Dividend Taxes and Share Prices: Evidence from Real Estate Investment Trusts
Authors: William M. Gentry, Deen Kemsley and Christopher J. Mayer
Theoretical and Empirical Background
3 lines of Theory
1. Investors impose a price penalty on high-dividend firms (Brennan 1970)
2. Prices may not reflect shareholder taxes because the marginal investor may be tax exempt (Miller and Scholes 1978)
3. Dividend taxes should be capitalized for all firms that will pay a dividend in the future (King 1977, Auerbach 1979, Bradford 1981)
Empirical investigation required b/c of potential strengths and weaknesses of arguments.
Empirical Investigations – attempting to isolate tax affects of dividends on prices/value
1. Long event windows used to examine hypothesis that dividend taxes increase pretax returns, indicating dividend taxes affect value/prices
a. No consensus of empirical studies, b/c hard to control for non-tax determinants of returns
2. Ex-dividend-day share prices behavior used to examine hypothesis that taxes result in less-than-dollar-for-dollar decline in share prices
a. Several questions remain unanswered
Fama and French (1998) used a cross sectional approach to answer remaining questions, focusing on share prices rather than returns.
1. Regression – equity market value on dividends
a. controlling for expected future returns and profitability
b. scale regression by book value of assets
2. Hypothesis: if tax penalty imposed, coefficient (dividends) should be negative
3. Findings: coefficient is positive on dividends
a. Concern: “imperfect controls for future profitability and the market value of assets may cause the signaling attributes of dividends to obscure tax effects.” (Basically no strong empirical conclusion possible)
Authors solution to empirical problems: Use REIT data!
Characteristics of REITs allow one to isolate tax effects and control for market value of assets and signaling attributes of dividends
1. Exempt from corporate taxes (conditional on paying out 95% of income)
a. REIT tax deductions pass through to shareholders by reducing the taxable component of dividends
2. Payouts in dividend form are required for REIT
a. This eliminates the signaling problems from previous studies
3. Market value of assets (real estate) readily available (appraisals of real estate)
a. Can control for fair market value of assets
i. Market value of properties allow one to control for much of REIT’s value but not for potential tax effects
Conclusions of the Paper
Paper hypothesis
1. Firm value is positively associated with the tax basis of REIT assets
a. Equivalent to finding investors capitalize shareholder-level dividend taxes into prices
b. Holds market value of assets constant
Conclusion: Investors assign a positive value to tax basis
1. Each dollar of tax basis is associated with an additional 9 to 27 cents of firm value
a. Estimates reflect the PV of future depreciation and other tax benefits
b. Results imply REIT prices appear to capitalize dividend taxes at high tax rates (non-zero tax rates)
“…result casts doubt on the tax irrelevance hypothesis that assets prices are determined by investors who are indifferent to taxes.”
2. Findings are consistent with tax interpretation for the ex-dividend day evidence
a. Results serve as a benchmark for assessing the impact of dividend taxes on share prices in a more general setting (i.e. with corporate taxes, dividend policy is discretionary and firms can use share repurchases in place of dividends)
Conclusion of the paper: Firm value is associated (positively) with tax basis, meaning share prices reflect future dividend taxes
II. Valuation Model – Derive the Equation
MVE – market value of equity
MVA – market value of assets (property appraisals that represent fair market value)
D - market value of debt
µ - all other factors influencing MVE (market value of equity)
MVE = MVA – D + µ (1)
Primary objective: specify how shareholder taxes affect the relationship between MVE and MVA – that is, how taxes affect the value of owning equity in an existing REIT relative to owning the assets directly as a new owner.
TB – tax basis
When an investor buys a property (not in REIT structure) TB = MVA. When investors buys shares in a REIT, TB< MVA, because REIT properties have already exhausted some of the deprecation tax deductions
INCTAX – present value of the incremental tax burden faced by the (new) REIT owners
MVE = MVA – D – INCTAX + µ (2)
INCTAX = τ (MVA – TB)
τ – is the capitalized effective tax rate for the marginal investor
Primary research questions: is τ > 0
Substituting above definition for INCATAX into equation (2) yields:
MVE = (1- τ)MVA + τTB – D + µ (3)
“Over time, tax depreciation expenses and changes in asset values can cause the REIT’s tax basis to diverge from the market value of its assets. Equation (3) implies that this divergence drives a wedge between the market value of equity and the market value of assets. If the marginal investor is taxable, equation (3) leads to the testable prediction that MVE should increase in the firm’s inside tax basis in its assets.”
REIT analysts report NAV – the market value of net assets, where NAV = MVA – D, so MVA = NAV + D, substituting this in to equation 3 for MVA results in:
MVE = (1- τ)NAV + τTB – τD + µ (4)
Equation (4) focuses on the manner in which shareholder taxes affect the difference between the value of a firm’s equity and the value of its assets.
III. Data and Empirical Specification
Data – 85 equity REITs from 1992 to 1999, 389 observations
Empirical Specifications
Change equation (4) into empirical equation:
MVE = (1- τ)NAV + τTB – τD + µ (4)
BVA – book value of assets (equal to the tax basis TB)
If investors capitalize future shareholder-level taxes into share prices, then we would expect to equal τ and to equal –τ, so that .
