Seasoned Equity Offerings and Payout Policy

Transcription

1 Seasoned Equity Offerings and Payout Policy Mark D. Walker a,*, Keven Yost b a North Carolina State University, Raleigh, North Carolina 27695, United States b Auburn University, Auburn, Alabama 36849, United States September 2017 * Corresponding author. Tel.: addresses: mdwalker@ncsu.edu (M. Walker),yostkev@auburn.edu (K. Yost)

2 Abstract We investigate firms that choose to issue seasoned equity following a period of paying dividends. Firms that engage in this roundtrip of equity and suspend dividends exhibit the strongest growth in total assets and investment in capital expenditures. Dividend suspension does not appear to be associated with future declines in operating cash flow, but rather is more associated with large absolute increases in future cash flow. Our findings are not consistent with firms issuing equity, and stopping dividend payments, to time the market or to compensate for poor future operating performance. Rather, equity-issuing firms that suspend dividends use the capital for investments in fixed assets consistent with market assessments of growth opportunities.

3 1. Introduction A firm s life cycle generally follows the pattern of capital flowing to the firm to fund projects then eventually back to investors as these projects generate cash flow. Over time, operating cash flows exceed the financing needs associated with new, valuable projects with the difference either being paid out to security holders or held as precautionary reserves. 1 Consistent with this view, previous studies on SEOs find that it is most often relatively younger and smaller firms that issue primary equity. However, not all firms follow this pattern. Some firms elect to issue equity concurrent with, or immediately following periods of payments to equity holders. Given floatation costs and potential signaling costs, this round trip of equity capital appears costly. In this study, we explore why firms engage in a round trip of equity capital. We focus on firms that have paid dividends in the year prior to engaging in an SEO and we explore three non-mutually exclusive reasons for these firms choosing to issue equity. First, a firm might be experiencing or anticipating poor operating performance and need equity financing in order to ensure the viability of the firm (viability motive). Second, the firm might have new unexpectedly large and valuable projects without sufficient cash flows or debt capacity to finance the investments (investment motive). Third, the firm might perceive an abnormally low cost of equity, possibly due to temporary overvaluation of the firm (timing motive). This view supposes that firms issue equity during periods of market euphoria either to fund projects with equity rather than debt or to facilitate a wealth transfer from new to existing shareholders. This timing motive is predicated on management s belief that the market will not sufficiently update the firm s valuation following the announcement of the firm s intention to issue equity. 2 1 This is described in DeAngelo, DeAngelo, and Stulz (2006), as well as in popular textbooks such as those by Ross, Westerfield, and Jaffe, or Brealey and Myers. 2 While controversial, this view does have some empirical support. See, for example, Baker and Wurgler (2002). 1

4 We find that SEO issuing firms engaging in equity roundtrips are larger, invest more, and have positive operating cash flows. In contrast, SEO firms that have not made equity payouts in the preceding five years have negative operating cash flows, but have higher cash balances prior to the SEO. We find that both groups have similarly high market-to-book ratios. Additionally, both groups have a positive cash burn rate in that their cash flows are smaller than their collective capital investment and dividend programs. Equity payout firms that decide to reverse course and raise equity capital have an entangled decision of whether to suspend dividend payments to fund a portion of the project or to maintain their dividend stream. Dividends have been described as financial slack in that dividends can be cut in order to provide cash flow to fund new projects. 3 Cutting dividends has the virtue of reducing the amount of new equity needed to fund the firm s investment program. However, dividends are generally viewed as a commitment; a costly signal of future cash flows, thus cutting dividends might represent a negative signal regarding the firm s future prospects as described by Bhattacharya (1979) and in Miller and Rock (1982). Consistent with this view, Brav, Graham, Harvey, and Michaely (2005) survey CFOs who state that cutting investment is preferable to cutting dividends. Roundtrip firms that choose to maintain versus discontinue their dividends have distinct characteristics. These two groups have similarly positive cash flows and capital investment programs. However, firms that discontinue their dividends (dividend-to-zero firms) are much smaller in terms of book assets and have a much higher ex ante dividends-to-assets ratio (16.4%) as compared to those firms that maintain a dividend program (4.0%) subsequent to the SEO. We calculate a burn rate that is dividends and capital expenditures minus cash flows. Firms that 3 For example, see Kaplan and Zingales (1997) and Bliss, Cheng, and Denis (2015). 2

