Do tax rates affect municipal bond yields?

Transcription

1 January 2018 Do tax rates affect municipal bond yields? Cadmus Hicks Manager of Performance and Risk Analysis Market Strategist Nuveen Asset Management Tax reform has caused people to wonder if tax-exempt municipal bonds would become less attractive if the maximum federal income tax rate were lowered significantly. Such concerns have diminished since a tax bill was enacted with a maximum personal income tax rate of 37%. However, this report discusses four elements of tax policy that may affect the relative pricing of tax-exempt bonds. See the Appendix for the effect of tax policy on income inequality. Elements of tax policy that may affect municipal bond pricing include: 1. Historically there is no consistent relationship between the top marginal tax rate and the ratio of tax-exempt to taxable yields. 2. The number of people subject to the alternative minimum tax (AMT) is expected to shrink as a result of changes in both the AMT and the regular tax rules, which could strengthen demand for municipal bonds subject to the AMT. 3. The limit on the ability to deduct state and local taxes should increase the attractiveness of bonds that are exempt from state income taxes. 4. The reduction in the corporate tax rate from 35% to 21% will likely reduce demand from banks and casualty insurance companies. HISTORICAL TAX RATES AND TAX-EXEMPT/TAXABLE RATIOS The Federal Reserve provides a convenient, publicly available source of data for this analysis, based on historical information on the yields of constant maturity Treasury securities and the Bond Buyer 20 Index of general obligation bonds maturing in 20 years. The chart below shows the ratio between the Bond Buyer 20 Index and 10-year and 30-year Treasury yields from 1982 through 2001, a period when the maximum tax rate changed several times. (The U.S. Treasury stopped issuing 30-year bonds in 2002 and resumed issuance in 2006.) 1 NOT FDIC INSURED MAY LOSE VALUE NO BANK GUARANTEE

2 Ratios have been quite volatile during the period under consideration, spiking during the debate over the Tax Reform Act of 1986 and shortly after the Act was passed, which lowered the top tax rate. Ratios eventually settled back down to levels consistent with where they had been before tax rates were lowered. Municipal-to-Treasury ratios have been volatile Yield ratio 120% 110% 100% 90% 80% 70% 30% Bond buyer 20 index/30-year U.S. treasury yield ratio Bond buyer 20 index/10-year U.S. treasury yield ratio Top federal tax rate Data source: Federal Reserve, 1 Jan Dec The table below shows how the average ratio of tax-exempt to taxable yields has varied under different tax regimes. Ratios vary based on the maximum tax rate Bond Buyer 20 / treasury ratios Period Maximum tax rate BB20 / 10-Year 55% 50% 45% 40% 35% BB20 / 30-Year % 87.7% 87.4% % 91.0% 89.0% % 85.7% 85.3% % 90.2% 84.5% % 93.5% 88.8% Data source: Maximum Tax Rate, Walters Kluwer, CCH, WBOT2013/029IncomeTaxRates.asp. Past performance is no guarantee of future results. Indices are unmanaged and unavailable for direct investment. Top federal tax rate As a result of the Tax Reform Act of 1986, the maximum federal tax rate fell from 50% to 38.5% in the transition year of 1987, and then down to 33% for the next two years. The initial drop in the tax rate was accompanied by an increase in the ratios in 1987, which is what one would expect if the maximum tax rate affected the pricing of tax-free securities. However, when the tax rate fell to 33%, the average ratios declined to levels lower than when the top tax rate was 50%. The slight reduction in the tax rate in 1991 and 1992 was associated with an increase in the ratio versus 10-year Treasuries, but a modest decrease in the ratio to 30-year Treasuries. The difference between the 10-year and 30-year ratios resulted from a steepening of the yield curve as the fed funds rate was lowered from an average of 6.91% in January 1991 to 2.92% in December Finally, when the tax rate jumped to 39.6% in 1993, municipalto-treasury ratios also rose, instead of falling as one might expect. CORRELATION IS LOW BETWEEN TAX RATES AND RATIOS Why is there so little correlation between the highest marginal federal tax rate and the ratio between tax-exempt and taxable securities? Some have surmised that the marginal investor in tax-exempt bonds (i.e., the person who buys tax-exempt bonds but gets the least tax benefit from the tax exemption) is in a tax bracket lower than that of the top marginal rate, and thus the yield this investor requires to purchase a tax-exempt rather than a taxable bond is not much affected by such changes in the maximum tax rate. However, that model fails to account for the fact that the ratios of tax-exempt to taxable yields are usually higher in the longer maturities. For example, during the 10 years ending November 2017, the average ratio of 5-year AAA rated, general obligation municipal yields to 5-year Treasury yields was 87%, while the 10-year ratio was 96%, and the 30-year ratio was 104%. It is not likely that the tax bracket of the marginal 2

