Market structure, institutional development and the provision of housing finance

Economic Modelling 1994 11 (2) 215-227

Market structure, institutional development and the provision of housing finance A comparative study John Lomax

This paper compares housin9 finance systems in seven countries, focusin9, in particular, on institutional structure, and examines national variation in market structure in the context of a n industrial economics approach to competition infinancial markets. It is suggested that developments in the UK housin9 market after deregulation could be emulated, to some extent, in other countries as they deregulate. Keywords:Housingfinance; Deregulation

Housing finance institutions are characterized by a variety of different forms of corporate structure. Mutuality is popular; for example, the building societies in the UK, some of the savings and loan associations in the USA, and credit cooperatives in Germany and Japan. In some countries public sector mortgage banks have been established, combining the roles of lending and implementing government policies; the Cr6dit Foncier in France, the Housing Loan Corporation in Japan and the Mortgage Bank of Spain fall into this category. Additionally, in the UK (until the mid-1970s) local authorities used to undertake a significant amount of mortgage lending. Finally, many housing finance institutions have stock status. The essential aim of a housing finance system is to provide long-term loans to households for the purpose of house purchase. Such loans are typically secured on the value of the property. In terms of funding techniques, two broad systems of housing finance can be distinguished, based on retail and wholesale inflows respectively. Regulation has frequently constrained institutions to operate in only one of these markets. Gradually, deregulation has occurred and competitive pressures, for example movements in the structure of

The author is with Prudential Portfolio Managers, 1 Stephen Street, London WlP 2AP, UK. Final manuscript receivedSeptember 1992. 0264-9993/94/020215-13 © 1994 Butterworth-Heinemann Ltd

relative interest rates, have in many instances induced individual institutions to use mixed funding strategies. With the retail based system, private sector savings are recycled as mortgages. The establishment of a large number of distributive outlets is typically required to collect deposits, a corollary being that these can also 'be used to sell mortgages. Since deposits attract variable rates of interest, loans are typically also made on a floating rate basis. Among the principal types of existing mortgage lending institutions - the commercial banks, savings banks and mortgage banks - the first two rely most heavily on retail deposits. The general banks typically provide the complete range of banking services to both the personal and corporate sectors; savings banks can be either general retail banking institutions providing services to the personal sector or specialist providers of housing finance. In some countries, the commercial banks are the dominant lenders. For example in France, they have a market share of outstanding loans for house purchase of 63% (1989) and in Japan the equivalent figure is 41% (1990). Elsewhere, the market share of the commercial banks is smaller, for example 16% in Germany (1990) and 24% in Spain (1989). Nevertheless, the low shares in some instances are rather deceptive; organizational structure has adapted to accommodate systems of regulation. Thus, commercial banks have established specialist mortgage lending subsidiaries in many countries, for example mortgage 215

Market structure, institutional development and housingfinance. J. Lomax T a b l e 1. S u p e r v i s o r y bodies.

UK Japan USA

Germany Spain France

Italy

Building societiesare supervisedby the BuildingSocieties Commission. Banks are supervised by the Bank of England. Specialist mortgage lenders and insurance companies are supervised by the Ministry of Finance (MOF). Banks are supervised by MOF and the Bank of Japan. Nationally chartered banks are supervised by the Office of the Comptroller of the Currency, state chartered banks by the state authorities, and bank holder companies by the Federal Reserve.The Federal Reserve has some supervisory rights over all members of the Federal ReserveSystem,and the Federal Deposit Insurance Corporation (FDIC) over all FDIC insured banks. The Federal thrift supervisor is the Office of Thrift Supervision; the insurer with supervisory rights is the FDIC. All mortgage lenders are supervised by the Federal Bank Supervisory Office. The mortgage bond market is supervised by the Finance Ministry. Banks are regulated by the Bank of Spain. The Commission Bancaire and its various sections (Conseil National de Cr6dit, Comit~des Etablissements de Cr6dit and Comit6 de la R6glementation Bancaire) regulate mortgage lenders. Mortgage lenders are regulated by the Italian central bank.

banks and Bausparkassen in G e r m a n y and special credit institutions in Italy. However, everywhere, to a greater or lesser extent, the direct share of the commercial banks in the flow of lending is increasing. The main losers have been the traditional mortgage lending specialists, the Bausparkassen and mortgage banks in Germany, the caisses d'~pargne in France, the savings associations in the USA and, to a lesser extent, the building societies in the UK. The public sector mortgage lenders (the Cr6dit Foncier, the Mortgage Bank of Spain and the Housing Loan Corporation in Japan) have also significantly scaled back activities. A subcategory of the deposit-based systems is the contractual savings scheme (offered by the Bausparkassen in West G e r m a n y and through the plans d'~pargne-logement in France). With these, regular savings are made over a number of years, receiving interest at below the market level, following which the investor becomes entitled to a loan at a preferential rate. The second general method of providing housing finance is via the money markets. In some instances, wholesale inflows are used in an equivalent manner to retail deposits - that is simply as a source of funds - and mortgage assets are retained on the lending institution's books. This is the case, for example, with the banks and building societies in the UK, the mortgage banks in G e r m a n y and the special credit institutions in Italy. By contrast, where extensive secondary markets have developed, assets can be securitized and sold to other investors such as life

216

insurance companies. The mortgage banks in the USA operate in this manner, as broadly do the specialized lenders in the UK. Since assets are moved off balance sheets, securitization enables lending institutions to conserve capital. As assets and wholesale liabilities can be directly matched, loans can be offered at both floating and fixed rates of interest. The supervisory framework varies between countries (Table 1). Typically, mortgage lenders are supervised by the same authorities as other classes of financial institution; this is, for example, the case in Japan, Germany, France, Spain and Italy. In the UK, the banks are regulated by the Bank of England and the building societies by the Building Societies Commission. The USA has a wider range of supervisory bodies and the banks and thrifts are currently monitored by different agencies. It is noteworthy, however, that both mutual and incorporated thrifts are supervised by the Office of Thrift Supervision.

