Future housing demands and sources of mortgage money
Alan R. Winger
In the eyes of most housing experts ten years ago, the 1980s was to be a decade in which homeownership demands would accelerate and the Nation’s money and capital would be hard-pressed to meet the financing needs of American home buyers. Our actual experience during the first three years of the decade raises questions about the strength of housing demands. Housing starts have averaged only a little over one million units per year, about half of what many expected. On the other hand, questions about the ability of our money and capital markets to provide for the home financing needs of Americans during the 1980s remain.
That demands have seemingly disappeared but funds needs have not present a paradox that needs an explanation. How much residential mortgage financing will be demanded during the remaining years of the 1980s? Will those dollars be forthcoming? What role will savings and loans play in satisfying these needs? These are the questions of concern in this article. HOUSING DEMANDS IN THE EIGHTIES New Unit Demands: “Need” Factors
Additions to the household population, removal of units from the stock, and added vacancies to accommodate mobility needs are the underlying demand factors in the new housing market. They give rise to the “need” for new dwellings. Of the three, population change is the most important, accounting for about 80 percent of new construction in the 1960s and 1970s. Most of the significant changes in new housing construction have also been a consequence of demographic developments. The sharply rising level of new housing construction in the 1970s, for example, was largely due to a rapid increase in net household formations.
Expectations for new housing construction were high coming into the 1980s, largely because of anticipated population developments. The “baby boom” of the 1940s and 1950s was still adding significantly to the age cohorts in which most new household formations occur. The upward trend in the formation of households by unrelated individuals was also expected to keep net household formations in the 1.5 to 2.0 million range. Given this as a base, it was not difficult to build a case for new housing starts of in excess of two million per year throughout the Eighties. New Unit Demands: Enabling Conditions
Market demand, of course, is more than a reflection of need. Casting dollar votes in some market not only reflects need or desire, the dollar votes must be there. Enabling conditions or ability to pay is a key element in demand.
Whether people form households depends on their economic circumstances. Most people, when they come to a certain point in their life cycle, seem to prefer a separate dwelling, and this preference usually comes to the foreground when they think about marrying, divorcing, graduating from college, getting their first job, etc. People, however, have the option of doubling up or living with others. What they do is importantly influenced by their economic situation. Unfavorable economic conditions can affect the timing of the formation of a new household. If unfavorable enough, they can also lead to a decision not to form a household or lead to the dissolution of an existing household. The enabling conditions are particularly important to the decisions individuals make about establishing a separate living arrangement since it is much easier for them to find living accommodations with others. Historical Perspective
Household formations in the 1960s averaged about 1 million per year. In the 1970s, they averaged about 1.6 million per year. Knowing why household formations increased 60 percent in the 1970s provides insight into earlier expectations about the 1980s and why it might be best now to revise these expectations somewhat downward.
Preliminaries: The Determinats of Household Formations. Studies of the determinants of household formations generally partition sources of household growth into three groups: 1) population growth, 2) changes in the age structure of the population, and 3) other factors working through headship rates (the proportion of individuals in a subgroup of the population who are a household head). The importance of population growth is apparent. Age distribution is potentially important because subgroup headship rates have differences. They increase with age, jumping sharply during the early, years of adulthood. If population increases irregularly, household formations will be affected. They will increase as the “bulge” works its way through the age cohorts where headship rates increase. If the population growth rate falls, household formation will diminish when those born during the period of declining growth began to reach these cohorts. The other factors include both sociological and economic variables. The sociological variables include all those influences that shape the values of people which affect their decisions to establish separate living quarters. Attitudes toward marriage is one of these. Attitudes toward privacy is another. The economic variables are primarily income and consumption cost factors that influence the ability of individuals to establish a household or separate living quarters.
By working with headship rates, we can calculate the changes in household population attributable to population growth, changes in the age structure and changes in the headship rate. Table 1 shows the results of these calculations for the decades of the Sixties and Seventies.
Net Household formations in the 1960s. Table 1 indicates population growth as the primary source of household formations in the Sixties. Changing age distribution of the population shows up as a negative factor, which is not surprising. The population moving into the 20- to 35-year age group–the age group in which headship rates increase sharply–was born in the 1930s and 1940s, a period in which birth rates were at low levels. There was, on the other hand, a good deal of growth in the Nation’s household population because of rising headship rates. Significantly, the impetus here was through the formation of households by unrelated individuals. Table 2 clearly shows the sharp increase in headship rates of individuals over this period; those of families changed very little. While the Sixties was not the beginning of the movement of individuals into separate living quarters, it was a period in which the trend accelerated. The cause was the rapid growth in real income. Real per capita income jumped sharply in the Sixties–see Table 3–putting many individuals in a position to live in a separate residence if they chose to do so. The sharp increase in the headship rate indicates that many chose to do so.