5 ultimately cut their dividends had a positive burn rate that suggests that all cash reserves would have been exhausted in 1.2 years without the SEO and dividend cessation, as compared to 2.9 years for the roundtrip SEO firms that maintain positive dividends and 2.4 years for non-roundtrip SEO firms. Following the SEO, the dividend-to-zero firms bring their burn rate to near balance. To test the investment motive, we examine firms investment programs around the time of the SEO. Prior to the issue, the capital investment programs for roundtrip firms are around 11% of their asset base per year as compared to 7% for non-roundtrip SEO firms. Following the SEO, the dividend-to-zero firms exhibit the largest increase in capital investment: 25% larger one year subsequent to the SEO than one year prior. The increase in investment continues as by the third year following the SEO the dividend-to-zero firms invest over 40% more than in the year prior to the SEO. The three-year change is a more modest 15% increase for roundtrip firms that maintain dividends. For non-roundtrip SEO firms, the increase is 21%. We also observe these larger increases in investment programs for dividend-to-zero firms in multivariate analyses. We conclude that SEO firms, and more so for the dividend-to-zero firms, use the newly raised capital for large investment programs rather than to change capital structure or hold high cash balances. Turning to the viability motive, we look at operating cash flows around the SEO. Dividendto-zero firms have chosen to access equity capital via cessation of dividends that might also signal poor future cash flows. Operating cash flows are 20% higher in the year subsequent to the SEO relative to the year prior and 46% higher three years subsequent to the SEO relative to year -1. As a ratio to book assets, which grows substantially, the dividend-to-zero firms exhibit a modest decline in cash flow from 9.3% to 8.7% in year +1 and to 7.8% in year +3 relative to year -1. The evidence does not appear to conform to dividend cessation signaling future poor performance for these SEO firms. 3

6 While our evidence as to why firms raise additional capital appears to be most consistent with investment opportunities rather than anticipating poor operating performance, it does not point to why these firms engage in an equity roundtrip rather than simply issuing debt. Recently, some research, such as DeAngelo, DeAngelo, and Whited (2011), and Denis and McKeon (2012), has argued that maintaining financial flexibility is a primary determinant in the debt versus equity choice. Firms maintain debt capacity in order to provide financial flexibility to embark on investment programs. Using debt rather than equity to finance new projects mitigates costs associated with information asymmetries and agency issues. The roundtrip dividend-paying SEO sample has a 23% leverage ratio in the year prior to the SEO, which in and of itself does not appear to preclude a debt issue. Farre-Mensa, Michaely, and Schmalz (2015) show that many firms finance their dividend payouts, but mainly do so with debt rather than equity. In our sample of firms that have chosen to engage in an equity roundtrip, factors in favor of equity include their positive burn rate as well as the size of the issue. The roundtrip firms raise capital that is equal to 80% of the book value of the firm in the prior year. These firms do also increase their debt. Roundtrip firms that maintain dividends hold 25% more debt in year +1 than in year -1. The roundtrip, dividend-to-zero firms have 39% more debt. The total amount of capital raised (debt and equity) and the cash burn rate for these firms might make raising capital exclusively with debt prohibitively expensive. The choice to issue equity brings us to our third, non-mutually exclusive explanation for engaging in an equity roundtrip: timing equity issues during periods of relatively lower cost of equity. Prior evidence in favor of the timing motive suggests that firms tend to issue equity when valuations are high, then experience poor stock returns in the several years following the issue. 4 4 See Loughran and Ritter (1995). 4

7 The timing-based explanation for equity issues remains controversial, as a variety of alternative explanations for these observed facts include the rational adjustment of prices associated with converting growth options into assets, agency based explanations, and mis-specified models. 5 The timing motive suggests that there is a disconnection between valuation and growth opportunities as perceived by the firm s management in that the market has a more optimistic view of the firm s prospects than management. We attempt to discern whether the sample firms make investment choices that conform to the market s perception of the firm s growth opportunities that would suggest that management and the market have similar beliefs. We find that investment growth for non-roundtrip SEO firms is related to the firms current market-to-book ratios (as a proxy for Q) for at least two years following the issue. We also find that for roundtrip, dividendto-zero firms, increases in investment are related to Q for all three post-seo years that we examine. These firms investment patterns appear to be connected to the relative valuation of the firm. We cannot rule out that both the market and firm were overly optimistic in their assessments of the value of the firm s growth opportunities. The positive relation between ex ante valuations and increases in capital investment is more consistent with management agreeing with the market s assessment of growth opportunities rather than management trying to take advantage of a temporary over-valuation. However, we do not find a significant relation between Q and increases in investment for roundtrip firms that maintain dividends. We continue our exploration of the timing motive by examining firm acquisitions following the SEO. We expect that firms that issue over valued equity to engage in more acquisitions using stock as payment rather than cash. In other words, firms that knowingly hold over valued equity 5 For converting growth options into assets, see Carlson, Fisher, and Giammarino (2006). For agency based explanations, see Kim, Jung, and Stulz (1996) and Kim and Purnanandam (2014). Carlson et al. (2006) also suggests prior models are mis-specified. See Brav, Geczy, and Gompers (2000) for further evidence of model misspecification relative to uniquely poor performance subsequent to equity issues. 5