Market structure The market structure of an industry depends to a very significant degree on the incidence of barriers to entry. In the absence of regulatory barriers it could be argued that housing finance exhibits some of the characteristics of a contestable market. As is well known, in such a market prices are held down to the levels of marginal costs by the threat of potential entry: thus, even if there are only a very small number of firms operating, price and quantity tend towards the perfectly competitive equilibrium. One important condition necessary for market contestability is the absence of sunk costs. These are fixed costs which cannot be recouped, the existence of which makes exit from a market costly. It is sometimes contended that, for housing finance systems based on retail savings, the requirement to invest in a branch network could involve such costs. A counter argument is, however, that branches (although not necessarily the associated equipment) can easily be deployed to other uses, particularly by institutions offering a range of financial services. For example, it is not very costly to switch out of mortgages and into other consumer or business loans. In addition, branches can be sold. A necessity to invest in branches could, however, have been more successful in discouraging nonfinancial or new firms from entering the mortgage market. In this context, an additional argument is that incumbent institutions with large distributive networks can threaten potential entrants with an aggressive reaction (for example, price cutting) should they enter, making such a move unprofitable. To the extent that this strategy can be credibly implemented, entry could be deterred.

E C O N O M I C M O D E L L I N G April 1994

Market structure, &stitutionaldevelopment and hous&yfinance: J. Lomax It has also been argued that the need to acquire a reputation is an important barrier to entering financial markets. Assets can only be funded when lending institutions can convince the suppliers of funds that their deposits will be safe. Moreover, borrowers have to be satisfied that loans are not of inferior quality (perhaps, for example, susceptible to higher interest rates over a period of time). On the other hand, reputation is neither institution nor market specific. It can, to some extent, either be transferred between markets, or purchased (assets can, for example, be insured). In this context, the ability to securitize assets has increased the contestability of the mortgage market. The quality of asset backed debt is entirely separable from that of the issuing institution; holders of a mortgage backed security will have first call on the collateral provided by the underlying mortgage pool. The reputation of the assets, which may have been upgraded through credit enhancement, replaces that of the lending institution. The vehicles through which issues are made are bankruptcy remote. Holders of other wholesale market instruments are not protected to the same degree, although they do usually rank highly in the list of creditors should bankruptcy occur. To some extent, therefore, dependence on a branch network is not an exogenous barrier to entering the mortgage market, but associated with a particular regulatory framework. Where non-retail funding, and particularly securitization, is permitted, or indeed encouraged, the need for a decentralized method of gathering funds diminishes; the corollary is that loans are made through intermediaries or related entities with extensive retail outlets such as estate agents and insurance companies or even by post. Nevertheless, the legal fees, credit enhancement and systems which need to be in place before a mortgage backed issue can be launched continue to represent quite significant sunk costs. The principal regulatory barriers to entry have risen from the limits placed on the markets in which particular financial institutions are able to participate. Constraints have been imposed for a number of different reasons. Regulations have been designed to restrain competition in order to limit risk and protect depositors. In some instances they have had the additional objective of supporting a social policy to encourage residential housing. Restraints imposed for monetary policy purposes have also sometimes barred entry to the mortgage market. Moves towards deregulation have applied widely, driven for the most part by a range of common factors. These include the following: high rates of inflation and nominal interest rates, increased fluctuations in nominal interest rates, government deficits, supply ECONOMIC MODELLING April ! 994

shocks, advances in computer and communications technology and changes in political ideology. The impact of these different pressures has, however, varied markedly, leading to sharp differences in the pace of financial liberalization between countries. Broadly speaking, a distinction can be drawn between developments in the USA and UK and those in other countries which continue to be characterized by higher regulatory barriers. To the extent that economic rather than political factors have driven the financial liberalization process, attempts to reverse developments - even if desired - will prove more difficult. Consequences of liberalization and innovation The main consequence of the reduction in the barriers to entry to the housing finance industry has been an increase in competition in most countries, accompanied by some blurring of the historial distinctions between commercial banks and specialist mortgage institutions. Mortgage loans are increasingly provided in all countries by multipurpose financial institutions. Furthermore, there has been a decline in credit rationing (Table 2). Part of the explanation lies in the relaxation of regulatory frameworks, and an easing of the terms on which institutions are able to extend credit. In addition, however, as the mortgage market has become more integrated with other loan markets, 'club' behaviour has broken down. In this context, financial markets are characterized by acute information asymmetries between the buyers and sellers of financial services. Traditionally, one means of improving information flows has involved institutions organizing as clubs (see Goodhart [11]) and operating as quasi or de facto cartels. These clubs have often been coordinated by regulatory and other public sector bodies. One consequence of this was that financial institutions were willing to follow the authorities' guidance, even if not in their short-run interests, given the long-run benefits perceived from club membership. Accordingly, they have sometimes been persuaded to meet social or political objectives; for housing finance clubs, these frequently involved exercising restraint in the mortgage market. In addition, the commercial banks have frequently not been encouraged to offer mortgage loans. Partly as a result, prices have historically tended not to respond strongly to excess demand. Financial deregulation has, however, made this type of market structure increasingly difficult to sustain. Without entry barriers, informal rules of conduct cannot be enforced, nor can quality standards be maintained. Free riders cannot be excluded. New entrants have no reason to obey the rules of the club, particularly since, with the breakdown of market segmentation, there might be different types of

217

4~

,-,t

>

f~

t"

O

U p to 100% Loans in excess of 80% generally have to be supported by an insurance company guarantee

25 years

Typically m a x i m u m is 3.5 times income

Loan maturity (typical m a x i m u m maturity level)

Loan/income ratios

UK

Loans to value ratios

Table 2. Credit rationing.

Loans are normally limited so that monthly repayments do not exceed 40% of the borrowers net cash flow

to 12 years Mortgage bond funded loans up to 30 years Loan repayments do not normally exceed 30-35% of net cash flow

Loan repayments do not normally exceed 33% of net cash flow

10-25 years

10-15 years

Vary considerably.

15 years

Normal limit is 50%; this rises to 75% for new houses and 90% for repairs

Free sector loans are for a m a x i m u m advance of 80% if they are to be eligible for refinancing on the secondary market

Savings banks offer loans for a m a x i m u m of 80% of the value of mortgaged property Banco Hipotecario offers loans for up to 60% A m a x i m u m of 70% applies ifa loan is to be eligible for refinancing on the secondary market.