3 investor varies by maturity, so a different factor must be involved. In our view, a more likely explanation is that tax-exempt bonds do not compete with fully taxable bonds but with taxable bonds purchased for tax-deferred accounts such as pension plans, insurance policies and individual retirement accounts. This model better explains the upwardly sloping ratio curve of tax-exempt to taxable yields, since a longer time horizon provides a greater benefit from deferring taxes. In this model, the effective tax rate on taxdeferred income is determined, not by the amount of taxes one would pay on interest income in the current year, but by the present value of taxes to be paid many years in the future. The effect of a decline in the current tax rate is muted by the fact that the effective tax rate is already lower by virtue of the deferral. Hence, the effective tax rate does not fall as much as the current tax rate (see Technical Note at the end of this report). In other words, a drop in the tax rate lowers both the value of tax-exemption and the value of tax-deferral. Another consideration for tax-deferred investments is that the tax rate may be higher in later years within the investor s time horizon and, more importantly, when earnings are taxed upon withdrawal from the accounts. Therefore, a reduction in the current tax rate may not have much effect on how investors value taxexemption relative to tax-deferral. While lowering the top marginal tax rate would not have had much impact on the benefit of tax-exemption relative to tax-deferral, the relationship would have been affected if Congress had opted to limit the benefits of exclusions and deductions to what they would be for someone in a lower tax bracket. The last few budgets proposed by the Obama administration would have provided a 28% credit instead of allowing deductions and exclusions for upper income taxpayers. That provision would have effectively levied an 11.9% tax on tax-exempt bonds held by someone in the 39.6% bracket. In our model, the effect of such a 28% limitation on the relative value of municipal bonds would be much greater than the effect of lowering the maximum rate to 28% because the limitation would not affect the value of tax-deferral, but would reduce the value of tax-exemption. Regardless of one s theory about the factors that affect how tax-exempt bonds are priced relative to taxable securities, the historical record indicates it is not safe to assume that a lower top marginal tax rate will lead to higher ratios of taxexempt to taxable yields. THE ALTERNATIVE MINIMUM TAX While the new bill does not repeal the alternative minimum tax, it makes it somewhat less onerous in two respects. Relative to the prior law, the amount of income that is exempted from the AMT in 2018 will be increased from $55,400 to $70,300 for single taxpayers, and from $86,200 to $109,400 for joint returns, which will broaden the pool of investors who can purchase bonds subject to the AMT. Two additional factors that will reduce the number of taxpayers subject to the AMT are the limitation on the amount of state and local income taxes that can be deducted, and the elimination of personal exemptions, both of which would have been added to regular income when calculating AMT income under the prior law. Since tax liability is based on the greater of the regular tax or the AMT, increasing the regular tax decreases the number of taxpayers subject to the AMT. The second change is the increase in the threshold at which the benefit of the AMT exemption starts to be phased out. For single taxpayers, the threshold jumps from $123,100 to $500,000, and for joint returns it climbs from $164,100 to $1,000,000. This phase-out increases AMT income by 25% of the amount above the threshold, which produces a maximum effective tax rate of 35% (35% = 28% + (0.25 x 28%)). In 2015, 3.22 million taxpayers with adjusted gross income of between $200,000 and $500,000 paid the AMT, compared to 407,046 AMT taxpayers with income between $500,000 and $1 million, and 84,488 with income over $1 million. 3

