Published on 
December 5, 2021

Our loan tape analysis and visualization techniques

Yujia Li

December 5, 2021

Loan Tape Analysis and Dashboard Visualization

Our investors and partners know that we put a lot of work into assessment and rigorous analysis of the loan portfolios of the borrowers we work with. Our methodology and loan tape assessment techniques allow us to obtain a deep understanding of the loan portfolio through analysing each loan that has been issued by the borrower, its performance, risk of default and predict its behavior for the next 12 months.
In this short video, Yujia Li, our Data and Credit Risk Analyst, talks about the basics of the loan tape analysis that we perform, the way we visualize it and present it to our investors via the dashboard.

Loan tape includes millions of data points, analysis of which helps us evaluate the portfolio. To have a full picture, we request the following information from the borrowers: scheduled payments (payment date, principal, interest), actual payment (payment date, principal, interest), loan details (type, issue date, amount, tenure, status, etc.), information about the end borrower.

We start with data cleaning and standardization, a crucial part of the process, that allows us to later reuse the code for calculation and prediction. The processed data is then input and run in Julia, a programming language created specifically for data science, which has a number of great features, including a system of smart code which allows us to go through the loan tape with a few clicks.

We perform a 4-step analysis of the prepared data using a code in Julia, as outlined below:

  1. Static description
  2. Delinquency analysis
  3. Parameter estimation
  4. Cashflow modelling

The key parameters we estimate during such analysis are: probability of default, recovery rate and future cash flow. The outcomes of the analysis are then presented graphically in the dashboard - which is a convenient way to grasp and understand the loan portfolio through visualization.

The charts demonstrate the following:

  • distribution and time series progression of the loan tenure, APR and disbursements
  • delinquency by vintage
  • repayment forecast
  • days till break-even
  • observed default rate
  • outstanding balance.

The dashboard is a great way to present our assessment to investors and use the outcomes to set the right covenants with the borrowers.  Consequently, our technology and code allows us to perform monthly monitoring of the loan book and covenants, and react to the potential risks identified in a timely manner.

Kilde is a regulated investment platform for alternatives. We operate as a two-sided platform connecting institutions / HNWI with securitised private investments. Our main alternative asset classes are private debt, venture debt, and recurring revenue financing. Kilde has partnered with leading non-banking consumer & SME lending firms to give investors safe and controlled access to consumer lending assets. Our unfair advantage is vast accumulated data on consumer & SME assets performance as well as scalable investment and securitisation tech platform. Thanks to Kilde’s license for dealing in securities, we securitize alternative investments into digital securities.

About the author

Yujia Li

Yujia Li

Yujia is a Data and Credit Risk Analyst at KILDE, a regulated platform for alternative investments. KILDE is powering digital lending firms with debt capital to reach underbanked customers in South East Asia.

Want to access our exclusive deals?