The Federal Reserve Board eagle logo links to home page


Remarks by Vice Chairman Roger W. Ferguson, Jr.
At America's Community Bankers Annual Convention, Seattle, Washington
October 31, 2000

Opportunities and Challenges in Community Banking

It is a pleasure to join America's Community Bankers for your 2000 Annual Convention. Today, I would like to discuss certain aspects of your most important markets--the residential mortgage, consumer, and small business markets. I would like to put these comments in a broader context of developments in our economy.

With the passage of more than a decade since the savings and loan crisis, we have abundant evidence that depository institutions that allocate most of their resources to housing finance are viable and can produce an ongoing, high level of profitability. This year, earnings by savings institutions during the second quarter, though not a record, remain at the high level that has persisted now for more than five years. In addition, almost 61 percent of those savings institutions insured by the FDIC showed improved earnings from a year ago, and only 8 percent were unprofitable.

The saving industry's continued low level of noncurrent loans is also encouraging. Furthermore, despite the concerns of some that depository institutions may be unprepared for the higher delinquency and default rates that could result should economic growth slow substantially, the saving industry's ratio of loan-loss reserves to noncurrent loans rose to a record high in the second quarter. Although this is good news, the industry still faces challenges and opportunities.

Macroeconomic Backdrop
Any discussion of current and future conditions facing community banking and thrift institutions must start with an understanding of the current macroeconomic backdrop. The U.S. economy is experiencing the longest expansion on record. Unemployment is the lowest it has been for more than a generation, and payroll employment has increased by 14 million in the past five years. This increase in labor utilization has been more than matched by a sharp rise in capital investment, with that investment concentrated in the high-technology areas of communications equipment and computer hardware and software. The boom in capital investment, together with greater efficiencies in the ways capital and labor are used to produce goods and services, has raised the growth rate of labor productivity to almost 3 percent per year over the past five years. And the gains in labor productivity in the most recent quarters have been even more impressive than that.

As growth in the aggregate supply potential of the economy picked up, we also experienced an even more rapid acceleration in aggregate demand. The monetary policy actions of the past year and a half have been geared to bring the growth in aggregate demand back into line with the growth in aggregate supply as an essential step in preventing the economy from overheating. Although it is still too early to declare victory, I think our actions have been reasonably successful. At this time, economic growth appears to have indeed moderated to a more sustainable pace, raising the odds that we have avoided the development of economic imbalances that have led to the end of past economic expansions. To be sure, headline inflation has risen over the past year, mainly because of the rise in energy costs. Inflation rates excluding food and energy, core inflation rates, have risen as well, reflecting in part the pass-through of the higher fuel costs into the prices of other goods and services, such as airfares. However, to date, an inflation psychology has not emerged in response to the higher energy prices and unusually taut labor markets, though we need to watch developments in these areas very closely.

One important sector of the economy that is of particular interest to you is the housing market. The late 1990s witnessed an impressive boom in sales of new homes, sales of existing homes, and housing starts. Following a remarkable pace of building activity last year, single-family housing starts fell to a 2- year low in July, as a result of somewhat higher interest rates and slower growth of income and employment. Nonetheless, activity is still relatively strong by historical standards, and the most recent indicators of housing demand point to a bit of a pickup over the summer, probably in response to the softening of home mortgage rates. Single-family housing starts have edged up since July, and incoming data point to a continued high level of home sales. Thus, though the housing markets have cooled a bit, they still remain quite healthy.

Business Opportunities
Depository institutions that specialize in housing are under constant pressure to perform in what is now a mature market. Perhaps no other asset on the books of savings banks and saving associations is more analyzed and better understood than the 30-year fixed-rate conforming mortgage. For years, some have argued that these savings institutions needed to take advantage of the most obvious opportunity available to them, the opportunity to diversify away from home mortgages. However, diversification, particularly into the type of lending that occurs in already well-established markets, has proven difficult. Commercial and industrial loans still account for only 3 percent of savings institutions' assets. Even expansion into consumer lending has been limited. Consumer loans to individuals have risen from less than 4 percent of assets in the early 1990s to a little more than 5 percent today, but this level is only a bit above that reached by savings institutions in the 1980s.

