FIGURE 38 AVERAGE LIFE OF GENERAL OBLIGATION AND REVENUE ISSUES YEARS YEARS 20 15 Source: Figure 3B-1 FIGURE 3B-1.-AVERAGE MATURITY REVENUE AND GENERAL OBLIGATION BONDS $1,000,000 AND LARGER, JANUARY JULY 1963 I indicated above that the longer the maturity and the lower the credit rating of a bond, the greater is the risk to the investor and the higher are the yields on which investors insist if they are to purchase the securities. The effect of maturity and quality rating differences on market yields of outstanding securities is illustrated in Figures 4A and 4B. These same differences exist in reoffering yields to investors of new issues for both revenues and G.O.'s FIGURE 4A FACTORS AFFECTING INTEREST YIELDS EFFECT OF MATURITY WEEKLY INDEXES AND AVERAGE OF INDEXES, JAN.-JULY, 1963 The Daily Bond Buyer prepares a weekly index of yields of twenty investment quality bonds which have a twenty year maturity. Moody's Investors Service publishes a weekly index of yields of Aa bonds with a ten year maturity. A comparison of these two indexes will indicate the effect of the ten year difference of maturities on bond yields. Since Moody's index involves Aa bonds and the Bond Buyer index involves high grade bonds with an average quality only slightly below Aa, any effect of quality rating differences on yields has been minimized. In Figure 4A the two weekly indexes for the seven month period, January-July, 1963 and the seven month average of the weekly indexes have been plotted. The yield index for the twenty year maturity is consistently higher than the ten year maturity index and by about the same margin from week to week. The two bars at the right show the seven month averages of the indexes. The average of the twenty year maturity index is 3.13% and of the ten year maturity index is 2.67%, a difference of 0.46%. FIGURE 48 FACTORS AFFECTING INTEREST YIELDS RE-OFFERING YIELDS OF 20 YEAR MATURITY OF ISSUES-$1,000,000 AND LARGER, JAN.-JULY, 1963 (AVERAGED BY NUMBER OF ISSUES) Figure 4B portrays the effect on yields of differences of quality ratings. Data for issues of $1,000,000 and larger in the seven months January-July, 1963 have been used. The average reoffering yields of the twenty year maturity bond have been calculated for Aa, A, and Baa issues for general obligations and revenues separately. Because the averages relate only to the single maturity, twenty years, it is possible to observe the full effect of quality rating differences without the influence of maurity differences. The resulting averages are set forth in Figure 4B. It is readily apparent that investors demand higher yields to compensate for the greater risk of lower quality securities. This systematic variation of yield with quality ratings is apparent for both general obligation and revenue bonds. The purpose of those last two charts is merely to show the nature or direction of the effect on yields of differences of maturity and quality rating of bonds; they do not aim at measuring the magnitude of the effects. For this reason, it did not seem necessary to prepare similar charts based on data for recent months to show the same relationships. Because we know that the average revenue issue is of longer maturity and lower quality than the average general obligation issue-features which would not be changed by the addition of commercial banks as underwriters-the yield to investors and the cost of money to issuers of the average revenue bond will be higher as a result of these features. And in order to analyze the effect on yields and thereby on costs of money to issuers of allowing commercial banks to compete as underwriters for revenue issues, it is necessary to adjust for the effect on yields of these systematic differences of maturity and quality. But, if yields, cost of money, and spreads are compared for issues of similar quality ratings, similar maturities, similar methods of issue, and similar market conditions at time of issue, then differences between general obligation and revenue bonds are sharply reduced or eliminated altogether. FIGURE 5.-COMPARATIVE YIELDS ON REVENUE AND GENERAL OBLIGATION BONDS OF 10-YEAR MATURITY FIGURE 6.-COMPARATIVE YIELDS ON REVENUE AND GENERAL OBLIGATION BONDS OF 20-YEAR MATURITY These conclusions are based on a series of studies covering data over the past dozen years, but in this presentation data are set forth only for three recent seven month periods. The Committee for Study of Revenue Bond Financing compiled data for the period January through July, 1963, in connection with a study made for the House Banking and Currency Committee in 1963. This compilation covered the seventy-four new issues of revenue bonds: and the three hundred fifty-two new issues of general obligation bonds, of $1,000,000 or more, having ratings of Aa, A and Baa from Moody's Investor Service, and for which appropriate price information was available from reports furnished by the Daily Bond Buyer. We did not include bonds where there was a partial loss of tax exemption resulting from an excessively "deep discount," where on the twenty year maturity the interest rate differed from the yield rate to an extent whereby the dollar price was below ninety-five. These data were supplemented for