But beginning in the 1970s, market forces and innovations began to erode the original advantage of specialized thrifts. Indeed, the special nature of the savings and loan--originating and holding long-term assets financed with short-term liabilities--could flourish only in a noninflationary environment, where interest rates were low and stable and where the yield curve rarely varied from its traditional upward slope. But as inflationary forces began to surface, market pressures on the conventional thrift model began to build. The need to adjust average asset maturities in the face of rising interest rates spurred development and expansion of a secondary mortgage market.

The mortgage banking industry, which during the 1950s and 1960s had focused almost exclusively on the origination of FHA and VA loans, emerged from being a fringe player to being a primary supplier of loans because of the expansion of the secondary mortgage market. This, along with the development of mortgage-backed securities and the technological revolution facilitated by the computer, hastened the evolution of the original thrifts from institutions with mismatched maturities of assets and liabilities to the highly viable institutions you represent today.

The mortgage-backed securities market grew dramatically, beginning in 1970 with the issuance of the first Ginnie Mae pass-through security and followed by Freddie Mac's sale of mortgage-backed securities backed by conventional loans in 1971, reflecting the wide acceptance of these securities by the investor community. This was also an era when the principal mortgage lenders, savings and loans, were sometimes constrained from satisfying mortgage demands by binding Regulation Q ceilings that eroded their deposit base when interest rates rose. In those difficult times, the development of the mortgage-backed securities market helped to provide a safety valve and, as a result, the standard residential mortgage today need no longer be funded or originated by specialized financial institutions.

By the mid-1980s, the channels for mortgage originations and holdings had become quite diverse, with the traditional depositories competing against mortgage brokers and mortgage bankers, who sold their loans not only to Fannie Mae and Freddie Mac but to others as well. This greater institutional diversity in the sources of mortgage finance played a key role in maintaining the uninterrupted flow of mortgage credit during the then-biggest financial debacle since the Great Depression--the S&L crisis of the late 1980s. The resiliency of the mortgage credit market during that period highlights the value of having a diverse set of financial institutions and financial markets that serve a key sector of the economy, such as housing.

To repeat, the 1990s have generally been a time of robust growth for the mortgage markets. But who at the beginning of the decade would have foreseen years when mortgage originations exceeded $1 trillion? While most people not involved in financial markets tend to think of the mortgage markets in terms of new home construction and ownership, it is, of course, predominately a market that dynamically finances the existing stock of housing. Moreover, owing to the simple arithmetic of our population growth, it is almost certain to become increasingly dedicated to existing home purchases and refinancing in the 21st century.

Over the past five years, for example, mortgage loan extensions on newly constructed homes averaged about $140 billion annually and comprised only about one-sixth of total mortgage loan originations on single-family dwellings. Three decades earlier it was more than 40 percent. Almost surely three decades hence, new home extensions will be even lower as a share of the total market than they have been in recent years.

The reason, of course, is that existing home sales reflect the turnover of the existing housing stock that, in turn, parallels the level of the population or, more directly, the number of households, while sales of new homes, more or less, are driven by the growth in population or, still more closely, household formation. Short of a significant acceleration in immigration or a remarkable surge in the birthrate, few demographers would project a continuing rise in the rate of population growth to match the inexorable rise in the level of population. To put it briefly, the rate of increase of both our population and the number of households appears destined to slow in the coming years. So assuming a continuation of the relatively stable turnover rate for our existing housing stock, existing home sales will continue to grow, though perhaps at a declining rate of increase.

Newly constructed homes, however, are tied to growth in the number of households. In the future, that growth is likely to be flat at best, with very little upside potential. Indeed, over the 1990s, single family starts averaged 1.10 million units, actually slightly lower than the average of 1.14 million units during the 1970s. Sales of existing homes, of course, reflecting the growth in population, averaged about 4 million during the 1990s, compared with a much smaller number--2.8 million-- during the 1970s.

This trend has important implications for economic activity beyond its impact on housebuilding, because the sale of a newly constructed home does not generate capital gains financed through the mortgage market. Sales of existing homes almost always do, and the purchasing power released through converting home equity to unencumbered cash can affect overall consumer demand and the economy, just as the stock market gains of recent years have boosted consumption.