Savings institutions, for now, remain mortgage finance specialists. However, this focus does not appear to have limited their business opportunities. Today, the mortgage market is, in fact, a variety of markets, and a plethora of tools can help manage the risks associated with mortgages, including credit and interest-rate risks. Depository institutions now can specialize in the risks they take; banks, thrifts, and mortgage bankers can originate mortgages, and then choose to hold or not hold the risks associated with mortgages. This specialization allows those with stable funding bases or special investment needs, to provide mortgage credit without participating in the origination of the credit, and it allows originators to create the mortgage without funding it.

The shifting of risks, as illustrated by the mortgage markets, is possible because financial intermediation facilitates diversification of risk and its redistribution among market participants with different attitudes toward risk. Indeed, all of the added value from new financial instruments derives from reallocating risk in a manner that makes it more tolerable. Insurance such as private mortgage insurance is the purest form of this service. But other elements of the mortgage market, such as collateralized mortgage obligations with their different tranches, contribute economic value by unbundling risks in a highly calibrated manner, allowing them to be reallocated.

At the same time, whether risks are unbundled or not, changing technology has facilitated the increasing use of risk-management models. Though I am a great fan of such models, the flight to quality by investors in the fall of 1998 and the abrupt disappearance of liquidity for almost all financial instruments during those months should prompt careful consideration of the effectiveness of risk-management models. A key question is whether the modeling of risks and the associated unbundling of risks have sufficiently matured to withstand the often temporary, but severe, disruptions that arise when the pessimism of some investors becomes pandemic. Here, the experience of management plays a critical role because, though the timing of financial crises may be unknowable, the appropriate safeguards to mitigate the damages from such experiences are well known.

The conservative nature of bankers, especially community bankers, often leads them to question the assumptions underlying their models and to set aside somewhat higher contingency resources--reserves or capital--to cover the losses that will emerge from time to time when investors suffer a loss of confidence. These reserves may appear to be a suboptimal use of resources--just like fire insurance premiums--particularly if other market participants do not prepare for the disruptions that periodically roil financial markets. But if market participants appropriately account for these possibilities, then the cost of the insurance premiums will simply be another cost of business, and the risk-spreads of financial instruments will appropriately reflect these concerns.

When looking at mortgage spreads over a long period, one can have little doubt that the dramatic improvement in information technology in recent years has altered our approach to risk and created new opportunities for bankers active in the mortgage markets. For mortgage originators, new technologies help maximize returns by providing up-to-date information about the market's valuation of mortgages and about mortgage pool characteristics. Indeed, information technology is profoundly changing the mortgage markets by providing information critical to the evaluation of risk. The greater availability of real-time information on mortgage valuations and the greater volume of information on mortgage borrowers has reduced the uncertainties and thereby lowered the variances used to guide portfolio decisions. It also has lowered the cost to many institutions of learning about, and entering, almost any aspect of the mortgage market and has thus created additional potential competitors.

One new technology that has opened possibilities for the evaluation and unbundling of mortgage risks is credit scoring. Credit scoring can make underwriting less subjective, lowers the costs of originating loans, and increases the speed and consistency of underwriting decisions. Perhaps most important, all available evidence indicates that scoring increases the accuracy of risk assessment. These features clearly benefit lenders, but they can also serve the interest of borrowers by expanding credit opportunities and improving the efficiency of the credit review process. In fact, the growing use of credit scoring--by contributing to the formalization of the analysis of risk and pricing--may have contributed in important ways to the growth in lending to lower-income households over the past decade.

