Corporate Taxes and Stock Volatility Erin E. Syron∗ February 1, 2012

Abstract How do dividend taxes affect stock volatility? If a risk-averse executive faces price risk through his incentive contract, changes in stock volatility due to dividend taxes may increase agency costs and therefore decrease overall welfare. In this paper, I use a decrease in dividend taxes as a natural experiment to identify the impact of dividend taxes on firm price volatility. I first create an agency model where the executive puts in effort that increases firm value and the variance of firm output. The model leads to several predictions including an increase in dividend taxes will increase stock volatility especially for firms with highly incentivized executives. I test these predictions empirically using the 2003 dividend tax cut. Stock volatility decreased after the tax cut for firms where an executive has large holdings of shares and options. Therefore, with a risk-averse executive and risk-neutral shareholders, dividend taxes may exacerbate agency costs. The increase in agency costs will decrease shareholder welfare.

∗

Email: [email protected]. I thank Alan Auerbach, Kevin Hassett, Ulrike Malmendier, and James Wilcox for their comments and discussions. I also thank Matt Jensen for excellent research assistance. I also thank participants at the UC Berkeley Public Finance Seminar.

1

Introduction

Often when we discuss the effects of taxes, we focus on the effects on the mean of firm value and de-emphasize the effects on the variance of the distribution. The existing literature on dividend taxation has analyzed the effect on firm stock price (Auerbach and Hassett (2005) and Amromin, Harrison and Sharpe (2008)). However, there are two important moments for the stock: the level of the stock price and the volatility of the stock. In a world with risk averse executives, if a dividend tax change sufficiently increases volatility, it could lower the utility of the shareholder even if it increases the mean of the share price. It is therefore important to consider the effects on the volatility of the stock. Taxes may affect the uncertainty of the market on any given day, which would affect the volatility of the stock. An important exception to Modigliani and Miller (1958), taxes may also change the financial structure of a firm. A decrease in the dividend tax may make debt less desirable and equity more desirable (Graham (1996) and MacKie-Mason (1990)). If firms move toward more equity, the volatility of their equity might decrease. In addition, if dividends are used as a signal, a decrease in taxes on dividends decreases the price of the signal (Bernheim and Wantz (1995) and Bernheim and Redding (2001)). Also, if the executive can change the variance of output of the firm by changing the investment strategy of the firm (Lambert (1986), Hirshleifer and Suh (1992), Knopf, Nam and Thorton (2002), and Coles, Daniel and Naveen (2006)), the volatility of the stock may change if the relative amounts of incentive pay versus salary changes. In many models of stock valuation, general shareholders are able to diversify the idiosyncratic risk of a stock and are thus only affected by systematic risk. However, one important class of shareholders is often unable to diversify the idiosyncratic risk: executives. A board of directors will set optimal incentives for their executives that include undiversifiable risk so that executives have an incentive to increase the firm’s value. However, when executives are unable to diversify the idiosyncratic risk, they will be negatively affected by an increase in the volatility due to a change in the tax environment. I model shareholders in corporations as risk-neutral and managers as risk-averse. In a world where taxes change the second moment, absent any changes in company policy, the welfare of the shareholders would be unaffected. Because the welfare of the managers would be affected by the second moment effects, the tax change could affect the principal-agent costs associated with the hiring of manager.

1

The principal-agent costs could be particularly large if an increase in investment leads to an increase in variance in the distribution of firm output. In this case, the risk-averse executive would benefit from putting in investment but would have a cost of an increase in the variance.1 The shareholders would only have the benefit from the investment. Therefore, they would want the executive to increase the investment, but the executive would increase the investment by less than the shareholder’s optimal amount due to the additional variance costs. This difference would exacerbate the original agency costs. The Jobs and Growth Tax Relief Reconciliation Act of 2003 changed many aspects of the US tax code. President Bush proposed the legislation January 7, 2003, and Congress passed the bill by the end of May. It was made retroactive to January 1, 2003. In particular, it temporarily decreased both the long-term capital gains tax rate and the dividend tax rate. The capital gains tax rate was reduced from a maximum rate of 20% to 15%. The larger change was in the dividend tax: the qualified dividend tax rate fell from a maximum rate of 35% to a maximum rate of 15%.2 This law disproportionately decreased the dividend tax rate for individuals with high income, as the dividend tax rate decreased from their rate on personal income (35%) to a maximum of 15%. It is therefore a change which can be used to study the effects of dividend taxation on stock volatility, through the effect on executive behavior. The rhetoric before the dividend tax cut followed two different molds. The first said that the dividend tax cut would increase investment for all firms as the cost of capital decreases. Investors will therefore invest more in the firms which will increase the value of all firms. However, if investment leads to larger volatility, then it should also increase the volatility of the stocks. The second mold was that the dividend tax would decrease overinvestment in non-value enhancing projects by the executives and increase corporate governance by increasing the incentive to disperse earnings through dividends. This would decrease volatility. It would also particularly decrease volatility for the executives who increased dividends the most. Therefore, although the two stories would both increase firm value, they have different effects on volatility. In Section 3, I develop an agency model where a risk-averse executive puts in effort that increases the expectation and variance of firm value. In order to incentivize the executive to put in costly effort, shareholders give the executive part of the firm’s equity as incentive pay. However, if the firm is risky, the executive then of course bears some of the risk of the firm. Because the effort 1 While the executive’s options portfolio would increase in value with increased stock variance, this is a second-order effect. 2 See Auten, Carroll and Gee (2008) for a detailed description of all the tax changes associated with the 2003 law.

2

increases firm value but also the variance of the firm, the increase in the executive’s utility from an increased expected firm value is partly counteracted by the corresponding increase in variance. I assume that the executive faces the dividend tax on dividends paid by the firm.3 The shareholders are taxed at a dividend tax rate on dividends. I find that a decrease in the taxes on dividends should decrease volatility to a greater degree for executives who own a large fraction of the firm before the tax change. I test these predictions in Section 5 empirically, using the data described in Section 4. A measure of variance is the daily volatility of the stock. However, this volatility comprises two pieces. The first is the firm-specific idiosyncratic volatility. The second is the systematic volatility, associated with overall market movements. I separate these two pieces by looking at the error from a daily Fama-French Three-Factor Model regression. I then look at the quarterly volatility of this error to examine whether the volatility increased after the 2003 tax cut. The model suggests that volatility should have decreased for all firms, but more for firms with a highly incentivized executive. Therefore, I take a difference-in-differences approach where I compare the volatility of firms who have executives with incentives in the lowest-quintile to firms who have executives with incentives in the highest-quintile. I find that firms with executives in the highest-quintile of incentives had the largest decrease in both systematic and idiosyncratic volatility after the 2003 tax change. The results are robust to specifying the volatility in returns or in price. I also examine incentive packages for executives. The model predicts that the level of incentives should not change following the 2003 dividend tax cut. I use the Core and Guay (2002a) measures of sensitivity to price as my incentive variable. The first measure is the sensitivity to price of new stock grants and new option grants. The second is the sensitivity to price for all stocks and options owned by the executive. I find a decrease in the sensitivity to price for all stocks and options owned, but not for new grants. This provides evidence in support of the model if shareholders adjust yearly packages rather than the overall incentive level. Section 6 discusses the results and concludes. Dividend taxes appear to affect firm volatility. Because they may exacerbate agency costs, we should consider these affects when changing tax policies. 3

See Hall and Liebman (2000) for a description of executive compensation taxation.

3

2

Literature Review

There is a large literature on the effects of the dividend tax change of 2003 that strongly relates to this paper. Amromin et al. (2008) and Auerbach and Hassett (2005) use event studies to show that firm value increases with a decrease in the dividend tax for high dividend yield firms. Auerbach and Hassett (2005) also find an increase for non-dividend paying firms. Both these papers demonstrate that the dividend tax change increased the first moment of the stock price. In this paper, I primarily focus on the second moment, the volatility of the stock price. In addition to the effect on the stock price, several papers look at the direct effect on dividends themselves. Chetty and Saez (2005) find that more firms initiated dividends after the tax change, and firms that were paying dividends increased the amount of the dividend. They find the largest effect for firms that have an executive who owns a large fraction of the firm. Nam, Wang and Zhang (2004) further show that dividend payments and managerial stock options are negatively correlated, as option holders have an incentive to keep cash in the firm for future payouts. Blouin, Raedy and Shackelford (2007) find that firms changed their payout policy from share repurchases to paying dividends. Brown, Liang and Weisbenner (2007) also demonstrate an increase in dividends at the expense of share repurchase, especially at firms with high executive ownership. Dividends increased relatively quickly, especially for firms with high executive stock ownership. Traditional models of dividend taxation would not predict the speed at which dividends increased. They would also not predict a differential response based on executive stock ownership. Chetty and Saez (2010) and Gordon and Dietz (2006) adapt the models by adding agency conflicts to show that agency costs are an important factor to consider when looking at dividend payout policy. Korinek and Stiglitz (2009) argue that it is intertemporal arbitrage rather than agency costs that lead to an increase in dividends as the firms were taking advantage of the temporary dividend tax cut. In this paper, I take into account some agency costs. I am however not looking at a life-cycle model of the firm, so I will not address intertemporal arbitrage costs. I consider the effects of a change in the dividend tax, and it is therefore important to understand why firms pay dividends. According to Miller and Modigliani (1961), absent taxes, dividends should not affect firm price. However there is a large literature suggesting they do. Allen and Michaely (2003) critically surveys the literature and identifies two main hypotheses, with implications for stock volatility. The first is a signaling hypothesis, which states that since insiders have more information about future firm cash flows, a dividend today means that the insiders are not worried

4

about the ability to pay for future investment out of future earnings. Bernheim and Wantz (1995) and Bernheim and Redding (2001) find evidence in support of the signaling hypothesis. A second hypothesis is that dividends are used to decrease agency costs at the firm. If the firm pays dividends, then it will have less cash on hand to use for overinvestment by the executive. Christie and Nanda (1994) find lower dividend growth in firms with higher agency costs as the firms had high dividends originally to mitigate some of the high agency costs. La Porta, de Silanes, Schleifer and Vishny (2000) and Fenn and Liang (2001) also find evidence in support of the agency theory. If dividends are used to decrease agency costs by decreasing overinvestment at a firm, an increase in the dividend tax may increase agency costs and lead to overinvestment. Chetty and Saez (2010) use this type of overinvestment model to study the dividend tax change. Another way that dividends could affect stock volatility is through a change in the firm’s leverage or capital structure. Graham (2003) and Auerbach (2002) survey the theoretical and empirical literature for this effect. Modigliani and Miller (1963) show theoretically that debt financing should increase with the corporate tax rate and that firm value will increase with the use of debt until the firm has exhausted the tax benefits of debt. This prediction has proven hard to demonstrate due to several endogeneity issues. Graham (2000) finds evidence that the tax benefit of debt is about 9% of firm value. MacKie-Mason (1990) and Graham (1996) finds a positive relationship between tax rates and new debt issuance. The theoretical underpinnings are thus much stronger than the empirical ones. If an increase in the dividend tax rate increases the tax benefit of issuing debt, then the firm’s leverage might increase with the tax increase. This increase in leverage might therefore increase the firm’s stock price volatility as fewer cash flows remain for the stockholders. This paper examines the effects of a tax change on idiosyncratic stock volatility, and therefore relates to a large literature that examines the changes in idiosyncratic volatility over time. Campbell, Lettau, Malkiel and Xu (2001) document an increase in idiosyncratic volatility from 1962 to 1997, with a particularly large increase in the 1990s. However, they note that the overall market volatility did not increase as much. Several papers attempt to explain this increase through different channels. In the empirical section, I take into account many of the explanations by adding the appropriate explanatory variables so the regressions are not biased by these omitted variables. Bennett, Sias and Starks (2003) hypothesize that the preferences of institutional investors changed to prefer smaller and riskier firms, therefore increasing the concentration of small and volatile firms in the market. Xu and Malkiel (2003) postulate that increasing institutional ownership increased demand for high growth and hence highly volatile firms. Fink, Fink, Grullon and Weston (2010) 5

shows that a decrease in firm age lead to an increase in overall volatility. Brown and Kapadia (2007) also suggest that the historical tendancy for firms to remain private is deteriorating. Wei and Zhang (2006) find that firm earnings worsened over this time period thereby increasing volatility, especially for newly public firms; an effect likely due to higher bankruptcy risks. Irvine and Pontiff (2009) hypothesize that an increase in the volatility cash flows due an increase in competition among firms lead to an increase in stock volatility. Comin and Philippon (2005) find a similar result due to deregulation. Rajgopal and Venkatachalam (2011) hypothesize that a decrease in accounting earnings quality over this time period lead to more volatile stock prices. They use the Dechow and Dichev (2002) accrual measure to control for earnings quality changes over time. The papers discussed above attempt to explain an increase in volatility during the 1990s. Brandt, Brav, Graham and Kumar (2010) documents a decrease in firm specific volatility during the 2000s, and suggest that the earlier increase was due to changes in retail customers and was primarily among low priced stocks. The sample in this paper is different from the sample in Brandt et al. (2010) as I require executive compensation data. Therefore, I only consider firms in the S&P 1500. These firms have a large institutional ownership so therefore fewer retail customers. However, I still find a decrease in firm specific volatility absent any controls in the 2000s. These papers highlight the need for a strong identification strategy as many factors influence volatility. This paper also relates to the literature on the relationship between firm specific risk and compensation. There is a large theoretical literature starting with Jensen and Meckling (1976) that demonstrates the trade-off between incentives and risk. In a traditional agency model, the executive takes firm risk as given and incentives are negatively correlated with firm risk. Several papers have looked at this effect empirically. Prendergast (2002) surveys the literature and finds mixed results. A few papers consider measures that are close to what I am comparing (idiosyncratic volatility and stock and option incentives): Jin (2002) and Aggarwal and Samwick (1999) find that firms with the largest nonsystematic risk have the lowest incentive levels. Core and Guay (1999), however, find the opposite result, a positive relationship between nonsystematic risk and incentive levels. Although in Core and Guay (2002b), they explain that the difference is due to firm market values. The basic theoretical models do not take into account difference in firm size and the resulting difference in the executive’s influence on the firm. Several papers add to the basic theoretical model by allowing the executive to control the idiosyncratic volatility of the firm. Core, Guay and Larcker (2003) summarize the literature on how options and shares affect the variance of the executive’s take-home pay. Lambert (1986) 6

demonstrates that executive’s may under or overinvest in risky projects in equilibrium. Hirshleifer and Suh (1992) suggest that the options (as opposed to shares) are required to incentivize the executive to take on risky projects if there are risky growth opportunities which the shareholders would like the executive to undertake. Hirshleifer and Thakor (1992) use a career concerns model to show that if executives have a concern for their reputation they will invest in safe projects. These papers suggest that executive’s may underinvest in risky projects if they are risk-averse. I find a similar result in the model. There is also a literature that empirically examines if executives affect the idiosyncratic volatility of the firm when they have large incentives. This paper’s model assumes that executives have control over idiosyncratic volatility and that executives have an incentive to decrease volatility if they have large incentives.4 Knopf et al. (2002) find that as the sensitivity of take-home pay with respect to firm price increases firms hedging actions also increase. However, as the sensitivity with respect to volatility increases, firms hedge less. Coles et al. (2006) find a similar result by looking at risky investment and increases in firm leverage. Chava and Purnanandam (2010) also look at cash balances and leverage to find that executives with a high sensitivity to firm price hold more cash and have lower leverage. Therefore it appears that the executive takes actions to decrease firm volatility when the executive is highly affected by the volatility through the sensitivity to price of his take-home pay. This evidence motivates the consideration of how the effects of changing incentive pay through taxes affects overall firm volatility.

3

Model

In a standard agency model where the executive puts in costly effort, the executive’s effort does not affect the variance of the firm as it is assumed exogenous. However, if investment and volatility are correlated, then the executive will choose the optimal investment to maximize his expected utility, taking into account the associated change in utility due to the change in variance. This model includes the effects of an increase in investment on the variance of the firm. It includes the effects from a dividend tax. Dividends are taxed at the dividend tax rate. All other income is not taxed. 4

The share component of incentive packages decreases in value to the executive with increased volatility. The option value increases, but it is a second order effect.

7

3.1

Model Setup

In this model, the executive’s earnings are based on the output of the firm and a fixed salary. The executive is risk-averse while the shareholders are risk-neutral. The shareholders have bargaining power. The firm has cash on hand (X) which can either be paid out to shareholders though a dividend or invested in the firm (I). The output of the firm is given by P = Iθ, where I is the investment and θ is a normally distributed random variable with mean µ and variance σ 2 . Therefore, as the executive increases investment, the variance of the firm also increases. The shareholder’s utility is increased by investment without any negative effect, other than through the amount of payment the executive requires. However, the executive’s utility decreases with the increase in the variance if the executive receives part of his compensation as incentive pay. We assume the executives have exponential utility, given by: U (d, c, tp ) = − exp (−r(d(1 − td )D + dIθ + c − w(I)))

(1)

where d is the fraction of the firm paid to the executive, r is a coefficient of risk aversion, c is the salary, and td is the dividend tax rate. The cost of investment is w(I) = 12 kI 2 , where k is simply a parameter describing the cost of investment. The wage of the executive is dIθ + c. He also receives dividends that are taxed at the dividend tax rate, and the executive must also pay an untaxed cost of investment, 21 kb2 .

3.2

Model Solution

For a given set of parameters, the shareholders will first decide the optimal incentive scheme, a pair (c, d), to minimize the executive’s utility, given the executive’s participation constraint. The executive will then maximize his utility by choosing his effort taking into account the incentive scheme, and the dividend - investment trade-off. The executive will therefore maximize the following: max b

1 1 d(1 − td )(X − I) + d(Iµ − c) + c − kI 2 − rd2 I 2 σ 2 2 2

(2)

taking the contract structure (d and c) and tax rate as given. This leads to the amount of investment

8

that is optimal for the executive: I ∗ (d, c, tp ) =

d(µ − (1 − td )) k + rd2 σ 2

(3)

The executive decides to either pay out the cash through dividends, in which case he receives d(1 − td )D, or he can invest in the firm where he receives dIµ + c − 12 kI 2 − 12 rd2 I 2 σ 2 . If the dividend tax changes, the executive will choose to re-optimize based on the new dividend tax. The return from investing is µ minus the opportunity cost, 1 − td . The denominator takes into account the direct costs of investing. Assuming the shareholders have bargaining power, the shareholders will minimize the executive’s utility. This will place the executive at his participation constraint. The shareholders will then be setting the salary for any given incentive scheme such that the executive is at his reservation utility: ¯ = d(1 − td )(X − I ∗ )dI ∗ µ + c(1 − d) − 1 I ∗2 (k + rd2 σ 2 ) U 2 ¯ U 1 ∗ ⇒ (1 − d)c = dX(1 − td ) − dI (µ − (1 − td )) − 2

(4)

This expected utility takes into account the effect on investment of a change in d. Therefore, the I in the above equation is the optimal I for the given incentive scheme d. The shareholders will then choose the incentive scheme d to maximize their expected utility, taking into account how investment changes for any given change in the incentive scheme. max

E[U ] = (1 − d)(X − I)(1 − td ) + (1 − d)(I ∗ µ − c)

max

1 2 2 ¯ + (d − 2 d )(µ − (1 − td )) ) X(1 − td ) − U k + rd2 σ 2

d

d

(5)

d must therefore satisfy the following quadratic equation: 0 =k − kd − rσ 2 d2 Note that the incentive scheme is unaffected by the dividend tax.

