Loan Risk Analysis

Mortgage Risk Visualizations

Welcome to Loan Risk Analysis (LRA), our website was developed to allow exploration into the complex mortgage industry with dynamic visualization, using publicly available data. Below are three visualizations; each designed to analyze different aspects (the who, what, when, and where) of lending in the US.


Delinquency Spotter

Delinquency Spotter allows the user to explore the delinquency rates of every 3-zip code within the US across time, FICO, Loan to Value (LTV), and debt to income (DTI) of borrowers.

3Score Inspector

3Score Inspector allows the user to compare delinquency rates of lenders across time, FICO, LTV, and DTI.

Similarity Explorer

Similarity Explorer allows the user to see which lenders had lending practices most similar to other lenders by loan makeup across different channels and loan purpose.

Technology Labs

Accenture Technology Labs, the technology research and development (R&D) organization within Accenture, has been turning technology innovation into business results for 20 years. Our R&D team explores new and emerging technologies to create a vision of how technology will shape the future and invent the next wave of cutting-edge business solutions. Find out more about us on the Accenture Technology Labs Site, our blog, or by contacting Constantine Boyadjiev.


Why are you doing this?

This was developed through collaboration with Accenture Tech Labs, Accenture Analytics, and our partnership with Stevens Institute of Technology.  This is an example of the innovative work we are all doing, and how our partnership is accelerating the digital movement.

Why 3D graphs? Don’t most experts like Edward Tufte discourage 3D?

We agree, rarely is it a good idea to make something 3D, but in this case it’s actually effective.  There are definitely other ways to explore the inter-relation of 3 quantitative variables, but we felt this was a very intuitive and straightforward method for the problem at hand.

What is the graph visualization?

This visualization is a graph diagram showing the similarity of lenders’ credit mix based on FICO (credit scores), LTV (loan-to-value) ratio, and DTI (debt-to-income) ratio.  After bucketing the loans across each dimension, we used Bhattacharya distance to measure similarity.  There’s a diagram on the page that walks through the analytics behind this. Thus, if lenders are close to each other they are more similar, and if they are further the inverse is true.  In addition, we used a minimum spanning tree algorithm to find the smallest path between all the nodes, as well as connected each node to its five most similar neighbors.

Why would you use network analysis and graphs in finance?

This is a great question, one that we will answer more in-depth in the future.  We believe that innovation comes from not only discovering something new, but also applying existing solutions to new problems.  Graph and network analysis may not be used in many fields today, but we feel in the future, graph analysis will become much more pervasive.  Graph analysis already has shown promising results in cyber security, anomaly detection, and in social media, but we believe this is just the beginning.

Where did you get your data?

We are using public data from Fannie Mae and Freddie Mac. The population of loans are 30 year fixed rate mortgages that are fully amortizing with full documentation.

What browsers are supported?

This web app uses SVG images and WebGL which are supported in the current versions of Firefox, Internet Explorer (IE 11+), Opera, Safari and Google Chrome.  Delinquency Spotter and Similarity Explorer will also work in IE 10.

© Accenture 2015 All Rights Reserved