Mexico: land of promise?
As recently as mid-1994, prospects for development of the Mexican mortgage market appeared bright (see Stan Ross and Ronald Johansson, Mortgage Banking, July 1994, and Christopher Barry, Gonzalo Castaneda and Joseph Lipscomb, Journal of Housing Research, Vol. 5, No. 2, 1994.) Commercial banks had been increasing their mortgage lending; the Mexican government had taken a number of actions to reduce the costs of property transfer, appraisal and registration; and both inflation and interest rates had declined significantly. An important government goal was to put in place the legal and regulatory framework for a secondary mortgage market by the end of the year.
Much of this progress was undone by the financial and economic turmoil that has engulfed the country since the end of 1994. The devaluation of the Mexican peso in December 1994 undermined confidence in the economy and induced foreign capital flight from the country. The government lost control of the currency and was forced to let it float several days after the initial devaluation. During the period from December 1994 to March 1995, the peso lost more than 50 percent of its value against the dollar.
Although the international community put together an unprecedented aid package for Mexico, and the government has adopted a stringent austerity program, investor confidence in peso-denominated debt has remained weak. As a result, short-term interest rates have reached 70 percent, inflation is forecast to be 42 percent for 1995 and the economy is projected to have at least a 2 percent fall in real GDP.
These events have had a wrenching effect on the Mexican banking system. A mismatch between peso-denominated assets and dollar-denominated liabilities has seriously reduced bank income. Loan arrears have been rising sharply as firms and households are affected by the rapidly rising interest rates. The government has taken over several large banks, and several other banks, have received liquidity infusions from the Central Bank.
The commercial banks had reduced their volume of mortgage lending prior to the devaluation and are unlikely to extend new loans until the financial situation stabilizes. The government austerity measures suggest that no new on-budget housing assistance is likely.
Does this spell the end of the dream to create a modern housing finance system in Mexico - one that can deliver credit to moderate- and middle- as well as upper-income households? Not necessarily. The Mexican government has made significant progress in stabilizing the economy in 1995. Interest rates have fallen; the peso has strengthened; and investor confidence has begun to return. There is still active discussion about creating a secondary mortgage market. However, many fundamental reforms must be undertaken if the dream of expansion of private sector housing finance is to become a reality.
From 1980 to 1990 the population of Mexico increased almost 22 percent. This growth rate, almost 2 percent (compounded) per year, has put a strain on the country's ability to adequately house its population. According to SEDESOL, the housing ministry, the total housing stock for the country in 1992 was 17.152 million units for 83.9 million citizens. It estimates that the country would require an additional 3 million units to adequately house this population. The housing deficit is only getting worse. SEDESOL estimates that housing needs are growing by more than 500,000 units per year, whereas new production is adding fewer than 400,000 units per year to the existing stock.
Housing finance needs span the entire income spectrum but are particularly [TABULAR DATA FOR FIGURE 1 OMITTED] acute in the low- to moderate-income range. Figure 1 provides a breakdown of the income distribution into minimum wage (MW) multiples along with estimates of the size and the primary housing finance programs serving the needs of the households in each segment. (In Mexico, the income distribution is usually expressed in terms of multiples of the minimum wage - the 1994 minimum wage was approximately N$5,500 [new peso] per year or US$1,620 per year using November 1994 exchange rates or less than $900 per year using second-quarter 1995 exchange rates.)
The very low and low-moderate segments of the population, making up more than 90 percent of the total population, obtain housing finance almost entirely through government programs. These programs include loans from pension funds (INFONAVIT for private sector workers and FOVISSSTE for public sector workers), a central bank lending program (FOVI) and an assortment of programs from federal and state agencies.
Government programs and strategy
The Mexican government has stated that affordable housing is a top priority. To this end it has significantly increased its investment in affordable housing. ln 1994, the estimated investment in housing increased 71 percent in real terms to N$16.5 billion for 486,597 units from N$9.7 billion and 300,000 units in 1992. The government also has taken significant steps to streamline regulatory and bureaucratic impediments to the production and financing of housing.