Mortgage bond funded loans cannot exceed 60%; this is also the case for loans offered by the savings banks, Bausparkassen loans cannot exceed 80%

Bausparkassen up

Italy

France

Spain

Germany

The standard m a x i m u m is 3 times income

28 years

To be eligible for purchase by F N M A loans must not exceed 95%; those in excess of 80% require insurance

USA

25 years fixed interest rate; 30 years variable interest rate

Typically up to a m a x i m u m of 60"/0

Japan

2

%

Market structure, institutional development and housingfinance: J. Lomax Table 3. Deposit insurance

UK

Japan

USA

Germany Spain France Italy

Deposits with banks are insured through the deposit protection fund. The first 75% of any loss up to a maximum of £20000 is refundable. A similar scheme is operated for the building societies where up to 90% of any loss not exceeding £20 000 is insured. Deposits with banks are insured by the deposit insurance fund. A maximum of ¥10 million per depositor is refundable. Specialist mortgage lenders are not permitted to offer deposits. The FDIC operates separate deposit insurance funds for banks and thrifts. Insurance covers 100% of deposits of not more than US$100000. There is some flexibility, however, and sometimes insurance may cover more than the legal limit. The Deposit Protection Fund covers liabilities to each non-bank depositor up to an aggregate amount of 30% of the equity capital of the member bank. Bank deposits are insured up to Ptsl.5 million per deposit. Up to FF400000 covered. Up to L3000 million covered.

institution with objectives different from those of the traditional mortgage lenders in the same market. Thus, as a consequence of increasing liberalization and innovation, much greater reliance has been placed on prices as a means of clearing markets. Correspondingly, there has been less resort to quantitative rationing. A further aspect has been the significant expansion of the range of mortgage, savings and wholesale instruments available and the consequent enhancement of consumer choice. At the same time, however, under certain conditions deregulation has the potential to exert an adverse rather than benign impact on the performance of housing finance institutions. This is primarily because of the existence of deposit insurance (Table 3). In all of the major seven countries, a significant proportion of loans for house purchase - which are by their nature illiquid and long-term - is financed using short-term liquid retail deposits. This creates a potential for destabilizing runs on the banking systems if savers rush to withdraw funds and assets have to be sold at a loss (Diamond and Dybvig [10-1). The rationale for deposit insurance is, therefore, to exert a stabilizing influence on the financial system by countering the possibility of such developments. Nevertheless, to the extent that costs of deposit insurance are not borne by the institutions and/or premiums are not related to risk, such schemes create a moral hazard problem; institutions have an incentive to take excessive risks since the financial gains resulting from successful gambles accrue to owners and managers, whereas the deposit insurers bear the loss over and above capital if the gamble fails. Inasmuch as deregulation allows financial institutions to adopt such behaviour, an increased rate of bankruptcy is likely. E C O N O M I C M O D E L L I N G April 1994

The geographical pattern of developments in market structure In the Anglo-Saxon countries the regulatory segmentation of the retail banking market has been to a large extent eliminated. The removal of the corset allowed the clearing banks into the housing finance market in the UK, while the deregulation of interest rates (and, in particular, the ending of the funding cost advantage the thrifts enjoyed) coupled with government support for the market in mortgage backed securities had the same effect in the USA. In order to permit a competitive response on the part of the housing finance specialists (the building societies and savings associations respectively) their asset and liability powers were also broadened through the 1986 Building Societies Act (UK) and Depository Institutions Deregulation and Monetary Control (1980) and G a r n St Germain (1982) Acts (USA). Moreover, in both countries, institutions were free to convert to stock status should they wish to operate outside the remaining constraints. As regards spatial competition, building societies have always been permitted to operate nationally; this is not, however, true of the savings associations in the USA. The continental European and Japanese housing finance systems have also developed, albeit at a less rapid pace. Geographical restrictions on the operation of the Spanish savings banks (1988) and G e r m a n mortgage banks (1991) have been lifted. Although savings in the G e r m a n Bausparkassen continue to attract a government bonus, this has declined. While the special credit institutions remain the dominant mortgage lenders in Italy, other institutions have increasingly been permitted to offer loans for house purchase; 8% of the commercial banks deposits are now eligible for long-term loans; the equivalent figure for the savings banks is 30%. The commercial banks have also increased their share of the mortgage market in France with the abolition of credit ceilings in 1987. In the same year the savings banks were permitted to lend to small and medium-sized companies. In both France and Spain secondary mortgage markets have been established in order to facilitate the refinancing of loans. The process of liberalization in Japan has been manifested in a rather different form from the European and US experience. In particular, while Japanese banks have historically not participated in the mortgage market, this has reflected broad administrative guidance more than the impingement of a precise regulatory framework: Consumer and mortgage credit were low priority and though financial institutions were not explicitly restricted from serving the borrowing needs of households, it was under-

219

Market structure, institutional development and housing l~nance. J. Lomax

stood that financial institutions were primarily responsible for meeting the financial needs of businesses especially the large corporations (Cargill and Royama [5]). Since in this context the constraints on banks balance sheets have been informal, it is extremely difficult to judge the pace and extent of which these have been eased. The evidence including, the lending figures in Table 8 - suggest, however, that some degree of liberalization has occurred. The trend towards securitization has also been most pronounced in the Anglo-Saxon countries, again largely for regulatory and other government policy reasons. In the USA, which has been the scene of the most dramatic developments, the establishment of markets in mortgage backed securities was in part a response to restrictive legislation which prohibited interstate branching by savings and loan associations. By contrast, mortgage banks were able to compete nationwide: regional imbalances in the supply and demand for mortgage finance could be evened out and the risks to lenders of regional recession spread. The process of securitization was aided by the establishment of federal and quasigovernmental guarantor companies (principally the Government National Mortgage Association, the Federal National Mortgage Association and the Federal Home Loan Mortgage Corporation), whose imprimatur effectively made mortgage bonds as safe as government securities. The UK has developed the most sophisticated secondary market outside the USA, although on a very substantially smaller scale. Not only is the mainstream housing finance industry less geographically fragmented, but in addition the authorities have refrained from guaranteeing mortgage backed issues. The development of securitization techniques elsewhere has been more constrained by regulation. In Japan, for example, banks are permitted to repackage debts and sell them to investors in the form of certificates. However, these are not classified as securities and there are severe limits on tradability; thus, there is virtually no secondary market. In addition, there are severe restrictions on the use of special purpose vehicles of the type used in the Anglo-Saxon countries to issue asset backed debt. The latter constraint also applied in France, for example, prior to 1988. In Italy, deals have to be structured as syndicated loans. The EC Solvency Ratios Directive, if it comes into force as it currently stands on 1 January 1993, could also reduce the likelihood of widespread mortgage securitization in Europe; this gives mortgage backed securities a weighting of 100% in the capital requirements arithmetic, unlike mortgages which have a weighting of 50%, raising doubts about the participation of the commercial banks, which are currently the dominant investors. At the same time, since 220

