capital experienced a high profile trading glitch due to a software snafu affecting trading activity [21144].
Impacted Organization 1. Goldman Sachs 2. NYSE Euronext's options exchange 3. Nasdaq OMX's options markets 4. Chicago Board Options Exchange 5. Knight Capital [21144]
Software Causes 1. Technical glitch in an internal computer system at Goldman Sachs causing incorrect equity options orders to be issued [21144] 2. Software snafu at Knight Capital affecting trading activity in nearly 150 NYSE-listed stocks [21144]
Non-software Causes 1. High speed trading dominated by computers and algorithms [21144] 2. Human error such as a "fat fingered" trader typing incorrect values [21144]
Impacts 1. Incorrect equity options orders were issued to various options exchanges by Goldman Sachs due to a technical glitch in an internal computer system, leading to erroneous trades [21144]. 2. The exact number of trades affected and the total losses incurred were still being calculated and were not disclosed by Goldman Sachs [21144]. 3. The software failure incident added to the list of trading glitches in recent years, highlighting the risks associated with algorithmic trading and the potential for significant financial losses [21144].
Preventions 1. Implementing rigorous testing procedures: Conducting thorough testing, including unit testing, integration testing, and system testing, could have helped identify any potential issues in the software before it went live [21144]. 2. Implementing proper risk management protocols: Having robust risk management protocols in place could have helped in quickly identifying and mitigating the impact of the glitch [21144]. 3. Implementing stricter controls on algorithmic trading: Enforcing stricter controls and oversight on algorithmic trading activities could have reduced the likelihood of errors or glitches causing significant market disruptions [21144].
Fixes 1. Implementing stricter testing protocols for software updates and changes to prevent errors from going live [21144]. 2. Conducting thorough code reviews and quality assurance checks to catch any potential issues before they impact trading activities [21144]. 3. Enhancing training and oversight for employees using algorithm software to ensure accurate inputs and prevent costly mistakes [21144].
References 1. Goldman Sachs spokesperson David Wells [Article 21144] 2. NYSE Euronext's options exchange [Article 21144] 3. Nasdaq OMX's options markets [Article 21144] 4. Chicago Board Options Exchange (CBOE) [Article 21144] 5. Dennis Dick of Premarketinfo.com [Article 21144]

