The different mortgage markets

Roche (1999) notes how debt financing has emerged as the 'nearly universal alternative' to paying for a house outright, obtaining the generally valued benefits of home ownership. She evokes the sense of an international phenomenon nicely by citing examples of the many names given to mortgage debt throughout the world, financiaciĆ³n de la vivienda in Mexico, credit au logement in France.2 Internationally there are a variety of mortgage instruments. This variety reflects the degree of competition in the respective mortgage markets, together with the institutional and regulatory features of the housing finance systems which delimit the risks faced by lenders and borrowers. This book adopts the view that the fundamental nature of mortgage contracts, their variety and household choices, can also be understood in theoretical terms, and that the extant empirical studies offer insights along with methodologies which can be applicable in many economies. However, we do need to note the significant differences between the main housing finance systems.

This section of the book motivates our study of mortgage market economics by noting the importance of variations in the form and use of mortgage finance in different parts of the world. Because most econometric research is located in North America (the US and Canada) and the UK, there will be an emphasis upon comparisons of the housing finance systems of these countries, with the UK market placed in its European context. Identifying some of the key characteristics of these markets is important for the interpretation of empirical studies. Finally, mortgage markets are not static but ever changing and produce heterogeneous mortgage products. The reasons for the heterogeneity of mortgage contracts, with the implications of this variety, form another key theme of this book. Though mortgage markets are dynamic and complex, the conceptual tools presented should assist in the continuing analysis and study of this important marketplace.

In the UK in 1998, mortgage debt constituted 57% of Gross National Product. This figure was as high as 69% in Denmark, and 65% in the Netherlands. Austria, however, had mortgage debt of just 5% of GNP3 demonstrating that there can be significant international differences in the importance of this market (see Table 1.1). For the US the volume of residential mortgages is estimated at $3 trillion dollars (2000), representing a doubling of the figure over 10 years. This is compared to a US figure for government borrowing of $5 trillion.4 Outside of Europe and the US, Canada has mortgage debt on residential properties of 33% of household net worth (1999).5 In Australia in 1999 there were 2.3 million homeowners with a mortgage (that is 31% of all Australian households).6 Thus for many economies the mortgage market is very large and important. For some

Table 1.1 Outstanding Residential Mortgage Debt In the European Union

Outstanding residential Trend in outstanding residential

Table 1.1 Outstanding Residential Mortgage Debt In the European Union

Outstanding residential Trend in outstanding residential


mortgage debt/GDP 19981

mortgage debt (1990-1998)2











































United Kingdom







1 The figures for Austria include commercial mortgages

2 The figures for Germany and Sweden relate to the trend growth in residential and commercial mortgages

1 The figures for Austria include commercial mortgages

2 The figures for Germany and Sweden relate to the trend growth in residential and commercial mortgages

Source: Hypostat, 1988-1998, Mortgage and Property Markets in the European Union and Norway, European Mortgage Federation (Tables 3 and Table 4).

countries beginning at a low base, such as Greece and Spain, the trend rate of growth in this market is high (see Table 1.1).

Several factors explain international differences in the size and economic significance of mortgage markets. Explanations include differences in the size and quality of the housing stock, variations in the proportion of houses that are owner-occupied properties, together with the range of mortgage instruments available. Variations in the legal and regulatory frameworks can also limit or encourage this market effecting the range and type of mortgage design. Differences in the general economic environment (e.g. monetary and fiscal factors including the subsidisation of mortgage interest payments), and the history of the housing finance systems can also create variety. The location and amount of research on the economics of mortgage choice reflects the sophistication and size of the mortgage markets concerned. Thus the US has provided the most prolific output of research in this area, a factor reflected in most of the literature referred to in this book.

Why differences in housing finance systems are important

Lea (2000) describes an evolutionary path for a housing finance system. It begins with informal lending for house purchase with this role eventually adopted by specialised financial intermediaries. These intermediaries may be mutual societies, owned by shareholders, or 'special circuits' provided through government-backed agencies. The final phase is mortgage market securitisation and the integration of the housing finance system with other capital markets. Different economies will be at different stages in this process so we will observe significant differences in housing finance systems, including prevalent mortgage designs. For Europe, Maclennan et al. (1998, p. 62) observes that '[h]ousing finance lenders within Europe have evolved within national boundaries, and reflect the influence of localised origins and national policies'. Even with financial deregulation post-1980, these histories have been modified to different degrees. So different housing finance systems provide the background to mortgage choices.

