Analytics, big data and financial governance

Shishir Shahnawaz | Published: August 06, 2019 21:48:39 | Updated: August 09, 2019 21:32:54

What if Bangladesh's financial sector had an access to information and data on its clients? Lending and deposit taking institutions would then systematically understand how clients behave at every financial touch-point, from opening an account to the time accounts become dormant, default or downgraded. Financial institutions would then highlight their best practices in making better decisions through big data and analytics. Year-end results: improved financial planning and risk hedging. Ultimate goal: more predictable 'liquid assets', among others, in the national income!

Such solutions are embedded in big data and analytics. In other words, data analytics can help in making better decisions. In a layman's term, big data is extremely large data sets that may be analysed computationally to reveal patterns, trends and associations. Big data grows exponentially with time. It is so large and complex that none of the traditional data management tools are able to store it or process it efficiently. A number of advanced economies are adopting big data analytics to particularly understand behavioural insights and human interactions.

Is big data important for Bangladesh? For financial wellbeing, fraud fighting, and avoiding financial vulnerabilities, big data and analytics can improve results with informed decisions. Data analytics can then take the 'guesswork' out of existing financial products and practices, and enable financial institutions to improve financial governance. Each institution will then be well equipped to identify the most at-risk clients as well as address any mischievous activities, such as identity thefts or intentionally skipping scheduled instalment payments. It is time that Bangladesh's financial sector plans to institutionalise data analytics to serve as a streamlined, more cost-effective platform for data-driven solutions while maximising resources and budgets.

Employees in this sector have already been doing analyses on the basis of data through an expanded provision of credit information and ratings, whereas an emergence of new credit information companies are providing relevant data to the financial sector. But there are shortfalls.

Most information on customer credit worthiness in Bangladesh offer miniscule information on consumer behaviour and postal code or geographic location. There are two organisations providing credit information: one is Credit Rating Information & Services Limited (CRISL), a private company that rates companies. The other one is Bangladesh Bank's credit information bureau (CIB).CRISL's sources for information are usually the banks, audited annual reports, and other materials available to the public through Securities and Exchange Commission (SEC). There is a systemic resentment about CRISL ratings, especially among those that have an economic or political advantage in Bangladesh. Some companies have raised questions about the fairness, acceptance, and accuracy of the ratings although CRISL claims to perform prudently by engaging specialists in the analysis of risks and all forms of financial intermediation.

On the other hand, CIB, managed by the Bangladesh Bank, provides loan data, including information on collateral and guarantees, but limits the information that the Bangladesh Bank can share. The government treats CIB as the key instrument for disciplining the lending process. Therefore, CIB posits great power to restrict lending to clients closely associated with companies with defaulting loans. Separating CIB from its parent Bangladesh Bank is complex. Also, the Bangladesh Bank has the right to collect whatever data it wants from banks along with the obligation to keep these confidential.

Data analytics can play a crucial role at this critical juncture of Bangladesh's economy. It can offer timely, reliable, efficient, and relevant insights about clients in order to support informed decision making. It will do the balancing act by considering trade-offs between each dimension of localness, granularity, timeliness, and frequency and improve understanding of financial consumerism.

Data analytics will require liaison with external partners and stakeholders, such as, Bangladesh Bureau of Statistics, Ministry of Finance, National Board of Revenue, SEC, among others that now need to develop data strategies to make available granular data. The intent is to enable data analysis to advance complex financial decision making on the basis of building, strengthening, and mobilising partnerships to achieve better outcomes. It is amazing what can be achieved when data sources that have never been integrated suddenly start talking to each other.

Does this sound like a public-private-partnership? Well, that is the beauty of big data analytics. There are risks abound, such as operational and reputational risks. But a good partnership or at least data sharing or data mining mechanism can mitigate such risks. Government entities can develop disclosure legislations, and financial institutions can provide implementation solutions to employ knowledge and practice of multiple lines of defence. By using data mining techniques as well as using various predictive modelling, financial institutions will be actively instrumental in identifying several risks. These will include risk of leak/disclosure of sensitive information; risk of producing and releasing inaccurate information, survey results, in-house research, responses, advice or analysis, risk of damage to the sector's relations with stakeholders and government; risk of business intervention, risk of not having adequate human resources to meet operational needs and risk of financial institutions being sued in relation to failure on the basis that those institutions individually have or have not done related homework.

Financial institutions in Bangladesh will find data analytics a pre-emptive premise for micro prudential policy analysis. Such analysis will encourage multi-pronged analyses and socio-economic research on a number of multi-dimensional issues, value chain, and financial governance. Capitalising big data from multiple sources and applying business analytics will offer Bangladesh's financial sector true insights. The sector can then move beyond simple black and white rules, allow its institutions to hold adequate capital as defined by the Bangladesh Bank's capital requirements, and encourage precautionary steps in order to avoid bad debts and foreclosures in the financial sector.


Shishir Shahnawaz, PhD is an Adjunct Professor in the School of International Development & Global Studies at University of Ottawa sshahnaw@uottawa.ca

Share if you like