8 want to exchange stock for real assets before valuations return to appropriate levels. We find that SEO firms do engage in more acquisitions following the SEO, but roundtrip firms do not increase their use of stock for payment. Roundtrip firms that engage in acquisitions do not appear to behave in a manner that is consistent with knowingly having over-valued equity. In total, our evidence suggests that firms with a recent history of dividends choose to issue equity to continue and expand investment programs rather than to compensate for poor cash flows or to affect a wealth transfer from new shareholders. Cooper and Lambertides (2017) show that dividend-increasing firms are often more mature firms with declining reinvestment. Our evidence suggests that these dividend-cutting firms are the opposite; firms that have a positive growth opportunity shock and are moving back to a growth stage. Collectively, our evidence suggests that dividends can be an important source of financial slack to fund new, valuable growth opportunities. We continue in section two by describing the sample. In section three, we report our main results. We further discuss our findings and conclude in section four. 2. Our Sample and Descriptive Statistics We obtain our data from multiple sources. We start with Thomson SDC to obtain our initial set of firms that conduct SEOs. We use Thomson SDC for repurchase programs and COMPUSTAT for dividend payments. We require COMPUSTAT data for a three-year period surrounding the SEO year (years -1 to +1), although we also conduct tests looking out through year +3. We exclude 26 SEO firms that have engaged in a repurchase program in the five-year period prior to the SEO as these firms do not readily fit into our analysis. 6 We utilize data from 6 These firms are arguably roundtrip firms but do not have the same ongoing commitment as do firms that pay dividends. Moreover, the relative scarcity of these firms does not allow for a systematic examination. 6

9 1990 to We limit our sample to 1994 through 2012 in order to check whether our sample firms conducted a repurchase in the five years preceding the SEO and allowing us to examine firm investment characteristics for three years following the SEO. We have 1,699 SEOs that meet our data requirements. We summarize our sample in Table 1. We define our roundtrip (RT) SEO firms as those that paid dividends to common stock in the year prior to the SEO. Out of our 1,699 sample SEOs, 156 of the firms paid common dividends in the year prior to the SEO. The average size of the SEO is larger than recent payout activity. For dividend payers, the average SEO was $112 million as compared to $23 million in dividends in the prior year. The relative amounts reveal that not having paid dividends over the several years prior to the issue would have comprised much of the funds that were subsequently raised in the SEO. In panel B of Table 1, we take a closer look at the dividend paying firms. Our interest here is in reporting how many of these firms discontinue their dividend programs in the period around their SEO. By the year following the SEO, 71 out of the 156 dividend paying SEO firms discontinue paying common dividends. We categorize these 71 firms as dividend-to-zero firms (DIV to 0). In Table 2, we report the time trend of the SEO issues. We divide our sample into three groups: roundtrip firms that maintain positive dividends, roundtrip firms that discontinue paying dividends, and non-roundtrip firms. Equity issues do have periods of greater or lesser activity. The mid-to-late 1990s had relatively more equity issues and relatively higher proportion of those issues were from firms that discontinued dividends. Roundtrip firms that maintained dividends do not appear to exhibit any notable time clustering. 7

10 In Table 3, we divide our roundtrip SEO firms into separate groups to examine their ex ante characteristics. In panel A, we report univariate statistics for those firms issuing seasoned equity that have been paying dividends (Roundtrip) relative to non-roundtrip SEO firms. All of our variables are from the year preceding this issue and denominated by book assets (A). We report the raw statistics and we report industry-adjusted statistics, where industry adjustments are made by subtracting the median values of each variable for firms in the same 3 digit SIC code each year. Roundtrip SEO firms are larger, invest more in capital expenditures, and less in research and development (RD/A) as compared to non-roundtrip firms. Roundtrip firms hold less cash (CASH/A), have more fixed assets (PPE/A), and more debt. The issue size of the SEO (IS/A) is somewhat larger for Non-roundtrip firms relative to the ex ante book assets of the firm. The relative valuations for roundtrip firms are similar to the non-roundtrip group. Industry does account for a portion of these differences. On an industry-adjusted basis, the roundtrip firms have significantly greater capital expenditures, less research and development, and less cash than the non-roundtrip SEO firms. In panel B, we examine the same variables between subsets of the roundtrip firms; those firms that maintain positive dividends following the SEO and dividend-to-zero firms. Roundtrip firms that maintain dividends are larger, have less cash, more fixed assets, and more debt. Roundtrip firms that maintain dividends also have lower valuations in terms of market-to-book ratios. Current investment programs appear to be of a similar size, however. Firms that stop paying dividends do have, counter-intuitively, more cash; however, the higher valuations indicate that these firms also have more valuable uses for their cash. The differences between these two groupings of roundtrip firms can almost entirely be explained by industry as all differences are 8

