Prepayment Risk of Public Housing Mortgages

Article excerpt

Executive Summary. This paper examines the determinants of prepayment risk of mortgages for Singapore public housing, which are adjustable rate mortgages. An understanding of the prepayment risk of mortgages is important because prepayments affect investor return from investing in mortgage-backed securities. The findings show that prepayment is positively related to market sentiments and the interest rate of public mortgages. Conversely, it is negatively related to income growth and the relative difference between private and public housing prices. The findings further imply that the prepayment rate of public mortgages is directly affected by any policies that enhance the affordability of residents of public housing to move to private housing.


Since 1998, the Government of Singapore has been pressing for reforms in the financial sector to establish Singapore as an international financial hub (Business Times, 2000). This is especially in view of economic globalization, accessibility of the Internet and the removal of trade barriers. On July 23, 2002, the Government of Singapore made a major policy change by giving lenders the first claim on Central Provident Fund1 (CPF)-financed home loans, effective September 1, 2002, which provided an opportunity for banks to securitize mortgage loans (Rashiwala, 2002).

Given that 86% of the population lives in flats provided by the Housing Development Board (HDB) (HDB Annual Report 1998/1999) and the total amount of public loans stands at $12,439.3 million, the potential associated with the securitization of mortgages is very great. Although the quantum of public loans may be dwarfed by private housing loans (see Exhibit 1), HDB, being a Statutory Board, is backed by the government and necessary financial muscle to guarantee the performance of its mortgages. Furthermore, a government led initiative in issuing mortgage-back securities will form the benchmark and attract entrants into the market (Wong, 1998). From the government's perspective, securitization of mortgages allows the HDB to obtain funds, which can be invested in assets with higher yield, and enable HDB to further diversify its portfolio. Hence, the increased income could offset the required government grants. The resultant savings in grants can also be ploughed back into other developments or financial subsidies.

Despite its attractiveness, real estate securitization is still in its infancy stage in Singapore. Singapore still lacks a well-developed, comprehensive legal system and a strong financial body. Moreover, a true mortgage-backed security has yet to emerge. Hence, many aspects of real estate securitization have not been fully explored yet. One of the more important aspects is prepayment risk of HDB mortgages.

Prepayment risk is very important as it has serious implications on the return to the mortgage-backed securities investor and the lending institutions. Curly and Guttentag (1974), using data from 1950s and 1960s, illustrate two complications that hamper estimates of mortgage yields. First, early termination will harm investors who purchase mortgages at premium. Second, asymmetry in yield and termination point relationships means that underestimating prepayment is not the same as overestimating prepayment. Hence, portfolio lenders, firms engaged in mortgage servicing and mortgage investors, might incur large losses when prepayment rates exceed expectations. In addition, Hendershott, Shilling and Villiani (1983) and Milonas (1987) also demonstrated the relevance of prepayment to market-based ex-ante yield spreads.

Although the literature on the prepayment risk of the debt market in the United States is well-established, most studies concentrate on the prepayment risk of Fixed-Rate Mortgages (FRM). A FRM is a fully amortizing loan with fixed coupons, and hence, fixed monthly payments. Unlike the FRMs, an Adjustable Rate Mortgage (ARM) is a loan with a varying interest rate that may change, usually in response to changes in the Treasury bill rate or the prime rate. …