9

(6)

3.3

Comparative Statistics

When the executive is deciding between dispersing cash on hand through dividends or investing in firm, he takes into account the benefits and costs of each. By dispersing cash through dividends, he receives the cash today so there is no variance or cost associated with it. However, he has to pay the dividend tax and he does not receive a return on the cash. By keeping the cash to invest in the firm, he receives a return on the cash and does not pay any taxes; however, he now faces the personal costs of investing and there is an increase in the firms variance. Therefore, the executive will trade off the costs and benefits to find the optimal balance for him. If the tax on dividends increases, then the benefits of dispersing cash through dividends decreases. The benefits to investment will not change. Therefore, the executive will increase investment and decrease dividends. This change will lead to an increase in volatility. Proposition 1. An increase (decrease) in the dividend tax rate increases (decreases) the volatility of the stock. The dividend tax change differentially affects volatility by level of executive incentive. The effects may not be the same over the entire distribution even if the sign is always positive. Proposition 2. The increase (decrease) in the volatility of a stock after an increase (decrease) in the dividend tax rate is larger in magnitude for firms with larger incentives. If the executive has a large holding of shares, he faces a larger cost of dispersing dividends if the dividend tax increases. Therefore, he will decrease dividends more than an executive with fewer shares. Therefore, he will increase investment more which will increase volatility most for firms who have highly incentivized executives. The investment is affected by the the dividend tax rate. However, the incentive scheme is not directly affected by the dividend tax. If dividend taxes increase, the volatility of the stock will change through the effects of the dividend tax change on investment. See Appendix for proofs. Proposition 3. An increase (decrease) in the dividend tax rate does not affect the incentive scheme. The incentive scheme is not affected by the dividend tax as both the executive and the shareholders face the dividend tax in the same manor. The dividend tax also does not directly affect agency costs through either the effects on the variance or the cost of investment. It only affects those costs through the investment decision. Therefore, the incentive schemes will remain the same.

10

As the overall incentive schemes are independent of dividend taxes, any change must also be independent of the original incentive scheme. Proposition 4. The change in the the incentive scheme after an increase (decrease) in the dividend tax rate is independent of the original incentive scheme incentives. There should be no effect on incentive schemes, and there should be no differential effect by original level of incentives.

4

Data

4.1

Data Sources

The primary data source for this paper is from the Center for Research in Security Prices (CRSP). This data source includes daily prices for stocks. It also includes the timing of the payment of dividends and the type of dividend paid for each day. I merge this data with data from Compustat on firm finances. The last data source is the ExecuComp Database, which includes information on compensation for the five highest paid executives of each firm in the S&P 1500. This limits my data to firms in the S&P 1500. There is partial data for 1992, but the full sample begins in 1993, so I consider all firms in the S&P 1500 from 1993 to 2009. In order to avoid changes in sample composition over time, firms must be in the Execucomp database in 2000, and must have several variables including total assets and net income. Those requirements limit the sample to 1,465 firms. Compustat includes quarterly financial data, so there are a total of 56,476 firm-quarter observations in the sample. The S&P 1500 contains 500 large market capitalization firms, 400 middle market capitalization firms, and 600 small market capitalization firms. Market Capitalization is defined as the number of common shares outstanding at the end of the quarter times the closing price at the end of the quarter. Table 1 includes summary statistics for the firms in the sample. Price is the price variable from CRSP at the close of the market on the last day of the quarter. As expected, there is a large dispersion in prices, with prices ranging from 14 cents to almost a thousand dollars. However, the mean is about 34 dollars, and most firms do not have extreme prices. Also, as expected for the S&P 1500 there is a large dispersion in market capitalization. The smallest firm has a market capitalization of just above 2.5 million dollars, while the largest firm has a market capitalization of over 571 billion dollars. Profit margin is defined as yearly net income divided by yearly sales

11

for the last 12 months. Most firms make a positive profit of about 6 cents per dollar of sales. The firms earn on average 49 cents per share, but this is much more dispersed than profit margin. Sales per share is defined on a yearly basis. There is a large dispersion in the overall sales per share. The market-to-book ratio5 averages 1.85 but ranges from 0.3 to 74.5. The leverage ratio is the long term debt divided by total assets. The average firm has a leverage ratio of 0.2, with some firms not financed at all by debt and other firms heavily financed by debt. Government Surplus is as a percentage of GDP at the end of the calendar year for the United States. Most years the United States runs a deficit, but for several years in the early 2000s there was a positive surplus. The Earnings Quality measure is the measure used in Rajgopal and Venkatachalam (2011)6 as they find that it can significantly affect firm volatility. This measure varies significantly for the firms-quarters in this sample.

4.2 4.2.1

Definitions Volatility Measures

An executive’s utility includes the value of his take-home pay, which may include incentive pay. If the incentive pay varies with the stock market, then the take-home pay of the executive will vary with the stock market. Further, if he is risk-averse, then an increase in the volatility of takehome pay would negatively impact his utility. The standard deviation of his take-home pay can be approximated as the volatility of number of shares that he owns through options and stock times the price. I have four possible proxies for the volatility of take-home pay. Two measures consider the standard deviation of the stock return over a quarter. The other two measures consider the standard deviation of the stock price over a quarter. If the number of shares is negatively correlated with the price of the shares, then the standard deviation of the return may be the best measure to use.7 However, if the number of shares is not negatively correlated with price, then the volatility 5 As in Malmendier and Tate (2009), I define market to book value as the ratio of Market Value of Assets to Book Value of Assets. Book value of assets is total assets at the end of the quarter. Market value of assets is defined as Book Value of Assets plus market equity minus book equity. Market equity is defined as common shares outstanding at the end of the quarter times quarter closing price. Book equity is calculated as stockholders equity at the end of the quarter [or the first available of common equity outstanding plus preferred stock par value or total assets minus total liabilities ] minus preferred stock liquidating value. 6 The authors use a measure developed by Dechow and Dichev (2002) which is based on an estimate for total accruals. 7 Consider two stocks, a high-priced stock and a low-priced stock, that increase by exactly one-percent every other day and decrease to the original prices (approximately a 0.99% decrease) on the other days. After the quarter (assuming an even number of days again), the price of the stocks is equal to the price at the start of the quarter. The overall return is zero. Both the stock with a low price and the stock with a high price will have the same standard deviation of returns. However, the stock with a low price will have a much higher standard deviation of price than

12

of price may be the best measure to use.8 In the data, I find a correlation between stock price and the level of incentives to be 0.2. It is highly significant with a t-statistic of 39.0. Therefore, the price and the overall level of incentives seems to be positively correlated, and the volatility of price is the best measure to use. I focus on this measure but also include the standard deviation of the stock price return. In addition to the price-return breakdown, I also consider the difference between the standard deviation and the excess standard deviation. The standard deviation of either the price or the return can be broken into two pieces. One piece is the systematic volatility. The other piece is the idiosyncratic volatility. The overall market may increase or decrease due to general market conditions. A firm may either move with or against the market (or not with the market at all). The executive may have little control over this piece of the volatility. However, the firm also will have an idiosyncratic error, which the executive may have more control over.9 In addition, the type of firm (whether it is a pro-cyclical or anti-cyclical firm) may have been previously determined. Therefore, although the executive’s income will increase or decrease with the overall volatility of the market, the piece of the volatility that he may have control over is the excess volatility. I define all four measures of the standard deviation: the standard deviation of the price of the stock, the standard deviation of the stock return, the standard deviation of the stock’s excess return, and the excess standard deviation of the stock price.10 All four are measured over the course of the quarter and consider daily changes in the stock’s value. the high priced stock. If the executives of both the high and low priced stock have the same number of share of the high and low stock, the volatility of both executive’s income is higher for the executive with the high priced share as a one-percent increase is larger for a high priced share than a low priced share. However, if the executive of the low priced stock has the number of shares required to make his shares value the same amount as the high priced stock executive, then the two would have the same volatility of income 8 Now consider two stocks that increase by exactly a dollar every other day. On the days when they do not increase by a dollar, they decrease by exactly a dollar. After a quarter (assuming an even number of days), the stocks would have the same prices as at the beginning of the quarter. If one stock has a high price and one has a low price, the volatility of the stock price will be exactly the same for the two stocks. However, the volatility of the return would be much higher for the stock with a low price compared to the stock with a high price. If the executives of both the high and low priced stock have the same number of share of the high and low priced stock, the volatility of both executive’s income is the same. However, if the executive of the low priced stock has more shares, the volatility of his income is higher. 9 Say he can choose either a high risk or low risk project. Then the volatility would increase or decrease but it may or may not be correlated with the overall market. The executive may also be able to increase or decrease overall risk by changing the size of the investment. 10 The model does not differentiate between idiosyncratic and systematic volatility for tractability. All volatility in the model is idiosyncratic. A possible extension of the model would allow for the systematic volatility.

13

The first measure of volatility is the standard deviation of the price. It is defined as:

sd(p)iqy

v u T PT uX pitqy 2 t = (pitqy − t=1 ) T

(7)

t=1

where pitqy is the price of stock i at day t in fiscal quarter q in fiscal year y. This measure includes both the systematic and idiosyncratic volatility. The second measure of volatility is the standard deviation of the return. It is defined as:

sd(R)iqy

v u T PT uX Ritqy 2 t = (Ritqy − t=1 ) T

(8)

t=1

where Ritqy is the daily stock price return for stock i for day t in fiscal quarter q in fiscal year y. This measure includes both the systematic and idiosyncratic volatility. It measures the return rather than the price. The third and fourth measure of volatility both consider the excess standard deviation of the firm. In order to determine the excess standard deviation, one first needs to determine either the excess price changes or the excess return. In order to determine both, I estimate the expected return for the stock using the Fama-French Three-Factor Model.11 I estimate the following equation using ordinary least squares for each fiscal year.

Ritqy − Rftqy = αiy + β1iy (Rmtqy − Rftqy ) + β2iy SM Btqy + β3iy HM Ltqy + itqy

(9)

where Ritqy is the daily stock price return for stock i for day t in fiscal quarter q in fiscal year y, Rftqy is the risk free rate for day t in fiscal quarter q in fiscal year y, Rmtqy is the market return for day t is fiscal quarter q in fiscal year y, SM Btqy is the Fama-French Small Minus Big factor for day t is fiscal quarter q in fiscal year y, and HM Ltqy is the Fama-French High Minus Low factor for day t is fiscal quarter q in fiscal year y. I then use these estimated betas to find the predicted return: ˆ (Rmtqy + Rftqy ) − β2iy ˆ SM Btqy + β3iy ˆ HM Ltqy ˆ = Rftqy + αˆiy + β1iy Ritqy 11

See Fama and French (1993) for a full description of the model.

14

(10)

and the predicted price: ˆ (Rmtqy − Rftqy ) − β2iy ˆ SM Btqy − β3iy ˆ HM Ltqy ) pitqy ˆ = pi(t−1)qy (1 + Rftqy + αˆiy − β1iy

(11)

The excess return is therefore ˆ (Rmtqy − Rftqy ) − β2iy ˆ SM Btqy − β3iy ˆ HM Ltqy itqy ˆ = Ritqy − Rftqy − αˆiy − β1iy

(12)

and the excess price is ∆pitqy = pitqy − pitqy ˆ

(13)

Now that we have measures of excess return and excess price, one can find the standard deviation of both. The third measure of volatility is the standard deviation of the return error from the FamaFrench regressions:

sd(ˆ )iqy

v u T PT uX itqy ˆ 2 t (itqy ˆ − t=1 ) = T

(14)

t=1

This measure includes only the idiosyncratic volatility. It also is measured by the return rather than the price. Finally the last measure of volatility is the standard deviation of the difference between price and expected price based on the Fama-French model:

sd(∆p)iqy

v u T PT uX ∆pitqy 2 t = (∆pitqy − t=1 ) T

(15)

t=1

This measure also only includes idiosyncratic volatility. It is measured by the price rather than the return. I focus on this measure as there is a positive correlation between executive incentive levels and stock price. Table 2 includes the average for each of the four measures in each year between 1993 and 2009. For the basic measure, the quarterly standard deviation of the daily price, the measure increases 3.65 in 1993 to 5.66 in 2000 as documented in the previous literature. It then begins to fall to 2.28 in 2003. At which point it slowly increases to 4.03 at the height of the financial crisis in 2008. It drops back to about the 2006 level in 2009. This measure does not control for any market changes, and 15

only considers overall volatility of the price. After controlling for some market changes by looking at the the standard deviation of the price error, the pattern is the same. Each of the volatilities is smaller however. The volatilities are statistically different at the one-percent level from 1996 on. The standard deviation of the return is much smaller than the standard deviation of price, but it follows a very similar pattern until 2003. After 2003 it continues to decrease until 2007 before jumping in 2008. Again this measure does not control for any market changes. If we control for market changes by looking at the standard deviation of the return error, the volatility decreases dramatically. It is about a magnitude smaller in most years, but it still displays the same pattern as the standard deviation of the return. 4.2.2

Incentive Measures

The above volatility measures estimate the overall and idiosyncratic volatility of the stock price. As discussed above, the optimal choice of the volatility measure depends on the executive’s pay package. Therefore, I need a measure of how the executive’s take-home pay increases with either a dollar increase in stock price or a one-percent increase in the stock price. For many firms, the executive’s pay package includes a salary, stock grants, and stock options. Both the stock grants’ and the stock options’ value depend upon the stock price. Therefore a one-dollar increase in the stock price will increase both the value of the stock grants and the stock options. For each share that the executive owns, a one-dollar increase in the stock price will increase the value of the executive’s stock portfolio by one-dollar. The change in the value of an option for a dollar increase in the value of the underlying stock varies based on the time remaining for the option and how far the stock price is away from the strike price of the option if the option is valued using the Black-Scholes formula. Therefore, it does not have a dollar for dollar increase in the value of the option for each dollar increase in the stock price. One needs to combine both the increase from the stock grants and the increase in the value from the option grants. To accurately estimate this value, one needs information on each of the options the executive owns. This information includes the strike price and the time remaining. The yearly proxy statement that a public firm is required to submit to the Securities and Exchange Commission unfortunately does not include this information. It does include information on the value of the options that are in the money, value of the options that are out of the money, the number of options that are in the money, and the number of options that are out of the money. In addition it includes information on the stock options granted that year.

16

Core and Guay (2002a) developed a methodology for estimating the effect of a one-percent increase in the stock price on the value of an executive’s overall portfolio, called the sensitivity to price. I use this methodology to estimate the effects of a stock price change on the executives take-home pay. If the stock price is more volatile and the executive has a large sensitivity to price, the the executive’s take-home pay will be more volatile than the take-home pay of an executive with either a smaller stock price volatility or a smaller sensitivity to price. This measure captures a percent change in the stock price not a dollar increase in the stock price, so it may align best with volatility of returns rather than the volatility of price. However, if executives at firms with a high price are also executives who have larger stock and option grants, the volatility of price may still be the best measure. In addition to the effect of a change in the volatility of the stock on the volatility of the takehome pay, the volatility of the stock enters directly into the valuation of a stock option through the Black-Scholes formula. If a stock has a high volatility, it is more likely to have a day where it is valued far above the strike price. Therefore, the value of an option increases with the volatility of the stock. Core and Guay (2002a) also develop a methodology for estimating the effect of a 0.01 increase in the stock-return volatility on the valuation of all options in the executive’s portfolio, called the sensitivity to volatility. If an executive owns a large number of options relative to the number of share of stock, an increase in the volatility of the stock will increase the value of the executive’s portfolio more than an executive who has fewer options in his incentive portfolio. For much of the empirical section, the sensitivities are broken into quintiles for 2000 for the executive with the highest sensitivity at the firm. How does the sensitivity increase by quintile? In Table 3, the sensitivity to price increases from zero (the executive owns no stock or options) to 585,400. The first quintile has executives who do not have much sensitivity to price. A one-percent increase in the stock price would increase the average executive’s take-home pay by 15 dollars, and the executive with the largest effect would only have take-home pay increased by 40 dollars. Compared to average executive salaries, these numbers are very small. The next quintile includes executives whose take-home pay would increase from 40 dollars to 137 dollars. Again, these are not large relative to overall compensation. The average in the third quintile is still only 244 dollars. In the fourth quintile take-home pay begins to be affected more, hut the maximum is still only 1,161 dollars. The large effects are only found in the last quintile. Here a 1 percent increase in price increases the take-home by almost 10,000 dollars on average. This quintile has large incentive pay, and executives in it may act differently than the other executives. 17

Table 3 also includes information on the sensitivity to volatility. 507 of the firms do not include any options in their pay packages so the executive’s take-home pay is not influenced directly by volatility. Again, only a small fraction of the executives have a large increase in the value of takehome pay for a 0.01 increase in volatility. The bottom four quintiles are all less than 55 dollars. The top quintile ranges from 55 dollars to 4,536 dollars. The effects here are also pretty small compared to the effects for sensitivity to price. Given that there are two measures based on how the executives are paid, the sensitivity to price and sensitivity to volatility might be strongly correlated if options are driving the sensitivity to price. However, the correlation is only 0.03. The correlation between quintiles is 0.44. Table 4 includes the number of firms in each quintile for sensitivity to price broken down by what quintile of sensitivity to volatility they are in. There is some correlation in the table. There are no firms that have executives with the highest sensitivity to volatility but the lowest sensitivity to price. Given that firms that have executives with a high sensitivity to volatility also must have a large number of options, this is understandable. On the other hand, the firms who have executives with a high sensitivity to price are relatively spread out by volatility quintile for the first four quintiles. There are more in the fifth quintile as expected. Therefore, there is some correlation between the two but it is not a one-for-one movement.

5

Empirical Strategy and Results

The model presented in Section 3 leads to several propositions which can be tested empirically. Propositions 1 imply that a decrease in the dividend tax rate in 2003 would lead the volatility of all stocks to decrease after the tax change. One can look at this in the time series. However, there may be other confounding economic factors which are also moving at the same point in time. As discussed in Section 2, there is a large literature explaining the increase in firm volatility during the 1990s. None of these explanations are tax related. Therefore, there are important economic variables which can cause a change in volatility in the time series. In the following analysis, I include many of the factors discussed in the literature. However, there may be other non-tax reasons for a change in the time-series of idiosyncratic volatility which I am unable to control for. Therefore, in the following analysis, I will primarily focus on Propositions 2 and 4. Both Proposition 2 and Proposition 4 rely on differences in pre-reform incentive schemes. Proposition 2 states that a decrease in the dividend tax will cause a decrease in the volatility of the stock

18

that is most exaggerated for firms with executives who have larger incentives to begin with. Proposition 4 states that a decrease in the dividend tax will not change the incentive schemes differentially by original executive incentive level. To evaluate these propositions empirically, I use the measure of the sensitivity of executives stock and option holdings to price defined in Section 4.2.2. Following the methodology from Chetty and Saez (2010), for each firm I consider the five executives in the ExecuComp Database in 2000. I then compare the sensitivity to price for each of the five executives. I find the executive who has the largest sensitivity to price for the firm. The year 2000 was chosen because it was before President Bush was elected. President Bush was one of the major supporters of the dividend tax change and discussed it before pursuing the legislation. Therefore, the board of directors may not have taken into account the tax change by 2000. I am interested in the volatility of take-home pay for the executives. Take-home pay for the executives depends upon both the valuation of the stock and the valuation of options. Volatility depends on the stock price, and the volatility of take-home pay depends on the volatility of price. Therefore, the overall sensitivity of take-home pay to stock price is the optimal measure here. As in Chetty and Saez (2010), I consider quintiles of sensitivity to price. I group the firms into five groups based on the maximum sensitivity to price for all five executives in the ExecuComp database in 2000. Using these quintiles interacted with time leads to a difference-in-differences approach to test Propositions 4 and 2 from the model. This avoids several of the time-series identification problems as I can add time fixed effects. If there are time-series changes that are associated with the incentive scheme, this difference-in-differences approach will not be identified. In addition, if there are other contemporaneous effects, I will be unable to separate the two. In Section 5.3, I provide evidence that the timing of the effects occurs at the beginning of 2003, exactly at the tax change.

