The collated data from various MFI is a storehouse of raw material waiting to be refined into predictive analysis that could help these organisations to source out potential clients, decide on their level of credibility and future potential too. ‘The rising role of data analytics and algorithms in micro finance’ explains the logic behind the computing...

Agri-commodity | 18 July 2017

The rising role of data analytics and algorithms in micro finance

Microfinance is all about empowering the most underprivileged sections of society. Micro Finance Institutions (MFIs) offer loans to people from disadvantaged backgrounds, especially women, who have no access to loans from formal financial channels. Facilitating this segment of the population, with loans, hasthe indirecteffect of created self-employment opportunities and economic growth too.  Yet MFIs are businesses in their own right and must be financially viable if they seek to be sustainable.


Credibility Conundrum

The biggest challenge MFIs have faced so far is not availability of funds to lend; it has been assessing the credit-worthiness of borrowers. The target group of MFIs comprises people who usually cannot afford collateral and do not have either a formal credit history or cash-flows, which have been monitored and documented.


The popular model adopted so far has been to insist on lending to borrowers as a group, without any security, but with the condition that repayment is also done collectively. In this model, the sole factor that MFIs fall back onto pre-empt defaults is societal pressure and support within the group. While this has worked to a great extent, there have been instances of dissent within the group or defiance of the entire group, which leads to a breakdown in the system.


Data-driven solutions

The availability of proxy data and techniques to analyse it has considerably facilitated working parameters of MFIs. These organisations have begun to use data analytics at three broad levels:


Identification of potential clients – ‘Geospacial analytics’, or demographic data of a particular area or region,helps uncover potential customers residing there, through human population forecasting, by filtering out relevant data and applying it to provide accurate trend analysis, modelling and predictions.Customer Sourcing Executives can use this analytical technique to expand their customer

base by targeting those who have not been tapped yet.


Disbursement decisions - Predictive analytics is a technique which studies the behavioural patterns of potential borrowers to determine whether they are likely to default. So, for instance, the KYC information along with other non-financial personal information of a potential client can be run through an algorithm-based programme to determine the likelihood of default. The base information for creation of the programme could be collated KYC and other data on existing clients of numerous MFI, mapped against the credit performance of those clients. Patterns, linking certain habits and situations to the chance of default, emerge when the data sample is large enough. MFIs could primarily disburse loans to those whose parameters fall within the broad category of acceptance of the algorithm-based programme. This reduces the chance of delinquencies to a large extent while at least giving the potential borrower a chance to be considered for a loan.


Nurturing clients – The last aspect of data analysis is the ability to gauge which existing clients should be encouraged to borrow further and expand. Profitability analysis on an existing client’s Key Performance Indicators (KPIs) enables a MFI to gauge his/her potential growth and encourage such clients to borrow more to expand and grow.


Crucial Caveat

While it is possible for an algorithm to discard outliers (i.e., data that seems inaccurate since it lies too far from the average), there is always the concern that large chunks of the data may be altered or unreliable. This impacts the machine learning process, which could result in inaccurate predictions. Nevertheless, as the use of data analytics increases and the relevance of using accurate data is appreciated, data reporting will become more reliable and could help correct past data sets too.

*, |

Copyrights © 2019 Inditrade Capital Ltd. All Right Reserved.

Designed,Developed & Content Powered by  Accord Fintech Pvt. Ltd.