عوامل موثر بر تفاوت های میزان بازده وام مسکن: اوراق بهادارسازی دارایی

This paper examines securitization and mortgage yield spread differentials among four groups of Australian mortgage providers: mortgage corporations, commercial banks, building societies and credit unions. The dataset includes over 2000 observations of standard adjustable rate mortgages from 180 institutions. Several studies on the impact of securitization (e.g., reduced funding costs, information costs, agency costs, regulatory tax and improved mortgage marketability and liquidity) can be found, but there is a little empirical evidence, particularly in an Australian context. Our results suggest that mortgage corporations that securitize all their loans have both the narrowest nominal and effective yield spreads among the four groups, consistent with the existing securitization and mortgage cost literature. The regression findings also indicate that both minimum and maximum amounts of loans are significantly associated with spreads.

Over the last three decades advances in technology, deregulation, re-regulation, liberalization and globalization have brought substantial changes and innovations to the financial sector. This has driven financial institutions increasingly to the capital market for funding, particularly for mortgage lenders. In an effort to meet the demand for housing loans, these institutions began to securitize these loans into the capital markets. This innovation has helped mortgage finance to become a major part of the capital market and the economy. Mortgage securitization in the U.S. and more recently in Australia plays an increasing role in mortgage finance. It has been a chief funding channel for lenders. It constitutes the largest part of all fixed income securities, soaring to US$ 5.1 trillion in 2002 and over 60% of all types of mortgages in the U.S. are securitized. Australia's mortgage securitization market is second to the U.S., with over AU$ 100 billion outstanding in 2003 (Liu et al., 2004). The Australian secondary mortgage market is different from that of the U.S. due to its reliance on the private sector activity instead of government agencies. Over 50% of mortgage-backed securities in Australia are issued to global capital markets. Around 80% of these mortgages have adjustable rates. Therefore, any mortgage rate changes immediately effect borrower costs of not only newly originated loans but also existing ones. In recent years, particular attention has been given by regulators, policy makers, academics, the public and the industry to securitization and its costs and benefits. Most of literature concentrates on the U.S. market where three government-sponsored enterprises (GSEs)1 play a key role in the secondary mortgage market. It may be difficult for researchers to distinguish the GSEs’ contribution from that of securitization itself in reducing mortgage costs. Most other securitization research on private label and asset-backed securities (ABS) transactions2 suggest that securitization can help reduce a range of costs, such as funding costs, information costs, agency costs, regulatory tax and improved marketability and liquidity. However, there is a little empirical evidence on its effects, in particular in an Australian context. This paper attempts to fill this gap. Objectives of this paper are threefold. Firstly, its empirical results should help policymakers and regulators to understand the impact of securitization. Secondly, it should help lenders to understand what factors statistically affect their yield spreads in forming credit policy. Thirdly, it should assist borrowers in choosing which type lenders and when they should invest. This paper makes several contributions to literature. It is the first paper to examine mortgage yield spread differentials across different groups of lenders (mortgage corporations,3 commercial banks, building societies and credit unions) in Australia. It is also the first to investigate the impact of credit criteria factors on yield spreads in Australia. Moreover, we discompose cost composition of different types of lenders to understand how securitization can reduce costs for lenders that involve securitization. These factors should help explain why mortgages corporations have consistently offered cheaper prices than depository institutions. We identify three research questions: Are mortgage yield spreads of mortgage corporations (that securitize almost all their loans) significantly different from other three lending groups? What key factors in origination have an impact on yield spreads? Are seasonal effects significant to yield spreads? To answer these three questions we use t-tests and two regression models to test monthly standard mortgage rate and fee data, 2262 observations from 180 institutions in 2001. The results of the t-tests and regressions suggest that mortgage corporations have both the narrowest nominal and effective yield spreads among the four groups. Mortgage corporations require over 20 basis points less than other three groups of depository institutions. This is consistent with the previous securitization theoretical work (e.g., Greenbaum and Thakor, 1987, Hess and Smith, 1988, Schwarcz, 1994, Hill, 1997 and Iacobucci and Winter, 2003) and empirical work (e.g., Hendershott and Shilling, 1989, Merriken, 1988, Rhoades, 1992, Ambrose et al., 2002a and Ambrose et al., 2002b). 4 The regression findings also indicate that both minimum and maximum amounts of loans are positively associated with spreads, consistent with early studies. Large dollar loans may have higher credit risk (e.g., Hendershott and Shilling, 1989). In contrast small loans lack economies of scale and so have higher transaction costs (e.g., Benston, 1992 and Hill, 1997). In addition seasonal effects, market default risk and prepayment risk are found significantly to effect spreads, consistent with previous literature (e.g., Hendershott and Shilling, 1989, Kolari et al., 1998, Ambrose et al., 2002a and Ambrose et al., 2002b). The rest of this paper is organized as follows. In Section 2 we review the relevant literature on benefits and advantages associated with securitization and its relations with mortgage yield spreads or rates. Section 3 we discuss the data and define variables at lenders’ level and market level used for regression models. In Section 4, we conduct empirical tests and discuss the results. The conclusions are drawn in Section 5.

There is extensive literature on securitization, and its effects on mortgage yield spreads or rates. However, this is concentrated on the U.S. market where the three GSEs have dominated the secondary mortgage market. Their direct involvements make it more difficult for researchers to distinguish GSEs’ contributions from those of securitization in reducing mortgage costs. Other research on private label and other asset securitization suggests that securitization itself affords many cost advantages. They support that the credit rating rule in securitization is shifted from being institution-based to asset-based by separating the ownership of the assets from originators and so the whole securitization process helps reduce information asymmetric and agency problems. However, there is a little such evidence in an Australian context. To fill the gap we use t-tests and two regression models to examine securitization and mortgage yield spread differentials for 2001 among one hundred eighty mortgage providers in Australia. Results of t-tests and regressions suggest that mortgage corporations have both the narrowest nominal and effective yield spreads among the four groups. These are consistent with existing securitization literature (e.g., Greenbaum and Thakor, 1987, Hess and Smith, 1988, Schwarcz, 1994, Hill, 1997 and Iacobucci and Winter, 2003) and empirical work (e.g., Hendershott and Shilling, 1989, Rhoades, 1992, Ambrose et al., 2002a and Ambrose et al., 2002b). Mortgage corporations require over 20 basis points less than other three groups of depository institutions. The evidence supports cost reductions and cheaper funding inherent in securitization technology. Because mortgage costs for the borrower are more sensitive to any decline in mortgage prices in Australia where ARMs dominate the market, this evidence suggests that policymakers and regulators should further promote securitization. The findings are also relevant to depository institutions in their securitization decisions. The findings in this paper further confirm that securitization can really help reduce a range of costs for originators, as well as borrowers. The regression findings also indicate that both small and large loans are positively associated with spreads for all groups. The largest loans may have higher credit risk, requiring over 20 basis points higher than its reference. In contrast the smallest loans do not have economies of scale and so are more costly, requiring around 10 basis points higher than its reference. This links to credit policies of the lender in origination. In addition we find that the seasonal effects, market default risk and prepayment risk are significant to spreads. January, March, April, September and November have over 20 basis points higher than December. March is the most expensive month with about 40 basis points higher. The results are important to both mortgagees in borrowing decisions and mortgagors in lending decisions