5.1

Volatility

What happened to stock volatility after the dividend tax decrease of 2003? Proposition 1 implies a decrease in the dividend tax rate in 2003 would lead to the volatility of all stocks decreasing. As discussed above, there are many time-series effects that influence the overall idiosyncratic volatility of all stocks. Proposition 2 states that the decrease in the volatility following a decrease in the tax rate on dividends should be largest in magnitude for firms which previously had a large incentive scheme. I test this proposition by examining the effect of the decrease in the dividend tax rate in 2003. Using the quintiles of highest sensitivity to price for the top five executives per firm in 2000, I have a measure of the incentive schemes for the executives with the largest incentive schemes. 19

Therefore, the firms with executives who are in Quintile 5 should experience a greater negative effect on volatility than the firms with executives in other quintiles. Quintile 4 should be more negative than Quintile 3 which should be more negative than Quintile 2 which should be more negative than Quintile 1. Therefore, one can look at the changes in volatility for the firms, and one can compare across quintiles. Section 4.2.1 describes four volatility measures. The standard deviation of the excess price is the optimal measure. Therefore, I focus on the standard deviation of price measure. First, however, I consider the log of the standard deviation of the stock price. This measure includes both idiosyncratic volatility and systematic volatility. It is a volatility that the executive is affected by, but it may not be the volatility that the executive has control over if the executive cannot control systematic volatility. Table 5 estimates the following equation:

log(sd(p)iqy ) =γi + β1 log(piqy ) + β2 Xiqy + ηd +

5 X

(β3n Dqy ∗ QPin ) +

n=1

5 X

(16)

(β4l Dqy ∗ QVil ) + iqy

l=1

where Xiqy are firm variables that the previous literature has indicated affect volatility, γi is a firm fixed effect, ηd is a date fixed effect, QPin is an indicator variable which is one if the firm is in price sensitivity quintile n, QVin is an indicator variable which is one if the firm is in volatility sensitivity quintile n, and Dqy is an indicator variable which is one if the year y is 2003 or later. The coefficients of interest is β3n where n defines the quintile. These are the coefficient on the interaction of the after tax change dummy and quintile n dummy. I add controls from the literature explaining the increase in firm volatility during the 1990s. I control for price here as the standard deviation of price should increase with the price of the stock. In Table 5, I find a positive and statistically significant coefficient on price but it is not one. This difference may be that volatility is largest for prices close to zero, ie firms closer to bankruptcy. The coefficient on price is statistically significant at the one-percent level. It is also statistically different than 1 at the one-percent level. Therefore, there does seem to be an increase in the volatility of price as prices increase, but it does not increase one-for-one. The previous literature showed that some of the increases in volatility over the 1990s were due to an increase in the number of smaller firms. Therefore, I include the log of assets and log of market capitalization. I find a statistically significant positive coefficient on assets in most specifications while a statistically 20

significant negative coefficient on market capitalization. Assets measures all tangible assets of the firm. This could be financed by either equity or debt. Market capitalization only includes equity. Therefore, I find the bigger the firm, the higher the volatility but the higher equity financed the lower the volatility. I may be finding a measure of leverage here. I also include the log of leverage directly. However, this has the opposite sign than expected. I find a negative and statistically significant coefficient. Theory would suggest that leverage should increase volatility as less cash flows are available to equity holders. Previous literature showed that the increase in leverage over the 1990s increased overall volatility during that time period. In the same vein, I find that market to book values are positive statistically significant. This may be due to newer firms who are riskier (ie high growth firms). I do not find an effect for profit margin when the other variables are included. Earnings per share is negatively associated with price as expected, but sales per share is positively related. If a firm has high sales per share but low earnings per share, the firm has low margins. This may increase volatility as the firm could be close to the break-even point often. As there may be many factors of the economy which change the overall level of volatility, I include date fixed effects. In addition, some firms or industries may naturally have higher volatility. Therefore, I include firm fixed effects.12 I include Government Fiscal Surplus as a measure of the overall level of the economy. I also include price-decile effects and allow the price-decile effects to change in 2003. These price-deciles are defined for the whole CRSP market, not just for the S&P 1500 to agree with the previous literature. I also include sensitivity to volatility quintiles times the tax change dummy. For executives who have a lot of options, the valuation of the take-home pay increases with an increase in volatility. Therefore, they have less of an incentive to decrease volatility if an executive has a large number of options. I find that the coefficient on Sensitivity to Volatility Quintile 5 times after tax dummy is positive and statistically significant as expected. The coefficients decrease as one moves to lower quintiles, as expected, and Quintile 2 is more negative than Quintile 1. Quintile 1 only includes firms that do not have executives with options. In the regressions, I use multiple samples. Column (1) always includes all firms. Column (2) includes only firms based in the United States. However, since I am looking at S&P 1500 firms this is most of the firms. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend 12

In results not shown in the paper I included industry-time effects. The results are similar to the Tables in the paper. I also considered a balanced panel of firm. I found similar results.

21

in the quarter being examined or increased the regular dividend more that 20% in the quarter. Regulated firms may behave differently or may have limits on dividends. Therefore, I separate the two regulated industries. Also, firms who change dividend policy during that quarter may have transition effects. Therefore, I exclude the relevant observations in Column (6). Proposition 2 states that the decrease in the volatility following a decrease in the tax rate on dividends should be largest in magnitude for firms which previously had a large incentive scheme. Therefore, the firms with executives who are in Quintile 5 should experience a greater negative effect on volatility than the firms with executives in other quintiles. Quintile 4 should be more negative than Quintile 3 which should be more negative than Quintile 2 which should be more negative than Quintile 1. I find a negative and statistically significant coefficient on Price Quintile 5 times After Tax Change. The coefficients on each of the consecutive quintiles is smaller in magnitude than the previous as predicted in Proposition 2. Therefore, the coefficients align correctly. For firms in industries which are not regulated, the p-value for the difference between Quintile 5 and Quintile 4 is 0.000. The p-value for the difference between Quintile 4 and Quintile 3 is 0.0147. The p-value for the difference between Quintile 3 and Quintile 2 is 0.0053. Therefore, Quintile 5 is statistically different than Quintile 4. Quintile 4 is statistically different than Quintile 3, and Quintile 3 is statistically different than Quintile 2. Quintile 1 and 2 are not statistically different. However, I do not find a statistically significant decrease for firms in the financial industry. Firms in the utility industry behave similarly to firms in non-regulated industries. Firms who do not change dividend policy have very similar results as all firms (0.316 vs 0.318). Therefore, I find evidence that the volatility of all non-regulated firms and utilities decreased for firms with the highest incentives after the dividend tax change. The second measure I consider is the log of the standard deviation of the stock price error from a Fama-French Three-Factor Model. This measure only includes idiosyncratic volatility. It is therefore only part of the volatility that the executive is affected by. However, it may be the volatility that the executive has control over if the executive cannot control systematic volatility. Table 6 estimates the following equation:

log(sd(∆p)iqy ) =γi + β1 log(piqy ) + β2 Xiqy + ηd +

5 X

(β3n Dqy ∗ QPin ) +

n=1

5 X (β4l Dqy ∗ QVil ) + iqy l=1

22

(17)

where Xiqy are firm variables that previous literature has indicated affect volatility, γi is a firm fixed effect, ηd is a date fixed effect, QPin is an indicator variable which is one if the firm is in price sensitivity quintile n, QVin is an indicator variable which is one if the firm is in volatility sensitivity quintile n, and Dqy is an indicator variable which is one if the year y is 2003 or later. The coefficients of interest is β3n where n defines the quintile. These are the coefficients on the interaction of the after tax change dummy and quintile n dummy. As in Table 5, I find a negative and statistically significant coefficient on Price Quintile 5 times After Tax Change. The coefficients on each of the consecutive quintiles is smaller than the previous as predicted in Proposition 2. The coefficients are however smaller in magnitude that the coefficients on overall price volatility. This makes sense given the standard deviation of the price error is smaller on average than the volatility of price. In addition, the coefficient for financial firms is now significant at the ten-percent level. Firms in the utility industry behave similarly to firms in non-regulated industries. Firms who do not change dividend policy have very similar results as all firms (0.296 vs 0.297). Therefore, I find evidence that the volatility of firms with the highest incentives decreased after the dividend tax change. The control variables are similar in statistical significance to the overall price volatility measure. Therefore, the volatility of the price and the volatility of the price error yield similar results. Both support Proposition 2 to show that volatility decreased most for executives who had large sensitivity to price in their contracts. Therefore, evidence from the optimal measure supports Proposition 2. I also consider two measures of the volatility of returns. While they are not the optimal measure, they many provide further evidence in support of Proposition 2. The third measure I consider is the log of the standard deviation of the stock price return. This measure includes both the idiosyncratic volatility and the systematic volatility. The executive is affected by this volatility as he is affected by prices. However, it may not be the volatility that the executive has control over if the executive cannot control systematic volatility. Table 7 estimates the following equation:

log(sd(R)iqy ) =γi + β1 log(piqy ) + β2 Xiqy + ηd +

5 X

(β3n Dqy ∗ QPin ) +

n=1

(18)

5 X (β4l Dqy ∗ QVil ) + iqy l=1

where Xiqy are firm variables that previous literature has indicated affect volatility, γi is a firm fixed effect, ηd is a date fixed effect, QPin is an indicator variable which is one if the firm is in 23

price sensitivity quintile n, QVin is an indicator variable which is one if the firm is in volatility sensitivity quintile n, and Dqy is an indicator variable which is one if the year y is 2003 or later. The coefficients of interest is β3n where n defines the quintile. These are the coefficients on the interaction of the after tax change dummy and quintile n dummy. Unlike the previous two regressions, this looks at the volatility of the return; therefore, it is not the optimal measure in the current economic environment. Like the previous two tables, I find negative and statistically significant coefficients on Price Quintile 5 times After Tax Change. The coefficients decrease in magnitude and significance as the quintile decreases. This is similar to the price volatilities. The coefficient is not statistically significant for financial firms. I also found a not statistically significant coefficient for financial firms in the price volatility regressions. Most of the controls have similar significance and sign. The log of price is now close to zero and not statistically significant. Therefore, firms with bigger prices do not have either larger or smaller standard deviation of return once you add the other controls. Overall the results are similar to the price regressions. Therefore, both the price and return overall volatilities decreased after the tax cut for firms with the largest incentives. This provides further evidence in support of Proposition 2. The last measure I consider is the log of the standard deviation of the stock price return error from a Fama-French Three-Factor Model. This measure includes only the idiosyncratic volatility, and is only part of the volatility that the executive is affected by as the executive is also affected by the systematic volatility and price levels. However, it may be the volatility that the executive has control over if the executive cannot control systematic volatility. Table 8 estimates the following equation:

log(sd(ˆ )iqy ) =γi + β1 log(piqy ) + β2 Xiqy + ηd +

5 X

(β3n Dqy ∗ QPin ) +

n=1

(19)

5 X (β4l Dqy ∗ QVil ) + iqy l=1

where Xiqy are firm variables that previous literature has indicated affect volatility, γi is a firm fixed effect, ηd is a date fixed effect, QPin is an indicator variable which is one if the firm is in price sensitivity quintile n, QVin is an indicator variable which is one if the firm is in volatility sensitivity quintile n, and Dqy is an indicator variable which is one if the year y is 2003 or later. The coefficients of interest is β3n where n defines the quintile. These are the coefficient on the 24

interaction of the after tax change dummy and quintile n dummy. Like the previous three tables, I find negative and statistically significant coefficients on Price Quintile 5 times After Tax Change. The coefficients decrease in magnitude and significance as the quintiles decrease. This is similar to the results for price volatilities. The coefficient is not statistically significant for financial firms as in the price volatility equation. Most of the controls have similar significance and sign. The log of price is now negative and statistically significant. This negative coefficient is expected if firms with close to zero price are closer to bankruptcy. Overall the results are similar to the price regressions and the overall return regression. Therefore, when looking at idiosyncratic volatility, both the price and return decreased in volatility after the tax cut for firms with the largest incentives. All four sets of regressions show that firms with larger incentives had the largest decrease in volatility. This finding agrees with Proposition 2. It appears that after the tax change, volatility decreased most for firms with executives who had the largest incentives. Agency costs associated with volatility might be important to consider when deciding on optimal tax policy.

5.2

Incentive Changes

What happened to incentives schemes after the dividend tax decrease of 2003? Proposition 3 states a decrease in the dividend tax rate in 2003 would not change the incentive schemes for all executives. As discussed above, there are many time-series effects that influence the overall incentive level of executives. Proposition 4 is more testable. I consider the effect of the decrease in the dividend tax rate in 2003. Proposition 4 states that any change in the incentive schemes following a decrease in the tax rate on dividends should be independent of the previous incentive scheme. Using the quintiles of highest sensitivity to price for the top five executives in 2000, I have a measure of the incentive schemes for the executives from each firm with the largest incentive schemes. Therefore, the firms with executives who are in Quintile 5 should have a similar change in incentive packages to other quintiles. Quintile 4 should be similar to Quintile 3 which should be similar Quintile 2 which should be similar Quintile 1. Therefore, one can look at the changes in incentives and compare across quintiles. Proposition 4 discusses changes in incentive pay schemes due to tax changes. There are two measures of incentive pay in Core and Guay (2002a) that are applicable here. The first measure is the sensitivity of pay to price changes for new stock and option grants. The second measure is the sensitivity of pay to price changes for all outstanding stock and option grants. As Proposition 4 25

considers changes to the incentive pay, the changes in grants might be the most applicable. Table 9 estimates:

log(SP Giy ) =γi + β1 log(piy ) + β2 Xiy + ηd +

5 X

(β3n Dy ∗ QPin ) +

n=1

(20)

5 X (β4l Dy ∗ QVil ) + iy l=1

where SP Giy is the largest sensitivity to price for new grants for a top five executive at firm i in year y. I add the same controls as in the volatility tables. I am controlling for price as higher priced stocks may have higher (or lower incentives). I do not find a statistically significant relationship with new grants of stock. I also do not find one for assets. However larger firms as measured by market capitalization have higher grants. This result is expected and consistent with the literature. Larger firms tend to have higher sensitivity to price. The other control which has a statistically significant relationship is the Log of Sales per share. This may also be a measure of the size of the company. No other control variable is statistically significant in more than one sample. As there may be many factors of the economy which change new stock and option grants, I include date fixed effects. In addition, some firms or industries may naturally have higher sensitivity to price. Therefore, I include firm fixed effects.13 I also include price-decile effects and allow the price-decile effects to change in 2003. These price-deciles are defined for the whole CRSP market, not just for the S&P 1500 to agree with the previous literature. I also include sensitivity to volatility quintiles times the tax change dummy. The coefficient on Quint 5 times After Tax Change is statistically significant and negative in Columns (1), (2), (4), and (6). It has the largest magnitude for utilities. For the firms in non regulated industries, the coefficient is negative, but much smaller in magnitude and not statistically significant. Based on Proposition 4, one would expect all the coefficients to be zero. Given that in non regulated industries the coefficient is not statistically significant, this is the expected result. Grants are only one part of compensation. If shareholders and the board of directors increase grants to give a certain overall level of incentives, then overall sensitivity to price may be the variable to consider. It is important to therefore see the effects on overall sensitivity to price. Table 13

In results not shown in the paper I included industry-time effects. The results are similar to the Tables in the paper.

26

10 estimates the following equation: log(SP Aiy ) =γi + β1 log(piy ) + β2 Xiy + ηd +

5 X

(β3n Dy ∗ QPin ) +

n=1

5 X

(21)

(β4l Dy ∗ QVil ) + iy

l=1

where SP Aiy is the largest sensitivity to price for all stocks and options for a top five executive at firm i in year y. If the shareholders are trying to keep the executives at the same level of incentives, Proposition 4 would lead to an expectation of coefficient on Quint 5 * After Tax Change to be zero. However, if shareholders use new grants to incentivize executives and all options and stock are a vestige of history and market changes (as in the case when the market has declined so no one sells), the coefficient might not be zero. In Table 10, I find negative and statistically significant coefficients for all specifications. Unlike the grants, the sensitivity to all shares and options has the smallest and least statistically significant coefficient for firms in the utility industry. Therefore, there may have been some change at this point for the utility industry which led executives to have new grants and buy out shares and options. However, for the other industries, I find results inconsistent with Proposition 4. Quintile 5 has the most negative result, followed by Quintile 4 and then Quintile 3. Quintile 2 is the least negative coefficient with Quintile 1 as the omitted quintile. The sensitivity of all stock and option to price does not provide evidence in support of Proposition 4. If shareholders are trying to set an optimal level of overall incentives this is the important measure to consider. However, it might lead executives to sell shares and options, but executives send a negative signal to the market when they sell shares. Therefore, the sensitivity of stock and options to price may not be a large concern.

5.3

Robustness Checks

As the first robustness check, I include an earnings quality measure as defined in Dechow and Dichev (2002). I include this as a robustness check rather than as the original specification since I am only able to define this variable for two-thirds of the overall sample, and so the sample is much smaller. There are also very few financial firms for which I am able to define this variable. I find very similar results for all Tables. For the stock and option grant table, I find less significant results. However, the significance returns for all stock and options. As the second robustness check, I allow the timing to vary. I look at the effect on each quarter 27

from 7 quarters before January 2003 to 7 quarters after. At that point I assume it has stabilized. I primarily consider the effect for the 5th Quintile, and only include the results for the 5th Quintile in the Tables in Appendix A as there are too many variables for a single page. For the log standard deviation of stock price, the effect seems to take place in Quarter 1 where Quarter 0 is defined as the quarter ending in either January, February or March of 2003 depending on the firm fiscal year. The coefficient is negative and statistically significant from Quarter 1 on. It increases in Quarter 3 in magnitude and stabilizes there. From this, it appears that the second quarter of 2003 is the important quarter for the stock price volatility. If I only consider the idiosyncratic price volatility, the pattern remains the same. The second quarter of 2003 is the quarter with the change. Therefore, I find clear timing for the optimal measure of volatility The results are murkier when you consider the standard deviation of stock return. Here, it appears the first effects enter in quarter 5, the second quarter of 2004 and increases in magnitude over time. However, once you look at the return error, the effects re-enter in quarter 1. Here, though, we also have some negative and statistically significant results in quarters before the start of 2003. For price volatility there is no clear pattern before 2003. Although the coefficients on return are statistically significant, the exact timing is not as clear. These measures are not the optimal measure given the positive correlation between incentives and stock price. If one looks at the timing of the effects for sensitivity to price for new stock and option grants, there is no clear pattern. This is not unexpected since I would expect no effect here. The timing of overall stock and option price sensitivity seems to begin in quarter -1. The sensitivity to price data is annual rather than quarterly. Therefore, the decrease in the last quarter of 2002. The timing for the level of incentives appears to start before President Bush announced his tax plan. In addition, many executive contracts are for several years. Therefore, they take several years to adjust. The timing should be later not earlier. Therefore, the changes to incentive packages may not be due to the tax change. Next, to consider how incentives changed over time for the fifth quintile, I consider the yearly interaction by quintile. I find that starting in 2002, incentives began to fall. They fell every year through 2007 for the fifth quintile relative to the first quintile. Therefore, other things may be occurring that are changing incentives at the top during this time period. This is consistent with Proposition 4. As a third robustness check, I compare firms that issue dividends to firms that never issue dividends. I would not expect the firms who do not issue dividends to have any effect of the tax change as they are not affected by the tax. Therefore, I take a triple difference approach. I find the 28

volatility results hold only for firms who issue dividends at some point. The coefficients on Quintile 5 times after the tax times the firm has issued dividends is negative and statistically significant for all measures of variance. The same is true for Quintile 4 and 3, but they are smaller in magnitude than Quintile 5. This is the expected result. I do not find statistically significant coefficients on Quintile 5 times after tax. Therefore, the volatility results hold for firms who issue dividends but not for firms who do not issue dividends. These are the expected results. I also compare firms who do not issue dividends to firms who do when I consider the changes to incentives. I find the negative affects here do not differentially affect firms who pay dividends. Therefore, it seems that something other than the dividend tax cut is affecting the overall level of incentives. This is consistent with Proposition 4.

