7 DESCRIPCIÓN DEL DISEÑO Y EL FUNCIONAMIENTO
7.6 Componentes principales
The financial service system has several major pools of data which are held by different stakeholders/parties. Data are classified into three major categories:
1. Structured data. This refers to information with a high degree of organisation, such that inclusion in a relational database is seamless and readily searchable by simple, straightforward search engine algorithms or other search operations. Examples of financial structured data sources are:
o Trading systems (transaction data)
o Account systems (data on account holdings and movements)
o Market data from external providers (e.g. Interactive Data, Reuters, Bloomberg) o Securities reference data
o Price information o Technical indicators
o Reference data (identifiers)
o Instrument master data
o Human Resources systems (employee data) o Customers’ profiles
o Portfolios
o Financial vocabularies (e.g., reputation vocabulary) and ontologies
2. Unstructured data. Although the financial industry has previously focused on high velocity market data, it is now moving towards unstructured data to changing trading dynamics. Unstructured text data usually does have some structure but it does not follow a record layout, nor are there any embedded metadata tags describing attributes. Examples of financial unstructured data are:
o Daily stock feeds
o Company announcements (ad-hoc news) o Online news media
o Articles/Blogs: people writing answers to questions, explaining business operations and strategies, describing processes and accounting for situations, incidents, and events that create risks, gains and/or losses.
o Documents
o Social networking media: websites, chats, tweets. o Multimedia (audio, video)
o Customers’ feedback/experiences.
3. Semi-structured data. It is a form of structured data that does not conform to the formal structure of data models associated with relational databases or other forms of data tables, but even so contains tags or markers to separate semantic elements and enforce hierarchies of records and fields within the data. Examples of semi-structured data are expressed in meta-languages (mostly XML-based) such as:
o Financial products Markup Language (FpML) for complex financial products. o Financial Information eXchange (FIX) for international real-time exchange of
information related to the securities transactions and markets.
o Interactive Financial Exchange (IFX) for electronic bill presentment and payment, business to business payments, business to business banking (such as balance and transaction reporting, remittance information), automated teller machine communications, consumer to business payments, and consumer to business banking.
o Market Data Definition Language (MDDL) is a XML specification to enable interchange of data necessary to account for, to analyze, and to trade instruments of the world's financial markets.
o Financial Electronic Data Interchange (FEDI) for exchange of payments between businesses, customers and vendors.
o Open Financial Exchange (OFX) for the electronic exchange of financial data between financial institutions, businesses and consumers via the Internet.
o eXtensible Business Reporting Language (XBRL) to describe financial information for public and private companies and other organisations.
o SWIFTStandards to support transactions in the financial markets for payments, securities, treasury and trade services.
Nowadays the amount of unstructured information in enterprises is around 80-85%. The financial and insurance industry has vast repositories of structured data in comparison to other industries, as shown in the figure below. A large amount of this information has its origin inside the organisations.
Figure 15: Industry Comparisons on Sources and structure of Data Source: Tata consultancy services.
The bulk of data is around financial, economic and trade data, nonetheless a growing segment of data constitute regulatory and compliance data generated and collected across the firms, as shown in the figure below.
Source: Squirro. Use of Unstructured Data in Financial Services 2014.
4.4.4 Drivers and Constraints
4.4.4.1
Drivers
Data Growth
Perhaps the most obvious driver is that financial transaction volumes are growing leading to data growth in financial services firms. In Capital Markets, the presence of electronic trading has led to a decrease in the value of individual trades and an increase in the number of trades. The advent of high turnover trading strategies generates considerable order flow and an even larger stream of price quotes. Data growth is not limited to capital markets businesses. The Capgemini/RBS Global Payments study for 2012 (Capgemini, 2012) estimates that the global number of electronic payment transactions is about 260 billion and growing between 15 and 22% for developing countries. As devices that consumers can use to initiate core transactions multiply, so too do the number of transactions they make. Not only is the transaction volume increasing, the data stored for each transaction are also expanding.
Increasing scrutiny from Regulators
Regulators of the industry now require a more transparent and accurate view of financial and insurance businesses, this means that they no longer want reports; they need raw data. Therefore financial institutions need to ensure that they are able to analyse their raw data at the same level of granularity that regulators will be. Financial Institutions need to be able to identify any issues before the regulators do. This means, financial institutions in fact have no choice, but to deal with Big Data.