4 Thus, the number of taxpayers subject to the 35% effective AMT rate will likely be reduced as the phase-out moves into higher, less populated, income ranges, while lowering the effective tax rate for those with AGI of $200,000 to $500,000 reduces the likelihood that their AMT liability will be greater than the regular tax liability. LIMITED DEDUCTION OF STATE AND LOCAL TAXES The legislation limits to $10,000 the amount of state and local taxes of any sort that taxpayers can deduct from their federally-taxable income. One effect of such a limit is to make bonds that are exempt from state taxation more attractive. For a taxpayer subject to California s 11.3% tax rate, and assuming a 35% federal rate, eliminating the deduction would increase the taxable equivalent yield of a 2.00% Californiaexempt bond from 3.47% to 3.72% (see Technical Notes). Eliminating the state and local deduction could also discourage steeply progressive tax rates, which tend to increase the volatility of income tax revenue since the incomes of upper income taxpayers are more volatile, due to the variable nature of capital gains, proprietor s income and bonuses. THE LOWER CORPORATE TAX RATE As of 30 Sep 2017, banks held securities and loans equal to 14.9% of the $3.803 trillion of municipal debt outstanding, while property and casualty insurance companies accounted for 9.0% and life insurers another 4.9% of the total. Thus, three material sources of demand for tax-exempt bonds could be reduced as a result of the reduction in the corporate tax rate from 35% to 21%. Even before the tax reform bill was enacted, corporate demand for tax-exempt bonds was subject to numerous constraints. Under Section 265 of the tax code, corporations are not allowed to deduct interest on indebtedness incurred or continued to purchase or carry obligations the interest on which is wholly exempt from federal income taxes. IRS rules, however, allow corporations to deduct full interest expense as long as their holdings of tax-exempt bonds do not exceed 2% of the average value of total assets held in the active conduct of the business. In addition to the 2% de minimis allowance, banks are permitted to deduct 80% of the interest expense for debt incurred to buy bankqualified bonds, which are issued for general government purposes by entities that issue less than $10 million of tax-exempt bonds in any given year. The after-tax yield of bank-qualified paper thus depends of a bank s cost of funding as well as its tax rate, and will decline as the cost of funds increases. Assuming a cost of funds of 1.00%, and a 35% tax rate, the taxable-equivalent yield of a bankqualified bond yielding 2.00% would be 2.97%, while the taxable-equivalent yield would decline to 2.48% assuming a 21% tax rate (see Technical Notes). Another factor affecting banks decision to hold tax-exempt bonds is that under current rules municipal bonds are not defined as high quality liquid assets for purposes of computing the liquidity coverage ratio. Insurance companies also find that the benefit of tax-exemption is limited by tax rules. Since insurance companies are permitted to deduct amounts credited to reserves for losses, and those reserves could be funded in part from tax-exempt interest income, the tax code has previously required that 15% of an insurer s tax-exempt income must be included in taxable income. With a 35% tax rate, a proration amount equal to 15% of tax-exempt income results in an effective tax rate of 5.25% (0.15 x 35% = 5.25%). The new law increases that proration to 25%, which combines with a 21% rate to produce the same effective tax rate of 5.25% (0.25 x 21% = 5.25%). Reducing the tax rate from 35% to 21% lowers the taxableequivalent yield of a 2.00% tax-exempt bond held by a casualty insurance company from 2.92% to 2.40% (see Technical Notes). The taxation of tax-exempt bonds held by life insurers has been simplified. Previously, life insurance companies did not know how income from municipal bonds would be taxed 4

5 until they had completed other calculations for their final tax bill. Going forward, 30% of the investment income received from municipal bonds will be taxed. The 21% tax rate means an effective tax rate of 6.3% on the income from tax-exempt bonds. Uncertainty regarding how tax-exempt interest would be treated has historically discouraged life insurance companies from investing in municipal bonds, so this simplification of the rules may increase life insurers willingness to consider municipal debt as a viable investment option. On the other side of the equation, the attractiveness of tax-exempt bonds relative to corporate bonds may increase if corporate yields decline as a result of a reduction in corporate debt issuance in response to the provision that limits the deduction for interest expenses to 30% of earnings