Estimates by the staff of the Federal Reserve indicate that about 40 percent of the growth in outstanding home mortgage debt during the past five years originated as financing the extraction of home equity. Such borrowing has largely reflected the extraction of realized capital gains, which almost automatically takes place on the sale of our stock of existing homes, as the new owners take on a much larger mortgage than the seller had at the time of sale.

The average nine-year period of ownership brings a substantial increase in market values even when the average annual increase in home prices has been modest. Obviously, were home prices to fall for a protracted period, no capital gains would be available to extract. Mortgage financing in such an environment, I am rash enough to say, would have been less than pleasant.

We estimate that, over the past five years, the average capital gain on the sale of an existing home net of transaction costs was more than $25,000, almost a fifth of the average purchase price, and roughly equal to the equity extraction financed by debt. But not all equity extraction reflects the capital gains realized from the sale of an existing home. A substantial part, approximately half, in recent years was the result of unrealized capital gains being drawn out through home equity loans and cash-out refinancing. Presumably even normal amortized equity that did not come from higher home prices was extracted in this manner.

To the average seller of an existing home, the equity extracted net of transaction costs generally far exceeds the down payment on his or her next home purchase. The unencumbered cash to the seller, while financed by debt, is not the seller's debt. Indeed, it appears that a significant amount is spent on consumer goods, especially big ticket items in a manner not materially different from windfall income in general. Home equity loans and cash-out refinancings, of course, finance both consumer purchases and debt consolidation.

Although, as I indicated in an earlier speech on this subject, the appreciation of stock prices has been vastly greater than that of home prices, most estimates suggest that stock market gains are consumed only gradually, with the level of consumer outlays lifted permanently by around 3 to 4 percent of the wealth generated by the stock market gain. The permanent increase in spending out of housing wealth is somewhat higher, perhaps in the neighborhood of 5 percent, and is financed in a different manner.

The major reason for these significant differences in spending out of household wealth is doubtless that, while home prices do on occasion decline, large declines are rare; the general experience of homeowners is a modest, but persistent, rise in home values that is perceived to be largely permanent. This experience contrasts markedly from volatile and often-ephemeral gains in stock market wealth. Moreover, most stock market wealth effects are associated with the highest income groups where the marginal propensity to spend is thought to be lower relative to somewhat lower-income groups where the preponderance of housing capital gains are realized.

Stock prices and existing home sales are somewhat correlated, a not altogether unexpected result, because each is affected by interest rates and presumably the gains from each help finance the other. This correlation makes it difficult to disentangle gains in overall consumer spending that are attributable to home equity extractions and to increases in stock prices. Nonetheless, the evidence suggests that, in recent years, about a sixth of the so-called wealth effect--that is, the impact of capital gains on consumer spending--stems from equity extracted from the stock of existing homes.

As the stock of dwelling units increases with population, the home mortgage market almost surely will become increasingly dominated by the financing of capital gains and the extraction of equity. An element, however, in the translation of total housing units into owner-occupied units will, of course, be the trend of home ownership, since the financing of rental housing has significantly different sources and economic consequences.

After rising steadily for nearly a half-century, the homeownership rate was stuck during the 1980s and early 1990s at a little more than 64 percent. The recent rise in the homeownership rate to over 67 percent in the third quarter of this year owes, in part, to the healthy economic expansion with its robust job growth. But part of the gains have also come about because innovative lenders, like you, have created a far broader spectrum of mortgage products and have increased the efficiency of loan originations and underwriting. Ongoing progress in streamlining the loan application and origination process and in tailoring mortgages to individual homebuyers is needed to continue these gains in homeownership. Lowering the costs of homeownership is particularly important for increasing homeownership rates among young adults. Recent progress in this area has been encouraging. For example, homeownership for adults from ages 25 to 29 has risen from about one-third to about 36 percent over the past several years. But in the early 1970's, the ownership rate for this age group was 44 percent. Putting today's youth on a higher ownership trajectory would be in the best interests of both your industry and of the country.

Community banking epitomizes the flexibility and resourcefulness required to adjust to, and exploit, demographic changes and technological breakthroughs, and to create new forms of mortgage finance that promote homeownership. As for the Federal Reserve, we are striving to assist you by providing a stable platform for business generally and for housing and mortgage activity. Our shared objective is to maintain a strong economy and to provide the setting so that homeownership becomes a reality for all who desire it.