11 insignificant after industry-adjustment. The only difference that retains significance is the lower valuations for the maintain dividends firms relative to the dividend to zero firms. Broadly, we can view these differences in the context of our explanations for why these firms might choose to eliminate dividends. Two of our explanations center on their high valuations. A rational story suggests that these high valuations correspond to valuable growth opportunities that need funding. The firms with the highest valuations are those that cease paying dividends, which is consistent with shifting to more (valuable) growth, which is the investment motive. In contrast, if the firm has higher market valuations as compared to an internal assessment, it might desire to move the firm s financial structure further towards equity by issuing equity regardless of the true nature of the firm s growth prospects (timing motive). We continue our analysis by investigating how these firms use the newly raised capital as well as examining other firm characteristics before and after the SEO to garner further insight differentiating our hypotheses. 3. Ex Post Use of Funds and Valuation This section comprises our main tests, where we focus on the use of funds and operating performance following the SEO for our roundtrip dividend firms and our non-roundtrip firms. We begin with a univariate analysis of the uses and sources of cash for our groups of firms. We then estimate multivariate models starting with cash flows in order to examine whether dividend cessation is predictive of poor future cash flows. We follow with estimates of the connection between the market s ex ante assessments of the firms growth opportunities and actual capital expenditures. We finish with an examination of acquisitions and the method of payment of those acquisitions. 9

12 3.1. Sources and Uses of Funds Our variables of interest are the increase in book assets (A), operating cash flow from the statement of cash flows (CF), investment in capital expenditures (CX), and common dividends (DIV). We denominate these variables by the firm s book assets. Using these variables, we calculate a burn rate. A burn rate is a term taken from the venture capital industry to examine the use of cash relative to cash inflows. We follow Denis and McKeon (2016) and calculate the burn rate as DIV plus CX minus CF. Finally, we include cash (CASH/A) and leverage (D/A). In Table 4, we report the uses and sources of funds by our three groups: roundtrip/maintain dividend firms, roundtrip/dividend-to-zero firms, and non-roundtrip firms. We examine our variables of interest in two manners. In panel A, we use the firm s concurrent asset base. Doing so allows us to account for relative changes in the firm s asset size. However, it is important to note that book assets substantively increase with an SEO, therefore a relative decline as reflected by the ratio still likely reflects an absolute increase in the numerator. Thus, in Panel B, we use the asset base in the year prior to the SEO to denominate all of our variables of interest. This allows us to observe the investment in each of these areas over time relative to a common starting point. We discuss panel A and panel B together. Not surprisingly following an SEO, all three groups exhibit significant increases in their asset base. The roundtrip firms that discontinue their dividends exhibit the most dramatic growth. The average book assets for these firms nearly quadruple from $191 million to $763 million from year -1 to year +1. The roundtrip firms that maintain their dividends increase at a more modest 45% from year -1 to +1. Non-roundtrip firms grow, on average, by more than 90% over the two years surrounding the SEO. All three groups continue to grow at more modest rates from years +1 to

13 Roundtrip SEO firms have much higher operating cash flow as compared to non-roundtrip SEO firms. Non-roundtrip SEO firms have a -7.9% CF/A ratio in the year prior to the SEO, which is significantly different from each of the groups of roundtrip SEO firms. The CF/A ratio modestly improves to -6.3% by year +3. Roundtrip firms that continue to pay dividends have a CF/A ratio of 10.7% in year -1 and have a modest decrease to 9.3% by year +3. Similarly, dividend-to-zero roundtrip firms have a 9.3% cash flow to assets ratio in year -1, decreasing to 7.8% by year +3. However, these statistics mask a large increase in cash flow that is evident when we examine the absolute increase in CF (relative to a fixed base of year -1 assets). In panel B, we show that the non-roundtrip firms change cash flows only slightly. By year +1, the increase is 0.7% and by year +3, cash flows increase by 5.9%. Roundtrip firms that maintain their dividends exhibit larger increases in cash flow as compared to non-roundtrip firms. Their cash flows increase by 7.2% through year +1 and 12.9% by year +3. The roundtrip, dividend-to-zero firms show large increases in their cash flows that are significantly greater than both the roundtrip firms that maintain dividends and non-roundtrip firms. The dividend-to-zero firms increase their cash flows by 20.1% through year +1 and 45.6% through year +3. This univariate evidence is not consistent with the viability motive, which suggests that dividend cessation should predict poor operating performance. The investment in capital expenditures generally follows this same pattern. All three groups exhibit small declines in capital expenditures relative to their increased assets. In absolute terms, the changes in capital expenditures show increases for all groups, with roundtrip firms that discontinue dividends having the largest increase. As shown in panel B, the increase in capital expenditures (relative to year -1 assets) is 4.7% by year +1 and 15.0% by year +3 for roundtrip firms that maintain dividends. The increase for non-roundtrip firms is larger, but not statistically 11