6

Conclusion

I analyze stock volatility for firms before and after the dividend tax cut in January of 2003. I find that stock volatility decreased most for firms with highly incentivized executives. The results are the strongest and have the clearest timing when using the idiosyncratic volatility of stock price. This price measure of volatility is the optimal measure if overall take-home incentive pay for an executive is strongly correlated with stock price. This correlation exists when there is no strong negative correlation between the stock price and the number of shares and stock options the executive receives. In the data I find that there is a positive correlation between overall take-home pay and stock price which supports using this measure. There are two pieces of volatility, the systematic volatility and the idiosyncratic volatility. An executive may have little control over systematic risk, but he may be able to exert control over the firm-specific risk. Idiosyncratic volatility is the measure of volatility that an executive is most likely able to control. I find that the idiosyncratic stock price volatility decreased most for executives who have large incentive schemes. These executives are the ones whose take-home pay is the most affected by the price volatility. In traditional models, where an executive does not control volatility, the volatility change should be constant for all levels of incentive pay. Often when we talk about changing a tax, we only talk about the effect on the mean. We do not often talk about the effect on the variance. However, when discussing corporate taxes, the variance can be important, especially in a world with agency costs due to hiring of a manager. If the manager is risk-averse, he will be negatively affected by the variance. If he has a high level of incentives,

29

he will want to take actions to reduce the variance or decrease his take-home pay’s dependence on the stock price. By decreasing his dependence on take-home incentive pay, he also has less of an incentive to increase firm value, thereby increasing agency costs. If decreasing variance can only be done in a firm value destructing way, then this will also lead to larger agency costs. My findings are in line with an agency cost model. I find the tax cut decreased variance unevenly across firms. Therefore, corporate taxes may affect the volatility of some firms more than others. Future work in optimal tax policy should consider the agency costs associated with changes to the variance of a stock price. Changes to personal taxes as well as corporate taxes could affect the agency costs. Therefore, the effects on the variance of a stock should be included in tax-code optimization.

30

References Aggarwal, Rajesh and Andrew A. Samwick, “The Other Side of the Trade-Off: The Impact of Risk on Executive Compensation,” Journal of Political Economy, February 1999, 107 (1), 65–105. Allen, Franklin and Roni Michaely, “Payout Policy,” in George Constantinides, Milton Harris, and Rene Stulz, eds., Handbook of the Economics of Finance, North-Holland, 2003. Amromin, Gene, Paul Harrison, and Steven Sharpe, “How Did the 2003 Dividend Tax Cut Affect Stock Prices?,” Financial Management, Winter 2008, 37 (4), 625–646. Auerbach, Alan J., “Taxation and Corporate Financial Policy,” in Alan J. Auerbach and Martin Feldstein, eds., Handbook of Public Economics, Vol. 3, Elsevier Science B.V., 2002, chapter 19, pp. 1251–1292. and Kevin A. Hassett, “The 2003 Dividend Tax Cuts and the Value of the Firm: An Event Study,” NBER Working Paper 11449, June 2005. Auten, Gerald, Robert Carroll, and Geoffrey Gee, “The 2001 and 2003 Tax Rate Reductions: An Overview and Estimate of the Taxable Income Response,” National Tax Journal, September 2008, 61 (3), 345–364. Bennett, James A., Richard W. Sias, and Laura T. Starks, “Greener Pastures and the Impact of Dynamic Institutional Preferences,” The Review of Financial Studies, Winter 2003, 16 (4), 1203–1238. Bernheim, B. Douglas and Adam Wantz, “A Tax-Based Test of the Dividend Signaling Hypothesis,” The American Economic Review, June 1995, 85 (3), 532–551. and Lee S. Redding, “Optimal Money Burning: Theory and Application to Corporate Dividends,” Journal of Economics and Management Strategy, Winter 2001, 10 (4), 463–507. Blouin, Jennifer L., Jana Smith Raedy, and Douglas A. Shackelford, “Did Firms Substitute Dividends for Share Repurchases After the 2003 Reductions in Shareholder Tax Rates?,” NBER Working Paper 13601, November 2007. Brandt, Michael W., Alon Brav, John R. Graham, and Alok Kumar, “The Idiosyncratic Volatility Puzzle: Time Trend or Speculative Episodes?,” The Review of Financial Studies, February 2010, 23 (2), 863–899. Brown, Gregory and Nishad Kapadia, “Firm-Specific Risk and Equity Market Development,” Journal of Financial Economics, May 2007, 84 (2), 358–388. Brown, Jeffrey R., Nellie Liang, and Scott Weisbenner, “Executive Financial Incentives and Payout Policy: Firm Responses to the 2003 Dividend Tax Cut,” The Journal of Finance, August 2007, 62 (4), 1935–1965. Campbell, John Y., Martin Lettau, Burton G. Malkiel, and Yexiao Xu, “Have Individual Stocks Become More Volatile? An Empirical Exploration of Idiosyncratic Risk,” The Journal of Finance, February 2001, 56 (1), 1–43.

31

Chava, Sudheer and Amiyatosh Purnanandam, “CEOs versus CFOs: Inventives and Corporate Policies,” Journal of Financial Economics, August 2010, 97 (2), 263–278. Chetty, Raj and Emmanuel Saez, “Dividend Taxes and Corporate Behavior: Evidence from the 2003 Tax Cut,” Quarterly Journal of Economics, August 2005, 120 (3), 791–833. and , “Dividend and Corporate Taxation in an Agency Model of the Firm,” American Economic Journal: Economic Policy, August 2010, 2 (3), 1–31. Christie, William G. and Vikram Nanda, “Free Cash Flow, Shareholder Value, and the Undistributed Profits Tax of 1936 and 1937,” The Journal of Finance, December 1994, 49 (5), 1727–1754. Coles, Jeffrey L., Naveen D. Daniel, and Lalitha Naveen, “Managerial Incentives and Risk-Taking,” Journal of Financial Economics, February 2006, 79 (2), 431–468. Comin, Diego A. and Thomas Philippon, “The Rise in Firm-Level Volatility: Causes and Consequences,” NBER Macroeconomics Annual, 2005. Core, John E. and Wayne R. Guay, “The Use of Equity Grants to Manage Optimal Equity Incentive Levels,” Journal of Accounting and Economics, December 1999, 28 (2), 151–184. and , “Estimating the Value of Employee Stock Option Portforlios and Their Sensitivities to Price and Volatility,” Journal of Accounting Research, June 2002, 40 (3), 613–630. and , “The Other Side of the Trade-Off: The Impact of Risk on Executive Compensation - A Revised Comment,” 2002. Working Paper. , , and David F. Larcker, “Executive Equity Compensation and Incentives: A Survey,” Federal Reserve Bank of New York Economic Policy Review, April 2003, pp. 27–50. Dechow, Patricia M. and Ilia D. Dichev, “The Quality of Accruals and Earnings: The Role of Accrual Estimation Errors,” The Accounting Review, 2002, 77 (s-1), 35–59. Fama, Eugene F. and Kenneth R. French, “Common Risk Factors in the Returns on Stocks and Bonds,” Journal of Financial Economics, February 1993, 33 (1), 3–56. Fenn, George W. and Nellie Liang, “Corporate Payout Policy and Managerial Stock Incentives,” Journal of Financial Economics, April 2001, 60 (1), 45–72. Fink, Jason, Kristin E. Fink, Gustavo Grullon, and James P. Weston, “What Drove the Increase in Idiosyncratic Volatility during the Internet Boom?,” Journal of Financial and Quantitative Analysis, October 2010, 45 (5), 1253–1278. Frydman, Carola and Dirk Jenter, “CEO Compensation,” Annual Review of Financial Economics, December 2010, 2 (1), 75–102. Gordon, Roger and Martin Dietz, “Dividend and Taxes,” NBER Working Paper 12292, June 2006. Graham, John R., “Debt and the Marginal Tax Rate,” Journal of Financial Economics, May 1996, 41 (1), 41–74.

32

, “How Big are the Tax Benefits of Debt,” The Journal of Finance, October 2000, 55 (5), 1901–1941. , “Taxes and Corporate Finance: A Review,” The Review of Financial Studies, Winter 2003, 16 (4), 1075–1129. Hall, Brian H. and Jeffrey B. Liebman, “The Taxation of Executive Compensation,” Tax Policy and the Economy, 2000, 14, 1–44. Hirshleifer, David and Anjan V. Thakor, “Managerial Conservatism, Project Choice, and Debt,” The Review of Financial Studies, Fall 1992, 5 (3), 437–470. and Yoon Suh, “Risk, Managerial Effort, and Project Choice,” Journal of Financial Intermediations, September 1992, 2 (3), 308–345. Irvine, Paul J. and Jeffrey Pontiff, “Idiosyncratic Return Volatility, Cash Flows, and Product Market Competition,” The Review of Financial Studies, March 2009, 22 (3), 1149–1177. Jensen, Michael C. and William H. Meckling, “Theory of the Firm: Managerial Behavior, Agency Costs, and Ownership Structure,” Journal of Financial Economics, October 1976, 3 (4), 305–360. Jin, Li, “CEO Compensation, Diversification, and Incentives,” Journal of Financial Economics, October 2002, 66 (1), 29–63. Knopf, John D., Jouahn Nam, and John H. Thorton Jr., “The Volatility and Price Sensitivites of Managerial Stock Option Portfolios and Corporate Hedging,” The Journal of Finance, April 2002, 57 (2), 801–813. Korinek, Anton and Joseph E. Stiglitz, “Dividend Taxation and Intertemporal Tax Arbitrage,” Journal of Public Economics, February 2009, 93 (1-2), 142–159. La Porta, Rafael, Florencio Lopez de Silanes, Andrei Schleifer, and Robert W. Vishny, “Agency Problems and Dividend Policies around the World,” The Journal of Finance, February 2000, 55 (1), 1–33. Lambert, Richard A., “Executive Effort and Selection of Risky Projects,” RAND Journal of Economics, Spring 1986, 17 (1), 77–88. MacKie-Mason, Jeffrey K., “Do Taxes Affect Corporate Financing Decisions?,” The Journal of Finance, December 1990, 45 (5), 1471–1493. Malmendier, Ulrike and Geoffrey Tate, “Superstar CEOs,” Quarterly Journal of Economics, November 2009, 124 (4), 1593–1638. Miller, Merton H. and Franco Modigliani, “Dividend Policy, Growth, and the Valuation of Shares,” The Journal of Business, October 1961, 34 (4), 411–433. Modigliani, Franco and Merton H. Miller, “The Cost of Capital, Corporation Finance and the Theory of Investment,” The American Economic Review, June 1958, 48 (3), 261–297. and , “Corporate Income Taxes and the Cost of Capital: A Correction (in Communications),” The American Economic Association, June 1963, 53 (3), 433–443. 33

Murphy, Kevin J., “Executive Compensation,” in Orley C. Ashenfelter and David Card, eds., Handbook of Labor Economics, Vol. 3B, Elsevier Science B.V., 1999, chapter 38, pp. 2485–2566. Nam, Jouahn, Jun Wang, and Ge Zhang, “The Impact of Dividend Tax Cut and Managerial Stock Holdings on Firm’s Dividend Policy,” EFMA 2004 Basel Meetings Paper, January 2004. Prendergast, Canice, “The Tenuous Trade-off between Risk and Incentives,” Journal of Political Economy, October 2002, 110 (5), 1071–1102. Rajgopal, Shiva and Mohan Venkatachalam, “Financial Reporting Quality and Idiosyncratic Return Volatility,” Journal of Accounting and Economics, February 2011, 51 (1-2), 1–20. Stein, Jeremy C., “Agency, Information, and Corporate Investment,” NBER Working Paper 8342, June 2001. Tirole, Jean, The Theory of Corporate Finance, Princeton, New Jersey: Princeton University Press, 2006. Wei, Steven X. and Chu Zhang, “Why did Individual Stocks Become More Volatile,” The Journal of Business, January 2006, 79 (1), 259–292. Xu, Yexiao and Burton G. Malkiel, “Investigating the Behavior of Idiosyncratic Volatility,” The Journal of Business, October 2003, 76 (4), 613–645.

34

7

Figures and Tables Table 1: Firm Summary Statistics

Price Total Assets Market Cap Profit Margin Earnings per Share Sales per Share Return on Assets Market to Book Leverage Ratio Government Surplus Earnings Quality

Count

Mean

Std. Dev.

Min

Max

56,476 56,475 56,468 51,654 54,684 52,260 52,218 56,464 56,047 56,476 35,636

33.77 16,640 7,704 0.06 0.49 23.32 0.02 1.86 0.20 -1.53 1.24E+6

32.48 85,730 23,571 0.16 1.22 34.96 0.08 1.53 0.18 2.52 8.14E+6

0.14 3.11 2.66 -1.00 -0.99 0.00 -0.98 0.30 0.00 -8.30 0.00E+0

983.02 2,358,266 571,614 7.27 196.57 1178.47 2.17 74.46 4.39 2.70 2.29E+8

Table 2: Volatility Measures Summary Statistics Year

Std Dev Price

Std Dev Price Error

Std Dev Return

Std Dev Error Return

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

3.65 3.059 3.783 2.703 3.355 4.011 4.224 5.66 3.399 2.69 2.278 2.449 2.635 2.805 3.073 4.03 2.821

2.951 2.302 2.942 2.131 2.68 3.252 3.348 4.234 2.672 2.112 1.744 1.929 2.102 2.228 2.444 2.904 2.176

0.16 0.163 0.165 0.179 0.201 0.274 0.269 0.381 0.284 0.323 0.23 0.169 0.168 0.167 0.182 0.525 0.358

0.02 0.0202 0.0202 0.0215 0.0232 0.0269 0.0313 0.0369 0.0316 0.0284 0.0213 0.0186 0.0185 0.0188 0.0183 0.0311 0.0296

35

Table 3: Sensitivity Measures Summary Statistics Sensitivity to Price by Sensitivity to Price Qntiles Quant 1 2 3 4 5

Count

Mean

Std Dev

Min

Max

293 293 293 293 293

14.94 85.13 244.28 688.39 9,783.45

11.48 28.13 67.17 220.42 42,167.10

0.00 40.58 137.74 377.17 1,164.54

39.64 137.52 371.03 1,161.54 585,399.80

Sensitivity to Volatility by Sensitivity to Vol Qntiles Quant 1 2 3 4 5

Count 507 79 293 293 293

Mean 0.00 0.29 6.79 31.26 184.50

Std Dev 0.00 0.31 4.10 12.04 297.22

Min 0.00 1.05E-30 1.04 14.61 54.54

Max 0.00 0.99 14.52 54.50 4,536.16

Table 4: Sensitivity Quantile Summary Statistics Price 1 2 3 4 5 Total

1 219 108 64 59 57 507

2 15 18 15 16 15 79

Volatility 3 4 55 4 63 87 66 81 50 67 59 54 293 293

36

5 0 17 67 101 108 293

Total 293 293 293 293 293 1,465

Table 5: Log Standard Deviation of Stock Price

Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assets Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Vol Qnt 2 * After Tax Vol Qnt 3 * After Tax Vol Qnt 4 * After Tax Vol Qnt 5 * After Tax Govt Fiscal Surplus Constant

(1)

(2)

(3)

(4)

(5)

(6)

-0.316*** (0.0372) -0.176*** (0.0360) -0.0925*** (0.0339) -0.00674 (0.0333) 0.630*** (0.0278) 0.224*** (0.0249) -0.127*** (0.0265) 6.17e-05 (0.0227) -0.0368*** (0.00879) 0.117*** (0.0179) -0.0994 (0.0612) 0.800*** (0.0374) -0.174*** (0.0600) -0.112** (0.0536) 0.0834*** (0.0302) 0.127*** (0.0275) 0.146*** (0.0278) 0.675*** (0.0770) -0.0190 (0.421)

-0.315*** (0.0376) -0.177*** (0.0366) -0.0896*** (0.0341) -0.00484 (0.0335) 0.635*** (0.0281) 0.223*** (0.0252) -0.125*** (0.0268) 0.00117 (0.0231) -0.0390*** (0.00890) 0.116*** (0.0182) -0.101 (0.0616) 0.793*** (0.0376) -0.172*** (0.0604) -0.111** (0.0537) 0.0847*** (0.0305) 0.132*** (0.0278) 0.145*** (0.0283) 0.674*** (0.0765) -0.00737 (0.420)

-0.214 (0.153) -0.122 (0.148) -0.0478 (0.147) -0.115 (0.152) 0.553*** (0.0550) 0.0695 (0.0501) 0.0139 (0.0552) 0.0137 (0.0627) -0.0737*** (0.0270) 0.168*** (0.0435) -0.288 (0.349) 1.059*** (0.174) -0.142 (0.149) -0.163 (0.0988) -0.0713 (0.0865) -0.136 (0.0830) -0.152* (0.0910) 0.0132* (0.00710) -3.890*** (0.312)

-0.345*** (0.113) -0.200* (0.103) -0.205** (0.0882) -0.00142 (0.0751) 0.693*** (0.101) 0.470*** (0.0883) -0.434*** (0.0993) -0.448** (0.204) 0.0332 (0.0246) -0.0465 (0.0366) 2.249* (1.221) 2.358*** (0.303) -0.836*** (0.267)

-0.330*** (0.0414) -0.176*** (0.0404) -0.0999** (0.0389) -0.00920 (0.0377) 0.628*** (0.0331) 0.240*** (0.0297) -0.139*** (0.0321) 0.0122 (0.0238) -0.0331*** (0.00982) 0.126*** (0.0207) -0.131** (0.0622) 0.796*** (0.0434) -0.173*** (0.0657) -0.139** (0.0629) 0.0839** (0.0360) 0.162*** (0.0309) 0.184*** (0.0320) 0.660*** (0.0958) -0.0561 (0.509)

-0.318*** (0.0370) -0.176*** (0.0358) -0.0961*** (0.0336) -0.00493 (0.0331) 0.621*** (0.0279) 0.220*** (0.0246) -0.122*** (0.0263) 0.00403 (0.0227) -0.0387*** (0.00880) 0.118*** (0.0177) -0.0913 (0.0607) 0.794*** (0.0374) -0.159*** (0.0590) -0.114** (0.0544) 0.0858*** (0.0301) 0.132*** (0.0278) 0.146*** (0.0278) 0.671*** (0.0770) 0.0163 (0.420)

0.00637 (0.0888) 0.0215 (0.0954) 0.0500 (0.0948) -0.0120 (0.00977) -2.972*** (0.452)

Observations 49,846 48,920 7,024 4,189 38,633 48,796 R-squared 0.462 0.460 0.421 0.482 0.485 0.463 Number of Firms 1,300 1,279 164 82 1,054 1,300 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

37

Table 6: Log Standard Deviation of Stock Price Error

Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assets Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Vol Qnt 2 * After Tax Vol Qnt 3 * After Tax Vol Qnt 4 * After Tax Vol Qnt 5 * After Tax Govt Fiscal Surplus Constant

(1)

(2)

(3)

(4)

(5)

(6)

-0.296*** (0.0356) -0.170*** (0.0342) -0.0959*** (0.0322) -0.0181 (0.0317) 0.694*** (0.0263) 0.261*** (0.0235) -0.179*** (0.0246) -0.0104 (0.0223) -0.0378*** (0.00811) 0.0602*** (0.0163) -0.119** (0.0573) 0.788*** (0.0351) -0.161*** (0.0552) -0.108** (0.0492) 0.0791*** (0.0287) 0.114*** (0.0259) 0.131*** (0.0260) 0.212 (0.216) -2.348** (1.084)

-0.295*** (0.0359) -0.170*** (0.0347) -0.0931*** (0.0324) -0.0153 (0.0319) 0.700*** (0.0266) 0.259*** (0.0239) -0.178*** (0.0249) -0.0102 (0.0228) -0.0391*** (0.00821) 0.0591*** (0.0166) -0.120** (0.0575) 0.781*** (0.0353) -0.160*** (0.0555) -0.107** (0.0494) 0.0807*** (0.0289) 0.120*** (0.0263) 0.130*** (0.0264) 0.212 (0.216) -2.335** (1.087)

-0.225* (0.131) -0.124 (0.126) -0.0791 (0.125) -0.118 (0.130) 0.586*** (0.0522) 0.0886* (0.0467) -0.0119 (0.0484) 0.00332 (0.0611) -0.0839*** (0.0264) 0.136*** (0.0386) -0.312 (0.344) 0.984*** (0.172) -0.155 (0.145) -0.163* (0.0962) -0.0634 (0.0843) -0.114 (0.0837) -0.136 (0.0901) 0.0358*** (0.00647) -3.803*** (0.311)