Advancements in technology means increased activity.
Thanks largely to the digitization of financial products and services, the ease and affordability of executing financial transactions online has led to ever-increasing activity and expansion into new markets. Individuals can make more trades, more often, across more types of accounts, because they can do so with the click of a button in the comfort of their own homes. Instead of visiting a broker to carry out to obtain insurance quotations, users are able to do this on the move with their mobile device. Increased access and ease of use translates into increased activity, which in turn translates into rapidly growing data volumes.
Changing business models
Driven by the aforementioned factors, financial institutions find themselves in a market that is fundamentally different from the market of even a few years ago. Successful organisations must be able to quickly apply changes and build agility into their business models or risk losing market share and the confidence of customers. Adoption of Big Data analytics is also necessary to help build business models for financial institutions geared towards retention of market share from the increasing competition coming from other sectors. For example ecommerce businesses have leveraged Big Data to start supplying consumer credit, if financial institutions do not adapt their business models, they risk losing their customer base to organisations who traditionally do not provide financial services.
Customer insight
Up until a decade ago, it may be said that banks owned the relationship with consumers, as primary source of the consumer’s identity for all financial, and many non-financial transactions. Today the relationship is reversed: consumers now have transient relationships with multiple banks. Moreover, even collectively, financial institutions no longer monopolize a consumer’s financial transactions. Banks no longer have a complete view of their customer’s preferences, buying patterns and behaviours. Gaining full understanding of a customer’s preferences and interests are prerequisites for ensuring that banks can address customer satisfaction (in order to prevent customer churn) and for building more extensive and complete propensity models. Banks must therefore bring in external sources of information, information that is often unstructured. Big Data technologies therefore play a focal role in enabling customer centricity in this new paradigm.
4.4.4.2
Constraints
Old culture and infrastructures.
Many banks still depend on old rigid IT infrastructure, with data siloes and a great many legacy systems. Big Data, therefore, is an add-on, rather than a completely new standalone initiative. Banks need to extract the legacy data out, streamline it, build the traceability and lineage. There may be a lack of organisational agility and skills training, while the key requirement for customer permission to use data must also be addressed.
Here regulation plays also a critical role. Just as banks may wish to streamline their processes, they are being asked to add additional process stages, to take up higher capital burdens and generally act in ways that are opposed to business agility.
Culture is even a bigger barrier to Big Data deployment. Many financial organisations fail to implement Big Data programs because they are unable to appreciate how data analytics can improve their core business.
A lack of skills.
A common theme can be noted across the industry regarding the challenge of having access to the right level of skills. Some organisations have recognised the data and the opportunities the data presents; however they lack human capital with the right level of skills to be able to bridge the gap between data and potential opportunity. The skills which are ‘missing’ are those of a data scientist. A data scientist can be identified as having three keys skills:
Commercial – The ability to translate business problems into technology solutions.
Analytical – Strong statistical/analytical skills with a background particularly geared towards unstructured data mining. These are not to be confused with the skills of a statistician.
Technical – Strong scientific or technical skills for example to be able to write scripts and really extract the core value from data.
These skills do exist in isolation in the industry, but in depth re-skilling is required to produce the human capital who can really extract value from Big Data. Some financial services institutes have sought the skills from outside their own organisation by partnering with specialists in this area; however an observation can be made to suggest resources possessing all three core skills are not in abundance.
Data ‘Actionability’.
The next main challenge can be seen in making Big Data actionable. Big Data technology and analytical techniques enable financial services institutes to get deep insight into customer behaviour and patterns, but the challenge still lies in organisations being able to take specific action based on this data. A hypothetical example: if an organisations’ Big Data initiative can identify in real time a specific target group of customers who are 90% likely for the uptake of a new offering, but the organisation does not have the technology and the processes to be able to leverage that information in real time, the information is rendered useless. So the challenge in organisations is still integrating intelligence from Big Data into operations.
Data privacy and security.
Customer data is a continuing cause for concern. Regulation remains a big unknown: What is and is not legally permissible in the ownership and use of customer data remains ill-defined, and that is an inhibiting factor to rapid and large-scale adoption. Here, regulations still differ from place to place, while practice changes according to generations. Members of “Generation Y” are more open to sharing information and there is a growing move towards the concept of data portability (similar to mobile industry). This might seem to encourage churn but it might also encourage higher levels of trust between customers and banks.