Hostname: page-component-78c5997874-m6dg7 Total loading time: 0 Render date: 2024-11-15T12:44:02.119Z Has data issue: false hasContentIssue false
  • English
  • Français

Determinants of Municipal Bond Yields

Published online by Cambridge University Press:  19 October 2009

K. Larry Hastie
Get access Rights & Permissions [Opens in a new window]

Extract

Since World War II, the significance of the municipal bond market has increased dramatically with state and local government debt growing much more rapidly than public and private debt, federal debt, or the gross national product. Between 1960 and 1970, the annual value of new issues of state and local government bonds increased 110 percent, and there is every indication that the total will continue its rapid rise. In 1969 and 1970, the values of state and local government bond new issues were second only to those of the corporate bond market. Despite the size of the state and local bond market, investors and researchers have devoted their attention to the markets for corporate and U.S. government securities; the interest in these markets has tended to overshadow activity in the municipal bond market.

Type
Research Article
Information
Journal of Financial and Quantitative Analysis , Volume 7 , Issue 3 , June 1972 , pp. 1729 - 1748
Copyright
Copyright © School of Business Administration, University of Washington 1972

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 Total new issues increased from $7.2 billion in 1960 to $15.2 billion in 1970. Moody's Municipal and Government Manual (February 1971), p. a20.

2 In 1970, the $15.2 billion of new issues of state and local bonds compared with $14.8 billion of new issues of U.S. government securities, $30.3 billion of corporate bonds, and $7.3 billion of corporate equities.

3 Overall debt rather than net debt is used because it considers the debts of overlapping jurisdictions, and therefore, it is a more effective measure of the resident's total debt burden. The true value of taxable property is preferred to assessed value because it is a more accurate indicator of the community's ability to pay. Since assessed valuation is determined by a wide range of percentages of full value, it would be difficult to compare the differences in ability to pay by using this factor. In addition, overall debt/true value is desirable because it recognizes the importance of property taxes as the principal source of local government revenue.

4 The “minimum requirements approach”a develops an index which approximates the minimum percentage of the labor force that is required by various sectors of the city's economy in order to maintain economic viability. When the minimum percentage is subtracted from the actual employment in a particular sector, we have the excess employment of the sector. The measure of economic diversity is a weighted average of the excess employment in each sector. A city with a low value of the measure will be described as more diversified than a city with a large value. See Ullman, Edward and Dacey, Michael F., “The Minimum Requirements Approach to the Urban Economic Base,” Papers and Proceedings of the Regional Science Association, Volume 6 (1960), p. 176.Google Scholar

5 A perfect or highly efficient market is, among other things, a market in which the buyer or seller has no effect on the market price; no matter how few or how many bonds the investor wants to buy or sell, the market will absorb them without affecting the price of the security.

6 Lawrence Fisher used total debt outstanding as the measure of marketability of corporate bonds. Fisher, Lawrence, “Determinants of Risk Premiums on Corporate Bonds,” Journal of Political Economy, Volume 57 (June 1959), pp. 217–237CrossRefGoogle Scholar. Lerner and Carleton used debt outstanding as a measure of marketability in their study of municipal bond ratings. Carleton, W. T. and Lerner, E. M., “Statistical Credit Scoring of Municipal Bonds,” Journal of Money, Credity and Banking, Volume 1 (November 1969) pp. 750–764.CrossRefGoogle Scholar

7 Fisher, “Determinants of Risk Premiums,” p. 229.

8 College students/population frequently takes zero values.

9 The observations in each cross section contain all noncallable, general obligation municipal bonds listed in the Blue List that have maturities between 15 and 25 years and that were new quotations during the cross sections. Old quotations were not considered. Since most municipal general obligations are serial bonds, more than one quotation may appear for a single municipality. So that the larger issuers would not swamp our sample, no more than five quotations of a single issuer appear in any cross section. If more than five observations appeared in the initial sample, five observations were randomly selected for study.