this presentation by comparable data for the seven months February 1966 through August 1966, a period of severely disturbed market conditions which culminated in August 1966 in the highest interest rates since the 1930's and for the seven months, September 1966 through March 1967, a period of improving but erratic market conditions. By hindsight the 1963 period now appears to have been one of relatively stable market conditions. In analyzing these data a common denominator had to be found to assure comparability. In order to eliminate the systematic differences which would result from comparison of interest yields on issues having different average maturities, the public reoffering yield of only the twenty year maturity and ten year maturity of each issue was tabulated, thus providing a basis for comparison uniform as to maturity. In order to make allowance for the lack of comparability which would result from comparing data of differing offering dates, in the face of constantly changing general market conditions, we tabulated the number of yield points difference (plus or minus) between the reoffering yield on the twenty year maturity of each issue and the current index of twenty year municipal bond yields published each Friday by the Daily Bond Buyer, and the difference between the reoffering yield on the ten year maturity of each issue and the current index of ten year municipal bond yields published by Moody's Investors Service. These differences were averaged for each quality class of bonds for the seven month period, and thus reflect the average number of "plus or minus" yield points off the Bond Buyer index and the Moody's index for the stated class of securities. The results of the study are summarized in Figures 5 and 6. Figures 5 and 6 indicate that in each of the three periods, for each rating class, and for each maturity more yield was needed to attract the investor into revenue bonds than into general obligation bonds. The average additional yield needed on revenue bonds in the relatively stable market of 1963 was distinctly less than either in the deteriorating market of the first half of 1966 or in the recovering but erratic market of late 1966 and early 1967. On ten year bonds the figures are 0.11 yield points v. 0.17 and 0.15 yield points respectively and for twenty year bonds the comparisons are 0.10 yield points v. 0.18 and 0.20 yield points respectively. The average additional yield needed on twenty year as compared to ten year bonds was very modest and appeared only for the two recent periods of unstable markets. Finally, the added yield needed for lower quality versus higher quality bonds is also clear, but it more marked in the periods of unstable markets. This set of relationships is quite similar to that shown in the Staff Study of the Board of Governors of the Federal Reserve System. That study, however, yielded results more nearly comparable with those of the 1963 sub-period. The Federal Reserve study was based on data of three six-month periods, January to June, of 1964, 1965, and 1966. The two earlier periods enjoyed relatively stable markets comparable to 1963, and the final period ending in June 1966 did not include the most extreme high interest cost months. I believe that the two studies reinforce each other in measuring the remaining differential that exists in yields necessary to attract investors into the two types of bonds under relatively stable market conditions after adjustment for differences in rating class, maturity, and market conditions at time of issue. The higher differentials in our study for the two periods of disturbed markets are not unexpected. They appear to reflect elements of investor preference for the two types of security which are unrelated to whether or not commercial banks were underwriters or dealers. There is no question but that markets for municipal securities were thinner during these disturbed periods, but this thinness prevailed both for G.O.'s and revenues. I have not been able to uncover any reliable evidence of revenue bond markets becoming thinner to a greater degree than G.O. bond markets. There is no evidence to indicate an inadequacy of the dealer organization for revenues to handle efficiently and economically the volume of transactions that arose. Two factors appear to be of significance as partial explanations of the changes in investor preferences even though we have not been able to measure their effects on market yield differentials. The first of these is the "call" privilege. Most revenue bonds have a call feature set forth in the indenture which enables the issuer to retire or refund the bonds prior to their stated maturity. This to the investor would mean that he could not count on the interest return set by the original coupon rates for the entire life of the bond; new and lower coupon rates could be substituted at the will of the issuer. Sometimes this feature is unlimited and in other instances calls can be made only after a five year or ten year period following the issue date. On the other hand, the existence of call features in G.O. issues is rare. The absence of the call feature can be of significant value to investors and especially when interest rates are very high |