The implicit subsidisation of fixed rate debt via Federal guarantees in the US, with the role of government sponsored agencies in standardising products for securitisation, might explain the prevalence and popularity of the fixed rate mortgage (FRM). The subsidisation of life insurance premiums on UK endowment mortgages, up to 1984, and tax relief on mortgage interest payments at source (MIRAS), until April 2000, might explain the initial popularity of the endowment mortgage. The UK has seen several shifts in the nature of property taxation; prior to the current council tax there was a shift from a system of rates to a poll tax, which was estimated to have had a significant impact upon housing demand (Rosenthal 1999). Stephens (2000) notes how offering cheap funds to some financial intermediaries prevents the convergence of European mortgage markets, also different legal systems inhibit the emergence of a standard mortgage product.

The theory of mortgage demand to be discussed in Chapter 2 will suggest a common decision-making framework for utility maximising consumers. In a perfectly competitive no arbitrage economy, with perfect capital markets and no information problems, a comprehensive model of mortgage choices might suggest a common choice of mortgage instrument, or at least an indifferent choice. The reasons why we do not observe a single common contract design are explored throughout this book, and in Chapter 7 in particular. However, differences in housing finance systems, and government involvement, clearly offer part of the explanation for the different contracts that we observe together with the borrowers' menu of choices. Housing finance systems will differ in efficiency and these differences will effect borrowers' choices, investors' opportunities and the degree of integration with other capital markets.

There is a body of research that has been concerned with the comparative efficiency of housing finance systems (Diamond & Lea 1992; Lea etal. 1997; Stephens 2000). This efficiency can be viewed in the strict economic sense of cost minimisation. Alternatively, it can be viewed as intermediation efficiency (Diamond & Lea 1992; Stephens 2000). The measure of intermediation efficiency used by Diamond & Lea is the difference between the costs to society of mortgage finance, measured by mortgage interest payments plus origination fees, less the theoretical minimum cost, which is the rate of interest on government debt of the same maturity. This margin reflects the extant market distortions evident in the competitive structure of the mortgage market, and any taxes and subsidies.

There are clearly technical difficulties associated with calculating intermediation efficiency and using it to compare housing finance systems. For example, it is difficult to find comparable mortgage contracts with which to compare interest rates. Other factors such as the underlying market conditions (e.g. competitiveness) that facilitate efficiency are also difficult to evaluate. Diamond & Lea incorporate a number of qualitative judgements to ascertain the impact of such factors. From the perspective of this book the measure of intermediation efficiency also leaves out an important dimension. That is the fact that through regulation and market distortion some housing finance systems exclude particular contracts from the consumer's menu of choices, or discourage them, e.g. long-run fixed rate mortgages in the UK. In principle, the measurement of the consumer's willingness to pay for excluded contract designs would provide a measure of any welfare loss (Diamond & Lea 1992).

The overall efficiency of a housing finance system can be judged by how well it matches the preferences of both borrowers and investors. Thus a mortgage finance system should be able to repackage the mortgage debt of borrowers to create securities of interest to investors. Investors may be better placed to bear the risk of holding mortgages than borrowers. The extent to which this risk is repackaged for investors is another perspective on housing finance efficiency (Jaffee & Renaud 1998). In the perfectly efficient system the propositions of Modigliani & Miller hold and borrowers would be indifferent between the choice of gearing and other features of mortgage contracts. In fact, with complete markets even the securitisation of mortgage debt cannot be justified, as investors could package their own securities. With incomplete markets, and markets with liquidity and information problems, we need to make a general judgement on how well borrowers' and investors' needs are met.

To take an example of how an efficient housing finance system, with some constraints, might operate, we consider the following. Residential loans with high loan-to-value ratios and higher risk of default could be sold to investors who wish to have some exposure to real estate markets. Low loan-to-value ratio loans, with lower risk of default, would be sold to investors such as pension funds who wanted a better match with their liabilities. Even if borrowers were constrained by facing a standard loan-to-value ratio, investment bankers could still package the mortgages and issue junior and senior debt to match the preferences of each client group. This would meet the need to repackage risk for investors, but a fully efficient housing finance system should also meet the needs of borrowers. When capital markets are imperfect this implies a variety of mortgage instruments to reflect risk positions and cash flow needs.

Diamond & Lea found both the US and UK housing finance systems to be high in comparative efficiency, though both systems have regulations and distortions. For example, regulations on capital adequacy have limited securitisation in the UK. In the US there are geographic limitations on setting mortgage underwriting criterion for loans to be securitised. However, financial deregulation and increased competition in both economies has led to a better match of contracts with consumer preferences. In the

UK the innovative use of credit derivatives has facilitated the supply of short-term fixed rate debt. In the US the emergence of the adjustable rate mortgage (ARM) has provided borrowers with a choice of risk exposure. Consumers needs appear to be well met in competitive mortgage markets.