-0.291*** (0.104) -0.179* (0.0981) -0.185** (0.0802) 0.00732 (0.0693) 0.787*** (0.0914) 0.509*** (0.0812) -0.479*** (0.0902) -0.419** (0.192) -0.0117 (0.0227) -0.0724** (0.0342) 2.349** (1.127) 2.251*** (0.271) -0.756*** (0.242)

-0.306*** (0.0397) -0.172*** (0.0382) -0.0996*** (0.0370) -0.0215 (0.0359) 0.695*** (0.0312) 0.276*** (0.0280) -0.194*** (0.0298) 0.00389 (0.0234) -0.0289*** (0.00895) 0.0635*** (0.0191) -0.160*** (0.0583) 0.782*** (0.0407) -0.156*** (0.0600) -0.125** (0.0583) 0.0807** (0.0343) 0.146*** (0.0293) 0.168*** (0.0297) 0.211 (0.196) -2.286** (0.986)

-0.297*** (0.0352) -0.169*** (0.0337) -0.0988*** (0.0318) -0.0157 (0.0314) 0.689*** (0.0264) 0.258*** (0.0234) -0.177*** (0.0245) -0.00624 (0.0224) -0.0395*** (0.00810) 0.0610*** (0.0162) -0.113** (0.0570) 0.782*** (0.0349) -0.147*** (0.0546) -0.108** (0.0501) 0.0816*** (0.0284) 0.118*** (0.0261) 0.133*** (0.0260) 0.209 (0.216) -2.309** (1.086)

0.0478 (0.0749) 0.0339 (0.0792) 0.0679 (0.0781) 0.0244*** (0.00855) -3.239*** (0.422)

Observations 49,846 48,920 7,024 4,189 38,633 48,796 R-squared 0.465 0.463 0.415 0.483 0.490 0.466 Number of Firms 1,300 1,279 164 82 1,054 1,300 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

38

Table 7: Log Standard Deviation of Stock Return

Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assets Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Vol Qnt 2 * After Tax Vol Qnt 3 * After Tax Vol Qnt 4 * After Tax Vol Qnt 5 * After Tax Govt Fiscal Surplus Constant

(1)

(2)

(3)

(4)

(5)

(6)

-0.231*** (0.0454) -0.133*** (0.0466) -0.0849** (0.0428) 0.0334 (0.0436) 0.0336 (0.0318) 0.306*** (0.0330) -0.282*** (0.0350) -0.0901*** (0.0293) -0.0313*** (0.00986) 0.0500** (0.0212) -0.183** (0.0711) 0.911*** (0.0501) -0.136 (0.0835) -0.0591 (0.0604) 0.0707** (0.0354) 0.117*** (0.0331) 0.161*** (0.0340) 0.721*** (0.184) 0.303 (0.957)

-0.228*** (0.0457) -0.127*** (0.0471) -0.0842* (0.0430) 0.0357 (0.0439) 0.0377 (0.0322) 0.306*** (0.0335) -0.283*** (0.0354) -0.0931*** (0.0301) -0.0337*** (0.00997) 0.0495** (0.0215) -0.191*** (0.0720) 0.904*** (0.0504) -0.132 (0.0841) -0.0604 (0.0605) 0.0727** (0.0355) 0.118*** (0.0335) 0.154*** (0.0344) 0.720*** (0.183) 0.291 (0.950)

-0.186 (0.173) -0.175 (0.172) -0.0693 (0.172) -0.160 (0.172) 0.00363 (0.0813) 0.246*** (0.0914) -0.211** (0.0861) -0.0402 (0.0844) -0.126*** (0.0295) 0.0809 (0.0664) -0.0968 (0.433) 1.250*** (0.221) 0.165 (0.234) -0.138 (0.133) -0.139 (0.0979) -0.208** (0.0949) -0.161 (0.101) -0.0659*** (0.0103) -3.723*** (0.565)

-0.369** (0.144) -0.0140 (0.160) -0.295** (0.140) 0.0636 (0.112) 0.118 (0.151) 0.644*** (0.106) -0.622*** (0.125) -0.357** (0.165) 0.0388 (0.0264) -0.121*** (0.0427) 1.475 (1.099) 2.679*** (0.523) -0.875** (0.415)

-0.250*** (0.0499) -0.137*** (0.0516) -0.0930* (0.0480) 0.0366 (0.0487) 0.0246 (0.0371) 0.302*** (0.0394) -0.283*** (0.0428) -0.0775** (0.0307) -0.0216** (0.0109) 0.0533** (0.0245) -0.208*** (0.0733) 0.887*** (0.0570) -0.129 (0.0919) -0.0820 (0.0691) 0.0884** (0.0410) 0.175*** (0.0365) 0.221*** (0.0389) 0.709*** (0.204) 0.407 (1.056)

-0.229*** (0.0448) -0.126*** (0.0463) -0.0860** (0.0420) 0.0376 (0.0430) 0.0229 (0.0320) 0.301*** (0.0330) -0.277*** (0.0349) -0.0865*** (0.0292) -0.0332*** (0.00980) 0.0517** (0.0212) -0.170** (0.0696) 0.905*** (0.0500) -0.119 (0.0829) -0.0614 (0.0616) 0.0679* (0.0350) 0.120*** (0.0333) 0.160*** (0.0342) 0.718*** (0.184) 0.320 (0.955)

-0.184 (0.164) -0.159 (0.171) -0.129 (0.173) -0.106*** (0.0143) -4.355*** (0.540)

Observations 49,559 48,635 6,985 4,173 38,401 48,514 R-squared 0.256 0.254 0.375 0.354 0.251 0.256 Number of Firms 1,300 1,279 164 82 1,054 1,300 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

39

Table 8: Log Standard Deviation of Stock Return Error

Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assets Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Vol Qnt 2 * After Tax Vol Qnt 3 * After Tax Vol Qnt 4 * After Tax Vol Qnt 5 * After Tax Govt Fiscal Surplus Constant

(1)

(2)

(3)

(4)

(5)

(6)

-0.201*** (0.0300) -0.100*** (0.0285) -0.0656** (0.0271) 0.000459 (0.0274) -0.227*** (0.0241) 0.171*** (0.0209) -0.157*** (0.0237) -0.0395** (0.0188) -0.0282*** (0.00592) 0.0278** (0.0141) -0.109** (0.0424) 0.712*** (0.0307) -0.176*** (0.0473) -0.0446 (0.0379) 0.0703*** (0.0244) 0.0960*** (0.0225) 0.0855*** (0.0237) 0.392** (0.164) -2.077** (0.833)

-0.201*** (0.0302) -0.0992*** (0.0289) -0.0629** (0.0273) 0.00469 (0.0276) -0.225*** (0.0244) 0.171*** (0.0212) -0.157*** (0.0240) -0.0409** (0.0192) -0.0275*** (0.00597) 0.0280* (0.0144) -0.109** (0.0426) 0.710*** (0.0309) -0.180*** (0.0476) -0.0433 (0.0380) 0.0718*** (0.0246) 0.0986*** (0.0227) 0.0855*** (0.0241) 0.391** (0.163) -2.086** (0.829)

-0.133 (0.0817) -0.0335 (0.0784) -0.0292 (0.0796) -0.105 (0.0965) -0.346*** (0.0453) 0.0726* (0.0401) -0.0223 (0.0429) 0.0234 (0.0564) -0.0646*** (0.0206) 0.0970*** (0.0340) -0.414* (0.229) 0.899*** (0.148) -0.234* (0.130) -0.199*** (0.0667) -0.0818 (0.0722) -0.0850 (0.0683) -0.146** (0.0736) 0.00940 (0.00719) -4.380*** (0.278)

-0.269*** (0.0940) -0.126 (0.0950) -0.144* (0.0834) 0.0230 (0.0766) -0.219*** (0.0732) 0.338*** (0.0708) -0.363*** (0.0778) -0.453*** (0.133) -0.00758 (0.0172) -0.0364 (0.0274) 1.742*** (0.636) 2.084*** (0.256) -0.551*** (0.202)

-0.222*** (0.0332) -0.116*** (0.0315) -0.0730** (0.0301) 0.00779 (0.0304) -0.219*** (0.0285) 0.162*** (0.0256) -0.155*** (0.0295) -0.0193 (0.0194) -0.0171*** (0.00629) 0.0260 (0.0164) -0.154*** (0.0443) 0.659*** (0.0345) -0.155*** (0.0513) -0.0374 (0.0453) 0.0910*** (0.0286) 0.130*** (0.0251) 0.116*** (0.0275) 0.393** (0.166) -1.914** (0.845)

-0.201*** (0.0295) -0.100*** (0.0282) -0.0679** (0.0267) 0.000325 (0.0272) -0.230*** (0.0244) 0.170*** (0.0210) -0.158*** (0.0238) -0.0380** (0.0186) -0.0290*** (0.00590) 0.0264* (0.0141) -0.102** (0.0420) 0.711*** (0.0308) -0.165*** (0.0476) -0.0446 (0.0385) 0.0715*** (0.0241) 0.0990*** (0.0226) 0.0865*** (0.0237) 0.389** (0.164) -2.052** (0.832)

-0.0189 (0.0570) -0.0301 (0.0618) 0.0202 (0.0565) -0.0578*** (0.00719) -4.489*** (0.398)

Observations 49,845 48,919 7,024 4,189 38,632 48,795 R-squared 0.431 0.430 0.551 0.569 0.420 0.433 Number of Firms 1,300 1,279 164 82 1,054 1,300 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

40

Table 9: Log of Sensitivity to Price for New Grants

Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assets Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Vol Qnt 2 * After Tax Vol Qnt 3 * After Tax Vol Qnt 4 * After Tax Vol Qnt 5 * After Tax Govt Fiscal Surplus Constant

(1)

(2)

(3)

(4)

(5)

(6)

-0.338*** (0.129) -0.110 (0.131) 0.00600 (0.124) 0.0227 (0.123) 0.183 (0.111) -0.184 (0.112) 0.740*** (0.120) -0.383 (0.345) -0.0490 (0.0395) 0.167** (0.0818) -0.437 (0.557) 0.149 (0.173) 0.0461 (0.235) 0.288 (0.190) 0.375** (0.162) 0.293* (0.163) 0.300* (0.171) 0.101 (0.136) 0.664 (0.702)

-0.334*** (0.128) -0.0995 (0.129) 0.0104 (0.123) 0.0226 (0.123) 0.169 (0.112) -0.161 (0.113) 0.744*** (0.120) -0.366 (0.349) -0.0557 (0.0395) 0.170** (0.0826) -0.424 (0.562) 0.120 (0.175) -0.00199 (0.237) 0.290 (0.190) 0.405** (0.164) 0.326** (0.164) 0.320* (0.170) 0.0723 (0.132) -0.121 (0.699)

-0.345 (0.240) -0.237 (0.241) -0.0130 (0.244)

-1.700*** (0.444) -1.801*** (0.510) -1.112** (0.486) -0.950*** (0.262) -0.527 (0.456) -0.481 (0.330) 1.419*** (0.424) -0.933 (2.205) -0.116 (0.130) 0.370* (0.192) 7.961 (6.666) -1.963 (1.436) -0.589 (1.197)

-0.219 (0.143) 0.0504 (0.144) 0.0898 (0.134) 0.129 (0.134) 0.267* (0.144) -0.0345 (0.161) 0.619*** (0.167) -0.448 (0.364) -0.0449 (0.0457) 0.129 (0.107) -0.350 (0.604) 0.366* (0.220) 0.00504 (0.270) 0.190 (0.237) 0.340* (0.174) 0.200 (0.177) 0.254 (0.187) 0.0403 (0.156) 0.335 (1.072)

-0.350** (0.138) -0.119 (0.138) -0.00962 (0.133) 0.00460 (0.134) 0.165 (0.113) -0.209* (0.114) 0.770*** (0.122) -0.439 (0.344) -0.0489 (0.0397) 0.179** (0.0818) -0.421 (0.556) 0.126 (0.176) 0.0607 (0.236) 0.339* (0.190) 0.428*** (0.154) 0.325** (0.154) 0.345** (0.161) 0.103 (0.142) 0.631 (0.729)

-0.0140 (0.363) -0.666** (0.279) 1.023*** (0.318) -0.888 (1.360) 0.0497 (0.102) 0.179 (0.328) -8.142* (4.410) -0.828 (0.751) 0.272 (0.782) 0.923 (0.619) 0.827 (0.626) 0.870 (0.607) 0.675 (0.620) 0.245*** (0.0487) 0.637 (1.505)

0.0999 (0.394) 0.578 (0.352) 0.830* (0.423) -0.143 (0.122) -1.453 (2.448)

Observations 5,379 5,285 945 399 4,035 5,226 R-squared 0.219 0.219 0.242 0.418 0.238 0.220 Number of Firms 1,018 1,000 151 75 792 1,016 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

41

Table 10: Log of Sensitivity to Price for All Stock and Options

Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assets Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Vol Qnt 2 * After Tax Vol Qnt 3 * After Tax Vol Qnt 4 * After Tax Vol Qnt 5 * After Tax Govt Fiscal Surplus Constant

(1)

(2)

(3)

(4)

(5)

(6)

-1.099*** (0.127) -0.825*** (0.117) -0.828*** (0.125) -0.540*** (0.110) 0.174 (0.112) 0.0837 (0.0840) 0.671*** (0.105) 0.147 (0.105) 0.0181 (0.0282) -0.00642 (0.0828) -0.166 (0.154) 0.612*** (0.140) 0.373* (0.197) -0.233 (0.147) 0.0547 (0.0881) 0.163* (0.0881) 0.321*** (0.0943) 0.171** (0.0841) -1.701** (0.675)

-1.105*** (0.129) -0.841*** (0.119) -0.830*** (0.126) -0.561*** (0.111) 0.158 (0.113) 0.0873 (0.0850) 0.681*** (0.106) 0.151 (0.108) 0.0193 (0.0284) -0.0101 (0.0842) -0.171 (0.155) 0.630*** (0.141) 0.383* (0.198) -0.233 (0.147) 0.0636 (0.0885) 0.185** (0.0887) 0.328*** (0.0955) 0.156* (0.0840) -1.706** (0.683)

-2.049*** (0.579) -1.835*** (0.581) -1.916*** (0.571) -1.605*** (0.551) 0.150 (0.296) -0.0614 (0.239) 0.787*** (0.275) 0.895 (0.818) -0.00341 (0.0823) 0.0554 (0.218) -3.070** (1.506) -0.109 (0.874) 1.817*** (0.620) 0.0742 (0.330) 0.102 (0.284) 0.205 (0.314) 0.149 (0.303) 0.124 (0.110) 0.149 (1.664)

-0.761* (0.416) -0.796** (0.380) -0.997*** (0.375) -0.702** (0.292) 0.484 (0.355) -0.368 (0.315) 1.204** (0.476) 0.475 (1.077) -0.0716 (0.130) 0.125 (0.129) -1.681 (3.542) -0.864 (1.459) 1.634 (1.527)

-1.037*** (0.140) -0.754*** (0.128) -0.741*** (0.141) -0.500*** (0.122) 0.197 (0.128) 0.130 (0.101) 0.604*** (0.124) 0.127 (0.106) 0.0252 (0.0306) -0.0864 (0.0966) -0.117 (0.151) 0.621*** (0.157) 0.246 (0.208) -0.310* (0.168) 0.0421 (0.0997) 0.109 (0.0989) 0.343*** (0.107) 0.157* (0.0911) -0.623 (0.601)

-1.105*** (0.128) -0.833*** (0.119) -0.839*** (0.126) -0.547*** (0.110) 0.174 (0.114) 0.120 (0.0834) 0.646*** (0.105) 0.146 (0.106) 0.0201 (0.0282) -0.00728 (0.0844) -0.179 (0.154) 0.652*** (0.140) 0.342* (0.198) -0.222 (0.150) 0.0779 (0.0889) 0.175** (0.0885) 0.341*** (0.0944) 0.173** (0.0845) -1.869*** (0.680)

-0.307 (0.328) 0.0639 (0.338) 0.0240 (0.354) 0.104 (0.111) -4.558*** (1.727)

Observations 10,266 10,106 1,289 680 8,297 10,066 R-squared 0.312 0.310 0.360 0.578 0.317 0.311 Number of Firms 1,274 1,253 160 81 1,033 1,274 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

42

A

Appendix

A.1

Proofs to Propositions

Proposition 1: An increase (decrease) in the dividend tax rate increases (decreases) the volatility of the stock. Proof: ∂V ol 2d2 σ 2 (µ − (1 − td )) 2Id = >0 = ∂td k + rd2 σ 2 (k + rd2 σ 2 )2

Proposition 2: The increase (decrease) in the volatility of a stock after an increase (decrease) in the dividend tax rate is larger in magnitude for firms with larger incentives. Proof: ol ∂ ∂V ∂td

∂d

=

4kd2 σ 2 (µ − (1 − td )) >0 (k + rd2 σ 2 )3

Proposition 3: An increase (decrease) in the dividend tax rate does not affect the incentive scheme. Proof: ∂d =0 ∂td

Proposition 4: An increase (decrease) in the dividend tax rate does not affect the incentive scheme differentially by initial incentive level. Proof: ∂d ∂ ∂t d

∂d

=0

43

Table Appendix 1: Log Standard Deviation of Stock Price With Earnings Quality

Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assetss Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Vol Qnt 2 * After Tax Vol Qnt 3 * After Tax Vol Qnt 4 * After Tax Vol Qnt 5 * After Tax Govt Fiscal Surplus Earnings Quality Constant

(1)

(2)

(3)

(4)

(5)

(6)

-0.359*** (0.0401) -0.202*** (0.0397) -0.154*** (0.0376) -0.0336 (0.0355) 0.592*** (0.0336) 0.263*** (0.0320) -0.113*** (0.0352) -0.00308 (0.0270) -0.0335*** (0.0104) 0.122*** (0.0223) -0.0386 (0.0702) 0.810*** (0.0487) -0.229*** (0.0676) -0.0996 (0.0619) 0.104*** (0.0340) 0.156*** (0.0291) 0.165*** (0.0304) 0.659*** (0.0779) -6.38e-10 (5.43e-10) -4.910*** (0.373)

-0.357*** (0.0404) -0.202*** (0.0402) -0.150*** (0.0377) -0.0336 (0.0358) 0.599*** (0.0339) 0.261*** (0.0323) -0.113*** (0.0356) -0.00150 (0.0278) -0.0363*** (0.0104) 0.119*** (0.0226) -0.0414 (0.0709) 0.804*** (0.0489) -0.227*** (0.0681) -0.0999 (0.0619) 0.106*** (0.0343) 0.159*** (0.0294) 0.162*** (0.0309) 0.659*** (0.0767) -4.21e-10 (5.55e-10) -4.939*** (0.379)

0.310 (0.269) 0.558* (0.301) -0.298*** (0.0996) -0.461** (0.170) -0.219 (0.236) 0.203 (0.193) 0.284 (0.204) 0.880 (1.034) 0.0353 (0.0897) 0.425** (0.180) -0.375 (0.697) 1.391*** (0.334) 0.124 (0.310) -0.929** (0.336) -0.433*** (0.122)

-0.417*** (0.140) -0.238** (0.110) -0.213** (0.0933) -0.0895 (0.0850) 0.650*** (0.106) 0.594*** (0.0710) -0.583*** (0.0875) -0.341 (0.227) 0.0305 (0.0283) -0.0641 (0.0399) 2.599* (1.447) 2.899*** (0.281) -0.820*** (0.260)

-0.359*** (0.0424) -0.202*** (0.0424) -0.146*** (0.0410) -0.0255 (0.0386) 0.580*** (0.0364) 0.232*** (0.0339) -0.0791** (0.0377) 0.00100 (0.0275) -0.0415*** (0.0112) 0.150*** (0.0249) -0.0584 (0.0707) 0.754*** (0.0499) -0.202*** (0.0698) -0.105 (0.0645) 0.102*** (0.0372) 0.164*** (0.0314) 0.173*** (0.0328) 0.653*** (0.0927) -6.06e-10 (5.38e-10) -4.926*** (0.403)

