According to the municipal analysts at Prescott Ball and Turben, another flaw is beginning to surface in tax-exempt housing bonds. This new fault is appearing in the old state FHA-insured, single-family mortgage programs.

The problem is not one of default, but rather one of a credit risk. This predicament arose from the fact that when these loan programs were originally issued in the 1970s, no one foresaw that interest rates would rise to the 20 percent level.

State agencies sold bond issues and the proceeds were loaned to individuals in the form of mortgage loans. The bond issues were structured with serial bonds that matured yearly from one to 15 years, with the remainder of the issue in longer-term bonds. Eighty-five percent of the bond proceeds were placed in mortgages while the balance was placed in various reserve funds which were invested in long Treasury securities. The revenues, or cash flow, which came from the mortgages in the form of interest, regular principal payments and principal prepayments were then scheduled to pay the debt service on the bond issue. In other words, the agencies structured the repayment of the bond principal to match as closely as possible the mortgage interest and principal repayments.

However, the high interest rates severely undermined this program in two ways. First, in structuring the bond maturity schedules, the programs generally assumed a 75 percent FHA termination or prepayment schedule on the individual mortgages in the program. Because of this, more bonds were able to be placed in the shorter serial maturities, which reduced the overall net interest cost on the entire bonds issue. When interest rates went sky-high, individuals decided not to prepay their mortgages of 7 to 10 percent and replace them with new mortgages at 14 to 16 percent. This interrupted and severed the cash flow to the debt service on the bond issue.

Secondly, in trying to subsidize the mortgage borrowers most of the state agencies probably did not fix the mortgage rates high enough to give them a cash cushion to completely offset a slowdown of mortgage prepayments. The overall net interest on the bond issue determined the mortgage rate that was charged. Under the law, the agency could add as much as 150 basis points to the interest cost on the bonds. For example, if the net interest on the bond issue was 7.5 percent, the agency could have charged up to 9 percent on the mortgage.

If we take the net interest cost of the bond issue (7.5 percent), add in 75 basis points for the agencies' operational costs, deduct this figure (8.25 percent) from an assumed mortgage rate (8.75 percent), we are left with a spread of 50 basis points. This spread may then be invested to form a cushion to cover cash shortfalls. In many cases the cushion was 25 basis points or less, while in other issues it ran as high as 75 basis points. As a result, some agencies like Wyoming were able to build up a cash cushion of $22 million by reinvesting the excess or spread after having set their mortgage rates a full 150 basis points above the interest cost on their bond issues.

Due to these two defects, it looks as if on a near-term basis the cash flow will not be available to retire some of the shorter maturities of various bond issues. On a worst-case scenario, this means missed debt service payments in lieu of some other source of cash external to the mortgage loan portfolio. However, on a long-term basis of 15 to 18 years, the funds will be available to retire the bond issue. This is little consolation if you own a particular bond due in 1986 that will not be paid off until the year 2000. So far the state agencies of Utah, New Mexico and Nevada are facing these problems. The best these state agencies can hope for is that interest rates will continue to decline and stay there for a long time.