Net Household Formations in the 1970s. In the 1970s, we added 60 percent more to the household population than in the Sixties. Population growth was a factor but was less important, both relatively and absolutely, than in the previous decade. The big difference was the impetus coming through changes in the age distribution of the population, which were important because of the movement of the “baby boom” population into the age cohorts where headship rates increased sharply. Household formations also increased because of the continued increase in the headship rates of individuals. Only in the Seventies did individuals continue to move into a separate residence more because of what happened to housing costs than incomes. While real per capita disposable income continued to rise, the rate of growth slipped. Productivity problems developed and the economy began to stagflate, slowing real income gains. Housing consumption costs, on the other hand, were being reduced, in part, because of a decline in real financing costs. (See Table 4.) For the homeowner, these costs also feel because of the rapid run-up in home prices, which sharply reduced the capital component of homeownership costs.
Household Formations in the Early 1980s. The sharp decline in new housing starts during the first three years of the 1980s, combined with a vacancy factor that did not fall very much, suggests the demographic variable is currently providing less impetus to demand than expected. While the facts are not yet available, there is indirect evidence of a diminution in the force of the demographic factor. Nothing happened to diminish the expected impact of population growth and the changing age distribution of the population. What has apparently happened is the demand potential stemming from headship rate changes has been reduced. Until early 1983, economic stagflation in the Eighties reduced real income gains to zero which had to impact negatively on decisions by individuals to establish separate living quarters. Equally important is what happened to real housing consumption costs. Real financing costs rose sharply following the rapid run-up in interest rates in 1980. (See Table 4.) The upward ascent in home prices also slowed, pushing up the capital costs of housing consumption in owner-occupied dwellings. Affordability became a major issue in the housing market. Separate living quarters apparently became a luxury that an increasing number of individuals decided they could no longer afford, including some who were already living in a household arrangement. When the figures become available, headship rates of individuals will probably show no increase during the first several years of the Eighties. Indeed, they may even show a decline. The Remaining Years of the 1980s: What Will the Demands Be?
Most earlier projections in the Eighties had net household formations running at 1.6 to 1.7 million per year. Obviously, we did not achieve this level during the first three years of the decade. Economic factors restrained a great number of people from moving out on their own–particularly individuals. If that restraint continues, household formations will continue to fall below earlier expectations, which implies housing demands that will not support the expected high levels of new construction. What happens to the economy, particularly as it influences the headship rates of individuals, is thus the key to housing demands over the remaining years of the 1980s.
The importance of the future of economic developments, as they affect these headship rates, can be illustrated with Census Bureau projections. The Bureau of the Census publishes a series of household formation projections that differ on the basis of certain assumptions made about key parameters. One of these is headship rates.
Three illustrative projections are shown in Table 5. The primary difference is in the assumptions made about headship rates. Projection B assumes that the growth rate of headship rates during the period, 1964-78, continues from 1980-95. Projection C assumes a short growth trend (1974-78) in headship rates to project to 1980 and uses headship rate trends from 1966-80 to project from 1980-95. Projection K assumes that headship rates observed in 1978 will continue to 1995. Since the headship rates for families changed little in the 1960s and 1970s, these differences in assumption are reflected in the headship rates for individuals.
Projections B and C assum, in effect, a continuation of the increasing trend in the headship rates of individuals. They have net household formations running between 1.6 and 1.7 million per year. Projection K assumes the headhsip rate of individuals will remain as they were in 1978 and yield a projection that is about 500,000 units per year less. If we assume demographic factors will account for 75 to 80 percent of the unit housing demands over the remainder of the decade (much as it has over the previous two decades), we are talking about a period in which housing starts will average about 1.5 million per year, a figure well below the earlier expected 2 million or more units per year.
Must we revise our expectations about housing demand in the 1980s downward? Clearly, if recent history provides insights, our views about future housing demand must reflect our forecasts of the economy. That this is so does not augur well for our ability to forecast housing demand. The record of the economic forecaster in periods of economic turmoil, e.g., the Seventies and early Eighties, is less than impressive.
Suppose the economic turmoil continues. We are talking about economic conditions that are unfavorable to the formation of new households by unrelated individuals. There will be economic constraints that keep the headship rates of individuals from increasing, which means household formations closer to projection K than the other two. Housing production, given the level of household formations, would be well below the 2.0 million-plus housing starts many expected as we came into the 1980s.
Suppose, on the other hand, the economy came out of the current recovery performing substantially better than most now expect, generating the incomes and financing costs that make it feasible for many more individuals to move out on their own. If this happens, earlier prophecies of housing demand in the Eighties could be fulfilled. Housing starts could rise to 2.0 million-plus.