14 distinct, at 13.6% and 21.2%. Again, roundtrip firms that discontinue dividends have a different pattern. These firms increase capital expenditures by 24.8% through year +1 and 40.9% through year +3. The differences for year +1 are statistically significant as compared to non-roundtrip firms. This evidence suggests that these firms may divert funds to capital expenditures that were previously destined for dividends. The pattern for dividends reflects our classification scheme. Non-roundtrip firms have no dividends in year -1 or in year +1. Dividend-to-zero firms have very high dividends relative to assets in year -1: 16.4%. Per our classification scheme, this becomes 0% in year +1. Roundtrip firms that maintain their dividends have a 4.0% dividend to assets ratio in year -1, declining to 2.3% in year +1. This decline reflects an increase in the firms asset bases as these firms increase their dividends in absolute terms by 9.4% through year +1, as shown in panel B. We now have the elements to estimate a burn rate following the method described in Denis and McKeon (2016). The burn rate is capital expenditures plus dividends minus operating cash flow, all denominated by book assets. A positive burn rate reflects a decline in cash position net of financing activities. The annual burn rate for roundtrip firms that choose to maintain dividends is a rather modest 3.2% in year -1, which is significantly smaller than both dividend-to-zero firms and non-roundtrip firms. The burn rate for dividend-to-zero firms is 18.9%, and for non-roundtrip firms it is 14.7%. The burn rates dramatically decline for dividend-to-zero firms to 1.8% in year +1. For these firms, the cessation of dividends puts the firm in near balance while still maintaining their investment programs. Roundtrip firms that maintain dividends are able to have a slight surplus by year +1 (negative burn rate). Non-roundtrip firms continue with a high burn rate throughout the period we investigate. These firms continue with negative operating cash flows and maintain large 12

15 investment programs, resulting in ongoing positive burn rates. The inflow of seasoned equity appears to restock capital to fund investment and negative operating cash flows. 7 An alternative way to examine firms cash positions and need for capital inflows at the time of the equity issue is to use another ratio borrowed from the venture capital industry. The burn rate relative to a firm s cash position can estimate how much runway the firm has available. Runway, in annual terms, is the firm s cash on hand divided by its burn rate. Like burn rate, cash on hand is scaled by assets. Prior to the SEO, the average runway of the dividend-to-zero firms was only 1.2 years (not shown in table) as compared to the maintain dividend group that had an average runway of 2.9 years and 2.4 years for the non-roundtrip firms. Perhaps unsurprisingly, the firms that choose to stop paying dividends have the least cash available relative to their cash outflows. We next report the cash and debt positions of our three groups. We define cash as cash and short-term securities. We define debt as long-term debt plus the short-term portion of long-term debt. Non-roundtrip firms maintain a similar cash-to-assets ratio from year -1 to year +3. As shown on panel B, this requires large increases in cash. These firms increase their cash an average of 61% by year +1 and 81.6% by year +3. They are able to do so, despite the high burn rate, through the equity issue, as well as corresponding, substantive increases in debt. These firms modestly increase their leverage ratios from 16.4% to 18.2% from year -1 to year +3. However, they increase their total debt by 59%. Roundtrip firms that maintain dividends follow a similar pattern. Their average cash holdings as a proportion of assets is much lower at 9.4% in year -1, and their average leverage ratio is higher, at 26.4% in year -1. As is the case for non-roundtrip firms, roundtrip firms maintaining dividends substantively increase their debt, 73.3% by year +3, in order to maintain a similar cash-to-assets 7 See Denis and McKeon (2016) for a discussion of cash holdings to fund ongoing operating cash flow deficits. 13

