Ch19: Data Flashcards
POPIA (Personal Information Act) eight conditions:
- Accountability
+ Party responsible for processing the data is also responsible for compliance with POPIA - Processing limitation
+ Information must be processed in a fair, lawful and relevant manner, after consent is given by the data subject - Purpose specification
+ Personal information must be collected for a specific purpose. Records should be destroyed when personal data is no longer relevant or authorized to be held. - Further processing limitation
+ Further processing must be compatible with the initial collection purpose - Information quality
+ Data completeness, accuracy and updates to be ensured by the holder of the data - Openness
+ Documentation to be maintained on all processing operations and maintaining transparency on data use - Security safeguards
+ Integrity and confidentiality of personal data must be secured . Notification to be done on security compromises - Data subject participation
+ Data subject may request conformation of personal data held and request correction or deletion of any inaccurate, misleading or outdated information held.
Personal data definition
Information in respect of an individual where individual can be identified, or data combined with other information could allow individual to be identified
Competition legislation, two examples
- Anti-competitive agreements
+ Data could be shared among a small number of companies to fix prices in a particular
market - Abuse of dominant market position
+ Imposing unfair trading terms, such as exclusivity
Three Big Data characteristics
- Very large datasets
- Data brought together from different sources
- Data which can analyzed quickly - such as in real time
Data governance definition
Data governance is the term used to describe the overall management of the availability, usability, integrity and security of the data employed in an organization
Data governance policy will set out guidelines with regards to: (6)
- Specific roles and responsibilities of individuals in the organization with regards to data
- How the organization will capture, analyze and process data
- Issues with respect to data security and privacy
- Controls that will be put in place to ensure that the required data standards are applied
- How the adequacy of the controls will be monitored on an ongoing basis with respect to data usability, accessibility, integrity and security.
- Provide mechanism for ensuring that relevant legal and regulatory requirements in relation to data management are met
Poor data governance risks (4)
- Legal and regulatory non-compliance
- Inability to reply on data for decision making
- Reputational issues
- Incurring additional costs (fines, legal costs)
Key risks relating to data: (6)
- Data are incomplete or inaccurate
- Data is not credible due to having insufficient volume, particularly for estimation of extreme outcomes
- Data is not sufficiently relevant to the intended purpose
- Past data may not reflect what will happen in the future
- Chosen data groups may not be optimal
- Data are not available in an appropriate form for its intended purpose
Algorithmic decision making description (5)
- Form of automated trading that involves buying or selling financial securities electronically to capitalize on price discrepancies for the same stock or asset in different markets
- Often many trades are carried out very quickly to take advantage of temporary price discrepancies, with aim of making small profits on each trade
- Trader will use a formula to decide whether a financial asset should be traded.
- Parameters underlying the algorithm used to determine when assets should be traded will need to be derived using data from appropriate sources
- Quality of these automated decisions depends on robustness of the programmed trading rules which in turn rely on the data used.
Benefits and risks relating to algorithmic trading (5&5)
Benefits:
- Potentially lead to quicker, more consistent and fairer decisions
- Potentially lead to lower dealing costs on trades
- Potentially facilitate execution of complex trading strategies that would not previously have been possible.
- Computer sciences advances allow for even greater amounts of data to be stored and analyzed much more quickly in the past
- Offered new tools for investment decisions, trading execution and risk management
Risks:
- Could be an error in the algorithm or the data used to parameterize the model could be wrong, leading to potential losses on each trade rather than expected profits. Issue when a large number of trades could be completed very quickly
- Algorithm may not operate properly in adverse conditions, e.g. could stop trading assets in turbulent markets, reducing liquidity of the asset and increasing volatility
- In very turbulent conditions, trading in individual stocks or even entire markets may be suspended before an algorithmic trade could have been completed
- Main risk is possible impact on financial system, large number of quick trades increases volatility of financial markets
- Increased integration between markets and asset classes means that a meltdown in one market could impact other markets and asset classes
Benefits of single integrated data system (5)
- Reduced chance of existing data becoming corrupted
- Reduced chance of inconsistent treatment of information, between products or over time
- Likely to be better level of control over those who may enter information or amend information
- Information will be easier to access. as it will not involve collating information form several systems
- Time will not need to be spent reconciling data from different systems
Data sources
- Publicly available data (published accounts, national statistics)
- Internal data
- External sources (reinsurer)
- Industry-wide data collection schemes
Possible reasons for heterogeneity in data when using industry-wide data (6)
- Companies operate in different geographical or socio-economic sections of the market
- Policies sold by different companies are not identical
- Sales methods are not identical
- Companies will have different practices e.g. underwriting or claim settlement standards
- Nature of data being stored by different companies will not always be the same
- Coding used for risk factors may vary between organizations
More problems with using industry-wide data may be: (4)
- Data will usually be less detailed or less flexible than those available internally
- External data often more out of date than internal data
- Data quality will depend on the quality of the data systems of all its contributors
- Not all organizations contribute, thus those who do, do not represent the market as a whole
