After the global financial crisis in 2007, we observed a boom in non-banking institutions proposing financial services as an alternative to the greedy and ailing banking system. The main idea of their business approach is to propose platforms (mostly online), which put borrowers and lenders in direct contact.
Institutions like LendingClub in the USA refer unsecured personal loans between $1,000 - $35,000 to borrowers with an average net annualized return (net of defaults and service fees) of 9.64% and default rate of 4%. This default rate may be higher for individual lenders, since each loan is a result of the contribution of several lenders and one lender may invest in several loans of defaulting borrowers.
It is therefore important for potential lenders to suitably diversify their investment portfolio in addition to taking into account individual creditworthiness of borrowers.
µDAI proposes a binomial mixture approach to estimate the concentration risk of borrowers in peer-to-peer lending based on publicly available credit data of LendingClub and macroeconomic indicators of the Federal Reserve Bank.
The objective of this research project is to develop a methodology to estimate concentration risk in peer-to-peer lending, that can serve lenders in their investments both in industrialized countries (i.e. Smava, Auxmoney in Germany) and (in an adapted form) in developing countries (i.e. Lendico in South Africa).