mortgages have only half the weight of consumer loans in the calculations of Cooke capital adequacy ratios, the incentive for banks to move the former off balance sheet is reduced. One consequence of deregulation and the associated increase in competition has been an expansion in the range of mortgage products available. Reflecting the geographical pattern of developments, this is most pronounced in the UK and USA. Among the plethora of instruments marketed are fixed rate, variable rate, capped, endowment, pension related and shared equity loans. Equally, mortgages tend to be granted at higher loan to value ratios (95% loans are frequently available, 100% advances to a lesser extent). In continental Europe and Japan, lower maximum loan to value ratios are often imposed by the authorities. For example, the German mortgage banks can only make loans funded by bond issue up to 60% of property value. Using alternative funding the ratio can be raised, although such lending cannot exceed 15% of the total value of their mortgage assets. In France and Spain, maximum loan-to-value ratios of 80% and 70% respectively apply to loans eligible for refinancing on the secondary market. The special credit institutions in Italy are only permitted to lend up to 50%, while Japanese savers also typically have to accumulate up to 40% of the purchase price as a down payment. Equally, maximum loan maturities are sometimes specified in the continental European countries, and these are generally shorter than is typical in the U K and USA. Nevertheless, the loan to value criteria cited above are not as binding as might initially be supposed, since borrowers can package together loans from a number of different sources. For example, German borrowers frequently acquire loans from a mortgage bank, a Bausparkasse and a commercial bank to finance house purchase. That said, in some countries the ability of the commercial banks to top up the loans made by more specialized lenders is limited (particularly in Germany and Italy). For prudential reasons long-term assets and liabilities have to be matched to at least some extent. Moreover, at the same time, the ability of the banks to use the money markets to fund mortgages long term is constrained. Equally, in France and Spain banks have been quite cautious in offering high percentage loans since they then lose the ability to refinance such assets on the secondary markets. For the most part, as a matter of lending policy, the commercial banks' collateral requirements have not been dissimilar to those of the housing finance specialists. In general, therefore, despite the existence of packaging techniques, mortgages remain rationed to a much greater extent in the continental European countries and Japan than in the UK and USA. E C O N O M I C M O D E L L I N G April 1994

Market structure, institutional development and housino finance: J. Lomax Table 4. restriction on funding of mortgages.

UK Germany

Spain Italy France

USA Japan

Building societies are not permitted to fund more than 40% of their lending through the wholesale market. Bausparkassen have to fund from retail deposits. Only mortgage banks can issue mortgage and communal bonds (the latter being bonds backed by communal loans). For banks long-term assets and liabilities must be to some extent matched. The banks are not permitted to issue certain types of mortgage bond. Loans eligible for refinancing on the secondary markets must meet certain criteria. Only 8% of the commercial banks deposits are eligible for longer-term lending. The equivalent figure for the saving banks is 30%. Loans eligible for refinancingmust meet certain criteria. For banks, prudential regulations require that longterm assets, 5 years and over, be funded at least 60% by similar term liabilities. Loans eligiblefor refinancingmust meet certain criteria. Some deposit rates are still regulated. Housing loan companies must borrow from parent institutions.

Table 5. Restrictions on lending.

UK Germany Spain

Italy

France USA

Japan

Most of building societies' lending has to be in form of first mortgages to owner occupiers. But up to 10% of their commercial assets can be unsecured. Bausparkassen can only lend against house purchase. The bonds mortgage banks issue have to be collateralized by mortgages or communal loans. Mortgage credit companies can only issue particular and preferred types of mortgage bond (cedulas hipotecarius) if 100% of the resulting loans are mortgages. The banks and savings banks are generally restricted to offering short-term loans. The special credit institutions undertake most mortgage and other long-term lending. Savings banks are not allowed to lend to listed companies. There is a qualified thrift lender test requiring savings associations to hold at least 70% of portfolio assets in qualified investments (mainly housing related). Commercial real estate loans are limited to four times capital. Thrifts may no longer purchase non-investment grade securities. Administrative guidance leads the banks to favour the corporate sector above persons.

Restrictions on the funding and lending of mortgages are summarized in Tables 4 and 5. Frequently, mortgage lenders - including the commercial banks - support required regulatory criteria with other voluntary limits, perhaps an example of the club-type behaviour referred to earlier. Thus, while in the Anglo-Saxon countries, loans tend to be granted up to a m a x i m u m of 3-3½ times income, m u c h more explicit attention is paid in continental Europe to borrowers' ability to sustain debt burdens. In particular, loans are normally limited so that monthly repayments of principal and interest do not exceed 3 0 - 4 0 % of the borrower's net cash flow. Nevertheless, some convergence of housing finance

ECONOMIC MODELLING

April 1994

systems now appears likely. The m o r e restrictive structures - which have already experienced some change - seem likely to undergo further development, particularly in E u r o p e where significant regulatory change is imminent. The abolition of exchange controls (France, Spain and Italy) has facilitated greater intra-EC competition in financial services. However, a variety of other obstacles to cross-border activity remain (Cecchini [8]). The Second Banking Co-ordination Directive which comes into force on 1 J a n u a r y 1993 should, however, ensure that these are eliminated. This will provide for a single E u r o p e a n licence; thus, if an institution is authorized to undertake one of the banking activities covered by the Directive (which includes mortgage lending) in its home member state, it m a y undertake that activity in every other state of the C o m m u n i t y without further authorization and regardless of whether the host countries' financial institutions also have such powers. Reduced barriers to crossing regional and national frontiers need not, however, necessarily imply an increased volume of cross-border services or entry into foreign markets. Instead, it is more likely that deregulation will encourage previously protected national banking sectors to adapt; heightened potential competition and the resulting changes in perception could induce substantial adjustment. In either case, however, it should add to existing pressures for convergence in patterns of housing finance across the EC. Alongside the on going deregulation of continental European housing finance systems a parallel reregulation of mortgage provision in the U S A is occurring. The liberalization of the savings associations was thought to have been excessive; accordingly the Financial Institutions Reform, Recovery and Enforcement Act (1989) restricted the type and extent of activities in which thrifts m a y engage. F o r example, the qualified thrift lender test was tightened, requiring thrifts to hold at least 70% - up from 60% - of portfolio assets in qualified investments (mainly housing related). As noted above, deregulation allied with the existence of deposit insurance had the potential to exert an adverse influence on the housing finance industry's performance. So far, this has been manifested principally in the U S A with the development of the thrift crisis, which could cost between $150 and $200 billion to rectify. Insolvent institutions - a significant part of the industry (Brumbaugh [3]) - tended to have the greatest incentives for risk taking behaviour, while deregulation afforded them increased opportunities for altering their conduct. 1 A degree of regulatory 1The liberalization process itself, and in particular the lag between the deregulation of interest rates on liabilities and assets paved the way for subsequent developments by allowing thrifts" margins to come under severe pressure.