before interest and taxes. Corporate debt outstanding could also be reduced to the extent that corporations use repatriated earnings from overseas to pay down debt. APPENDIX: Tax Policy and Income Inequality One of the reasons why the tax reform legislation only lowered the top marginal tax rate for individuals from 39.6% to 37% was due to the desire to keep the tax code from increasing the degree of income inequality in the United States. Indeed, many people look to the tax code to offset some of the macroeconomic forces that contribute to income inequality, but the record suggests that the impact of the tax code on income inequality is minor. For example, the U.S. Treasury Department concluded that changes to the tax code under Obama decreased income inequality by an amount that is about 8 percent of the increase in this measure of income inequality between 1979 and (See Reducing Income Inequality through Progressive Tax Policy, U.S. Department of the Treasury, September 26, 2016.) Even though, under current law in 2015, 4.5% of all taxpayers paid 58.8% of all personal income taxes (and reported 23.9% of all adjusted gross income), our research likewise shows that tax law has had only a small impact on after-tax income inequality relative to the increase in income inequality that has occurred over the past few decades. The Gini index is a measure of the extent to which the distribution of income is concentrated among households in the upper reaches of the income spectrum. A Gini value of 0 would indicate complete equality, with each household receiving as much income as every other household; a value of 1 would indicate that the majority of households have no income, while a minority of households takes it all. To calculate the Gini index, one can sort the population into different brackets based on income and compute the amount of income received by those in the lowest income bracket, and the cumulative amount by those in two lowest brackets, and so on until one reaches the highest income bracket, at which point the cumulative amount of income equals 100% of all income received by the entire population. The cumulative income for each bracket is then compared to the amount that those currently in each bracket would receive if income were evenly distributed (in which case the cumulative income would increase in direct proportion to the number of households observed). Income Inequality Is Usually Calculated before Taxes The following graph illustrates how the Gini index can be estimated using data from personal income tax returns as tabulated by the Internal Revenue Service for The Actual line shows how much cumulative income is attributable to all households with adjusted gross income below the indicated income level. The equal line is computed by multiplying the number of households in each income bracket by a constant amount equal to total income divided by total number of households. Because the number of observations varies from bracket to bracket, the Equal line rises at an angle that is close to, but not exactly, 45 degrees. The area between the two lines represents the divergence between the actual distribution of income and an equal distribution of income. 5

6 Cumulative income: Actual vs. equal for all $12000 in 2015, four had marginal income tax rates in excess of 8%, and six had tax rates of 6% or more, as illustrated in the following table. $ Billions $8000 $4000 Gini index by state Dividing the area between the two lines by the area under the Equal line yields a Gini ratio of According to the U.S. Census Bureau, the value of the Gini index for the United States rose from in 1967 to in The average absolute change in the index from year to year was The following chart shows how the Gini value has increased over time, and how the rise in income inequality seems to have been unaffected by changes in the highest marginal tax rate in the federal tax code. Top tax rates have not affected income inequality 80% 40% 0% $0 None Actual < 5k Tax Rate (L) < 10k Equal < 15k < 20k Gini (R) < 25k < 30k < 40k < 50k Data source: Internal Revenue Service, Most recent data available Data source: Internal Revenue Service, U.S. Census Bureau, < 75k < 100k < 200k Income Inequality Varies by State < 500k > 500k 0.5% 0.4% 0.3% An analysis of the Gini index by state reveals an interesting correlation: of the seven states with Gini values in excess of the average of State Gini index Top tax rate Washington, DC % New York % Connecticut % Louisiana % California % Massachusetts % New Jersey % Data source: U.S. Census Bureau. To the extent that there is any causal relationship between income inequality and income tax rates, it