Shaping Nepalese Banking Sector With AI
Nepal's GDP growth rate was around 6% as per World Bank data in 2019. One of the major contributors to GDP is the service sector,
with the banking business being the largest service industry in Nepal. The banking sector is considered a high-tech adopting industry
in Nepal and has embraced banking practices from the West.
By the end of the decade in 2019, a significant development was the rise in the use of Machine Learning and Artificial Intelligence.
However, Nepal lags behind in adopting these changes, and the banking industry has not updated its working platforms. The recent
trend of increasing interest rates on deposits and creating unhealthy competition in the overall industry has led to a decline in GDP
growth rate and has severely impacted small and medium sector businesses.
The banking sector should look forward to adding more dimensions to the business rather than competing for customers. The industry should
focus on satisfying customers by personalizing their services and incorporating more analytics into their data. This blog discusses a few
tools of Artificial Intelligence, Machine Learning Algorithms, and Deep Learning.
Deposit:
Deposits are the mainstay of the banking industry. Machine Learning Modules should be utilized by banks to identify depositor behavior.
A well-designed and trained module based on a bank's own past years' data can help answer questions such as: What are the chances of a
depositor adding more deposits? Will the depositor transfer the deposit to another bank? How does depositor behavior change based on demographics,
employment history, and family patterns? What are the chances of a fixed deposit being withdrawn? These and many more questions can be answered.
Loans & Advances:
Loans and advances are essentially the outcomes that generate revenue. Modules trained and tested on a bank's own data can predict customer behavior with an accuracy
exceeding 90% if the data are properly labeled. Techniques like
regression analysis based on specific industry growth and the use of multivariable regression analysis can help banks predict which industries to focus on.
Revenue Prediction:
The main source of revenue generation for Nepalese banks is interest income. This is where loan defaulters can be accurately predicted, and measures can be initiated
before the interest due date. Recurrent Neural Networks (RNNs) can
be used to create Deep Learning Neural Modules.
Other Sectors where Machine Learning needs to be used are:
i. Map Pattern Analysis of Deposits & Loans (Companies like ESRI and Near are world leaders).
ii. Card Fraud Detection.
iii. Personalized Financial Management.
iv. Consumer Spending Analysis.
v. Generating Customer Satisfaction Index.
In summary, banks need to lend money more based on data analytics and identify trends by deploying modules for time series forecasting. Banks need to understand their responsibility in nation-building and keep interest rates in check.
Click here to read the original article published by author in Blogspot.com