221

Market structure, institutional development and housing.finance: J. Loma.v Table 6. Increase in average house prices relative to retail price inflation (%).

Japan

UK

Mean increase (1975-80)

USA 2.5

Germany 3.4

7.1

1.8

-1,2

Mean increase (1981 89)

-1.2

--1.4

1.3

7.8

6.9

capture by the savings associations industry could have contributed to these developments. The liberalization of the building societies in the UK has been associated with a less pronounced increase in their financial fragility, not least because the societies' initial balance sheet position was sounder; in addition, prudential supervision has been firmer. Mortgage losses as a percentage of mortgage assets in the UK are much the same as in Japan (0.05% and 0.04% respectively in 1989), although in this country significant additional losses have been passed on to insurance companies offering top slice insurance. Finally, partly reflecting less extensive structural change, losses on continental European deposit insurance schemes have been extremely low. There are significant differences in the insurance schemes operated by different countries. That operated by the USA is clearly the most generous with US$100000 per account insured. The European systems offer lower degrees of compensation and typically provide only a single payout per individual regardless of the number of accounts held. UK deposit insurance offers only partial cover. Differences in coverage between the European countries could be destabilizing if depositors start to chase the best coverage. A further distinguishing feature of the continental European systems is that they are little publicized. In Germany, publicity is actually forbidden, the argument being that the announcement of deposit insurance could reduce confidence in the banking system. In conclusion, although markets are clearly at different stages of development, the housing finance sector in many countries is now tending much more closely towards the 'contestable' model. Features of this include an intensification of competition, a decline in rationing, an increase in the range of institutions active in the market and a development of securitization and wholesale funding.

Financial liberalization, house prices and personal sector gearing During the 1980s, it has been widely suggested that trends in house prices and, to a greater extent, personal sector gearing were associated with variations in the 222

Italy

incidence of financial liberalization. The relationships are, however, quite complex and often simultaneous. As regards the association between financial liberalization and house prices, it is perhaps useful to start by outlining international trends in the latter over the past 15 years. Data are not available for all of the countries considered in this survey and even where available, series are not directly comparable. The UK index is mix adjusted and refers to a sample of all housing market transactions; the US data are also mix adjusted but based on new one-family homes. The series for Italy and Germany are straight averages and drawn from Holmans [15]. For Japan, the figures refer to land rather than house prices. In addition, they represent assessed values (for tax purposes) rather than market values. Nevertheless, the various series offer a broad guide to cross-country developments, and a summary of real house price changes is presented in Table 6. On average throughout the five countries, real house price increases in the 1980s were much the same as in the second half of the 1970s, 2.2% per annum in comparison with 2%. Nevertheless, there were some significant national differences within the aggregate movement. Those countries which experienced the most rapid real price growth between 1981 and 1989 (UK and Japan) had witnessed real price falls over the preceding six years, and vice versa. At first blush, the association between financial liberalization and mean real house price increases appears tentative. While the UK and Japanese data are consistent with such a relationship, the sluggish growth of US house price is less so. Moreover, the slowdown in German and Italian real house prices occurred at a time when financial deregulation was taking place, albeit at a less rapid pace than in the Anglo-Saxon countries. Nevertheless, arguably the effects of financial liberalization on house prices would not be expected to show up very strongly in average data covering a number of years. It is difficult to see, for example, how the deregulation of the financial sector could induce a permanent rise in prices above that justified by fundamentals. On the other hand, there is some evidence to suggest that the housing market is prone to speculative bubbles or fads; for example, E C O N O M I C M O D E L L I N G April 1994

Market structure, institutional development and housino finance: J. Lomax

questionnaire surveys by Case and Shiller [7] investigating the behaviour of home buyers in four regional markets in the USA point to A market driven largely by expectations. People seem to form their expectations on the basis of past price movements rather than any knowledge of fundamentals. This increases the likelihood that price booms will persist as home buyers in essence become destabilizing speculators. There is no necessary reason why bubbles should be initiated by the deregulation of housing finance systems but they could well facilitate the resulting rapid price movements. Accordingly, we should be looking for a relationship between the volatility of house price changes, rather than their mean, and financial liberalization. More research is required in this area. In fact, a general problem with rational speculative bubble theories of house price movements is that it is difficult to explain why the bubble should start (see Tirole [22]). By contrast, fads can be driven by irrational social or psychological forces. In the context of house prices, the latter would seem to be more likely in situations supported by strong fundamentals. These include, for example, rising real incomes, a large proportion of the population concentrated in first time buyer age cohorts and a low user cost of home ownership. Government intervention is also widely thought to exert an influence on prices by affecting ownership decisions. The tax wedges for investment in owner occupation housing in the UK and USA are particularly large (Dean et al [9]). These countries allow the most generous deductibility of interest' payments on debts of up to US$1 million in the USA and £30000 in the UK; owner occupation rates are again high at 64% for the USA and 66% for the U K . 2 In the UK and Japan (where the owner occupation rate is 62%) rent controls have had the effect of reducing the supply of rented accommodation (Minford et al 1-21]), providing an additional incentive for home ownership. This has been reinforced by the low rates of capital gains tax applying to investment in owner occupied dwellings. At the same time tight planning controls in the two countries contributed to the pressure on prices by reducing land availability. By contrast, France and Germany have large rented sectors; rent controls have been significantly less pervasive (Hallett [12]); furthermore, there are fewer fiscal incentives to invest in owner occupied accommodation and more inducements to invest in rented housing. Thus, owner occupation rates are relatively low, at 51% and 41% respectively. Meanwhile, land 2 However, the tax wedge in the UK has decreased since 1985. Mortgage interest relief has not been index linked and thus it cannot account for subsequent house price increases.

ECONOMIC MODELLING April 1994

Table 7. Personal sector savings ratios (%).