Some contracts might be less prevalent due to labour market and housing market imperfections that inhibit residential mobility. For example, more mobile households might be less willing to pay a premium for the risk to the lender of prepayment (or risk incurring penalties) and thus choose an adjustable (variable) rate mortgage rather than a long-term fixed rate contract. Also, information asymmetries prevent the development of innovative contracts that require specialised underwriting skills. The lack of specialised underwriters then further exacerbates the information prob-lem.7 Primary mortgage markets require underwriting and property valuation skills to underpin any secondary mortgage market development. Thus observed and excluded mortgage contracts can reflect other market imperfections and information problems.

A source of variation in available mortgage contracts that has major implications for household choices is the maximum loan-to-value ratio. Diamond & Lea note that the down payment required of borrowers varies significantly between housing finance systems. In a life cycle decision-making framework a large minimum deposit can defer entry into owner occupation. This requirement can distort life cycle consumption as households save to purchase the house (Artle & Varaya 1978). Tax incentives to borrowing, in the form of a subsidy on the mortgage interest rate, encourage early entry into owner occupation, but this contradicts the desire to postpone consumption to accumulate a deposit (Slemrod 1982; Hayashi et al. 1988). Theoretical work has demonstrated the importance of down payment requirements for the timing of entry into own occupation, and its impact upon house price cycles (Stein 1995; Ortalo-Magne & Rady 1998, 1999, 2002). Combinations of loan-to-value ratios and interest rates can also form the basis of separating equilibrium in the mortgage market (Brueckner 2000). This source of variation in mortgage contract design will recur in discussions throughout this book.

In summary, housing finance systems exhibit different degrees of efficiency. In a general sense efficiency can be viewed as the capacity to meet both the needs of investors and borrowers. Incomplete markets, information problems, market distortions and subsidies explain both the prevalent contract choices that face consumers, along with the choices that are not made available to them. Mortgage choices must be viewed in the context of the menu of contract designs that households face, and the characteristics of the housing finance systems in which those choices are made. However, we shall find that competitive pressures in the major systems of housing finance have generated extensive mortgage contract menus. Studies of both the US and UK mortgage choices provide insights into the behaviour of borrowers in highly evolved financial systems, with broad menus of contract choice. We now consider the major and pertinent characteristics of the mortgage finance systems that feature most in this book, that is the US and Canada and the UK and European mortgage markets.

North American mortgage markets

The US market for residential mortgage debt has several important characteristics that will be referred to when evaluating research into mortgage choices. There are a variety of available mortgage designs, with the key instruments being fixed rate debt (the FRM) and the adjustable rate mortgage (the ARM). Fixed rate mortgages fix the rate for 15 or 30 years, while for adjustable rate debt the adjustment period can vary from 1 to 5 years. This compares with fixed rate debt in the UK and Canada which is typically fixed for 1 to 5 years, with some longer periods available, but which reverts back to a variable rate of interest after the period of fixity. These differences will be seen to be significant when we examine household behaviour regarding the choice of mortgage instrument and the impact of mortgage contract choice on housing demand. For example, UK borrowers appear to focus much more on expected movements in short-term interest rates than long-run interest rate expectations.

The mortgage banks that have generally superseded the Savings and Loans institutions dominate mortgage finance in the US. These mortgage banks originate loans, which are then sold to government sponsored agencies (GSEs). That is, the mortgage debt is passed on to the secondary mortgage market. The mortgage banks came to prominence after the Savings and Loan crises in the US. One advantage of this system is that the mortgage banks do not need large amounts of funding to conduct their business, as the debt is sold on. They are also able to pass on the risks of holding mortgage debt to investors, and the risks are largely outside any regulatory framework.

A critical factor in understanding US research is the importance of the securitised mortgage market, where mortgage loans are packaged into securities. Lenders in most economies traditionally originate, fund and service loans. The securitisation of debt separates the functions of originating, funding and servicing. The benefits of securitisation for lenders and borrowers will be briefly discussed below but it is useful to consider the main reason for the growth of this market in the US. During the 1980s there was a crisis for the Savings and Loans institutions that were the main providers of mortgage finance. These thrift institutions borrowed short-term and loaned money for long periods at fixed rates of interest. Subsequent high and volatile interest rates left the thrifts short of capital. It is perhaps not surprising that subsequent developments favoured funding by selling off packages of mortgages as securities (Coles 2001), increasing the lenders' capital base. This is also a prime lesson in the importance of risk sharing, a key feature of mortgage contract designs. For example, a move to adjustable rate mortgages shifted the risk of interest rate variatiations onto the borrower.