The Mexican government's housing strategy is threefold. First, it hopes to target its funding to affordable housing programs to meet critical housing needs. Second, it is attempting to integrate its housing loan programs with those of the private sector. Third, it is taking steps to improve competition in and efficiency of the housing market.
The largest source of housing finance in Mexico are the pension fund loan programs. The pension fund for private sector workers is INFONAVIT (Instituto del Fondo Nacional de la Vivienda Para Los Trabajadores). In 1993 INFONAVIT was responsible for approximately 62,000 new home loans (N$3.9 billion) and provided approximately 27,000 loans (N$1.4 billion) for home improvement and other housing-related items.
Employers pay a 5 percent contribution to the INFONAVIT fund on behalf of their workers. Once a worker meets a set of requirements based on age, total contribution to the fund, number of years worked and number of dependents, he or she is allowed to apply for a loan. INFONAVIT provides the loan directly to the plan participant.
FOVISSSTE (the pension fund for public sector workers) works similarly to INFONAVIT. In contrast to INFONAVIT, however, its lending is focused on home improvement. Although the pension funds target their loan programs to households that earn between one and 10 times the minimum wage the majority of lending goes to those earning between two and three times the minimum wage.
Another major government program is FOVI (Fondo de Operacion Y Financiamiento Bancario A La Vivienda). FOVI funds, which come from the Central Bank and the World Bank, are made available to housing developers through commercial banks via public auctions. Developers who obtain FOVI funds must build and sell homes in the price range of eight to 13 times the minimum wage. Commercial banks provide the end loans to the housing purchasers with FOVI either rediscounting or guaranteeing the loan. Although FOVI loans can be made to households earning up to 15 times the minimum wage, given the constraints on house prices, its efforts are focused on households earning between three and six times the minimum wage. It is estimated that FOVI funded the construction of approximately 43,000 homes with an investment of N$3.7 billion in 1994 - a significant increase over prior years.
Besides the pension funds and FOVI, many smaller agencies are projected to provide more than N$2.7 billion in 1994 for financing housing rehabilitation, progressive housing (building one room at a time) and land. FONHAPO (Fideicomiso Fondo Nacional De Habitaciones Populares) is the government's primary low-income housing agency. It targets non-salaried workers earning less than 2.5 times the minimum wage. Many other state agencies exist to provide funds for this lowest income segment. Typically, these programs do not provide for direct home purchase but rather for home improvement and progressive housing construction.
The second part of the Mexican government's housing strategy is the integration of the pension funds and FOVI with the activities of the private sector. A recent example of integration was a 1992 constitutional amendment to change the focus of INFONAVIT and FOVISSSTE from housing construction agencies to housing banks. Whereas these funds once were responsible for building housing, they now rely on private developers to build all housing.
More recently, FOVI began in 1993 to provide guarantees of 50 to 60 percent on loans originated by commercial banks. By providing guarantees instead of the actual loan funds, the government can leverage its scarce resources. There are plans to make FOVI an independent housing bank or merge it with an existing development bank by 1996. INFONAVIT is currently exploring plans to offer cooperative financing (i.e., second mortgages) with commercial banks.
The third part of the government's housing strategy involves changing the laws and regulations that affect many aspects of mortgage lending to create a more market-based system. The government brokered a major agreement in 1992 between federal and state agencies and various private sector groups to develop consistent building standards, reduce time limits for approving permits and licenses and reduce development fees and bureaucratic costs. In 1994, it created a new category of financial institution, the nonbank bank. These institutions can be set up with an initial capitalization of $3 million and can operate as mortgage banks.
To expand the flow of funds into housing, the government is attempting to facilitate the development of a secondary mortgage market. One way it is attempting to do so is through reform of laws that hinder the sale of mortgages. Currently a mortgage cannot be sold without the permission of the borrower. Furthermore, a sale requires re-recording, a process that can take three to six months to complete and cost 0.5 to 1 percent of the mortgage amount. Another area of attention is the foreclosure process, which favors the borrower and typically takes two or more years to complete.