16 ratio. By year +3, these firms have a 10.3% cash-to-assets ratio that reflects a 20.9% average increase in cash holdings. Roundtrip firms that stop paying dividends follow a somewhat different pattern. These firms exhibit a decline in their cash-to-assets ratio from 22.2% in year -1 to 13.3% in year +3. Further, they do so despite having a burn rate in near balance subsequent to the SEO and a massive increase in debt: more than doubling by year +3. These firms have engaged in such large growth to shift their assets towards more fixed assets and away from cash. In aggregate, all three groups maintain large investment programs. Non-roundtrip firms appear to be in an early stage growth pattern. They have poor operating cash flows and high investment, but maintain modest leverage ratios even while borrowing more. Roundtrip firms are separated into two distinct groups by their choices of payout policy. Roundtrip firms that maintain dividends tend to be larger, have growth through capital investments, and have relatively lower cash balances. Firms that choose to discontinue dividends face large burn rates that appear unsustainable, with a runway of just over one year. These firms have large SEOs and large increases in debt. Even with the elimination of dividend payments, these firms still experience a decline in their cash-to-assets ratios. These dividend-to-zero firms engage in very large investment programs and overall growth. In addition, they have positive operating cash flows that grow substantially in absolute terms and only modestly decline as a proportion of assets. These univariate statistics suggest that firms use dividend suspensions as a source of capital to grow the firm, per our investment motive, rather than prepare for poor future performance as suggested by the viability motive. We continue our investigation into the use of funds using multivariate models to examine further whether dividend suspension predicts poor performance 14

17 when accounting for other factors. We also examine the determinants of firm investment in a multivariate framework Cash Flows In this section, we more formally test the relation between cash flows and the choice to discontinue dividends. Our aim is to understand whether the reason for discontinuing dividends for SEO firms is to access internally generated capital in order to fund valuable investments (investment motive) or if these firms are reacting to anticipated declines in cash flows that often follow dividend omissions (viability motive). 8 The univariate evidence presented in Table 4 indicates that roundtrip firms that stop their dividends perform as ably as roundtrip firms that maintain their dividends based on operating cash flows subsequent to the SEO. We examine the ex post cash flow changes in two ways. First, we run OLS regression models of our SEO sample to investigate cross-sectional differences based on roundtrip status and dividend status. We then compare our SEO firms to non-seo firms using propensity score matching, providing us with a comparison group of firms with similar characteristics that chose not to issue. In panel A of Table 5, we begin with our regression models of SEO firms. The dependent variables in Table 5 are the firms operating cash flows denominated by assets for the three years following the SEO (CF/A). Our variables of interest are indicator variables for roundtrip firms (Roundtrip) and for dividend-to-zero firms (DIV to 0). Because DIV to 0 firms are a subset of all roundtrip firms, the coefficient for DIV to 0 firms is essentially an interactive term equal to 8 See Aharony and Swary (1980) and Brickley (1983) for examples of papers that connect dividend changes to signals of future profitability. Nissim and Ziv (2001) provide evidence that this signaling effect is present for dividend increases but not decreases. 15

18 Roundtrip * (DIV to 0). Therefore, to test the difference between dividend-to-zero firms relative to non-roundtrip firms, we estimate an F-statistic to test if the sum of these two coefficients is different than 0. We include a variety of control variables to capture the visible state of the firm at the time of the issue (year -1). We include market-to-book (M/B), our proxy for Q, as the market s estimation for the firm s growth prospects. We include the firm s operating cash flows (CF/A), the size of the firm as estimated by the natural log of the firm s book assets [LN(A)], cash and short term securities (CASH/A), and the firm s leverage ratio (D/A). We include fixed effects for year and cluster the error terms at the two-digit SIC level. Our first model tests the operating performance in year +1. The coefficient for Roundtrip is positive and significant. The coefficient for DIV to 0 is positive but not statistically significant. The F-test for [(Roundtrip + DIV to 0) 0] is significant at the 5% confidence level. Operating cash flows for dividend-to-zero firms appear to be higher than for non-roundtrip firms having controlled for various firm characteristics including prior operating cash flows. The tests for years +2 and +3 add further evidence that this effect is not fleeting. In years two and three, the F-test also indicates better operating cash flows for dividend-to-zero firms relative to non-roundtrip firms. These results provide further evidence that the cessation of dividends is driven by the investment motive and not the viability motive. The termination of dividends does not foretell poor operating performance, but rather is being used as a source of capital for investment. In panel B of Table 5, we turn to our propensity score tests that provide comparisons relative to non-equity issuing firms. Our choice of variables reflects our desire to match on some base characteristics of the assets of the firm. To provide a match, we use the following probit regression model: 16