The author’s view is that many of our economic problems will remain as we come out of the recovery period. The result will be economic conditions that restrain individuals from establishing a separate residence. The upward trend in headship rates will not resum; hence, household formations will be closer to K the B or C. However, they will not be as low as those projected in K. Housing starts will probably average about 1.7 million per year over the remaining years of the Eighties and will do so for two reasons. One is, while the economy may not perform well, that performance is likely to be better than it was in the latter part of the Seventies and early part of the Eighties. Recent actions taken to deal with our economic problems should bear some fruit, which should mean some improvement in income and financing conditions. But it will not be enough to lead to significantly higher levels of household formations. Secondly, the very low levels of housing starts during the first three years of the Eighties undoubtedly generated some “pent up” demand that will come into fruition as the recovery continues and we move into the next phase of the cycle.
Those, the author believes, who expect that housing starts will average 2.0 million-plus during the remainder of the Eighties will be disappointed. Demands are more likely to sustain a starts level of about 1.7 million units per year. While below the glowing expectations of many as we started the decade of the Eighties, housing construction at this level would still represent an activity level that generates a healthy demand for mortgage money. HOUSING FINANCE PROBLEMS IN THE EIGHTIES Sources of Funds: Background
Housing finance problems in the 1980s, as we looked at them in the 1970s, appeared to stem from the expected heavy mortgage demand in the face of strong capital demands from competing sectors and an archaic mortgage finance instrument. There was much discussion of the need to broaden the base of mortgage money through the development of new financing instruments, e.g., mortgage-baced securities and the enlargement of the secondary market for mortgage loans.
Until the 1970s, most housing finance was provided by four major financial intermediaries–savings and loans, commercial banks, life insurance companies, and mutual savings banks. These institutions accounted for in excess of 80 percent of the net acquisition of long-term residential mortgage loans made each year up through 1970. (See Table 6.) The mortgage funds flowing through these institutions since then have diminished. In 1975, net acquisitions totaled about 75 percent. In 1980, this figure was 66 percent. (See Table 6.)
Simultaneous with this diminution of the relative importance of the financial intermediary has been the rapid development of the secondary market for residential mortgages and the related rapid growth of the market for mortgage-backed securities. The growing importance of these markets is clearly reflected in Table 6. Loan sales, as a proportion of total residential mortgage loans originated, rose from about one-third to one-half from 1970 to 1980. The net acquisitions involved in pools of mortgages backing up securities increased from 4 percent of the total to 17 percent over the same period.
Many have argued for some time now that the development of the secondary and mortgage-backed securities markets is essential if the home mortgage borrower is to compete effectively in the Nation’s savings market. After the passage of the Depository Institutions Deregulation and Monetary Control Act of 1980 and the Garn-St Germain Depository Institutions Act of 1982, further expansion and development of the these markets is deemed even more important by most housing finance experts.
Dergulation implies that savings and loans–for years the major source of mortgage money to the American home buyer–will become less important. While thrifts will remain heavily committed to real estate, they now have the opportunity to move in other lending markets and are likely to do so for a number of reasons, not the least of which is to achieve more balance in the maturity or duration of their assets and liabilities. The removal of th rate advantage that thrifits have over commercial banks will also affect deposit inflows into thrifts, which could reduce their ability to add loans to their mortgage portfolio on the basis of funds acquired through deposits. Most expect the traditional commitment of thrifts to real estate lending to remain, but with operations that become increasingly involved with the secondary market in the sense og originating loans, forming mortgage pools, issuing mortgage-backed securities, and selling them to other investors. Savings and loans are thus expected to gear their operations in a way that brings them into the mainstream of developments taking place that restructure the mortgage market and, hence, intensify that development. Alternative Mortgages
Another important recent development in the mortgage market is the alternative mortgage financing arrangements that have flourished over the past four to five years. Some of these–the so-called creative financing arrangements such as builder buydowns, wrap-arounds, mortgage loan assumptions–have become less frequently used as mortgage rates have fallen to levels that reduce the dimensions of the “tilt” problem confronting potential home buyers. Other instruments, such as the adjustable rate and graduated payment mortgages, will remain.
The adjustable rate and adjustable payment mortgages are not new. They were just not used very much because of regulatory constraints. It was not until the inflation of the late 1970s and early 1980s created severe problems for both mortgage lenders and borrowers that these restrictions were eased. Now many lenders make both adjustable rate and adjustable payment mortgages, as well as certain other kinds of mortgages such as the price-level adjusted and shared appreciation mortgages.
While there is no indication that borrowers have lost interest in the fixed-rate mortgage, the current use of alternative mortgage forms will not disappear. How important that use becomes will depend partly on the performance of the economy. It the problems of inflation and instability persist, there will be economic conditions that sustain borrower and lender interest in the adjustable rate and adjustable payment mortgages. Given the view expressed earlier about the future economy, the interest in and use of these two mortgage forms should be sustained over the remaining years of the 1980s. No matter what happens to the economy, however, residential mortgage financing arrangements are likely to continue to take many forms in a deregulated setting, given the diversity in lender and borrowers needs. The mortgage market will never be like it was What Will the Mortgage Market Look Like?