-0.359*** (0.0397) -0.201*** (0.0393) -0.158*** (0.0372) -0.0319 (0.0354) 0.590*** (0.0339) 0.262*** (0.0320) -0.113*** (0.0351) -0.00223 (0.0270) -0.0350*** (0.0104) 0.122*** (0.0223) -0.0317 (0.0701) 0.803*** (0.0487) -0.228*** (0.0682) -0.0978 (0.0631) 0.107*** (0.0338) 0.162*** (0.0294) 0.165*** (0.0302) 0.658*** (0.0768) -6.58e-10 (5.65e-10) -1.764*** (0.271)

-1.250*** (0.365) -0.0155 (0.0692) -1.09e-07 (1.61e-07) -5.544*** (1.254)

-0.00922 (0.0892) 0.00640 (0.0930) -0.0231 (0.0948) 0.00650 (0.0199) -5.97e-10 (5.35e-09) -3.080*** (0.475)

Observations 33,980 33,383 513 2,965 30,502 33,311 R-squared 0.469 0.467 0.617 0.482 0.477 0.470 Number of Firms 1,103 1,085 16 79 1,008 1,103 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

44

Table Appendix 2: Log Standard Deviation of Stock Price Error With Earnings Quality

Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assetss Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Vol Qnt 2 * After Tax Vol Qnt 3 * After Tax Vol Qnt 4 * After Tax Vol Qnt 5 * After Tax Govt Fiscal Surplus Earnings Quality Constant

(1)

(2)

(3)

(4)

(5)

(6)

-0.323*** (0.0379) -0.200*** (0.0370) -0.150*** (0.0354) -0.0417 (0.0334) 0.679*** (0.0306) 0.304*** (0.0298) -0.183*** (0.0321) 0.000146 (0.0266) -0.0310*** (0.00948) 0.0595*** (0.0196) -0.0706 (0.0646) 0.800*** (0.0451) -0.210*** (0.0626) -0.0983* (0.0552) 0.0990*** (0.0320) 0.136*** (0.0274) 0.145*** (0.0278) 0.202 (0.217) -6.35e-11 (6.30e-10) -4.043*** (0.608)

-0.322*** (0.0382) -0.200*** (0.0375) -0.146*** (0.0356) -0.0412 (0.0337) 0.686*** (0.0309) 0.303*** (0.0301) -0.184*** (0.0325) 0.000824 (0.0274) -0.0330*** (0.00954) 0.0564*** (0.0198) -0.0715 (0.0651) 0.795*** (0.0453) -0.210*** (0.0630) -0.0985* (0.0551) 0.101*** (0.0323) 0.139*** (0.0277) 0.142*** (0.0283) 0.202 (0.218) 1.66e-10 (6.38e-10) -4.060*** (0.617)

0.304 (0.264) 0.710*** (0.185) -0.362*** (0.120) -0.449** (0.189) 0.114 (0.184) 0.204 (0.176) 0.162 (0.197) 1.521** (0.608) -0.0720 (0.0586) 0.379*** (0.121) -0.396 (0.537) 1.207*** (0.265) 0.0811 (0.487) -1.014*** (0.296) -0.215 (0.123)

-0.372*** (0.138) -0.226** (0.108) -0.188** (0.0858) -0.0809 (0.0806) 0.741*** (0.107) 0.630*** (0.0657) -0.599*** (0.0794) -0.265 (0.200) -0.0289 (0.0256) -0.0753** (0.0361) 2.552** (1.280) 2.735*** (0.250) -0.720*** (0.239)

-0.322*** (0.0401) -0.198*** (0.0396) -0.141*** (0.0387) -0.0344 (0.0365) 0.668*** (0.0329) 0.275*** (0.0316) -0.153*** (0.0342) 0.00307 (0.0270) -0.0346*** (0.0103) 0.0787*** (0.0222) -0.0942 (0.0650) 0.746*** (0.0463) -0.181*** (0.0643) -0.0995* (0.0576) 0.0945*** (0.0351) 0.142*** (0.0296) 0.150*** (0.0300) 0.200 (0.202) -0 (6.16e-10) -4.058*** (0.568)

-0.322*** (0.0374) -0.197*** (0.0364) -0.153*** (0.0350) -0.0388 (0.0332) 0.678*** (0.0308) 0.303*** (0.0298) -0.183*** (0.0321) 0.00100 (0.0265) -0.0323*** (0.00942) 0.0599*** (0.0196) -0.0657 (0.0647) 0.794*** (0.0450) -0.208*** (0.0632) -0.0960* (0.0567) 0.100*** (0.0316) 0.139*** (0.0276) 0.145*** (0.0277) 0.200 (0.219) -9.66e-11 (6.48e-10) -2.935*** (0.567)

-0.995** (0.378) 0.0588 (0.0600) -7.68e-08 (1.61e-07) -5.356*** (1.135)

0.0657 (0.0739) 0.0556 (0.0754) 0.0348 (0.0780) -0.0153 (0.0187) -1.18e-10 (4.97e-09) -3.296*** (0.469)

Observations 33,980 33,383 513 2,965 30,502 33,311 R-squared 0.476 0.474 0.637 0.471 0.486 0.477 Number of Firms 1,103 1,085 16 79 1,008 1,103 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

45

Table Appendix 3: Log Standard Deviation of Stock Return With Earnings Quality

Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assetss Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Vol Qnt 2 * After Tax Vol Qnt 3 * After Tax Vol Qnt 4 * After Tax Vol Qnt 5 * After Tax Govt Fiscal Surplus Earnings Quality Constant

(1)

(2)

(3)

(4)

(5)

(6)

-0.267*** (0.0468) -0.158*** (0.0496) -0.160*** (0.0443) 0.0107 (0.0451) 0.0352 (0.0403) 0.344*** (0.0403) -0.274*** (0.0444) -0.0703** (0.0320) -0.0210* (0.0115) 0.0315 (0.0256) -0.151** (0.0727) 0.905*** (0.0639) -0.155* (0.0911) -0.0224 (0.0659) 0.0880** (0.0388) 0.161*** (0.0347) 0.194*** (0.0370) 0.728*** (0.190) -6.47e-10 (8.44e-10) -4.950*** (0.604)

-0.266*** (0.0470) -0.150*** (0.0499) -0.158*** (0.0443) 0.0136 (0.0453) 0.0362 (0.0407) 0.342*** (0.0408) -0.272*** (0.0447) -0.0714** (0.0332) -0.0251** (0.0115) 0.0296 (0.0259) -0.160** (0.0741) 0.896*** (0.0642) -0.155* (0.0916) -0.0267 (0.0659) 0.0892** (0.0389) 0.157*** (0.0351) 0.183*** (0.0375) 0.726*** (0.188) -4.40e-10 (9.14e-10) -4.949*** (0.602)

0.0448 (0.405) -0.231 (0.434) -0.444** (0.189) -0.677*** (0.222) -0.905*** (0.251) 0.279 (0.239) 0.390 (0.315) 0.966 (0.978) -0.0734 (0.0684) 0.424 (0.260) -0.298 (0.406) 1.074** (0.475) 1.193** (0.546) -0.442 (0.364) -0.631*** (0.142)

-0.473*** (0.162) -0.0277 (0.155) -0.295** (0.139) -0.0210 (0.113) 0.196 (0.185) 0.720*** (0.0991) -0.790*** (0.108) -0.254 (0.202) 0.0281 (0.0284) -0.153*** (0.0508) 1.241 (1.392) 3.169*** (0.497) -1.033** (0.449)

-0.272*** (0.0502) -0.169*** (0.0530) -0.150*** (0.0481) 0.0240 (0.0492) 0.0185 (0.0425) 0.302*** (0.0420) -0.236*** (0.0470) -0.0668** (0.0326) -0.0236* (0.0123) 0.0557* (0.0289) -0.166** (0.0744) 0.839*** (0.0641) -0.122 (0.0938) -0.0338 (0.0681) 0.0909** (0.0426) 0.175*** (0.0374) 0.219*** (0.0399) 0.716*** (0.206) -6.69e-10 (8.39e-10) -4.815*** (0.672)

-0.263*** (0.0458) -0.151*** (0.0489) -0.160*** (0.0432) 0.0157 (0.0444) 0.0256 (0.0405) 0.341*** (0.0403) -0.270*** (0.0442) -0.0701** (0.0319) -0.0231** (0.0114) 0.0332 (0.0256) -0.138* (0.0718) 0.901*** (0.0639) -0.147 (0.0911) -0.0204 (0.0671) 0.0864** (0.0384) 0.165*** (0.0347) 0.191*** (0.0368) 0.726*** (0.190) -7.12e-10 (8.67e-10) -4.927*** (0.604)

-0.719* (0.377) -0.00789 (0.106) -1.95e-07 (1.80e-07) -6.684*** (1.995)

-0.219 (0.150) -0.206 (0.166) -0.277* (0.166) 0.0504** (0.0218) 2.94e-09 (9.94e-09) -4.567*** (0.718)

Observations 33,782 33,186 510 2,955 30,317 33,116 R-squared 0.241 0.239 0.519 0.392 0.241 0.240 Number of Firms 1,102 1,084 16 79 1,007 1,102 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

46

Table Appendix 4: Log Standard Deviation of Stock Return Error With Earnings Quality

Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assetss Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Vol Qnt 2 * After Tax Vol Qnt 3 * After Tax Vol Qnt 4 * After Tax Vol Qnt 5 * After Tax Govt Fiscal Surplus Earnings Quality Constant

(1)

(2)

(3)

(4)

(5)

(6)

-0.230*** (0.0310) -0.130*** (0.0297) -0.0983*** (0.0281) -0.00733 (0.0281) -0.230*** (0.0247) 0.175*** (0.0246) -0.136*** (0.0282) -0.0216 (0.0172) -0.0162** (0.00636) 0.0236 (0.0162) -0.137*** (0.0470) 0.674*** (0.0370) -0.165*** (0.0521) -0.0134 (0.0418) 0.0832*** (0.0262) 0.115*** (0.0234) 0.103*** (0.0250) 0.391** (0.158) -9.90e-10** (4.36e-10) -4.848*** (0.494)

-0.230*** (0.0312) -0.129*** (0.0301) -0.0949*** (0.0282) -0.00473 (0.0283) -0.229*** (0.0251) 0.175*** (0.0249) -0.136*** (0.0284) -0.0245 (0.0177) -0.0166*** (0.00640) 0.0231 (0.0164) -0.139*** (0.0477) 0.674*** (0.0373) -0.170*** (0.0525) -0.0130 (0.0419) 0.0848*** (0.0264) 0.115*** (0.0236) 0.103*** (0.0255) 0.391** (0.157) -8.93e-10* (4.65e-10) -4.899*** (0.494)

0.299 (0.193) -0.840*** (0.193) -0.0285 (0.0801) -0.191 (0.204) -0.525** (0.214) 0.0812 (0.139) 0.00710 (0.178) 0.877 (0.593) -0.0433 (0.0343) 0.109 (0.0967) -1.019* (0.490) 0.779*** (0.226) 0.129 (0.397) -0.765*** (0.243) -0.342*** (0.0948)

-0.306** (0.121) -0.128 (0.103) -0.123 (0.0872) -0.0155 (0.0848) -0.237*** (0.0809) 0.424*** (0.0629) -0.443*** (0.0669) -0.379** (0.151) -0.00475 (0.0163) -0.0418 (0.0288) 1.465** (0.692) 2.534*** (0.237) -0.485** (0.202)

-0.246*** (0.0326) -0.147*** (0.0313) -0.107*** (0.0303) -0.00344 (0.0302) -0.242*** (0.0261) 0.142*** (0.0246) -0.105*** (0.0290) -0.0138 (0.0170) -0.0190*** (0.00676) 0.0399** (0.0167) -0.158*** (0.0478) 0.611*** (0.0361) -0.147*** (0.0533) -0.00532 (0.0435) 0.0966*** (0.0285) 0.137*** (0.0246) 0.116*** (0.0271) 0.395** (0.168) -9.89e-10** (4.20e-10) -4.811*** (0.526)

-0.226*** (0.0304) -0.128*** (0.0293) -0.0993*** (0.0276) -0.00677 (0.0279) -0.232*** (0.0249) 0.174*** (0.0245) -0.138*** (0.0281) -0.0218 (0.0171) -0.0170*** (0.00627) 0.0222 (0.0162) -0.130*** (0.0464) 0.674*** (0.0368) -0.159*** (0.0527) -0.0133 (0.0426) 0.0835*** (0.0259) 0.117*** (0.0235) 0.101*** (0.0248) 0.390** (0.157) -1.06e-09** (4.29e-10) -2.949*** (0.409)

-0.709** (0.256) -0.0634 (0.0432) -9.81e-08 (1.24e-07) -4.707*** (0.701)

0.0180 (0.0610) 0.00295 (0.0638) 0.00296 (0.0595) -0.0991*** (0.0137) 9.80e-10 (4.83e-09) -4.288*** (0.378)

Observations 33,979 33,382 513 2,965 30,501 33,310 R-squared 0.437 0.437 0.639 0.602 0.435 0.438 Number of Firms 1,103 1,085 16 79 1,008 1,103 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

47

Table Appendix 5: Log of Sensitivity to Price for New Grants With Earnings Quality

Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assetss Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Vol Qnt 2 * After Tax Vol Qnt 3 * After Tax Vol Qnt 4 * After Tax Vol Qnt 5 * After Tax Govt Fiscal Surplus Earnings Quality Constant

(1)

(2)

(3)

(4)

(5)

(6)

-0.183 (0.132) -0.0450 (0.130) 0.0483 (0.124) 0.0528 (0.121) 0.169 (0.134) -0.139 (0.133) 0.729*** (0.141) -0.443 (0.371) -0.0461 (0.0466) 0.201** (0.0889) -0.390 (0.616) 0.193 (0.200) 0.0771 (0.264) 0.193 (0.216) 0.379** (0.176) 0.275 (0.177) 0.313* (0.187) -0.0255 (0.143) -1.56e-09 (2.30e-09) 0.545 (0.855)

-0.191 (0.132) -0.0359 (0.131) 0.0577 (0.125) 0.0575 (0.122) 0.154 (0.135) -0.129 (0.134) 0.732*** (0.142) -0.435 (0.374) -0.0446 (0.0467) 0.201** (0.0900) -0.397 (0.618) 0.173 (0.202) 0.0541 (0.267) 0.176 (0.219) 0.392** (0.179) 0.279 (0.179) 0.297 (0.190) -0.0449 (0.141) -1.07e-09 (2.31e-09) -0.969 (0.814)

0.344 (1.057)

-1.816** (0.698) -1.872*** (0.643) -1.035 (0.624) -0.976*** (0.278) -0.416 (0.543) -0.435 (0.366) 1.264** (0.483) -1.570 (2.598) -0.0353 (0.144) 0.308 (0.216) 6.516 (7.396) -1.421 (1.445) 0.579 (1.270)

-0.0769 (0.138) 0.0986 (0.135) 0.146 (0.127) 0.150 (0.127) 0.206 (0.152) -0.129 (0.166) 0.726*** (0.173) -0.389 (0.404) -0.0442 (0.0503) 0.173 (0.116) -0.486 (0.652) 0.218 (0.220) 0.0819 (0.284) 0.131 (0.232) 0.340* (0.188) 0.225 (0.189) 0.281 (0.199) -0.0753 (0.148) -1.43e-09 (2.34e-09) 0.185 (1.058)

-0.175 (0.141) -0.0299 (0.138) 0.0538 (0.135) 0.0491 (0.133) 0.153 (0.137) -0.181 (0.136) 0.773*** (0.143) -0.446 (0.364) -0.0420 (0.0465) 0.194** (0.0892) -0.459 (0.604) 0.130 (0.203) 0.0980 (0.268) 0.254 (0.218) 0.432*** (0.165) 0.309* (0.165) 0.352** (0.176) -0.0210 (0.149) -1.64e-09 (2.27e-09) -0.719 (0.789)

2.467** (0.960)

5.200** (2.159) 1.279 (1.594) -5.456 (3.534) 7.219 (14.32) 1.965 (1.697) -5.416* (2.716) -21.34 (14.90) 3.194 (2.546) -9.228** (3.955) -1.956* (0.891)

-0.0566 (0.148) -1.70e-05** (5.39e-06) 35.62** (13.97)

0.370 (0.422) 0.816** (0.342) 0.953 (0.605) -0.173 (0.124) -1.39e-08 (2.05e-08) -1.406 (2.678)

Observations 4,047 3,980 47 341 3,659 3,936 R-squared 0.234 0.233 0.919 0.396 0.240 0.235 Number of Firms 824 811 10 68 746 823 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

48

Table Appendix 6: Log of Sensitivity to Price for All Stock and Options With Earnings Quality

Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assetss Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Vol Qnt 2 * After Tax Vol Qnt 3 * After Tax Vol Qnt 4 * After Tax Vol Qnt 5 * After Tax Govt Fiscal Surplus Earnings Quality Constant

(1)

(2)

(3)

(4)

(5)

(6)

-1.061*** (0.137) -0.812*** (0.124) -0.802*** (0.134) -0.479*** (0.116) 0.111 (0.129) 0.140 (0.0965) 0.656*** (0.124) 0.186* (0.0994) -0.00248 (0.0336) 0.0415 (0.0911) -0.161 (0.160) 0.738*** (0.149) 0.292 (0.211) -0.291* (0.174) 0.114 (0.0934) 0.135 (0.0976) 0.358*** (0.107) 0.116 (0.0868) -2.95e-09 (3.10e-09) -3.713*** (0.928)

-1.071*** (0.138) -0.830*** (0.125) -0.801*** (0.134) -0.499*** (0.116) 0.0948 (0.131) 0.128 (0.0975) 0.673*** (0.125) 0.185* (0.101) -0.000777 (0.0336) 0.0449 (0.0922) -0.164 (0.161) 0.746*** (0.150) 0.306 (0.212) -0.294* (0.173) 0.125 (0.0939) 0.162* (0.0978) 0.360*** (0.109) 0.104 (0.0866) -7.48e-10 (2.53e-09) -2.996*** (0.792)

-1.797 (1.262)

-0.534 (0.417) -0.696 (0.433) -1.139*** (0.405) -0.523 (0.326) 0.297 (0.361) -0.453 (0.343) 1.213** (0.539) 0.878 (1.194) -0.0741 (0.146) 0.180 (0.142) -4.680 (4.257) -0.517 (1.660) 1.431 (1.727)

-1.025*** (0.145) -0.769*** (0.132) -0.753*** (0.146) -0.477*** (0.126) 0.147 (0.138) 0.170 (0.107) 0.592*** (0.136) 0.177* (0.0984) 0.00999 (0.0338) -0.0464 (0.0986) -0.139 (0.158) 0.719*** (0.155) 0.225 (0.215) -0.323* (0.178) 0.0972 (0.101) 0.109 (0.104) 0.351*** (0.115) 0.103 (0.0911) -2.81e-09 (3.16e-09) -1.323 (1.032)

-1.067*** (0.137) -0.820*** (0.126) -0.815*** (0.135) -0.494*** (0.116) 0.106 (0.131) 0.162* (0.0972) 0.643*** (0.125) 0.185* (0.0994) -0.00168 (0.0335) 0.0435 (0.0920) -0.174 (0.161) 0.762*** (0.150) 0.276 (0.213) -0.292* (0.177) 0.139 (0.0940) 0.150 (0.0981) 0.373*** (0.107) 0.123 (0.0879) -2.80e-09 (3.12e-09) -3.847*** (0.931)

0.353 (1.374) -1.366 (1.091) -2.563 (1.492) -0.110 (1.549) 2.593 (1.584) -7.489 (11.43) -0.389 (0.519) 2.083* (1.106) 3.372 (4.608) 1.996 (2.989) -4.307 (3.137) 1.088 (1.780) -1.948 (1.323)

0.00574 (2.296) 0.200 (0.319) -1.70e-07 (1.12e-06) -16.02* (8.260)