A further important feature of the US mortgage market is the role of the GSEs, the Federal National Mortgage Association (FNMA or Fannie Mae), the Government National Mortgage Association (GNMA or Ginny Mae) and the Federal Home Loan Corporation (FHLC or Freddie Mac). The standardisation of pools of mortgages by these agencies also encouraged the rapid growth in the secondary mortgage market. GNMA is on the federal budget and so receives direct government insurance against mortgage default, the other two agencies have implicit government guarantees. It is argued that these guarantees allow the issuing of mortgage-backed securities at lower interest rates.

The development of the GSEs has increased the liquidity and depth of the securitised mortgage market, underwriting a good deal of US mortgage lending, and possibly contributed to lower borrowing costs. A key research issue is the extent to which mortgage interest costs are lower for the borrower in the US, due securitisation per se, or to the underwriting by government agencies (Hendershott & Shilling 1989; Kolari et al. 1998; Cantor & Demsetz 1993). The continuing growth of the GSEs and the secondary mortgage market reflects a lower cost of cash flow management and mortgage funding (Van Order 2000). Jaffee & Renaud (1995) argue that the development of the securitised mortgage market is a 'revealed preference' for the lower cost of funds involved, and the ability of lenders to hedge against interest rate risk, that is falls in interest rates when lending at a fixed rate. The introduction and development of such agencies may be necessary to expand housing finance systems that wish to achieve efficiency through the development of a secondary mortgage market.

The role of the Federal Housing Administration (FHA) is important in placing US research on mortgage choices in context. The FHA insures the mortgage debt of the secondary mortgage market institutions against default. For conventional lending outside of this system Savings and Loan institutions and banks have recourse to private insurers. The interesting point is that the FHA prohibits variations in underwriting criterion (that is loan-to-value ratios) by geographical region, so that they cannot be changed to reflect local economic conditions, say, increased to choke off excess demand. This is not a constraint applied to conventional lending. Here we have an example of the securitisation of debt leading to the standardisation of mortgage contracts. This disparity has led to some interesting research on the existence of credit rationing in the US, an issue to be explored fully in Chapter 5 (Ambrose et al. 2002; Duca & Rosenthal 1991).

One important aspect of the valuation of mortgage-backed securities is the prepayment behaviour of borrowers. MBS investors are insured against default risk but not interest rate risk, that is the risk that borrowers refinance if interest rates fall. There are ways of creating different classes of security for investors that reflects this risk. However, the analysis of prepayment behaviour remains a major concern in mortgage market research. From the lender's perspective, prepayment risk can be partly covered by charging up-front points. This has led to financial institutions offering different combinations of points and interest rates, which can then attract different types of borrowers. This trade off has formed the major focus of research into asymmetric information and the signalling properties of different mortgage designs. In the UK prepayment risk has been covered primarily by redemption penalties, which penalise borrowers for 'prematurely' exiting from a deal, say on a discounted or fixed rate contract. Once again we will find that these different features of housing finance systems have a bearing upon observed household choices, though our understanding can emanate from a common theoretical base.

Canada is another economy where there has been substantive mortgage market research (see Zorn & Lea 1989; Jones 1993, 1994, 1995; Breslaw et al. 1996). The Canadian market is interesting in having so called 'roll over mortgages' which are essentially ARMs. Since the early 1980s Canadian fixed rates have typically been adjusted every three years. Canada also mirrors the UK market in having penalties for prepayment of debt, though these typically relate to partial prepayments of over 10%. When reviewing Canadian research we will have to recall a system with these characteristics and few long-term fixed rate mortgages. However, in common with other mortgage markets there has been considerable innovation in Canada. There are now many examples of fixed rate products, discounts, cash backs and flexible or 'open' mortgages. Mortgage finance in Canada is becoming increasingly separated from retail funding, and characterised by increasing use of securitisation.

A further interesting feature of the Canadian mortgage market is the fact that mortgage interest payments are not tax deductible. In the US there is tax relief for mortgage interest payments, though this is complicated to the extent that the US tax reform act of 1986 has meant that for many taxpayers tax gains on mortgage debt have been significantly reduced (Follain & Dunsky 1997). The point is that the tax treatment of mortgage interest payments differs between countries and can have a significant impact upon housing and mortgage demand. The UK is now like Canada in that tax relief on mortgage interest payments was abandoned in April 2000, though its presence will feature in a number of the studies referred to and discussed in later chapters. UK and European mortgage markets vary in a number of other key ways that bear upon the analysis of household choices.