Private sector mortgage finance
The commercial banks provide significantly more funds for housing than the government, with an estimated investment of N$21 billion for 138,000 units in 1994. The volume of bank lending in Mexico increased 600 percent from 1989 to 1992 but has remained constant in real terms since 1992. Commercial banks provided approximately 67 percent of the NS volume of mortgage loans in Mexico but only a 28 percent share of the total number of mortgages. Mortgage lending by the banks has been highly concentrated, with the four largest banks accounting for 80 percent of the volume done by commercial banks. In contrast, the government-sponsored programs provided 33 percent of the N$ volume of lending with a 72 percent share of the total number of mortgages originated.
The commercial banks cater mainly to the top end of the market as reflected in an average loan size in excess of N$152,000. The average commercial bank customer in Mexico earns N$100,000 per year (18 times the minimum wage) and owns a home in the N$240,000 range. The banks are beginning to make more loans to middle- and moderate-income households (the latter when FOVI guarantees are available). Nonetheless, their business remains concentrated in the top 5 percent of the household income distribution.
The commercial banks have experienced significant difficulties with loans originated in the 1990-1993 period. Many banks have high delinquency rates in their mortgage portfolios. This was the case both before and particularly after the devaluation.
There are several reasons for these problems. First, the increase in bank lending far outstripped their servicing and collection capabilities. Second, the expansion in lending came at a time when the economy was being opened to increased competition resulting in corporate downsizing and significant layoffs among middle-class households. Third, the structure of the mortgage instrument may have contributed to a poor portfolio performance in a volatile economy.
The principal mortgage instrument used by the commercial banks is the double-indexed mortgage (DIM). The original DIM design, which is used by FOVI, works by adding to principal any difference in the payment accruing based on a variable market interest rate (e.g., the average cost of funds of the banking system) and a lower payment due, initially set by the government and adjusted thereafter according to changes in wages (the minimum wage index). Thus, by design these loans are subject to negative amortization. FOVI loans have an initial term of 20 years with the possibility of extension to 30 years. The government will forgive any balance outstanding at the end of a maximum 30-year term. This loan performed well during the late 1980s when real interest rates were falling and real wages, as measured by the minimum wage index, were constant or rising.
The advantage of the DIM is that it eliminates the "tilt" problem inherent in loans with fixed nominal payments, whereby early payments have much higher real values to compensate for the erosion in the real value of later payments. ln its pure form it is a loan that can be both profitable to the lender and affordable to the borrower. However, any indexed mortgage changes the risk allocation relative to a fixed nominal-rate mortgage. The balance due will decline much more slowly, and the term can actually lengthen. This increases the chance that at some point the collateral value will decline below the loan balance. In addition, the bank will not have as high a cash flow coming from the loan to re-lend or to meet liquidity needs. In other words, the bank is taking on more credit risk, cash-flow risk and liquidity risk.
The commercial banks in Mexico use loan designs that are variants of the original DIM design. In particular, the loans made by the banks after privatization had much wider spreads than the FOVI loans (due both to the markup of the loan rate over the benchmark index and the mark down from the index to the payment rate), and also a maximum term (20 years). As a result of wide spreads between the payment and loan rates, the loans began accruing large amounts of negative amortization immediately. The negative amortization buildup was compounded by an economic environment in the early 1990s characterized by high real interest rates and declining real payments (as the minimum wage lagged inflation) as shown in Figure 2.
This led the banks to "restructure" many of their mortgages. The restructuring involved extension of the term, decreases in the spread between the accrual rate and index value, and a shift away from the minimum wage to the inflation rate as the index for payment adjustment.
