Capital Gains Taxes and Stock Return Volatility

This paper investigates the effect of capital gains taxes on stock return volatility, a topic that has been largely ignored by scholars and policymakers. The modest attention in the tax literature is surprising, because stock return volatility plays an important role in investment decisions for both investors and firms and is closely monitored by financial market regulators.

We predict that capital gains taxes should reduce stock return volatility because they allow the government to share in the gains and losses incurred by the investors. When the capital gains tax rate is high (low), investors' exposure to stocks' cash flow risk is low (high), leading to lower (higher) stock return volatility. We base this prediction on our extension of the theoretical framework of Sikes and Verrecchia (2012). We demonstrate that increased exposure to firms' cash flow risks increases the stock return volatility, leading to an inverse relation between capital gains tax rates and return volatility. Intuitively, when the government shares in the gains and losses in the firms' cash flow, the variability of the cash flows received by the investors is reduced. This sharing reduces the volatility of the stock prices of these firms because (from the investors' perspective) the value of these firms is their after-tax discounted cash flow.

For the empirical tests, we adopt a difference-in-differences approach and control for factors known to affect return volatility. To isolate the tax impact on volatility, we identify two cross-sectional variations in the relation between changes in capital gains taxes and changes in return volatility: unrealized capital gains and dividend distributions. We predict that the more stock returns are subject to capital gains taxation (such as stocks with larger unrealized capital gains and/or non-dividend-paying stocks), the greater the increase in return volatility following a capital gains tax rate cut due to reduced risk-sharing in firms' cash flows between shareholders and the government.

Consistent with this prediction, we find that return volatility of firms with more stock returns subject to capital gains taxes increased more following the Revenue Act of 1978 (hereinafter, RA) and the Taxpayer Relief Act of 1997 (hereinafter, TRA). We choose these two legislations because they are the two most recent tax acts that substantially reduced capital gains tax rates while affecting few other elements of the tax code. Finding consistent cross-sectional changes in return volatility following capital gains tax rate cuts from two distinct economic environments occurring nearly two decades apart provides added assurance that stock return volatility moves inversely with capital gains taxes.

Our specific findings are as follows: In our RA tests, we find that the monthly return volatility for portfolios with price appreciation in the upper quartile rose 56 basis points more than that for portfolios with price appreciation in the lower quartile. Further, the non-dividend-paying portfolios experienced a 29 basis points higher return volatility increase than did dividend-paying portfolios. These estimates are substantial, as they represent 17 percent and 9 percent of the average monthly return volatility of gain portfolios and dividend-paying portfolios, respectively. Similarly, in our TRA tests, we find that the monthly return volatility for portfolios with price appreciation in the upper quartile rose 96 basis points more than that for portfolios with price appreciation in the lower quartile. Non-dividend-paying portfolios show a 68 basis points higher return volatility increase than that for dividend-paying portfolios after the capital gains tax rate cut under TRA. As with the RA results, the estimated return volatility increases are non-trivial. They account for 18 percent and 17 percent of the average monthly return volatility of gain portfolios and dividend-paying portfolios, respectively.

Did the 2003 Tax Act Increase Capital Investments by Corporations?

This study investigates the role of shareholder-level taxes on corporate investment. We use the passage of the Jobs and Growth Tax Relief Reconciliation Act of 2003 (hereafter, 2003 Tax Act) as a quasi-natural experiment to answer our research questions. The 2003 Tax Act reduced shareholder-level taxes on dividends and capital gains. Relative to legislative changes in 1993 and 1997 that also impacted shareholder-level taxes, the 2003 Tax Act changed tax rates on both capital gains and dividends and led to a dramatic decrease in the average marginal shareholder-level tax rate. Furthermore, prior research suggests that the 2003 Tax Act increased firms' share prices (Auerbach and Hassett 2006) and reduced firms' cost of equity capital (Dhaliwal et al. 2007). We predict that firms increased investment in response to the Act.

Our study focuses on corporate investment for two reasons. First, a stated goal of the 2003 Tax Act was to encourage capital investment by corporations. Second, capital investment is a fundamental component of firm value (Hanlon and Heitzman 2010). Thus, our study is of interest to policymakers and academics. Whether changes in shareholder-level taxes will impact corporate investment is an open empirical question. Prior research suggests that firms could increase dividends in response to reductions in shareholder-level taxes instead of increasing investment (Chetty and Saez 2005). In addition, some firms may fund investment with internal funds and not equity, in which case, any reduction in the cost of equity capital derived from the 2003 Tax Act is unlikely to impact investment.