The United Kingdom and European mortgage markets

The UK mortgage market is one of the most sophisticated and liberal markets in the world. The UK economy and its housing market also have their own particular characteristics. There are high levels of owner occupation, high levels of indebtedness and high proportions of mortgages where the interest rate is variable rather than fixed. These features of the UK housing and mortgage market have led to high sensitivity of consumer spending, and savings decisions, to changes in interest rates (Earley 2000). However, high levels of owner occupation and indebtedness are features of other economies-recent reports for both Canada8 and Australia9 have featured large amounts of mortgage borrowing as posing a potential afford-ability problem for borrowers. This demonstrates the importance of understanding the basis of mortgage demand, the key focus of Chapters 2 and 3 of this book.

The funding of mortgage finance in the United Kingdom has been primarily through retail deposits though there has been an increase in off balance sheet financing and the securitisation of mortgage debt. Securitisation has been primarily used by so called centralised mortgage lenders who entered the UK market during the late 1980s (Pryke & Whitehead 1991; Pais 2002). Some of these lenders were of US origin, a competitive incursion to be repeated later with the development of the UK sub-prime lending market (e.g. the entry of Money Store in 1997). The centralised lenders were competitors to the traditional building societies and banks. Recently the share of lending by the centralised sector has declined having dropped below 10% during the 1990s,10 but increasing competitive pressures are leading more traditional providers to consider mortgage securitisation.

European mortgage markets offer more examples of size, role and funding of mortgage debt,11 and provide an interesting comparison with the UK. Germany and Denmark rely mostly upon mortgage banks which issue mortgage bonds, thus allowing longer-term fixed rate deals to what is a smaller mortgage sector. France has a whole range of legal restrictions and regulations that make its mortgage market less heterogeneous. France also relies upon retail savings to finance mortgage lending, as does the UK, the Irish Republic and Spain. The Netherlands has a system similar to the UK. The existence of such varied housing finance systems has created impediments to convergence towards a single European mortgage market (Munchau 1997; Stephens 2000), so that in empirical work we must note the features of the particular housing finance system concerned.

The UK mortgage market is an example of a mature and deregulated mortgage market where competition has generated a large variety of mortgage contracts. This competition has encouraged discounting of interest rates (teaser rates), cash back arrangements, indexing to the base rate of interest and the emergence of flexible/lifestyle mortgages. Here we have an important dimension of housing finance efficiency discussed above, which is matching consumer preferences. However, the basis of this heterogeneity is something of a puzzle. If lenders are risk-neutral and borrowers risk averse, then the FRM is the optimum mortgage instrument with the lender bearing interest rate risk (Brueckner & Arvan 1986; Brueckner 1993). Explanations lie with housing finance systems and imperfect markets.

We have already noted how differences in funding, regulations and underwriting risk might explain why some contracts, in this case the long-term FRM in the UK, may not be generally available. Housing finance systems vary according to how the risk of interest rate changes is shared. In the UK the borrower has typically borne this risk through the prevalence of variable rate debt. The management of credit risk, where the borrower may default, is also an important driver of differences in mortgage designs (Roche 1999). In France this has led to the imposition of low loan-to-value ratios. In the UK households are liable for any shortfall if they sell a property that is in negative equity, so that deliberate default may not be wealth maximising. The heterogeneity of mortgage contracts might also reflect affordability and information problems which again may vary by housing finance system.

One important aspect of most mortgage markets is that they are constantly subject to change. Considerations of mortgage market futures have even suggested the eventual demise of the specific mortgage instrument, as mortgage borrowing becomes indistinguishable from other forms of debt, with financial planning taking the form of a universal account (Roche 1999). There has been an observable trend towards combining mortgage debt with liquid bank accounts and other lines of credit, certainly in the UK and Australia. There may be other forces leading to a more common future, for example it has been argued that the Canadian mortgage market increasingly resembles that of the United States. Lending has moved away from the traditional banking sector and there is increasing securitisation of mortgage finance. The imperatives of securi-tisation might eventually impose a degree of uniformity on contracts across a number of countries. The role of secondary mortgage market development and the securitisation of mortgage debt is clearly critical to present and future contract choices.

First Time Home Buyers Guide

First Time Home Buyers Guide

If you are getting ready to purchase your first home or if you think you can't afford to purchase your first home, don't make another move until you have read this important information! Every year, Federal, State and Local government and community development programs help thousands of people obtain there first home.

Get My Free Ebook

Post a comment