Page 285 - Data Science Algorithms in a Week

266 Oloruntomi Joledo, Edgar Gutierrez and Hatim Bukhari

Once the borrower is screened, an interest rate is generated to reflect his risk profile.
A draw on the lookup table is used to generate an interest rate that corresponds to the
borrower. On receiving the interest rate, the borrower makes a decision to agree or
decline the terms of the loan. If he declines, he has an option to follow the
noToAgreement transition and go back to the PotentialBorrower state where he can
decide to remain or to leave the platform. If the borrower agrees to the terms of the loan,
he proceeds to the PostedProfile state via the yesToAgreement transition. The decision to
accept interest rate is internal and probabilistic based on the borrower’s risk preference
and personal goals. A call is made to the requestServiceB() function which communicates
the borrower profile to available lenders. If the borrower profile matches a given lender’s
risk aversion, he accepts and stores the id of the borrower along with his profile
information.
Once the lender agrees to fund the borrower, the borrower transitions to the Funded
state where it remains for a uniformly distributed period that reflects the time it takes to
fully fund the request. After which it transitions to the InRepayment state where it
remains for the term (usually 36 or 60 months). Thirty days after entering the
InRepayment state, the borrower starts to make payment every 27 to 31 days. This time
range reflects the fact that borrowers pay their bills early, on time or late.

Figure 4. Lender statechart.