We regress capital expenditures on an indicator variable equal to one for time periods after the Act, and controls for cross-sectional differences in capital expenditures. We first document that capital expenditures increase after the 2003 Tax Act. Because the economy was coming out of a recession around the 2003 Tax Act, it is possible that macroeconomic conditions unrelated to taxes caused investment to increase. To better link our findings to the 2003 Tax Act, we use a difference-in-differences research design to show that this increase in capital expenditures varies predictably with two shareholder-level tax-motivated hypotheses. First, we find that the increase in investment is smaller for firms largely held by investors that are less sensitive to shareholder-level taxes. Second, we find that the increase in investment is larger for firms most likely to fund investment from new equity issuances, rather than internal funds.

In additional analysis, we examine how firms' dividend payout policy impacts our results. We find that while the majority of firms increase investment after the tax cut, a small subset of larger, older, and cash-rich firms increased dividend payout instead. Overall, our results suggest that, consistent with the intent of policymakers, the shareholder-level tax rate reductions set forth in the 2003 Tax Act increased corporate investment. Our findings add further evidence to the question of whether taxes impact firms' value and investment.

Who Benefits from Tax Rate Transparency? Evidence from the Laboratory

In this paper, the phrase tax rate transparency describes the ease with which decision-makers can determine the rate at which their income is taxed. Prior experimental research has demonstrated that increased tax rate transparency results in better decisions (Rupert and Wright 1998; Rupert et al. 2003; Boylan and Frischmann 2006). This study contributes to our understanding of the effects of tax rate transparency by providing evidence on its distributional consequences. More specifically, this study provides evidence on who benefits from tax rate transparency, particularly when some individuals have access to better information about tax rates than others. Of particular interest is how those with limited access to information about tax rates fare in comparison to those who may have better information about relevant tax rates, and whether the private information that informed parties possess leaks out to the uninformed over time.

Understanding the effects of tax rate transparency is important because a lack of transparency can cause decision-makers to miscalculate the potential tax effects of planned transactions. This can lead to strategic and tactical errors, and potentially reduce profits. It also creates incentives for decision-makers to use scarce resources to attempt to overcome the lack of transparency, which, in turn, can create distributional consequences in an economy. In addition, a lack of transparency often creates calls for tax reform as a means of promoting increased fairness and reducing the burden associated with measuring one's tax liability.

This paper reports on a set of experimental markets in which participants were required to generate after-tax trading profits sufficiently large to generate a specific return on investment. Each market lasted eight periods, and the degree to which the relevant tax rate was transparent to participants varied across markets. Consistent with prior research, the results indicate that a lack of tax rate transparency had a negative effect on profits earned in the markets. Greater transparency led to higher profits for those who had access to the information about the relevant tax rate. However, the results from the experiment add to the existing literature by documenting that the effect of greater transparency spilled over to those who did not have access to the information about the relevant tax rate. It was those participants who benefitted the most over the course of the experiment—simply by participating in markets in which the tax rate was transparent to others.

More specifically, data from the experiment show that when the tax rate was transparent to at least some of the participants, those participants earned anywhere from 14 percent to 45.4 percent more profit than they would have in markets in which the rate was not transparent to anyone, with specific amounts depending on whether the rate was transparent to everyone or only to some participants, as well as on the stage of the markets (early periods, middle periods, or late periods). In markets in which there was a mix of informed and uninformed participants, the increased profits earned by participants for whom tax rates were made transparent spilled over to the uninformed participants. It was these uninformed individuals who gained the most over time. Initially, their profits were about 20–25 percent lower than the profits of those who were informed (but still about 9 percent higher than profits earned in markets in which tax rates were opaque for everybody), but by the late periods of the markets, their profits were virtually indistinguishable from the profits of those for whom tax rates were made transparent.

This research may be of interest to policymakers, who must wrestle with tradeoffs associated with adopting new tax rules. Some rules might influence behavior in ways that policymakers desire, but at the same time make tax rates less transparent and, hence, generate undesirable side effects. Alternatively, the effects of tax simplification will depend, at least in part, on the degree to which individuals are able to deal with complexity on their own. Policymakers should be aware of these tradeoffs when considering altering tax rules.

