Actually I already downloaded and installed markovchain package in R but I have difficulty following the instruction guide of the developer. I am estimating the transition probabilities of loan data from different states of delinquencies.
Thanks for answering my question. I will probably using a continuous data because i will be getting the proportions of loans at the different states of delinquencies. Can you help me further by giving me some examples if its okay with you? thanks.
I have a monthly data of accounts with outstanding loan balances and payment status. I wanted to forecast the outstanding balances distributions per delinquency type (quality of the portfolio). My problem actually is the estimation of the stable transition matrix. I downloaded the markovchain package but couldn't follow through on the examples given. Here are the details of my data.
Monthly loan account with the following loan statuses: Current, delayed less than 30 days, delayed more tha 30 but less than 60 days, delayed more than 60 but less than 90 days (substandard), delayed more than 90 but less than 180 days (doubtful) and delayed more than 180 days (loss).
I wanted to forecast the distribution of the loan portfolio.