may be that the political pressure for a steeply progressive tax rate structure is especially strong in states where income inequality is more pronounced, or it may be that pretax incomes need to be adjusted to attract highlycompensated employees to work in states with high tax rates. In the latter case, greater progressivity in the tax rate schedule would be reflected in greater concentration of pretax income among those in higher tax brackets. The state of Connecticut stands out because its Gini index has fallen by from to This reduction in income-inequality may reflect a loss of high income residents following the state s decision to raise the top marginal tax rate in 2010, 2012, and The only other of the seven states that experienced a decline in the Gini index was California. After-Tax Income Inequality Is Slightly Less The tax data used for estimating pretax Gini values can also be used to estimate after-tax income inequality. When we subtract income tax before credits from total income we get a Gini index value of 0.491, which indicates that the structure of the federal tax code reduces income inequality by (before giving effect to such credits as the earned income tax credit and forms of direct assistance such as 6

7 Medicaid, and subsidies for housing and food). That reduction in the inequality of after-tax income reflects an even greater inequality in tax burden, which is estimated to be using the methodology of the Gini index. The Treasury report mentioned earlier found that changes in tax rules enacted during the Obama administration reduced the Gini index by The reduction in after-tax inequality only partly offset the increase in the Gini index of pre-tax income between 2008 and The next question is how potential changes in tax policy might affect income inequality. To see what might happen if the interest income earned by individuals on all outstanding taxexempt bonds were to become taxable by the U.S. Government, we estimated average tax rates for each income bracket by dividing income tax before credits by total income. We then applied those average tax rates to the amount of taxexempt income reported for each bracket, and computed after-tax income that included tax on currently tax-exempt bonds. The result is that taxing all tax-exempt bonds would reduce income inequality by just (from to ). Eliminating all itemized deductions would reduce income inequality by to (from to ). Tax Rates Have Little Overall Effect on Inequality In summary, it appears that the tax code is not a very effective mechanism for reducing income inequality. Changes in the maximum federal tax rate over time have not had a perceptible effect on the trend of rising income inequality. Even the current progressive schedule of marginal tax rates does not have much impact on the aftertax Gini index value, while changes that merely alter the definition of taxable income, without modifying tax rates, would have an impact that is roughly equivalent to the average variation from year to year. And the effect on income inequality that would result from taxing all outstanding taxexempt bonds would be a rounding error in the Gini index values reported by the Census Bureau. Technical Notes THE BENEFIT OF TAX-DEFERRAL To illustrate how tax-deferral mitigates the benefit of a reduction in the tax rate, consider an investment of $100,000 that produces $4,000 of interest income each year. If that income is taxed at 35% in the year earned, the taxpayer will pay $1,400 to the federal government. If that income is not taxed until 20 years later, the present value of that tax liability will be $639 (assuming a 4% discount rate compounded annually). The result is an effective tax rate of 15.97% (639 / 4,000 = ). On the other hand, if the tax rate is 28%, the tax liability of $1,120 would have a present value of $511. This implies an effective tax rate of 12.78%, which is only 3.19 percentage points lower than the effective tax rate produced by a 35% nominal tax rate even though the nominal tax rate has been reduced by 7 percentage points. If the tax rate reverts to 35% when the earnings are withdrawn from the account, the effective tax rate would again be 15.97%. TAXABLE-EQUIVALENT YIELDS The taxable equivalent yield is the yield that a taxable bond must produce in order to provide the same after-tax yield as a tax-exempt bond. The taxable equivalent yield of a bond whose interest earnings are exempt from state as well as federal taxes will reflect the fact that a taxable bond would be subject to both taxes. The following computations show how to calculate the taxable equivalent yield of a bond yielding 2.00% that is exempt from federal and state income taxes depending on whether one can or cannot deduct the state tax. Taxable-equivalent yield if state taxes are deductible: 2.00% / (1 ( (0.113 x (1 0.35)))) = 2% / (1 ( )) = 2% / = 3.47% 7