Firm and Investor Responses to Uncertain Tax Benefit Disclosure Requirements

We examine whether proprietary costs affect disclosure quality and how investors react to disclosure quality in a new proprietary cost setting. Leuz (2004) notes, “there is little empirical evidence on the existence of proprietary costs and their importance in explaining firms' disclosure choices.” We test for evidence of Verrecchia's (1983) proprietary cost hypotheses in a new setting, the adoption of FASB Interpretation No. 48, Accounting for Uncertainty in Income Taxes (FIN 48), where proprietary costs result from beliefs that the new required FIN 48 disclosures could weaken a firm's competitive position with tax authorities. Our approach overcomes two important measurement issues that researchers face in the empirical proprietary cost literature. First, the FIN 48 disclosures are proprietary with respect to tax authorities and should increase in the firm's level of tax avoidance, a construct with numerous established proxies in the tax avoidance literature (see Hanlon and Heitzman 2010; Lisowsky et al. 2013). This overcomes the elusive nature of identifying and measuring competitively sensitive information (see Beyer et al. 2010). Second, the mandatory nature of the FIN 48 disclosures provides a benchmark for measuring disclosure choices—deviations from the requirements of FIN 48 imply that firm managers withhold information. Finally, and more generally, we improve the external validity of the proprietary cost hypothesis by offering empirical evidence in a new setting with different types of firms (Shadish et al. 2002). Much of the empirical proprietary cost literature is limited to proprietary costs related to product market competition (i.e., Botosan and Stanford 2005; Berger and Hahn 2003, 2007).

We first investigate whether we observe lower-quality FIN 48 adoption disclosures from firms that face higher proprietary costs. Adoption disclosures provide the most powerful setting for our study; ex ante perceived proprietary costs were significant—anecdotally, this perception declined ex post. We measure disclosure quality using a two-part disclosure score: (1) disclosure compliance (completeness), and (2) disclosure precision (clarity). We measure proprietary costs using a parsimonious measure constructed from multiple proxies for tax avoidance. Controlling for other determinants of disclosure quality, we find a negative association between proprietary costs and disclosure quality, consistent with the proprietary cost hypothesis. Moreover, we also predict and find that the relative significance of proprietary costs may differ across certain components of FIN 48 disclosures—in particular, the proprietary costs of disclosing forward-looking unrecognized tax benefit (UTB) changes are greater than the proprietary costs of disclosing current UTB changes.

Next, we examine whether the market reaction to firms' disclosed UTB amounts varies with disclosure quality. Verrecchia (1983) suggests that managerial concerns of revealing proprietary information rationally limit full disclosure, despite its apparent benefit, because investors no longer treat withheld information as unequivocally less favorable. Hence, in the presence of proprietary costs, investors' demand for full disclosure is unclear. If investors favor full disclosure (i.e., they are primarily concerned with transparency), they will reward high-quality FIN 48 disclosures. If investors do not favor full disclosure (i.e., they are primarily concerned about avoiding scrutiny from taxing authorities), they will penalize high-quality FIN 48 disclosures. We find evidence consistent with investors penalizing firms that make high-quality FIN 48 disclosures; our results are concentrated in small firms and firms with higher proprietary costs, consistent with Verrecchia (1983). This analysis complements a small, but growing, literature that documents a positive association between firm value and tax avoidance (e.g., Frischmann et al. 2008; Desai and Dharmapala 2009; Hanlon and Slemrod 2009; Koester 2011) by providing new evidence that investors appear willing to accept less transparent disclosure in order to “facilitate” firms' tax avoidance activities. This finding is particularly interesting because a primary motivation for the FIN 48 disclosure requirements was to encourage firms to provide investors with transparent and comparable disclosures about firm-specific tax uncertainties.

Asset and Business Valuation in Estate Tax Cases: The Role of the Courts

In the U.S. court system, estate tax cases often involve a dispute between the taxpayer and taxing authority over the value of various assets of the estate. Valuation can be a complex process based on many assumptions, especially when the asset being valued is a closely held business. Because the tax levied is a percentage of the value of the assets in the estate, the taxpayer and taxing authority have obvious incentives for a lower or higher valuation. However, it is not clear how the court arrives at its valuation decision.

In this study, we look for evidence of an association between the valuation decision of the court and various judge- and case-related factors. Englebrecht and Davison (1977) find that the court's valuation is related to the taxpayer/taxing authority mean. Other research, as well as comments in the popular press and even comments by Tax Court justices, have suggested that the court simply picks the arithmetic mean of the taxpayer and taxing authority positions. However, Greenaway (2010) suggests that various attributes of the case and the judge are related to court decisions. We empirically examine these claims.

We review 134 estate tax cases heard in the U.S. Tax Court and District Court from 1986–2010, involving the valuation of 181 different assets. We predict and find that the taxpayer's use of more appraisers is related to lower court valuations (favoring the taxpayer), but do not find a corresponding advantage for the taxing authority's use of appraisers. We also predict and find that assets subject to more possible discounts (such as businesses) tend to be decided in the taxpayer's favor. We find limited evidence that conservative judges rule in the taxpayer's favor.

Overall, our results are consistent with court valuation decisions being related to various attributes of the case and judge. Whereas prior literature suggests the court simply picks the mean between the contesting parties, our results suggest that the type of asset, choices by the contesting parties (such as number of appraisers), and ideological leanings of the judge all influence the final court valuation.