Business Challenges of Household Debt
Despite these clear benefits, however, the growing reliance on credit scoring raises a number of potential concerns. First, as I mentioned, assumptions in modeling are critical. Scoring models are built on the premise that past repayment performance is the best indicator of future repayment. Model developers build scoring models that rely on the relationships between loan and borrower characteristics and repayment performance observed in large databases of loans.

Additionally, most credit-scoring models in use in the mortgage market today are generic in the sense that they are built on the performance of borrowers nationwide. They, in general, are not focused on the performance of borrowers in a given state or local community, where economic conditions may be different or may respond differently to broader developments. Thus, a generic model used to assess the creditworthiness of an applicant from a small geographic area may fail to provide a reliable measure of the credit risk represented by applicants from that area. I am, to be sure, very aware that the strength of community bankers--a strength that too often larger institutions cannot replicate--is their understanding of, and closeness to, the local market and the circumstances of their customers. It would be a shame if that skill--that credit judgment--were discarded because of the evolving technologies. Credit scoring is a very useful tool, but it must be used as a complement to--and not a substitute for--credit judgment, particularly when applied to specific local communities.

Clearly one challenge is the application of credit-scoring models to smaller communities or the use of these models to predict future performance when the historical data cover only a period of economic expansion. Another challenge facing community bankers is understanding the debt dynamics of individual potential borrowers. Fueled by expectations of solid income growth and continued low unemployment rates, household debt has grown faster than disposable personal income in nearly every quarter over the past five years and has reached a new record high relative to income. Household debt has risen from an average of about 69 percent of disposable personal income in 1985 to 97 percent of income in the second quarter of 2000. The main driver of the rise in outstanding debt has been increased home mortgage borrowing, which has grown from an average of 43 percent of disposable personal income in 1985 to 67 percent. Clearly, it is the credit judgment of lenders that households have the ability to service this debt. But our experience with economic slowdowns is dated, and the rise in debt has led some analysts to question the economic and business implications of continued extensions of credit to households with already high ratios of debt to income.

Reinforcing this concern, the household debt-service burden--the ratio of required debt-service payments to disposable income--recently has risen to about 13- percent of disposable income, though it remains below levels reached in the 1980s, even with the rapid growth of debt outstanding. Importantly, despite the rise in mortgage debt, the contribution of home mortgage debt payments to the household debt-service burden has been little changed over the period from 1985 to 2000, fluctuating between 5 percent and 6- percent of disposable personal income. The reason for the stability in mortgage payments is that generally declining mortgage interest rates have about offset the effects of the rise in mortgage debt. The primary reason for the recent rise in debt-service burden is to be found in other forms of debt, namely consumer debt. Interestingly, the recent rise in debt-service burden appears to be faster for households in the bottom 25 percent of the income distribution than for other households. These households have tended to borrow disproportionately in the form of consumer loans, which have shorter maturities than mortgages and higher required payments.

More generally, the recent overall rise in household debt-service burden may not be alarming. Measures of credit quality in the second quarter were mixed and did not point uniformly to deterioration. Delinquency rates on closed-end consumer loans increased slightly after first-quarter declines, while several measures of credit card delinquencies declined. Home mortgage delinquencies edged up slightly but remained near very low levels. However, it is true that the household debt-service burden has proven helpful in predicting problems with household credit quality, such as consumer loan delinquencies, so its rise may portend some deterioration in measures of credit quality going forward.

Nevertheless, households have accumulated significant wealth over the past few years, with capital gains in both the housing and stock market. Thus, although some households--particularly households without significant holdings of wealth--may experience difficulties in servicing their debt if unexpected misfortune hits, many households may be well situated to carry increased debt burdens. Lenders appear to be responding to this rise in wealth. For example, despite the rise in debt that I referenced above, the August and May Senior Loan Officer Surveys suggested that mortgage lenders--like lenders generally--do not appear to have tightened credit to households. Moreover, debt, primarily mortgage debt, has in recent years grown most rapidly among upper-income households, the same group that experienced the largest increase in the value of its assets. This finding suggests that lenders are making distinctions among households. Good financial management by households and lenders implies that households continuing to accumulate debt have been well evaluated by lenders, who have determined that these households have the wherewithal to service their debt. A rising debt may suggest that both consumers and lenders expect good economic times to continue. If that's a reasoned judgment based on realistic expectations, then any problem should be readily absorbed. But if, instead, the decision to continue to lend is simply a mechanical extrapolation of the recent past, then there are reasons to emphasize the need for judgment.