-0.350 (0.360) -0.117 (0.338) -0.0641 (0.389) 0.109 (0.0845) 1.11e-09 (8.67e-09) -2.494 (1.929)

Observations 8,317 8,200 115 594 7,608 8,175 R-squared 0.320 0.317 0.629 0.615 0.318 0.319 Number of Firms 1,080 1,062 16 76 988 1,080 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

49

Table Appendix 7: Log Standard Deviation of Stock Price With Multiple Times

Price Quint 5 * Quarter -7 Price Quint 5 * Quarter -6 Price Quint 5 * Quarter -5 Price Quint 5 * Quarter -4 Price Quint 5 * Quarter -3 Price Quint 5 * Quarter -2 Price Quint 5 * Quarter -1 Price Quint 5 * Quarter 0 Price Quint 5 * Quarter 1 Price Quint 5 * Quarter 2 Price Quint 5 * Quarter 3 Price Quint 5 * Quarter 4 Price Quint 5 * Quarter 5 Price Quint 5 * Quarter 6 Price Quint 5 * Quarter 7 Price Quint 5 * 2+ Years Constant

(1)

(2)

(3)

(4)

(5)

(6)

0.0863 (0.0621) -0.0920 (0.0609) 0.123* (0.0734) -0.109* (0.0560) -0.191*** (0.0574) 0.0608 (0.0680) -0.0112 (0.0737) 0.0299 (0.0667) -0.172*** (0.0630) -0.159*** (0.0568) -0.250*** (0.0626) -0.244*** (0.0736) -0.278*** (0.0576) -0.338*** (0.0707) -0.229*** (0.0745) -0.347*** 0.0927 (0.644)

0.0865 (0.0625) -0.0966 (0.0613) 0.127* (0.0736) -0.113** (0.0564) -0.194*** (0.0577) 0.0389 (0.0676) -0.00405 (0.0749) 0.0303 (0.0678) -0.173*** (0.0636) -0.167*** (0.0573) -0.244*** (0.0629) -0.230*** (0.0739) -0.281*** (0.0581) -0.342*** (0.0713) -0.228*** (0.0755) -0.347*** 0.105 (0.647)

0.205 (0.224) 0.243 (0.201) 0.191 (0.267) -0.512* (0.260) 0.126 (0.253) 0.138 (0.239) -0.0745 (0.284) 0.00284 (0.190) 0.262 (0.239) -0.0698 (0.210) 0.0598 (0.229) -0.170 (0.420) -0.643** (0.272) -0.369 (0.336) -0.0134 (0.364) -0.216 -3.590*** (0.318)

0.0929 (0.169) 0.452* (0.242) 0.939** (0.426) 0.269 (0.236) 0.404* (0.225) 0.408 (0.462) 0.435 (0.298) -0.0347 (0.240) -0.154 (0.299) 0.127 (0.196) -0.380** (0.154) -0.0730 (0.238) 0.0511 (0.242) -0.248 (0.213) -0.220 (0.232) -0.240* -4.273*** (0.423)

0.0687 (0.0690) -0.126* (0.0686) 0.119 (0.0824) -0.0935 (0.0609) -0.206*** (0.0642) 0.0970 (0.0740) -0.00870 (0.0822) 0.0628 (0.0753) -0.179*** (0.0692) -0.162*** (0.0625) -0.259*** (0.0675) -0.249*** (0.0833) -0.252*** (0.0642) -0.310*** (0.0792) -0.176** (0.0856) -0.373*** 0.128 (0.669)

0.0924 (0.0624) -0.0882 (0.0606) 0.126* (0.0734) -0.104* (0.0562) -0.189*** (0.0574) 0.0615 (0.0679) -0.00850 (0.0741) 0.0445 (0.0680) -0.179*** (0.0626) -0.159*** (0.0558) -0.267*** (0.0635) -0.225*** (0.0745) -0.259*** (0.0576) -0.337*** (0.0719) -0.229*** (0.0754) -0.348*** 0.129 (0.617)

Observations 49,846 48,920 7,024 4,189 38,633 48,796 R-squared 0.469 0.467 0.440 0.522 0.492 0.470 Number of Firms 1,300 1,279 164 82 1,054 1,300 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

50

Table Appendix 8: Log Standard Deviation of Stock Price Error With Multiple Times

Price Quint 5 * Quarter -7 Price Quint 5 * Quarter -6 Price Quint 5 * Quarter -5 Price Quint 5 * Quarter -4 Price Quint 5 * Quarter -3 Price Quint 5 * Quarter -2 Price Quint 5 * Quarter -1 Price Quint 5 * Quarter 0 Price Quint 5 * Quarter 1 Price Quint 5 * Quarter 2 Price Quint 5 * Quarter 3 Price Quint 5 * Quarter 4 Price Quint 5 * Quarter 5 Price Quint 5 * Quarter 6 Price Quint 5 * Quarter 7 Price Quint 5 * 2+ Years Constant

(1)

(2)

(3)

(4)

(5)

(6)

0.0748 (0.0599) -0.0951 (0.0601) 0.178*** (0.0681) -0.132** (0.0552) -0.172*** (0.0545) 0.0503 (0.0666) -0.0586 (0.0678) 0.0629 (0.0638) -0.160*** (0.0608) -0.117** (0.0566) -0.264*** (0.0560) -0.245*** (0.0697) -0.257*** (0.0565) -0.313*** (0.0666) -0.231*** (0.0730) -0.326*** (0.0446) -2.281*** (0.865)

0.0757 (0.0603) -0.0994 (0.0605) 0.184*** (0.0685) -0.134** (0.0557) -0.175*** (0.0549) 0.0340 (0.0665) -0.0595 (0.0690) 0.0581 (0.0646) -0.163*** (0.0614) -0.125** (0.0570) -0.259*** (0.0563) -0.236*** (0.0702) -0.258*** (0.0570) -0.317*** (0.0673) -0.230*** (0.0739) -0.325*** (0.0450) -2.269*** (0.866)

0.160 (0.218) 0.0317 (0.208) 0.143 (0.235) -0.251 (0.229) 0.157 (0.257) 0.202 (0.228) -0.0471 (0.275) 0.290 (0.200) 0.243 (0.238) -0.0713 (0.206) -0.188 (0.211) -0.264 (0.362) -0.665** (0.257) -0.426 (0.273) -0.0122 (0.298) -0.189 (0.175) -3.399*** (0.320)

0.198 (0.144) 0.534* (0.278) 0.821** (0.324) 0.0462 (0.210) 0.435** (0.215) 0.409 (0.511) 0.271 (0.218) -0.0457 (0.254) -0.0796 (0.294) 0.0993 (0.212) -0.410** (0.192) 0.0280 (0.205) 0.0872 (0.255) -0.209 (0.173) -0.328 (0.216) -0.196 (0.119) -4.712*** (0.394)

0.0479 (0.0660) -0.115* (0.0671) 0.164** (0.0771) -0.121* (0.0618) -0.185*** (0.0610) 0.0832 (0.0714) -0.0330 (0.0767) 0.103 (0.0711) -0.172** (0.0670) -0.102 (0.0623) -0.233*** (0.0634) -0.246*** (0.0787) -0.229*** (0.0630) -0.260*** (0.0758) -0.197** (0.0846) -0.348*** (0.0506) -2.135** (0.836)

0.0804 (0.0603) -0.0884 (0.0598) 0.183*** (0.0680) -0.125** (0.0551) -0.170*** (0.0545) 0.0524 (0.0665) -0.0536 (0.0680) 0.0764 (0.0649) -0.165*** (0.0603) -0.121** (0.0556) -0.277*** (0.0566) -0.228*** (0.0707) -0.237*** (0.0564) -0.316*** (0.0677) -0.234*** (0.0727) -0.325*** (0.0441) -2.242** (0.892)

Observations 49,846 48,920 7,024 4,189 38,633 48,796 R-squared 0.471 0.469 0.432 0.518 0.497 0.472 Number of Firms 1,300 1,279 164 82 1,054 1,300 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

51

Table Appendix 9: Log Standard Deviation of Stock Return With Multiple Times

Price Quint 5 * Quarter -7 Price Quint 5 * Quarter -6 Price Quint 5 * Quarter -5 Price Quint 5 * Quarter -4 Price Quint 5 * Quarter -3 Price Quint 5 * Quarter -2 Price Quint 5 * Quarter -1 Price Quint 5 * Quarter 0 Price Quint 5 * Quarter 1 Price Quint 5 * Quarter 2 Price Quint 5 * Quarter 3 Price Quint 5 * Quarter 4 Price Quint 5 * Quarter 5 Price Quint 5 * Quarter 6 Price Quint 5 * Quarter 7 Price Quint 5 * 2+ Years Constant

(1)

(2)

(3)

(4)

(5)

(6)

0.162** (0.0698) 0.00446 (0.0671) -0.0239 (0.0671) -0.153** (0.0655) -0.0161 (0.0819) 0.0223 (0.0798) -0.0204 (0.0682) 0.0481 (0.0684) 0.0150 (0.0563) -0.0633 (0.0616) -0.0855 (0.0725) -0.0946 (0.0805) -0.149** (0.0694) -0.209*** (0.0706) -0.168** (0.0729) -0.273*** (0.0589) 0.250 (1.041)

0.165** (0.0702) 0.00757 (0.0675) -0.0201 (0.0675) -0.150** (0.0662) -0.0298 (0.0817) 0.00924 (0.0799) -0.0202 (0.0690) 0.0547 (0.0686) 0.0179 (0.0564) -0.0647 (0.0619) -0.0812 (0.0727) -0.0821 (0.0804) -0.149** (0.0699) -0.211*** (0.0711) -0.163** (0.0736) -0.271*** (0.0593) 0.221 (1.041)

0.322 (0.304) 0.341 (0.336) 0.199 (0.342) 0.0777 (0.303) -0.167 (0.364) 0.0512 (0.323) -0.129 (0.298) -0.193 (0.241) 0.0650 (0.285) -0.0384 (0.248) 0.261 (0.307) -0.0487 (0.331) 0.0488 (0.293) 0.120 (0.329) 0.160 (0.339) -0.251 (0.234) -3.399*** (0.567)

0.151 (0.279) 0.510 (0.344) 0.568 (0.348) 0.408 (0.286) 0.787 (0.608) 0.957 (0.652) 0.247 (0.338) 0.0145 (0.328) 0.0156 (0.215) 0.0360 (0.227) 0.126 (0.228) 0.226 (0.243) -0.00573 (0.502) -0.0475 (0.413) -0.121 (0.434) -0.277 (0.186) -3.343*** (0.696)

0.179** (0.0771) -0.00925 (0.0725) -0.0333 (0.0737) -0.169** (0.0726) -0.00208 (0.0885) 0.0248 (0.0865) 0.0187 (0.0749) 0.110 (0.0755) 0.0313 (0.0612) -0.0609 (0.0669) -0.0830 (0.0794) -0.0928 (0.0896) -0.133* (0.0768) -0.180** (0.0780) -0.146* (0.0806) -0.300*** (0.0657) 0.405 (1.074)

0.172** (0.0699) 0.0110 (0.0669) -0.0226 (0.0672) -0.146** (0.0658) -0.0126 (0.0819) 0.0233 (0.0797) -0.0164 (0.0683) 0.0575 (0.0697) 0.0249 (0.0556) -0.0614 (0.0588) -0.0958 (0.0738) -0.0973 (0.0818) -0.133* (0.0706) -0.233*** (0.0714) -0.169** (0.0732) -0.267*** (0.0579) 0.271 (1.007)

Observations 49,559 48,635 6,985 4,173 38,401 48,514 R-squared 0.268 0.266 0.395 0.408 0.264 0.268 Number of Firms 1,300 1,279 164 82 1,054 1,300 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

52

Table Appendix 10: Log Standard Deviation of Stock Return Error With Multiple Times

Price Quint 5 * Quarter -7 Price Quint 5 * Quarter -6 Price Quint 5 * Quarter -5 Price Quint 5 * Quarter -4 Price Quint 5 * Quarter -3 Price Quint 5 * Quarter -2 Price Quint 5 * Quarter -1 Price Quint 5 * Quarter 0 Price Quint 5 * Quarter 1 Price Quint 5 * Quarter 2 Price Quint 5 * Quarter 3 Price Quint 5 * Quarter 4 Price Quint 5 * Quarter 5 Price Quint 5 * Quarter 6 Price Quint 5 * Quarter 7 Price Quint 5 * 2+ Years Constant

(1)

(2)

(3)

(4)

(5)

(6)

-0.0272 (0.0353) -0.0622 (0.0381) -0.101*** (0.0385) -0.121** (0.0481) -0.153*** (0.0392) -0.126*** (0.0444) -0.0905** (0.0425) -0.0436 (0.0433) -0.118*** (0.0408) -0.113*** (0.0426) -0.157*** (0.0422) -0.159*** (0.0525) -0.230*** (0.0441) -0.261*** (0.0497) -0.202*** (0.0491) -0.244*** (0.0390) -1.772** (0.799)

-0.0300 (0.0355) -0.0650* (0.0384) -0.0993** (0.0387) -0.129*** (0.0482) -0.154*** (0.0396) -0.132*** (0.0446) -0.0940** (0.0429) -0.0507 (0.0442) -0.118*** (0.0411) -0.116*** (0.0430) -0.158*** (0.0425) -0.156*** (0.0529) -0.230*** (0.0445) -0.262*** (0.0500) -0.202*** (0.0494) -0.245*** (0.0393) -1.776** (0.800)

-0.174 (0.151) -0.0337 (0.141) -0.0745 (0.138) -0.364** (0.166) 0.0435 (0.150) -0.172 (0.154) -0.0930 (0.181) -0.175 (0.140) -0.0701 (0.126) -0.168 (0.215) -0.110 (0.167) 0.0581 (0.176) -0.462** (0.222) -0.298 (0.206) -0.206 (0.246) -0.218* (0.115) -3.977*** (0.284)

0.0424 (0.155) 0.253 (0.155) 0.319* (0.186) 0.195 (0.174) 0.244 (0.185) -0.0189 (0.242) -0.0105 (0.162) -0.128 (0.154) -0.217 (0.137) -0.244* (0.143) -0.292** (0.134) -0.102 (0.161) 0.237 (0.259) 0.0431 (0.237) -0.214 (0.268) -0.277** (0.107) -5.108*** (0.384)

-0.00375 (0.0377) -0.0460 (0.0421) -0.0820* (0.0428) -0.0915* (0.0531) -0.145*** (0.0427) -0.0834* (0.0479) -0.0213 (0.0451) 0.0145 (0.0476) -0.0980** (0.0437) -0.101** (0.0461) -0.163*** (0.0472) -0.159*** (0.0595) -0.235*** (0.0483) -0.245*** (0.0552) -0.165*** (0.0537) -0.263*** (0.0436) -1.588** (0.737)

-0.0235 (0.0353) -0.0577 (0.0379) -0.101*** (0.0385) -0.120** (0.0479) -0.149*** (0.0392) -0.125*** (0.0443) -0.0874** (0.0427) -0.0344 (0.0439) -0.120*** (0.0398) -0.120*** (0.0411) -0.174*** (0.0425) -0.161*** (0.0530) -0.216*** (0.0441) -0.260*** (0.0505) -0.202*** (0.0495) -0.241*** (0.0384) -1.731** (0.773)

Observations 49,845 48,919 7,024 4,189 38,632 48,795 R-squared 0.439 0.439 0.566 0.605 0.429 0.442 Number of Firms 1,300 1,279 164 82 1,054 1,300 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

53

Table Appendix 11: Log of Sensitivity to Price for New Grants With Multiple Times

Price Quint 5 * Quarter -7 Price Quint 5 * Quarter -6 Price Quint 5 * Quarter -5 Price Quint 5 * Quarter -4 Price Quint 5 * Quarter -3 Price Quint 5 * Quarter -2 Price Quint 5 * Quarter -1 Price Quint 5 * Quarter 0 Price Quint 5 * Quarter 1 Price Quint 5 * Quarter 2 Price Quint 5 * Quarter 3 Price Quint 5 * Quarter 4 Price Quint 5 * Quarter 5 Price Quint 5 * Quarter 6 Price Quint 5 * Quarter 7 Price Quint 5 * 2+ Years Constant

(1)

(2)

0.0295 (0.476) -0.724 (0.565) -1.195** (0.513) 0.335 (0.235) -0.471 (0.334) 0.103 (0.864) -0.859*** (0.255) -0.0495 (0.187) 0.912** (0.354) -0.323 (0.363) -0.722* (0.422) -0.398* (0.216) -0.490 (0.363) -0.475 (0.412) -1.275*** (0.375) -0.320* (0.167) -3.092*** (1.110)

0.0204 (0.475) -0.744 (0.562) -1.151** (0.517) 0.305 (0.232) -0.478 (0.335) 0.0877 (0.863) -0.835*** (0.249) -0.0743 (0.189) 0.897** (0.354) -0.340 (0.360) -0.680* (0.405) -0.401* (0.218) -0.503 (0.365) -0.478 (0.407) -1.216*** (0.359) -0.318* (0.165) -3.020*** (0.611)

(3)

(4)

-0.606 (0.614)

-1.620** (0.668)

-0.341 (0.326) -0.311 (0.230)

-0.730 (0.720)

-0.824** (0.401)

-1.864*** (0.543)

-1.128*** (0.244) -0.625 (0.448) 5.150*** (1.418)

-3.278*** (0.541) -1.555 (2.179)

(5)

(6)

0.0867 (0.485) -0.748 (0.547) -2.026*** (0.617) 0.454 (0.321) -0.305 (0.344) 0.197 (0.885) -1.194*** (0.298) -0.00493 (0.228) 0.940*** (0.354) -0.292 (0.374) -0.406 (0.351) -0.273 (0.247) -0.434 (0.379) -0.365 (0.391) -1.374*** (0.409) -0.131 (0.189) -0.744 (1.012)

0.0125 (0.484) -0.736 (0.567) -1.236** (0.510) 0.299 (0.245) -0.573* (0.337) 0.103 (0.864) -0.906*** (0.261) -0.178 (0.195) 0.604* (0.364) -0.323 (0.380) -0.635* (0.360) -0.367 (0.235) -0.543 (0.382) -0.546 (0.451) -1.242*** (0.380) -0.367** (0.180) -2.975*** (0.639)

Observations 5,379 5,285 945 399 4,035 5,226 R-squared 0.258 0.258 0.286 0.536 0.287 0.261 Number of Firms 1,018 1,000 151 75 792 1,016 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

54

Table Appendix 12: Log of Sensitivity to Price for All Stock and Options With Multiple Times

Price Quint 5 * Quarter -7 Price Quint 5 * Quarter -6 Price Quint 5 * Quarter -5 Price Quint 5 * Quarter -4 Price Quint 5 * Quarter -3 Price Quint 5 * Quarter -2 Price Quint 5 * Quarter -1 Price Quint 5 * Quarter 0 Price Quint 5 * Quarter 1 Price Quint 5 * Quarter 2 Price Quint 5 * Quarter 3 Price Quint 5 * Quarter 4 Price Quint 5 * Quarter 5 Price Quint 5 * Quarter 6 Price Quint 5 * Quarter 7 Price Quint 5 * 2+ Years Constant

(1)

(2)

0.184 (0.301) -0.157 (0.427) -1.021*** (0.354) -0.238 (0.159) -0.119 (0.432) -0.372 (0.503) -0.950*** (0.364) -0.837*** (0.197) -1.162*** (0.386) -0.707** (0.329) -0.897** (0.409) -1.100*** (0.185) -1.068*** (0.405) -1.338*** (0.364) -1.596*** (0.392) -1.449*** (0.166) -2.540*** (0.706)

0.174 (0.301) -0.171 (0.425) -1.026*** (0.357) -0.249 (0.160) -0.128 (0.433) -0.383 (0.501) -0.957*** (0.366) -0.829*** (0.197) -1.168*** (0.387) -0.717** (0.330) -0.864** (0.410) -1.110*** (0.187) -1.077*** (0.405) -1.343*** (0.364) -1.562*** (0.393) -1.456*** (0.168) -2.629*** (0.717)

(3)

(4)

