‘Guaranteed Trouble’ uncovers hidden risk that can build up in real estate developments. First the bank helps fund a real estate company to develop, lets say in this case, residential apartments. Once the apartments are ready for sale, the bank can also provide the home loans to the buyers, which extends the banks business and helps the real estate company make a sale.
Reminiscent of the US subprime crisis, the trouble begins when the real estate company can’t sell enough apartments. To boost sales they help investment customers secure loans they may otherwise not be able to afford, by guaranteeing them. From the buyers view there is little risk. It is an investment only and if house prices go down, they can simply walk away from the loan, handing back the risk to the real estate company as the Guarantor.
On the banks side, each home loan is assessed on its merit individually. One individual loan, guaranteed by the real estate company, by itself is not a significant risk. But what if the real estate company is actively guaranteeing most of the loans for a given development?
This anonymized visualization lets the bank see the whole relationship covering all the guarantees between the real estate developer and the loan customers. The dots (or nodes) represent home loan customers guaranteed by a real estate company. The lines (or edges) show the link between the guaranteed loans and the real estate company guarantor. The colors show different developments. Instantly we can see there are a number of real estate companies that have massive guarantees out supporting the loans to buyers of their developments.
‘Guaranteed Trouble’ enables the bank to visualize and monitor the true nature and size of the total exposure. It can isolate high-risk developments and take direct measures to manage and price for the risk being created. Action can be taken both with the real estate developer’s credit positions and any new home loans taken out on the development.
About the Analytics
This visualization shows a sigma graph created in the Teradata Aster Discovery Platform using Aster Lens. The data used includes guarantee enterprise ID, and guarantee contract information, guarantee number, guarantee amount, enterprise credit grade etc. Social network analytics was applied in order to find the core customer of concern and identify guarantee patterns. The analysis functions applied include: betweenness, degrees, PageRank, modularity, eigenvector centrality and local clustering coefficient.