Policy Challenges
At the aggregate level--the level of concern when conducting monetary policy--the link between household debt-service burden and consumer spending has been more difficult to document. However, logically enough, some models suggest that high debt burdens increase households' vulnerability to unexpected adverse external economic developments. According to these models, households with high debt burdens view their future income prospects less favorably when hit by such a shock. They tend to cut back on consumption more than they otherwise would have (and more than a standard consumption model might predict) because the debt payments they took on in good times now look too high. This potential for household distress after a negative economic event also may prompt lenders to cut back on their loans to households in such circumstances, thus deepening the fall in consumption. In this view, debt-service burdens have the potential to worsen the effect of the original shock and should be seen as a source of "imbalance" to the extent that households' (and lenders') forecasts of future income growth turn out to have been too optimistic. If this is the case, "balance" is restored only by a slowdown in spending and debt growth that lowers debt payments to a level more in line with the revised expectations of income growth. Nonetheless, the relationships between household debt-service burden, spending, and expectations of future income are complex, and more empirical work is needed before relying on these models very heavily.

I should also note that increased leverage is not exclusively concentrated in the household sector. Aggregate leverage ratios of domestic nonfinancial corporations--as measured by debt-to-assets or debt-to-equity--are rising as well. Larger, publicly traded corporations with investment-grade ratings and access to the capital markets should have the financial resources and alternatives to handle a rising debt burden even during times of slowing economic activity. But, for some speculative-grade and small unrated businesses whose financial condition likely has deteriorated as their debt-service burdens have risen rapidly over the past year, the consequences may be more serious. Recent data suggest that the pace of bank borrowing by nonfinancial businesses has slowed considerably on average over the past few months, although this moderation follows several quarters of unusually strong growth.

Though banks likely have become generally less accommodative in lending to nonfinancial businesses during the past several months, the bond market has become quite bifurcated. Risk spreads on high yield bonds have climbed back to the elevated levels of a few years ago. However, though investors have shown some reluctance recently to participate in the market for the fixed-income securities of lower-rated firms, larger issuances by investment-grade firms are still well received, and yields on investment-grade bonds are little changed. Therefore, the market for fixed-income securities appears to be functioning in a way that one might describe as more discriminating.

I remain optimistic about the prospects for sustainable growth in the U.S. and world economies. While there are a few signs of a deterioration of credit quality concentrated in a small number of consumer and business bank credit portfolios, banks' overall asset quality and earnings remain strong, putting the industry in a better position to withstand future problems than it has been in many years.

However, household and business debts have risen sharply, debt burdens are rising, and, because of the widespread use of credit scoring, more of the analysis underlying credit extensions may have been based on recent repayment experience in a growing economy. As a result, particularly as our economy settles down to a more sustainable, but somewhat slower growth pattern, this may well be the time for caution and judgment. The old adage that bad loans are made in good times reflects the reality that slowing income flows expose the inevitable mistakes. And such mistakes may induce both lenders and borrowers to overreact.

Conclusion
Members of America's Community Bankers bring to the table significant assets--their knowledge about local markets and skills and experience in operating in them. I have every reason to believe that you have deployed these assets wisely during the recent past and that you will continue to do so in the future. I thus expect that you have coupled your innovative finance and applications of new technology with your expertise and judgment. If you continue to apply judgment based on experience, both you and the economy will benefit. I am confident that you will do just that.

Return to topReturn to top

2000 Speeches