1.060*** (0.304) 0.251 (0.799)

0.734 (0.798)

0.713 (0.453)

0.911 (0.848)

-1.518 (0.965)

-0.125 (0.586)

-1.973*** (0.617) -1.914 (1.638)

-0.592 (0.698) -2.498 (1.974)

(5)

(6)

0.208 (0.307) -0.138 (0.430) -1.144*** (0.380) -0.0973 (0.173) -0.0745 (0.436) -0.317 (0.501) -1.124*** (0.411) -0.818*** (0.225) -1.115*** (0.392) -0.738** (0.330) -0.834* (0.453) -1.019*** (0.212) -1.028** (0.409) -1.362*** (0.373) -1.645*** (0.453) -1.365*** (0.185) -2.430** (1.003)

0.179 (0.302) -0.137 (0.427) -1.011*** (0.358) -0.257 (0.160) -0.149 (0.434) -0.350 (0.507) -0.905** (0.380) -0.841*** (0.200) -1.188*** (0.398) -0.719** (0.334) -0.899** (0.438) -1.069*** (0.186) -1.101*** (0.410) -1.332*** (0.366) -1.667*** (0.390) -1.458*** (0.167) -2.690*** (0.713)

Observations 10,266 10,106 1,289 680 8,297 10,066 R-squared 0.345 0.343 0.416 0.634 0.353 0.345 Number of Firms 1,274 1,253 160 81 1,033 1,274 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only includes utilities. Column (5) only considers firms that are not in regulated industries (utilities and financial firms). Column (6) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

55

Table Appendix 13: Log Standard Deviation of Stock Price Triple Difference

Price Qnt 5 * After Tax * Div Pay Price Qnt 4 * After Tax * Div Pay Price Qnt 3 * After Tax * Div Pay Price Qnt 2 * After Tax * Div Pay Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assets Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Constant

(1)

(2)

(3)

(4)

(5)

-0.339*** (0.0751) -0.223*** (0.0681) -0.115 (0.0709) -0.00629 (0.0630) -0.0435 (0.0607) 0.00142 (0.0521) -0.00876 (0.0564) -0.0277 (0.0460) 0.619*** (0.0272) 0.218*** (0.0244) -0.109*** (0.0261) -0.00267 (0.0211) -0.0337*** (0.00863) 0.119*** (0.0173) -0.0883 (0.0581) 0.766*** (0.0377) -0.162*** (0.0596) -3.405*** (0.150)

-0.343*** (0.0758) -0.227*** (0.0697) -0.116 (0.0713) -0.00876 (0.0638) -0.0379 (0.0613) 0.00550 (0.0538) -0.00404 (0.0567) -0.0260 (0.0465) 0.625*** (0.0275) 0.216*** (0.0247) -0.107*** (0.0265) -0.00151 (0.0215) -0.0359*** (0.00873) 0.118*** (0.0176) -0.0887 (0.0583) 0.758*** (0.0379) -0.160*** (0.0600) -3.383*** (0.153)

-0.806*** (0.151) -0.345** (0.133) -0.692*** (0.162) -0.385* (0.219) 0.365*** (0.0544)

-0.347*** (0.0800) -0.224*** (0.0741) -0.121 (0.0767) -0.000160 (0.0700) -0.0613 (0.0609) -0.00659 (0.0526) -0.0149 (0.0567) -0.0372 (0.0475) 0.614*** (0.0324) 0.237*** (0.0288) -0.122*** (0.0314) 0.00809 (0.0222) -0.0294*** (0.00960) 0.126*** (0.0200) -0.117** (0.0589) 0.767*** (0.0435) -0.156** (0.0653) -3.372*** (0.153)

-0.339*** (0.0753) -0.222*** (0.0681) -0.118* (0.0708) -0.00379 (0.0630) -0.0451 (0.0610) 0.00138 (0.0523) -0.00848 (0.0565) -0.0275 (0.0462) 0.610*** (0.0272) 0.214*** (0.0240) -0.104*** (0.0258) 0.00126 (0.0211) -0.0355*** (0.00864) 0.121*** (0.0171) -0.0793 (0.0577) 0.759*** (0.0377) -0.148** (0.0580) -3.351*** (0.148)

0.422*** (0.121) 0.0716 (0.164) 0.553*** (0.0559) 0.0655 (0.0499) 0.0130 (0.0552) 0.0159 (0.0624) -0.0742*** (0.0270) 0.166*** (0.0446) -0.229 (0.375) 1.065*** (0.178) -0.160 (0.142) -3.918*** (0.320)

Observations 49,846 48,920 7,024 38,633 48,796 R-squared 0.466 0.464 0.422 0.489 0.466 Number of Firms 1,300 1,279 164 1,054 1,300 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only considers firms that are not in regulated industries (utilities and financial firms). Column (5) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

56

Table Appendix 14: Log Standard Deviation of Stock Price Error Triple Difference Price Qnt 5 * After Tax * Div Pay Price Qnt 4 * After Tax * Div Pay Price Qnt 3 * After Tax * Div Pay Price Qnt 2 * After Tax * Div Pay Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assets Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Constant

(1) -0.331*** (0.0713) -0.208*** (0.0652) -0.142** (0.0664) -0.0184 (0.0602) -0.0340 (0.0574) -0.00938 (0.0503) 0.00698 (0.0525) -0.0321 (0.0443) 0.685*** (0.0257) 0.255*** (0.0230) -0.163*** (0.0241) -0.0128 (0.0209) -0.0348*** (0.00800) 0.0622*** (0.0157) -0.108** (0.0549) 0.755*** (0.0354) -0.149*** (0.0548) -3.421*** (0.148)

(2) -0.333*** (0.0720) -0.209*** (0.0669) -0.142** (0.0669) -0.0195 (0.0610) -0.0297 (0.0580) -0.00701 (0.0519) 0.0106 (0.0529) -0.0300 (0.0449) 0.690*** (0.0259) 0.253*** (0.0233) -0.162*** (0.0244) -0.0126 (0.0213) -0.0363*** (0.00809) 0.0610*** (0.0159) -0.108* (0.0551) 0.747*** (0.0356) -0.148*** (0.0551) -3.401*** (0.151)

(3) -0.638*** (0.145) -0.267** (0.121) -0.508*** (0.151) -0.195 (0.211) 0.265*** (0.0687)

0.286** (0.118) -0.0341 (0.167) 0.583*** (0.0526) 0.0876* (0.0469) -0.0126 (0.0483) 0.00456 (0.0612) -0.0842*** (0.0264) 0.136*** (0.0390) -0.267 (0.362) 0.996*** (0.173) -0.159 (0.138) -3.976*** (0.313)

(4) -0.341*** (0.0761) -0.217*** (0.0709) -0.152** (0.0720) -0.0240 (0.0669) -0.0476 (0.0573) -0.0152 (0.0507) 0.00357 (0.0526) -0.0363 (0.0456) 0.683*** (0.0303) 0.273*** (0.0271) -0.178*** (0.0290) 0.000163 (0.0220) -0.0254*** (0.00879) 0.0639*** (0.0182) -0.146*** (0.0556) 0.754*** (0.0408) -0.140** (0.0596) -3.357*** (0.151)

(5) -0.329*** (0.0712) -0.206*** (0.0650) -0.145** (0.0662) -0.0151 (0.0601) -0.0354 (0.0577) -0.00948 (0.0505) 0.00712 (0.0527) -0.0318 (0.0445) 0.678*** (0.0256) 0.252*** (0.0227) -0.159*** (0.0239) -0.00871 (0.0210) -0.0365*** (0.00799) 0.0632*** (0.0155) -0.101* (0.0547) 0.749*** (0.0352) -0.135** (0.0538) -3.365*** (0.148)

Observations 49,846 48,920 7,024 38,633 48,796 R-squared 0.469 0.467 0.416 0.495 0.470 Number of Firms 1,300 1,279 164 1,054 1,300 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only considers firms that are not in regulated industries (utilities and financial firms). Column (5) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

57

Table Appendix 15: Log Standard Deviation of Stock Return Triple Difference Price Qnt 5 * After Tax * Div Pay Price Qnt 4 * After Tax * Div Pay Price Qnt 3 * After Tax * Div Pay Price Qnt 2 * After Tax * Div Pay Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assets Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Constant

(1) -0.340*** (0.0866) -0.250*** (0.0859) -0.127 (0.0857) -0.0357 (0.0803) 0.0186 (0.0676) 0.0413 (0.0617) -0.0140 (0.0655) 0.0244 (0.0563) 0.0283 (0.0321) 0.307*** (0.0330) -0.274*** (0.0350) -0.0916*** (0.0284) -0.0288*** (0.00975) 0.0486** (0.0210) -0.173** (0.0686) 0.890*** (0.0501) -0.126 (0.0822) -3.339*** (0.204)

(2) -0.342*** (0.0872) -0.250*** (0.0871) -0.129 (0.0862) -0.0313 (0.0811) 0.0238 (0.0682) 0.0485 (0.0630) -0.0108 (0.0659) 0.0224 (0.0569) 0.0328 (0.0324) 0.306*** (0.0335) -0.275*** (0.0355) -0.0945*** (0.0292) -0.0313*** (0.00986) 0.0480** (0.0213) -0.180*** (0.0694) 0.881*** (0.0505) -0.122 (0.0827) -3.339*** (0.207)

(3) -0.685*** (0.209) -0.375* (0.197) -0.874*** (0.242) -0.308 (0.277) 0.295*** (0.0703)

0.602*** (0.146) -0.0154 (0.171) 0.00106 (0.0814) 0.246*** (0.0922) -0.212** (0.0857) -0.0370 (0.0833) -0.127*** (0.0294) 0.0812 (0.0668) -0.0328 (0.462) 1.280*** (0.231) 0.146 (0.229) -3.409*** (0.576)

(4) -0.362*** (0.0924) -0.263*** (0.0937) -0.135 (0.0925) -0.0327 (0.0895) 0.00427 (0.0684) 0.0348 (0.0623) -0.0222 (0.0664) 0.0214 (0.0585) 0.0172 (0.0374) 0.304*** (0.0391) -0.275*** (0.0427) -0.0802*** (0.0297) -0.0186* (0.0107) 0.0513** (0.0243) -0.196*** (0.0707) 0.870*** (0.0570) -0.117 (0.0903) -3.176*** (0.209)

(5) -0.339*** (0.0861) -0.243*** (0.0858) -0.130 (0.0850) -0.0313 (0.0797) 0.0195 (0.0677) 0.0431 (0.0619) -0.0120 (0.0655) 0.0258 (0.0564) 0.0175 (0.0323) 0.301*** (0.0330) -0.268*** (0.0349) -0.0881*** (0.0283) -0.0306*** (0.00970) 0.0504** (0.0210) -0.159** (0.0671) 0.883*** (0.0501) -0.109 (0.0812) -3.305*** (0.198)

Observations 49,559 48,635 6,985 38,401 48,514 R-squared 0.259 0.257 0.376 0.255 0.259 Number of Firms 1,300 1,279 164 1,054 1,300 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only considers firms that are not in regulated industries (utilities and financial firms). Column (5) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

58

Table Appendix 16: Log Standard Deviation of Stock Return Error Triple Difference Price Qnt 5 * After Tax * Div Pay Price Qnt 4 * After Tax * Div Pay Price Qnt 3 * After Tax * Div Pay Price Qnt 2 * After Tax * Div Pay Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assets Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Constant

(1) -0.237*** (0.0581) -0.167*** (0.0565) -0.0880 (0.0551) 0.0168 (0.0510) -0.0150 (0.0443) 0.0274 (0.0428) -0.00547 (0.0421) -0.0327 (0.0359) -0.233*** (0.0239) 0.169*** (0.0207) -0.146*** (0.0237) -0.0412** (0.0178) -0.0259*** (0.00582) 0.0286** (0.0141) -0.100** (0.0408) 0.690*** (0.0308) -0.166*** (0.0465) -4.055*** (0.145)

(2) -0.236*** (0.0586) -0.165*** (0.0578) -0.0841 (0.0554) 0.0187 (0.0516) -0.0142 (0.0448) 0.0283 (0.0442) -0.00425 (0.0423) -0.0303 (0.0364) -0.231*** (0.0242) 0.168*** (0.0210) -0.147*** (0.0240) -0.0425** (0.0181) -0.0253*** (0.00587) 0.0286** (0.0143) -0.0995** (0.0410) 0.688*** (0.0311) -0.169*** (0.0468) -4.056*** (0.147)

(3) -0.210** (0.0909) -0.0780 (0.0655) -0.318*** (0.103) -0.0245 (0.197) 0.0304 (0.0539)

0.243** (0.0992) -0.101 (0.181) -0.348*** (0.0453) 0.0746* (0.0405) -0.0229 (0.0426) 0.0240 (0.0562) -0.0646*** (0.0206) 0.0985*** (0.0341) -0.392* (0.235) 0.920*** (0.148) -0.235* (0.133) -4.449*** (0.274)

(4) -0.264*** (0.0619) -0.197*** (0.0603) -0.101* (0.0589) 0.0290 (0.0549) -0.0232 (0.0443) 0.0245 (0.0430) -0.00797 (0.0425) -0.0367 (0.0359) -0.228*** (0.0282) 0.162*** (0.0251) -0.144*** (0.0294) -0.0221 (0.0181) -0.0141** (0.00613) 0.0255 (0.0164) -0.143*** (0.0418) 0.639*** (0.0346) -0.142*** (0.0504) -3.906*** (0.149)

(5) -0.236*** (0.0577) -0.168*** (0.0562) -0.0912* (0.0547) 0.0165 (0.0508) -0.0153 (0.0443) 0.0279 (0.0429) -0.00498 (0.0421) -0.0325 (0.0360) -0.237*** (0.0241) 0.168*** (0.0208) -0.146*** (0.0238) -0.0397** (0.0176) -0.0266*** (0.00580) 0.0273* (0.0140) -0.0924** (0.0405) 0.689*** (0.0309) -0.155*** (0.0466) -4.018*** (0.144)

Observations 49,845 48,919 7,024 38,632 48,795 R-squared 0.435 0.434 0.551 0.426 0.438 Number of Firms 1,300 1,279 164 1,054 1,300 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only considers firms that are not in regulated industries (utilities and financial firms). Column (5) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

59

Table Appendix 17: Log of Sensitivity to Price for Grants Triple Difference

Price Qnt 5 * After Tax * Div Pay Price Qnt 4 * After Tax * Div Pay Price Qnt 3 * After Tax * Div Pay Price Qnt 2 * After Tax * Div Pay Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assets Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Constant

(1)

(2)

(3)

(4)

-0.944* (0.530) -0.433 (0.550) -0.597 (0.554) -1.169** (0.535) 0.579 (0.522) 0.292 (0.542) 0.570 (0.547) 1.143** (0.524) 0.174 (0.112) -0.192* (0.112) 0.761*** (0.120) -0.380 (0.344) -0.0513 (0.0395) 0.175** (0.0822) -0.409 (0.559) 0.137 (0.173) 0.0508 (0.233) -0.182 (0.630)

-0.852 (0.531) -0.347 (0.549) -0.507 (0.555) -1.073** (0.540) 0.494 (0.524) 0.220 (0.541) 0.486 (0.548) 1.049** (0.529) 0.160 (0.113) -0.170 (0.112) 0.765*** (0.121) -0.363 (0.349) -0.0579 (0.0396) 0.179** (0.0831) -0.397 (0.564) 0.110 (0.175) 0.00225 (0.235) -0.377 (0.632)

-0.868 (0.537) -0.265 (0.554) -0.546 (0.550) -1.116** (0.547) 0.625 (0.527) 0.288 (0.544) 0.600 (0.541) 1.189** (0.535) 0.250* (0.145) -0.0485 (0.161) 0.653*** (0.168) -0.430 (0.366) -0.0484 (0.0458) 0.143 (0.108) -0.330 (0.608) 0.351 (0.220) 0.0198 (0.268) -1.283* (0.743)

-0.965* (0.545) -0.449 (0.560) -0.623 (0.562) -1.197** (0.555) 0.584 (0.536) 0.295 (0.552) 0.575 (0.555) 1.146** (0.541) 0.155 (0.113) -0.219* (0.113) 0.792*** (0.122) -0.438 (0.343) -0.0510 (0.0398) 0.188** (0.0823) -0.392 (0.557) 0.113 (0.176) 0.0638 (0.234) -0.200 (0.645)

Observations 5,379 5,285 4,035 5,226 R-squared 0.222 0.223 0.243 0.224 Number of Firms 1,018 1,000 792 1,016 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only considers firms that are not in regulated industries (utilities and financial firms). Column (4) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

60

Table Appendix 18: Log of Sensitivity to Price for All Stock and Options Triple Difference

Price Qnt 5 * After Tax * Div Pay Price Qnt 4 * After Tax * Div Pay Price Qnt 3 * After Tax * Div Pay Price Qnt 2 * After Tax * Div Pay Price Qnt 5 * After Tax Price Qnt 4 * After Tax Price Qnt 3 * After Tax Price Qnt 2 * After Tax Log Price Log Assets Log Market Cap Log Profit Margin Log EPS Log Sales PS Log ROA Log Market to Book Log Leverage Constant

(1)

(2)

(3)

(4)

(5)

-0.399 (0.292) -0.393 (0.241) 0.249 (0.244) 0.0663 (0.212) -0.819*** (0.252) -0.569*** (0.193) -1.081*** (0.183) -0.630*** (0.160) 0.153 (0.112) 0.110 (0.0844) 0.673*** (0.106) 0.131 (0.105) 0.0214 (0.0284) -0.0127 (0.0844) -0.163 (0.153) 0.617*** (0.137) 0.378* (0.194) -1.510** (0.626)

-0.420 (0.296) -0.391 (0.243) 0.246 (0.246) 0.0693 (0.213) -0.804*** (0.254) -0.586*** (0.194) -1.080*** (0.184) -0.658*** (0.160) 0.137 (0.113) 0.110 (0.0854) 0.685*** (0.107) 0.136 (0.108) 0.0216 (0.0286) -0.0162 (0.0858) -0.167 (0.154) 0.630*** (0.138) 0.394** (0.195) -1.542** (0.633)

-2.152** (0.944) -2.767*** (0.894) -1.829* (1.018) -2.365** (0.946) -0.841*** (0.156)

-0.294 (0.309) -0.290 (0.255) 0.447* (0.267) 0.245 (0.230) -0.826*** (0.258) -0.562*** (0.195) -1.096*** (0.186) -0.686*** (0.166) 0.171 (0.128) 0.165* (0.100) 0.601*** (0.125) 0.109 (0.105) 0.0311 (0.0308) -0.0919 (0.0979) -0.110 (0.152) 0.635*** (0.154) 0.252 (0.204) -1.074 (0.705)

-0.406 (0.294) -0.404* (0.243) 0.236 (0.246) 0.0543 (0.213) -0.820*** (0.252) -0.568*** (0.193) -1.081*** (0.183) -0.627*** (0.160) 0.153 (0.114) 0.143* (0.0838) 0.650*** (0.106) 0.130 (0.106) 0.0235 (0.0283) -0.0127 (0.0859) -0.175 (0.153) 0.654*** (0.137) 0.344* (0.195) -1.670*** (0.631)

-1.010* (0.536) -0.416* (0.212) 0.200 (0.286) -0.145 (0.247) 0.787*** (0.271) 0.816 (0.849) -0.00271 (0.0830) 0.000288 (0.219) -2.963* (1.572) -0.365 (0.955) 1.823*** (0.590) 0.145 (1.800)

Observations 10,266 10,106 1,289 8,297 10,066 R-squared 0.323 0.321 0.376 0.330 0.321 Number of Firms 1,274 1,253 160 1,033 1,274 Notes: Clustered standard errors in parentheses. Column (1) includes all firms. Column (2) includes only firms based in the United States. Column (3) only includes financial firms. Column (4) only considers firms that are not in regulated industries (utilities and financial firms). Column (5) does not include quarterly observations for firms that initiated a regular dividend in the quarter being examined or increased the regular dividend more that 20% in the quarter. *** p<0.01, ** p<0.05, * p<0.1

61