Additional Variables
UPREIT – a dummy variable (1 if UPREIT, 0 otherwise)
• UPREITs may face certain conflicts of interest and restrictions on sales and refinancing that could affect firm value relative to NAV
BVPE – book value of preferred stock
• To capture unobservable differences in the few REITs that issue preferred stock
Additional Variables (Error Term Concerns)
REIT shares are much more liquid than real estate properties. If this liquidity has value, then REITs could be valued at a premium relative to NAV.
SMALL – a dummy variable for REITs with equity cap under $400 million
• Controls for size and liquidity
SPREAD – midpoint of bid/ask price of the month of Dec. for each year’s observation
To control for investor sentiment, include year-specific intercepts and allow the coefficient on NAV to vary by year. By estimating separate coefficients on NAVE for each year, we allow investor sentiment to affect the ration of average REIT share prices relative to NAV.
3 Models
OLS – simple linear regression
Random Effects Model – large population of levels and corresponding effects, factor effects used in a study may be considered as a representation of this population, treats categorical variables the same as the error term
Fixed Effects Model – fixed number of levels, we are interested in comparing the mean response for each combination of factor levels, treats categorical variables as part of the intercept term (Should be viewed with caution, because of degrees of freedom from small sample size)
OLS estimates are used as a naïve benchmark for both random and fixed-effects estimates for the primary empirical equation below.
Primary Empirical Equation
(6)
γt are the year-specific coefficients for NAV
Xjit are the observable firm specific characteristics (δj are the associated coefficients)
θi are the year-specific constants
are the firm-specific components of the error term (specified with either random of fixed effects models)
εit captures any remaining error
Testing Data
First test used to test if .
Second test uses BVCE as measure of the tax basis, expecting α1 to be positive and α2 to be zero, however debt and preferred stock is included to control for possible capital structure affects on firm value.
The actual magnitude of τ should depend on the identity of the marginal investor, as well as the applicable tax rate and the timing of the tax benefits to shareholders. In analysis no assumptions are made as to who the marginal investor is, but it is inferred from the data what the applicable tax rate would be.
IV. Empirical Results
Table I
MVE/NAV mean is 1.09 in contrast to the discount that is expected from tax factors alone, so it is important to control for nontax variables
High Dividend result of REIT requirement of cash distributions of income
Table 1
Summary Statistics for REIT Sample
Table II
BVA coefficient is positive and statistically significant at the 95 percent confidence level (OLS 0.17, REM 0.18, FEM 0.25)
This is consistent with the hypothesis that investors capitalize dividend tax savings from tax basis into share prices.
D coefficient is negative and statistically significant, further more we cannot reject the prediction that .
Panel B
Coefficients remain positive (negative) and statistically significant at the empirical level, but scaling the variables by NAV reduced the magnitude of the coefficients.
Evidence suggests that investors capitalize substantial shareholder taxes in REIT share prices, but the magnitudes of the estimates depend on the specifications used.
Table II
Regression Measuring Tax Basis with Property Value
Table III
Uses BVCE in place of BVA (BVCE captures tax basis in all assets, not just tax basis in real estate properties)
BVCE estimate is positive and statistically significant
In contrast to predictions (debt should have a coefficient of zero here), debt D is estimated to be positive, however it is negative in most specifications
Table III
Regression Measuring Tax Basis with Book Value of Common Equity
Table IV
The final two tests of the paper (pages 278 and 279) are simply to examine the possibility that ‘measurement error’ could drive the results of the study. Specifically, a FFO (Funds form Operations) test was preformed in the study as an example to see if correlation between measurement error in NAV and BVA, BVCE or D could bias the estimated tax rate results. Table IV shows that FFO does provide incremental value-related information that is not captured by NAV. Nevertheless, FFO in other specifications does not reduce the magnitude of the BVA and D coefficients. Some of the authors concern for the measurement error driving the results is mitigated from the FFO results.
Table IV
Regressions of Leading FFO on Book Value of Common Equity and NAV
The Pretax market value of equity:
ρ is the discount factor (i.e. 1 + appropriate discount rate)
π are the pretax cash flows
µ represents all other factors that may affect PMVE
Table IV – Question: does the BVCE have predictive power for future profitability after controlling for NAV? Table IV says “no”, and shows that even if the measurement error in NAV is material it does not appear to be correlated with BVCE.
Summary and Conclusions
Study exploits characteristics of REITs to estimate the influence of shareholder-level taxes on share prices. Evidence indicates that each dollar of tax basis increase REIT share prices by 9 to 26 cents, conditional on their fair market value of properties, with most estimates ranging from 9 to 20 cents. These estimates reflect the discounted value of future tax benefits from existing tax basis in properties. To the extent investors do , indeed, capitalize shareholder-level taxes into REIT share prices, REIT investment is essentially on equal footing with direct real estate investment, at least from a tax perspective. New REIT investors implicitly charge sellers for the tax depreciation deductions they have consumed, as well as for any unpaid taxes on property appreciation, via lower purchase prices for the stock. Although the potential effects of omitted nontax factors remain a concern, the estimates are robust to a variety of specifications. Investors capitalize a substantial amount of dividend taxes into prices when dividend policy is largely nondiscretionary, there are no corporate taxes and share repurchases do not offer a tax advantage relative to dividends. This result casts doubt on the tax irrelevance hypothesis that asset prices are determined by investors who are indifferent to taxes. The study provides a benchmark for future examinations of share price effects of dividend taxes in a more complex setting.