Japan Germany France USA UK

1982

1989

13.7 13.8 19.5 9.3 11.6

14.4 13.6 11.4 7.4 5.0

shortages and controls on land use are less pervasive than in the UK and Japan, and hence upward pressure on house prices has been less strong. In overall terms, to the extent that speculative bubbles or fads are caused by fiscal and other government incentives, but permitted by deregulated financial systems, then clearly financial liberalization is less likely to lead to rapid price rises when such incentives are removed or reduced. Moving on to the relationship between financial liberalization and savings behaviour (Table 7), in macroeconomic terms, those countries with the least developed systems of housing finance also have the highest personal sector savings ratios. Of course it is possible that this phenomenon is attributable to cultural factors. On the other hand, it is noteworthy that not only are savings ratios in the Anglo-Saxon countries lower, but they have also decreased in parallel with financial liberalization. There are several mechanisms whereby the nature of the housing finance system could in principle influence personal sector savings behaviour. It has been argued that in circumstances in which house prices are high, but mortgage markets undeveloped, personal savings will be higher. Larger down payment requirements mean that prospective buyers have to accumulate larger prepurchase stocks of financial assets, although this might be partially offset by a tendency to buy smaller houses to reduce the down payment burden and the inheritance of valuable estates. In addition, fiscal incentives associated with savings for house purchase could boost savings. Debate about the influence of undeveloped housing finance systems on personal savings behaviour has been most intense with regard to Japan. Using a lifecycle simulation analysis of Japan and the USA, Hayashi et al [13) argue that the contribution of the induced savings among the young in Japan due to down payment requirements is too small to explain much of the large differential in the savings rates between the two countries. Horioka [16] argues that demographics, and in particular the low share of the elderly in the total population, are much more important. By contrast, Balassa and Noland [1] argue that a special combination of high house prices, undeveloped mortgage markets and a strategic 223

Market structure, institutional development and housing finance. J. Lomax

Conclusions

Table 8. Mortgage debt/GDP ratios 1982-89 (%).

UK USA Germany Japan France

1982

1989

32.1 33.5 22.2 18.7 19.2

58.3 45.2 21.9 25.1 2 !.0

bequest motive on the part of the elderly provides the best explanation of the high Japanese savings rate. Horioka [16, 17] reports that while opinion surveys in the USA suggest old age as the most important motive for saving, surveys in Japan place saving to buy a house as more important. There has been less research into the effects of housing finance systems on savings behaviour in other countries. The counterpart to lower saving ratios in the Anglo-Saxon countries is that financial liberalization has also permitted significantly higher levels of household sector gearing (proxied by the ratio of mortgage debt to G D P ) see Table 8. In the first place, individuals have had greater ability to borrow in order to finance house purchase. Second, equity extraction, which can be defined as the difference between the net increase in the stock of house purchase loans and the private sector's net expenditure on housing, has become more widespread. In the UK, equity extraction rose from £1416 million in 1980 to £24663 million in 1988 (Miles [20]). Not all equity extraction can be directly attributable to financial liberalization, since over the period 1980-84 an average of 48% of the gross figure related to last time sales (Holmans [14]); nevertheless, the residual increase remains substantial. In the USA home equity financed credit has also grown rapidly. Canner et al [6] suggest that the strong rise in such loans was associated with the Tax Reform Act of 1986 (this inaugurated a phased elimination of the tax deductibility of interest; payments remained deductible with increased limits) and largely involved a substitution for other forms of consumer debt. This is, however, disputed by Manchester and Poterba [19], who estimate that 75% of second mortgage lending contributes to a reduction in individuals' net worth. Overall, it seems likely that not only has financial liberalization permitted house price inflation and directly raised gearing, but in addition, individuals have been able to borrow against resulting increases in housing equity which has induced further increases in personal indebtedness. Such developments are clearly possible elsewhere as the liberalization process becomes more widespread.

224

A qualitative analysis such as that presented here prevents very firm conclusions from being drawn, but nevertheless it is suggested that the patterns outlined do offer some potentially interesting implications. In particular, it is suggested that financial deregulation, itself driven by a range of exogenous factors, is affecting the housing finance systems in most countries, although the pace of change has been greatest in the Anglo-Saxon countries. In the latter, not only has there been an easing of the regulatory criteria applied to mortgage loans, but at the same time informal rules stemming from club behaviour have been abandoned as markets have become more contestable. Nevertheless, an increasing degree of convergence is in prospect, perhaps most notably between the U K and continental Europe when the Second Banking Co-ordination Directive comes into force. This convergence may have a number of consequences, notably further reductions in credit rationing, a sharp increase in household gearing levels, some fall in personal sector savings ratios and possibly upward pressure on house prices. The potential for such changes may be strongest in countries with tight regulations at present, and most incentives in favour of owner occupation.

References 1 B. Balassa and M. Noland, Japan in the World Economy, Institute for International Economies, Washington DC, 1988. 2 M. Boleat, International Housing Finance Factbook, International Organization for Housing Finance Institutions, 1987. 3 R.D. Brumbaugh, Thr(['ts under Siege, Ballinger, 1988. 4 G. Butterworth, Housing Finance in Europe, Council of Mortgage Lenders, London, 1990. 5 T.F. Cargil! and S. Royama, The Transition of Finance in Japan and the United States, Hoover Institution Press, Stanford University, 1988. 6 G.B. Canner, J. Fergus and L. Luckett, 'Home equity lines of credit', Federal Reserve Bulletin, June 1988. 7 K.E. Case and R.J. Shiller, 'The behaviour of home buyers in boom and post-boom markets', New England Economic Review, November/December 1988. 8 P. Cecchini, 1992: The Benefits o [ A Single Market, Wildwood House, 1988. 9 A. Dean, M. Durand, J. Fallon and P. Hoeller, Savings Trends and Behaviour in OECD Countries, OECD Working Paper No 67, 1989. 10 D. Diamond and P. Dybvig, 'Bank runs, deposit insurance and liquidity', Journal of Political Economy, 1983. 11 P. Goodhart, Money, It~[brmation and Uncertaino', Macmillan, 1989. 12 G. Hallett, Land and Housing Policies in Europe and the USA, Routledge, 1988. 13 F. Hayashi, T. Ho and J. Slemrod, Housing Finance

E C O N O M I C M O D E L L I N G April 1994

Market structure, institutional development and housing finance: J. Lomax Imperfections and Private Saving: A Comparative Simulation of the US and Japan, NBER Working Paper 15

No 2272, 1987. A.E. Holmans, Flow of Funds associated with House

Purchase for Owner Occupation in the United Kingdom 1977-84 and Equity Withdrawal for House Purchase Finance, GES Working Paper No 92, 1986. 15 A.E. Holmans, House Prices." Change Through Time at National and Sub-National Level, Government Economic Service Working Paper No 110, 1990. C. Horioka, Household Saving in Japan: The Importance of Target Savingfor Education and Housing, Washington DC 1985. 17 C. Horioka, 'Why is Japan's private savings rate so high?', Finance and Development, December 1986.