19 SEO FIRM = b0 + b1(ln(a)) + b2(cf/a) + b3(m/b) + ε (1) Our model includes size, cash flows, and a measure of valuation. Our potential matches are firms that did not issue equity. All independent variables are lagged by one year. We run this model three separate times, one for each of our SEO groupings, only including the relevant SEO firms for each regression specification. We match each firm to the closest match by the generated propensity score and compare the values of our variable of interest, change in cash flows to assets (CF/A) from year t relative to year -1: [(CFt /At) / (CF-1 /A-1) - 1]. All statistics reported are for the sample SEO firms net of the matched non-seo firms. As reported in panel B of Table 5, the SEO roundtrip firms that maintain dividends exhibit insignificantly better performance as compared to their matched firms over the three-year post SEO period. The SEO roundtrip, dividend-to-zero firms have an insignificantly better performance in year +1 and significantly better performance in years +2 and +3 as compared to their matched peers. In contrast, the non-roundtrip SEO firms exhibit significantly worse performance as compared to their matched firms in all three years. Prior research has found consistent evidence that dividend cuts predict decreases in cash flow. 9 Our subset of dividend cutting firms is different in that these firms seem to be signaling greater investment. The evidence indicates that these investments translate into higher cash flows. Ex ante, the market appear to agree with the wisdom of greater investment as reflected in their high valuations. Thus far, we have shown evidence suggesting that SEOs facilitate large increases in investment. This is particularly the case for firms that discontinue their dividend programs. These firms appear to be using capital freed up from dividend cessation to partially fund their investment 9 Such as in Grullon, Michaely, and Swaminathan (2002). 17

20 programs, in addition to increases in debt and proceeds from the SEO. Moreover, we do not find any evidence that suggests that cash from operations declines in the years following SEOs. Rather we find that these dividend-to-zero firms perform the best in regards to operating cash flows. These results support the investment motive, and not the viability motive, that firms are raising capital and possibly eliminating dividends due to poor expected cash flows. This evidence is also inconsistent with a strict version of the timing motive that suggests that firms will issue equity when over-valued in order to affect a wealth transfer by recapitalizing the firm with greater equity. We find no evidence that recapitalization is a primary motive to the issue relative to raising funds for investment Investment and Growth Opportunities We now aim to shed light on whether these SEO firms are engaging in valuable investment or if this investment appears to be more agency spending in nature. We are unable to directly test whether these investment programs are valuable or are being used for empire building. Some standard, indirect approaches include estimating valuation changes in the form of long run stock returns or through other relative valuation measures such market-to-book or a related transformation such as excess value. Unfortunately, valuations have multiple interpretations, particularly in the context of equity issues. An established empirical regularity is that firms that issue equity have high valuations at the time of the issue, but the valuations precipitously decline over the subsequent few years. 10 Explanations for the large declines in valuations include overvalued firms returning to more rational values, as described by Loughran and Ritter (1995), by 10 These phenomena have been documented by Loughran and Ritter (1995), Speiss and Affleck-Graves (1995), Baker and Wurgler (2000), and Henderson, Jagadeesh, and Weisbach (2006), among others. 18

21 firms successfully exercising their growth options, as discussed in Carlson, Fisher, and Giammarino (2006) or an increase in agency spending, as shown by Kim and Purnanandam (2014). Our strategy is to test whether firm investment is related to the market s assessment of the firm s growth prospects. In so doing, we are able to provide information on whether the firm s private assessment of growth opportunities is congruent with the market s assessment. We can infer that agency spending is less likely the main motivation for investment if investment is aligned with the market s assessment of growth opportunities in a manner described by Frank and Sanati (2016). 11 We use market-to-book as our estimate of Tobin s Q. Theory predicts that firm investment should be positively related to Q. Our tests are reported in Table 6. We include market-to-book (M/B) and its interaction with an indicator variable for Roundtrip and an indicator for dividend elimination (DIV to 0). As in Table 5, we include fixed effects for year and cluster the error terms at the two-digit SIC level. As stated previously, because DIV to 0 firms are a subset of all roundtrip firms and thus, the coefficient for DIV to 0 firms is essentially an interactive term equal to Roundtrip * (DIV to 0), we calculate F-statistics to test the differences between dividend-to-zero firms and non-roundtrip firms, summing the coefficients for M/B along with the coefficients for the relevant indicator variables and test relative to 0. We find that non-roundtrip firms have a positive and significant relation between M/B and capital expenditures for year +1 and year +2. In year +3, the coefficient is of similar magnitude, but is no longer significant at conventional levels. Roundtrip firms that maintain dividends do not have this same relation between M/B and capital investment. The F-test for [M/B + (M/B *Roundtrip)] is insignificantly different than 0 in all three years. In contrast, the F-test for 11 In contrast, various behavioral models suggest that firm financing and investment patterns can be in response to deviations from rational valuations. For example, see Stein (1996) and Baker, Stein, and Wurgler (2003). 19