While no one really knows what is going to happen in the future, it is not difficult to draw up scenarios that set forth reasonable possibilities. Four models that fall under this heading are: 1) a reversion-to-what-was model, 2) the retail industry model, 3) the mortgage-backed security model, and 4) a hybrid model.
A Reversion-to-What-Was Model. The key element in this model is the assumption that thrift institutions regain their position in the savings market and use savings deposits to make mortgage loans that are added to their mortgage portfolio.
The Retail Industry Model. This model assumes an ever-growing secondary market which becomes dominated by a smaller number of relatively large institution that operate nationwide and make only a few types of mortgage loans. A proliferation of mortgage forms is quickly found to be inconsistent with the requirements of a well-developed secondary market. A primary ingredient of such a market is a homogeneous product. A quick response to an offer to sell implies a ready knowledge of what is being purchased. If the product being exchanged is homogeneous, buyers and sellers can come to know it and, hence, respond decisively and with dispatch in purchase and sale. Expanding the number of products being traded was found to mean less homogeneity. For the secondary market to continue to grow, the number of mortgages traded had to be restricted and the remaining products standardized.
Simultaneous with this development is a rapid growth in the number of firms operating locally making specialized mortgage loans. These operations develop successfully in response to market demand. A population with diverse needs is not satisfied with the limited mortgage offerings of the secondary market. Locally owned firms step in and fill this need.
This model is called the retail industry model because of its parallels with the retail industry. That industry, of course, is dominated by the activities of large chain stores with nationwide operations. It is also populated with a great many locally owned speciality shops that meet needs not satisfied by the chains.
The Mortgage-Backed Security Model. This model assumes the mortgage-backed security comes to totally dominate the market. Most mortgage loans made are placed in pools which back up security issues that are sold to a wide range of investors. Many different types of loans are made and combined into these pools. Heterogeneity causes no problems in the nationwide distribution of the mortgage securities. The problem of risk evaluation when the pool is composed of heterogeneous mortgages is resolved through the application of insurance techniques that also foster the development of pooling techniques that better control the risk. Scale factors in this kind of operation lead to a mortgage market that is dominated by a small number of relatively large firms.
The Hybrid Model. The hybrid model is a combination of the other three. It assumes that thrifts continue to make mortgage loans for their own portfolios on the basis of funds they accumulate through savings deposits and borrowings from Federal Home Loan Banks. It also assumes further growht in the secondary market and a standardization of the mortgages involved in secondary market transactions. Some of the firms operating in this market grow quite large and deal with relatively few mortgage forms. Limitations on the type of mortgage loans made for sale in this market lead to a growth in the number of locally owned firms that make only special types of mortgage loans. Some of these are savings and loans that still accumulate deposits and borrow from Federal Home Loan Banks. The mortgage market is not totally dominated by either of these two groups, however. Mortgage-backed securities market experiments with, and achieves some success in, working with pools of heterogeneous mortgages, applying insurance techniques to deal with risk evaluation problems. The mortgage market does not become one great big mortgage-backed security, but the MBS market remains and grows in importance. What Will It Be?
While no one really knows what the future shape and form of the mortgage market will be, if change is evolutionary rather than revolutionary, the author believes the hybrid model should prevail. The secondary mortgage market and mortgage-backed securities market will expand further, and one important aspect of this expansion is likely to be the growth and development of a number of very large mortgage lending firms that operate nationwide. This development, however, will not preclude locally owned firms from making mortgage loans in local markets. Some of these loans will be very specialized; others will be more conventional, made on the basis of locally generated funds.
In a deregulated setting, a market with such diversity should evolve without difficulty, providing all of the funds residential mortgage borrowers want at a price (interes rate). That price is likely to be relatively high, however, if the economy does not perform well. Indeed, relatively high interest rates are one of the reasons why housing demand is expected to sustain a starts level of only 1.7 million units per year.
Thrifts in this setting will find plenty of opportunity to pursue their real estate lending goals. While they will not be able to operate in this market as they have in the past, they will be able to add loans to their portfolio through funds accumulated by deposit inflows and borrowings from such sources as the Federal Home Loan Banks. The continuation of a dominant role in real estate finance, however, will require that thrifts become more actively involved in the secondary market and with mortgage-backed securities operations. Thrift participation in the secondary market has increased sharply over the past five years; their involvement with mortgage-backed securities has moved more slowly. Thrifts will find more and more that to expand, indeed even maintain their position in real estate finance markets, they will have to work at establishing further links with a market that is evolving into a national market, the key elements of which at this time are the secondary and mortgage-backed securities markets.
COPYRIGHT 1984 U.S. Government Printing Office
COPYRIGHT 2004 Gale Group