The switch from the minimum wage to the inflation rate as the index for payment adjustment could not have come at a worse time. One of the major effects of the devaluation was a sharp increase in inflation (the March government forecast for 1995 was for an inflation rate of 42 percent with an increase in the minimum wage of only 10 percent). This has resulted in severe payment shock for most commercial bank mortgage borrowers. The default rates have also risen due to the increase in unemployment. The banks may be forced to further extend terms or reduce payment rates to avoid immediate mortgage losses.
What the future may hold
The first prerequisite for establishing a private market housing finance system is macroeconomic stability. Thus, it is essential for interest rates and inflation rates to return to their 1994 levels. In the longer term, it will be important for the government to help reduce the credit and liquidity risks faced by the banks in extending mortgage credit.
The credit risk in making mortgage loans in Mexico is high for several reasons. First, the foreclosure process is time-consuming and costly, reducing the access to collateral that can benefit the borrower and lender. Second, little historical information exists on borrower income or past credit history. Third, the volatility of the Mexican economy has subjected many borrowers to significant payment shock or negative equity.
The Mexican government has recognized these problems. The past administration proposed a reform of the foreclosure law. However, the proposal did not pass last year and is unlikely to go anywhere soon given the precarious state of many borrowers.
The formation of a credit agency to improve borrower information has been proposed. However, the large banks must agree to share borrower information to make this project a reality. The DIM instrument design as a way to manage volatility is sound in theory but difficult to implement. The banks changed the basic design for two reasons: lack of confidence in the minimum wage index and lack of a government takeout of remaining negative amortization at the end of the loan term.
The government can foster use of a sound mortgage design by publishing a wage index for different categories of workers. If this index is viewed as an accurate representation of income as opposed to a policy tool, it would be more credible for payment adjustment on DIM loans. A government guarantee to purchase any remaining balance at the end of the mortgage term (e.g., 30 years) could reduce the concern about negative amortization buildup.
The development of a secondary market is essential to allow lenders to manage the liquidity risk of mortgage loans. The reliance of banks on very short-term funding, combined with the prospects of significant negative amortization buildup on mortgage loans, has made them reluctant to expand their mortgage lending. In addition, the banks now suffer from significant capital adequacy problems. Two major prerequisites for a successful secondary market are credible guarantees to investors against default loss and potential buyers of long-term instruments.
The government has provided partial (50 to 60 percent) guarantees on loans meeting the guidelines of FOVI. However, because of the perceived credit risk, lenders were reluctant to extend credit to these households prior to devaluation even with the partial guarantee. The Mexican government has been reluctant to increase the insurance coverage because of the high monitoring costs associated with provision of 100 percent guarantees. Until the foreclosure process is streamlined and made more certain and the lending and servicing practices of the primary lenders improved, credit risk is likely to remain a major impediment to expanded provision of mortgage finance in Mexico.
Even with credit guarantees, the question remains as to who would purchase mortgage bonds or securities. There is little discretionary long-term savings in Mexico. The recent creation of the SAR (Sistema de Ahorro Para el Retiro), which is a mandatory 2 percent payroll deduction that will be eventually credited to individual worker accounts, is a step in the right direction. It is planned that these funds will be invested in various mutual funds, which in turn will invest in bonds, equities or money market instruments. Estimates are that the SAR is generating between $1 billion and $2 billion per year. Long-term bonds or mortgage-backed securities would be attractive investments for the SAR mutual funds.
In the long run, the Mexican economy will stabilize and a market-oriented housing finance system will take root. Given the housing needs of the country, it is essential for such a system to develop. In recent years, the government and the banks have been making great strides to modernize the system and reduce the risks of mortgage credit delivery. It is hoped this progress will continue once the current financial crisis has abated.
Dr. Michael Lea is president and Steven Bernstein is vice president of Cardiff Consulting Services in Cardiff, California. Lea also is director of research for the International Union of Housing Finance Institutions and editor of Housing Finance International.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Market Research & Housing Policy; Mexican mortgage market|
|Author:||Lea, Michael J.; Bernstein, Steven A.|
|Date:||Aug 1, 1995|
|Previous Article:||Job worries hammer new home market.|
|Next Article:||Exploring new territory.|