10 Seven days were used for 1957 to obtain a comparable sample size.

11 In the first six months of 1963 and 1967, commercial banks, holdings of state and local government bonds increased $3.2 billion and $5.9 billion respectively. In the first six months of 1957 and 1960, commercial banks' holdings increased $0.5 billion and decreased $0.2 billion respectively.

Source: Assets and Liabilities and Capital Accounts – Commercial and Mutual Savings Banks, Federal Deposit Insurance Corporation. Washington, D.C.

12 Overall debt/personal income (Xoi) was tested in place of overall debt/true value (Xo). In 1965, the coefficient for Xoi had the wrong sign while the coefficients for Xo had the correct sign in all cross sections. Moreover, when Xoi was substituted for Xo, the range of estimated coefficients for the other seven independent variables was larger in every case.

13 See equation (7) in Table 2.

14 In addition, two other specifications of block size were tested. The results of substituting block size (Xbs) for log of block size (Xs) yielded the correct sign, but the R2s and t-values were significantly larger only in one out of the five cross sections. In addition, when Xbs is used in place of Xs, the ranges of the coefficients of the other seven independent variables were larger. The results of substituting block size (Xbs) and block size squared (Xbs)2) in place of Xs yielded the hypothesized signs. However, the adjusted coefficient of multiple determination (adjusted R2) was lower in four out of five cross sections. Moreover, Xs is significant at the 025 level in four out of five cross sections; Xbs and (Xbs)2 are significant at the 025 level in three out of five cross sections, respectively.

15 Lerner and Carleton, “Statistical Credit Scoring,” p. 154.

16 Robinson, Roland, Postwar Market for State and Local Government Securities. (Princeton, N.J.: Princeton Press for the National Bureau of Economic Research, 1960), p. 185.CrossRefGoogle Scholar

17 As a result of the November 1964 change in Regulation Q, yields on municipal bonds declined about 20 basis points between october and February 1965. During the month of the cross section (March 1965), the yields regained most of what they had lost over the preceding five months. U.S. Board of Governors, Federal Reserve Bulletin, Volume 51 (August 1965), p. 1058.Google Scholar

The period of adjustment in 1960 was not as unusual. Moody's Aaa municipal bond index cropped 19 basis points in the first four months of 1960, and yields on state and local governments declined more than those outstanding corporate issues of similar quality. Federal Reserve Bulletin, Volume 46 (December 1960), p. 1325.

18 In order to determine the stability of the coefficients within the periods, the following statistic was calculated

where: ,

bit = the estimated partial regression coefficient for the ith variable in the tth cross section, and

sit = the standard error of the estimate – bit.

This test for stability was suggested by Lawrence Fisher. According to Fisher, the statistic, y, has approximately the X2 distribution with T – l degrees of freedom. Therefore, a high value of the statistic is reason for rejecting the hypothesis that all of the partial regression coefficients are estimates from the sample population. See Fisher, Lawrence, “Determinants of Risk Premiums on Corporate Bonds,” Journal of Political Economy, Volume 57 (June 1959), p. 229.Google Scholar

19 The following F–test was suggested by Gregory Chow to test the equality between sets of coefficients:

where:

SSRs = the sum of squared residuals based on the hypothesis that the coefficients are the same for all cross sections,

SSRd = the sum of squared residuals based on the hypothesis that all of the coefficients are different for all cross sections,

V2 = degrees of freedom on the first hypothesis (coefficients are the same for all cross sections), and

V1 = degrees of freedom on the first hypothesis minus the degrees of of freedom on the second hypothesis.

See Chow, Gregory C., “Tests of Equality between Sets of Coefficients in Two Linear Regressions,” Econometvica, Volume 28 (July 1960), pp. 591–605CrossRefGoogle Scholar; and David, Shaul Ben and Tomak, William, “Allowing for Slope and Intercept Changes in Regression Analysis” (unpublished paper, Cornell University, November 1965).Google Scholar