16

18

C. Horioka, 'Saving for house purchase in Japan', in

Global Rule of the Japanese Economy with Affluent Savings and Accumulated Wealth, Papers and Proceedings of the Fifth EPA International Symposium, Tokyo, 13-14 October 1988. 19 J. Manchester and J. Poterba, Second Mortgages and Household Savings, NBER Working Paper No 2853, 1989. 20 P. Miles, 'Housing market, consumption and financial liberalisation in the major economies', European Economic Review, June 1992. 21 P. Minford, M. Peel and P. Ashton, The Housing Morass, Hobart Paperback No 25, 1987. 22 J. Tirole, 'On the possibility of speculation under rational expectations', Econometrica, September 1982.

Appendix Housing finance institutions in each country USA Although their market share has significantly declined over the past two decades, the savings and loans or thrifts continue to play an important role in the mortgage market (38% share of loans outstanding in 1989). Passbook accounts were the principal retail savings instruments offered 15 years ago, accounting for 80% of the stock of liabilities. By 1985, however, these only accounted for some 12%, with certificates of deposit of various denominations accounting for most inflows. Restrictions on the distribution of savings institutions' assets have eased. In addition to making loans on

mortgage, thrifts are permitted to make loans to purchasers of mobile homes, home improvement loans, loans secured on savings accounts, loans for educational purposes and consumer loans generally. Thrifts can either be mutually owned or be incorporated. Largely because of the very large secondary market, banks and mortgage companies have gained a substantial market share (52% between them: see Table 9). It is also noteworthy that the savings and loan associations have securitized some of their mortgage books. Mortgage companies have no fixed constitution. A little over a third are owned by banks, about a quarter

are controlled by savings associations and other institutions. About 40% are independent organizations. They initially fund themselves short term; loans made are then pooled together and sold to investors in the secondary market. The mortgage bank usually continues to collect monthly repayments which are passed on to the investor purchasing the loan.

UK The building societies, which are mutual institutions, remain the principal mortgage lenders in the UK (Table 10), accounting for 61% of outstanding

Table 9. The US mortgage market - % shares of the major participants in the flow of lending

Savings and loan associations Commercial banks Life insurance companies

1980

1981

1982

1983

1984

1985

1986

1987

1988

46.3 31.7 22,0

55.9 30.1 14.0

2.5 67.2 30.3

54.6 31.3 14.2

55.7 35.0 9.3

51.9 33.8 14.3

41.1 45.7 13.1

50.0 36.1 13.9

38.9 46.5 14.6

Source: Statistical Abstract of the United States. Table 10. The UK mortgage market - % share of the major participants in the flow of lending.

Building societies Banks Miscellaneous financial institutions Other

1978

1979

1980

1981

1982

1987

1988

94.1 5.0

81.0 9.1

78.6 6.9

68.0 24.4

57.6 35.9

1983

76.3 24.7

87.6 12.3

1984

1985

78.6 22.7

1986

79.0 21.1

50.6 34.6

59.2 27.2

1989

71.2 21.3

1990

74.8 19.5

0 1.0

0 10.0

0 14.6

0 7.7

0 6.5

0 - 1.0

0 0.1

0 - 1.3

0 -0.1

13.5 1.3

12.5 1.0

7.6 - 1.0

6.4 -0.6

Source: Financial Statistics.

ECONOMIC MODELLING

April 1994

225

Market structure, institutional development and housing finance: J. Lomax mortgage loans (1990). There are 116 societies (March 1991), although the largest account for 80% (1990) of societies' total mortgage assets. The banks and specialized lenders (which fund exclusively from the wholesale markets) have gained increasing importance over time and their market shares are 30% and 6% respectively. The 1986 Building Societies Act laid down the regulatory framework under which societies operate. In this context, the largest institutions have evolved, to some extent, away from their traditional role of collecting retail savings and offering mortgage loans. On the liabilities side, societies now provide a range of money transmission services as well as savings products. Funds are also obtained from the wholesale market. On the asset side, societies are now permitted to make some unsecured loans (10% of their total asset portfolio). They offer a wide range of additional services; these include investment services for individuals, trusteeship, insurance, residential development, and estate agency. The UK has developed the most sophisticated market in mortgage backed securities outside the USA, although its scale remains small. Unlike the savings and loans associations in the USA, societies have so far found it unnecessary to securitize since they have faced fewer capital constraints. Indeed, recently societies have, via special subsidiaries

(appropriate mortgage companies), purchased a modest amount of mortgages from other lenders. Germany

No one type of institution has more than 30% of the total German mortgage market (Table 11), with the four biggest sectors being the mortgage banks, savings banks, Bausparkassen and the credit cooperatives. Far from being independent entities, however, these bodies have strong interrelationships. The savings banks (26% of outstanding mortgage loans in 1990) are major financial players, being the largest holders of personal deposits, and they also offer a complete banking service to industry and commerce, Long-term housing loans accounted for 24% of their assets at the end of September 1990, a relatively small proportion compared with the Liinder (state governments); they also own the central giro institutions. These have interests in the public mortgage banks and public Bausparkassen and undertake some direct lending themselves. The mortgage banks have 25% of the market, and at the end of September 1990 long-term loans for house purchase comprised 37% of their assets. They are not independent institutions but rather are subsidiaries of more general financial institutions. The banks obtain most of their funds from bearer

bonds and long-term borrowing. They are in principal not deposit taking institutions. There are 38 mortgage banks. Of these, 27 are private with the majority being subsidiaries of other financial institutions, especially the big three commercial banks. The public banks are owned either by federal government or by the central giros. The Bausparkassen (13 % share) operate a contractual savings scheme which involves people contracting to save a certain sum after which they are entitled to receive a loan. These accounts have the advantage of both tax-free interest and a government bonus. In the event of funds being limited, loans are distributed according to a system based on the time elapsed since the savings contract commenced and the amount of savings. The loans portfolios of the Bausparkassen are heavily focused on the housing market, since 91% of their assets are building loans. There are 17 private Bausparkassen which are owned wholly or in part by insurance companies, commercial banks and other financial institutions. There are also 11 which are publicly owned by the savings banks' regional giro organizations. France

The main French housing finance institutions (Table 12) are the banks, (63% market share) the Caisse des D6p6ts et des Consignations (CDC)

Table 11. The German mortgage market - % shares of the major participants in the flow of lending.

