Prepayment projections are at the center of all bond valuation and analysis. Since the pioneering development of prepayment model by FHA, many models have come to be widely used to obtain prepayment projections and indeed, are critical for valuation techniques such as OAS analysis. To value a bond or loan, it is necessary to project cashflow. The problem is that the cashflow is not known because of prepayment option. The only way is to make assumptions about the prepayment over the entire life of a bond or loan. Several conventions have been used as a benchmark for prepayment. Among them, most widely used are: (1) the Public Securities Association(PSA) standard and (2) Conditional Prepayment Rate(CPR). The PSA standard is very simple and based on the real 30-year prepayment rates. However, it is of no use anymore. The CPR is a very useful measurement and is still one of the most widely used methods.
At the same time, most market participants are well aware that projecting prepayment is not reliable. As with many econometric models, projecting prepayment covers only a partial range of economic and interest environments. And it also requires as many data as thirty year's of prepayment to track entire life of a bond or loan.
Some factors, like interest rates and turnover rates, are known to affect prepayment. However, each factors have shown that the overall economic and non-economic environments tend to decide the sensitivities of each factor. And data availability itself is a major factor of an exact modeling. In Korea, enough data are not available to do the entire life analysis.
The main objective of the analysis is to obtain the factors that influences housing loan prepayment in Korea. Since the housing loan have been made almost exclusively by the Housing & Commercial Bank at a lower rate than market interest rates, all data were abstracted from the H&C Bank.
The analysis suggests no direct relationship with market interest rates. However some patterns were found in interest rates, loan size, and loan period. The bigger the size is, the faster the prepayment. The higher the loan interest rates, the faster the prepayment. And finally the longer the loan period, the slower and the more constant the prepayment.
As a whole, prepayment rates in the first and second month after the loan are very high and those of later period(especially after 6~7 years after the loan) were very low.
these findings can be useful for the financial institutions, especially for the housing loan specializing institutions. Some kind of penalty rate can be applied to prevent the early period prepayments.