Software Taxonomy of Faults

Category Option Rationale
Recurring one_organization, multiple_organization (a) The software failure incident having happened again at one_organization: - Goldman Sachs experienced a technical glitch in an internal computer system that caused incorrect equity options orders to be issued, resulting in erroneous trades [21144]. - Knight Capital, another organization, also faced a high-profile trading glitch in the past, losing $440 million due to a software snafu affecting trading activity [21144]. (b) The software failure incident having happened again at multiple_organization: - The article mentions that smaller trading glitches have made their mark in 2012, citing examples like the botched IPOs of the BATS exchange and Facebook, highlighting the frailty of markets in a high-speed trading world [21144]. - The "flash crash" of 2010, which caused significant market disruption, was triggered by an unnamed investor using a computer algorithm, selling a large number of futures contracts in a short period of time [21144].
Phase (Design/Operation) design, operation (a) The software failure incident related to the design phase can be seen in the article where it mentions that a technical glitch in an internal computer system at Goldman Sachs caused the firm to issue incorrect equity options orders to various options exchanges [21144]. This indicates that the failure was due to contributing factors introduced during the system development or updates. (b) The software failure incident related to the operation phase is evident in the same article where it states that Knight Capital experienced a high-profile trading glitch last August, resulting in the firm losing $440 million after a software snafu affected trading activity in nearly 150 NYSE-listed stocks [21144]. This highlights a failure due to contributing factors introduced by the operation or misuse of the system.
Boundary (Internal/External) within_system, outside_system (a) within_system: The software failure incident at Goldman Sachs was caused by a technical glitch in an internal computer system, leading to the issuance of incorrect equity options orders to various options exchanges [21144]. This indicates that the failure originated from within the system itself. (b) outside_system: The article mentions that the "flash crash" in May 2010, which caused significant market disruptions, was triggered by an unnamed investor using a computer algorithm that sold a large number of futures contracts in a short period of time [21144]. This incident highlights how external factors, such as investor actions and market conditions, can also contribute to software failures.
Nature (Human/Non-human) non-human_actions, human_actions (a) The software failure incident occurring due to non-human actions: - The article mentions a technical glitch in an internal computer system at Goldman Sachs that caused incorrect equity options orders to be issued to various options exchanges [21144]. - The "flash crash" in May 2010, which caused a significant market disruption, was triggered by an unnamed investor using a computer algorithm that sold a large number of futures contracts in a short period of time [21144]. - Knight Capital's high profile trading glitch in August, which resulted in a loss of $440 million, was due to a software snafu affecting trading activity in nearly 150 NYSE-listed stocks [21144]. (b) The software failure incident occurring due to human actions: - The article mentions the possibility of a "fat-fingered" Citibank trader accidentally typing "billion" instead of "million" on the number of trades, contributing to the flash crash in May 2010 [21144]. - Dennis Dick of Premarketinfo.com highlighted the importance of being cautious when using algorithm software and the potential for costly mistakes due to human errors like key stroke errors [21144].
Dimension (Hardware/Software) hardware, software (a) The software failure incident related to hardware: - The article mentions a technical glitch in an internal computer system at Goldman Sachs that caused incorrect equity options orders to be issued to various options exchanges [21144]. - The incident at Knight Capital, where a software snafu affected trading activity in nearly 150 NYSE-listed stocks, resulted in the firm losing $440 million [21144]. (b) The software failure incident related to software: - The article discusses how the "flash crash" in May 2010 was triggered by an unnamed investor using a computer algorithm that sold a large number of futures contracts in a short period of time [21144]. - It is highlighted that smaller trading glitches in 2012, such as the botched IPOs of the BATS exchange and Facebook, emphasized the frailty of markets in a high-speed trading world [21144].
Objective (Malicious/Non-malicious) non-malicious (a) The software failure incident at Goldman Sachs was non-malicious. It was caused by a technical glitch in an internal computer system that led to incorrect equity options orders being issued to various options exchanges [21144]. The incident was described as a "snafu" and not as a deliberate act to harm the system.
Intent (Poor/Accidental Decisions) poor_decisions (a) The software failure incident related to the trading glitch at Goldman Sachs can be attributed to poor decisions. The incident was caused by a technical glitch in an internal computer system that led to the issuance of incorrect equity options orders to various options exchanges. This glitch resulted in erroneous trades, impacting the markets and causing financial losses. The incident highlights the importance of ensuring the accuracy and reliability of algorithm software before deployment to prevent such costly errors [21144]. (b) Additionally, the article mentions the "flash crash" of 2010, where a significant market disruption occurred due to an unnamed investor using a computer algorithm that sold a large number of futures contracts in a short period of time. This incident was not attributed to accidental decisions but rather to the use of algorithm software without proper oversight and risk management, leading to a drastic market event [21144].
Capability (Incompetence/Accidental) development_incompetence, accidental (a) The software failure incident related to development incompetence is evident in the article as it mentions the trading glitch at Goldman Sachs caused by a technical glitch in an internal computer system, leading to incorrect equity options orders being issued to various options exchanges [21144]. This indicates a failure due to contributing factors introduced due to lack of professional competence by humans or the development organization. (b) The software failure incident related to accidental factors is also highlighted in the article, particularly in the context of the "flash crash" of 2010, where an unnamed investor using a computer algorithm triggered a crash by selling a large number of futures contracts in a short period of time. This incident was not intentional but resulted from accidental actions within the automated trading system [21144].
Duration temporary The software failure incident reported in the articles can be categorized as a temporary failure. The incident at Goldman Sachs was caused by a technical glitch in an internal computer system that led to the issuance of incorrect equity options orders to various options exchanges [21144]. This glitch resulted in erroneous trades, and the exchanges were working to resolve the issue. The impact of the failure, in terms of dollars lost, was still being calculated, indicating a temporary disruption caused by specific circumstances rather than a permanent failure introduced by all circumstances.
Behaviour value, other (a) crash: The article mentions a software failure incident where Goldman Sachs issued incorrect equity options orders due to a technical glitch in an internal computer system, leading to erroneous trades on various options exchanges [21144]. (b) omission: The article does not specifically mention a case where the system omitted to perform its intended functions at an instance(s). (c) timing: The article does not mention a failure due to the system performing its intended functions correctly, but too late or too early. (d) value: The software failure incident described in the article falls under this category as the system performed its intended functions incorrectly by issuing incorrect equity options orders, resulting in erroneous trades [21144]. (e) byzantine: The article does not describe a failure due to the system behaving erroneously with inconsistent responses and interactions. (f) other: The other behavior observed in the software failure incident is the potential loss of dollars due to the glitch, which is still being calculated. The impact in terms of dollars lost is uncertain, and the exchanges are working to resolve the issue [21144].

IoT System Layer

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

Other Details

Category Option Rationale
Consequence property, theoretical_consequence (a) unknown (b) unknown (c) unknown (d) The software failure incident at Goldman Sachs caused the firm to issue incorrect equity options orders to various options exchanges, leading to erroneous trades and potential losses. The impact in terms of dollars lost was still being calculated [21144]. (e) unknown (f) unknown (g) unknown (h) The potential loss from the software failure incident at Goldman Sachs was stated not to be material to the financial condition of the firm [21144]. (i) The article mentions that the exchanges were working to resolve the issue caused by the software glitch, and it was unclear how many trades were affected. The incident was part of a series of trading glitches in recent years as trading becomes more dominated by computers and high-speed traders [21144].
Domain finance (a) The failed system in the incident reported in Article 21144 was related to the finance industry. Goldman Sachs experienced a technical glitch in an internal computer system that caused the firm to issue incorrect equity options orders, impacting trades on various options exchanges [21144]. The incident involved erroneous trades and the exchanges were working to resolve the issue, highlighting the impact on financial markets due to software failures in the finance sector.

Sources

Back to List