Mortgage banks

Bausparkassen Savings banks Commercial banks Other

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

29.8 24.7 22.9 4.6 18.1

31.9 21.6 28.2 3.2 15.0

18.1 12.1 29.4 10.9 29.5

20.8 13.6 31.4 13.7 20.5

22.4 2.0 27.3 18.3 29.9

26.8 -- 4.7 31.5 31.5 14.9

25.6 - 3.0 31.0 30.9 15.4

14.7 - 20.2 36.5 50.9 ! 8.1

12.2 - 5.9 26.7 40.3 16.6

11.0 7.3 26.7 43.2 11.7

Source: Statistiche Beihefte zu den Monatsberichten der Deutschen Bundesbank. Table 12. The French mortgage market - % shares of the major participants in the flow of lending. 1978

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

Banks and finance companies Cr6dit Foncier and Comptoir des Entrepreneurs Caisse des D6p6ts et Consignations and the

50.3

54.3

51.5

47.5

47.6

48.0

46.5

53.6

53.2

62.6

65.2

69.8

13.9

14.0

13.2

13.9

15.8

18.1

21.2

15.7

16.8

13.0

10.6

8.0

caisses d'~pargne

29,7 6,1

25.9 5.9

28.9 6.3

31.6 7.0

29.5 7.2

26.9 7.0

25.1 7.1

23.7 7.0

23.7 6.4

18.8 5.6

18.9 5.3

17.4 4.7

Other

Source: Bulletin Trimestiel.

226

ECONOMIC MODELLING

April 1994

Market structure, institutional development and housing finance: J. Lomax Table 13. The Spanish mortgage market - % shares of the major participants in the flow of lending.

Private banks Savings banks The Mortgage Bank of Spain Mortgage companies

1983

1984

1985

1986

1987

1988

1989

2.8 55.3 39.8 2.1

7.1 41.6 46.4 4.9

15.0 31.4 34.7 10.8

23.7 47.4 18.0 10.9

28.5 50.4 7.9 13.3

37.8 44.7 5.9 11.5

36.2 53.2 5.8 4.7

Source: Revista Espafiola de Financiacion a la Hvienda March 1991. Table 14. The Japanese mortgage market - % shares of the major participants in the flow of lending.

Bank accounts of all banks Trust accounts of all banks Housing loan companies The Housing Loan Corporation Other

1981

1982

1983

1984

1985

1986

1987

1988

1989

24.6 2.4 14.8 45.4 12.8

19.7 0.9 13.0 54.1 12.3

11.9 - 0.2 11.0 62.7 14.5

12.3 - 0.4 8.0 66.5 13.7

35.1 - 1.6 1.6 59.8 5.1

52.5 - 2.0 12.4 46.6 -9.5

54.2 0.2 13.0 30.2 2.4

47.7 - 0.4 8.5 40.0 4.2

42.9 1.9 16.2 27.8 11.2

Source: Economic Statistics Annual, Bank of Japan (19%) and the Crrdit Foncier (13%). The banking sector includes a variety of different types of institution including commercial banks, the Credit Agricole, and mutual credit organizations. The CDC is a public sector body which centralizes funds from the savings bank system, pension funds and other institutions and uses these to fund public capital investment and social housing. The Crrdit Foncier is a private company, but government related, which plays an important role in implementing official policy towards the housing market. One of the characteristics of the French housing finance system is the variety of types of loan available. In addition to market-determined products, these include PAP (prdts aides pour l'accession ~ la propribtb) loans which are state subsidized and available to low or middle income households; conventional loans (PCs) which are not subsidized but are made on terms (including a ceiling on the rate of interest) stipulated by the government; and plans d'bpargne logement which are made to those who have completed a contractual savings scheme. France has a secondary mortgage market which is supervised by the Credit Foncier. Before a loan can be refinanced on the market, it must meet a number of conditions. Nevertheless, a declining proportion of loans are refinanced in this way since the banks generally raise sufficient funds through their deposit taking activities to fund their lending.

ECONOMIC MODELLING

Spain The confederated savings banks are by far the largest mortgage lenders in Spain (53% of outstanding mortgage loans at the end of 1989) and these are also the major retail financial institutions (Table 13). They fund themselves largely through retail deposits and approximately half of their loans are accounted for by advances for house purchase. The other major operator in the Spanish mortgage market is the Mortgage Bank of Spain (17 % share). It is a public sector body owned and controlled by the government, making loans directly for house purchase and also acting as an agent in the implementation of government policy.

Italy In Italy the housing finance market is significantly less developed than that of most other Western European countries. Indeed, by some estimates 75 % of house purchase is financed by borrowers' own funds. The government plays a central role in the financial system, not least through its ownership of most of the major banks. Banks themselves are restricted to short-term loans (the rationale being that they fund themselves with short-term deposits). Thus, special credit institutes (SCIs) - which are generally owned by the banks undertake the majority of mortgage lending, typically funded by bond issues. Increasingly, however, the banks are

April 1994

being allowed access to the longer-term market.

Japan

The mainstream (city and regular) banks (Table 14) are the principal group of mortgage lenders (41% of outstanding mortgage loans at the end of 1990). These raise most of their deposits from the personal sector. The second largest provider of housing credit in Japan is the public sector (36% share) mainly through the Home Loans Corporation. The Corporation obtains almost all of its funds through an intermediary - the Trust Fund Bureau - from the postal savings system. Moreover, it does not lend money directly, but rather operates through approved financial institutions acting as agents. The board of the HLC is appointed by the government. Rates of interest charged are modest because of the non-commercial way in which the system operates. Because the government stipulates both the rates of interest at which the HLC lends and the rates of interest on postal savings, it can always ensure an adequate margin between the two. The third biggest category of mortgage lenders are the housing loan companies, which account for about 13% of the total market. These are owned by groups of financial institutions, including the banks, and they raise their funds generally by borrowing from parent institutions.

227