8 Taxable-equivalent yield if state taxes are not deductible: 2.00% / (1 ( )) = 2% / ( ) = 2% / = 3.72%. The following equations show how changing the corporate tax rate affects the taxable equivalent yields of tax-exempt bonds held by banks and casualty insurance companies. Bank-qualified taxable-equivalent yields assuming a 1.00% cost of funds: (2.00% - (0.2 x 1.00% x 0.35)) / (1 0.35) = 2.97% (2.00% - (0.2 x 1.00% x 0.21)) / (1 0.21) = 2.48% Casualty insurance company taxable-equivalent yields: (2.00% x ( )) / (1 0.35) = 2.92% (2.00% x ( )) / (1 0.21) = 2.40% Life insurance company taxable-equivalent yields: (2.00% x ( )) / (1-0.21) = 2.37% 8

9 For more information, please visit nuveen.com Sources Federal Reserve Board Interest Rates ( The Federal Reserve recently stopped maintaining its series of the Bond Buyer 20 Index on their website. In the future, a subscription to The Bond Buyer will be required to obtain this information. MMD Municipal Yield Curves Thomson Reuters (Subscription required) Conference Report on the Tax Cuts and Jobs Act Internal Revenue Code Section Reducing Income Inequality through Progressive Tax Policy U.S. Department of the Treasury, September 26, Personal Income Tax Data U.S. Department of the Treasury, Internal Revenue Service Gini Index from 1967 to 2016 U.S. Census Bureau Maximum marginal U.S. Personal Income Tax Rate Walters Kluwer, CCH Gini Index by State U.S. Census Bureau This material is not intended to be a recommendation or investment advice, does not constitute a solicitation to buy or sell securities, and is not provided in a fiduciary capacity. The information provided does not take into account the specific objectives or circumstances of any particular investor, or suggest any specific course of action. Investment decisions should be made based on an investor s objectives and circumstances and in consultation with his or her advisors. A word on risk Investing involves risk; principal loss is possible. Debt or fixed income securities are subject to market risk, credit risk, interest rate risk, call risk, derivatives risk, dollar roll transaction risk and income risk, including the possible loss of principal. The value of the portfolio will fluctuate based on the value of the underlying securities. As interest rates rise, bond prices fall. Below investment grade or high yield debt securities are subject to liquidity risk and heightened credit risk. Preferred securities are subordinated to bonds and other debt instruments in a company s capital structure and therefore are subject to greater credit risk. Foreign investments involve additional risks, including currency fluctuation, political and economic instability, lack of liquidity and differing legal and accounting standards. Asset-backed and mortgage-backed securities are subject to additional risks such as prepayment risk, liquidity risk, default risk and adverse economic developments. Clients should contact their tax advisor regarding the suitability of tax-exempt investments in their portfolio. If sold prior to maturity, municipal securities are subject to gain/losses based on the level of interest rates, market conditions and the credit quality of the issuer. Income may be subject to the alternative minimum tax (AMT) and/or state and local taxes, based on state of residence. Income from municipal bonds held by a portfolio could be declared taxable because of unfavorable changes in tax laws, adverse interpretations by the Internal Revenue Service or state tax authorities, or noncompliant conduct of a bond issuer. It is important to review your investment objectives, risk tolerance and liquidity needs before choosing an investment style or manager. This information represents the opinion of Nuveen Asset Management, LLC and is not intended to be a forecast of future events and this is no guarantee of any future result. Information was obtained from third party sources which we believe to be reliable but are not guaranteed as to their accuracy or completeness. There is no assurance that an investment will provide positive performance over any period of time. Neither Nuveen, LLC nor any of its affiliates or their employees provide legal or tax advice. You must consult with your personal legal or tax advisor regarding your personal circumstances. Nuveen Asset Management, LLC is a registered investment adviser and affiliate of Nuveen, LLC. Nuveen 730 Third Avenue New York, NY nuveen.com GPE-NTXINC-0118P INV-AN-01/19