22 roundtrip firms that terminate dividends, [M/B + (M/B *Roundtrip) + (M/B *DIV to 0)], is significant at the 1% confidence level in all three years examined. The non-roundtrip firms and the roundtrip, dividend-to-zero firms appear to broadly agree with the market s assessment of their growth opportunities and have responded with investment correlated with their valuation. This evidence is consistent with an efficient investment pattern, or at a minimum, fails to provide evidence of an inefficient investment pattern. The regression models indicate that the investment policies of roundtrip SEO firms that maintain dividends are not aligned with the market s view of the quality of their available projects. One possibility is that these firms are appropriately pursuing valuable projects that they have not been able to credibly convey to the market and rejecting projects in which the market has greater confidence than the firm. Another possibility is managerial agency spending. We have shown in these regression models that investment patterns are generally aligned with valuations for two out our three groups. We finish this section using propensity score matches to further verify that we are observing actual increases in investment associated with the SEO. We showed in Table 4 that these SEO firms exhibit large increases in their investment programs subsequent to the SEO. Here, as with cash flows, we compare the investment patterns with matched, non-equity issuing firms. We track the percent increase in capital expenditures (denominated by year -1 assets) from year -1 to years +1, +2, and +3. We follow the same propensity score matching procedure as before, using Equation 1 to assign our matches, and all statistics reported are net of the matched firms. As shown in panel B, all three SEO groups significantly increase their investment in year +1. However, the magnitudes are quite different. The roundtrip, dividend-to-zero SEO firms increase their investment the most, followed by the non-roundtrip SEO firms, with the roundtrip, maintain dividend SEO firms having the smallest 20

23 increase. The large increases in investment continue for non-roundtrip firms in years +2 and +3, and for dividend-to-zero firms in year +2. We have found that non-roundtrip SEO firms and roundtrip, dividend-to-zero SEO firms both increase their investment relative to similar non-issuing firms and that this investment increase is aligned with market valuations. The evidence suggests that dividend cessations for SEO firms are being directed towards projects that are viewed by the capital markets as valuable Acquisitions Our evidence thus far is more consistent with the investment motive than the viability motive, that new investment opportunities rather than poor operating performance appear to explain the decision to issue seasoned equity, as well as the decision to forego dividend payments. The evidence is not consistent with the timing motive, that SEO firms engage in a wealth transfer by timing the market for equity. We next turn to acquisition activity in order to provide further evidence on whether these firms are attempting to sell overvalued equity. Acquisitions potentially allow a knowingly overvalued firm to exchange inflated equity for real assets, assuming that the market does not recognize the efforts to do so. We expect firms using acquisitions as part of a timing strategy will engage in more acquisitions and use relatively more stock to fund those acquisitions. We use Thomson as our source for M&A data. In Table 7 we report statistics describing acquisition activity for the time period around the SEO. We include year -1 to give a baseline and years 0 and +1 to capture acquisitions that may have been in the planning stages associated with the decision to issue equity. We first report the percentage of firms that engage in an acquisition during each of these years. Roundtrip firms that maintain dividends show a temporary increase 21

24 from 8.1% in year -1 to 15.1% in year 0, falling back to 9.3% in year +1. The increase in proportion of firms engaging in acquisitions for roundtrip firms that stop dividends has an upward trend from 11.3% in year -1, to 14.1% in year 0, jumping to 22.5% in year +1. The percentage of nonroundtrip firms acquiring other firms remains relatively flat, ranging from 9.2% in year -1 to a high of 11.9% in year 0. The total number of deals to firms follows the same pattern. The total value of deals also follows a similar pattern, with a notable exception for the dividend to zero roundtrip firms. These firms exhibit a large spike in year 0 and a decline in year +1, suggesting that deals done in year 0 tend to be larger. We next report how these acquisitions are financed. Roundtrip firms that maintain dividends and non-roundtrip firms do exhibit an increase in the proportion paid with stock in year 0. For these two groups, the evidence of payment method is consistent with using overvalued equity in exchange for real assets. However, neither of these groups exhibit large increases in acquisitions, as measured by relative value of deals. The dividend-to-zero firms that do show an increase in deal value to assets actually exhibit a decline in the use of stock as payment, inconsistent with the timing motive. Finally, we conduct an event study on our acquisition announcements in order to provide a greater understanding on whether the market views the increased investment for these SEO firms favorably. The event study uses a five-day window centered on the announcement date. We employ a standard market model using the CRSP EW index as the market with parameters estimated from -50 to Importantly, the evidence that we present here is suggestive, given the relatively small number of firms that engage in acquisitions in the near period following an SEO, and not reported in a table. For years 0 through +2 relative to the equity issue, the nonroundtrip SEO firms average abnormal announcement return for acquisitions was -0.8% and for 22

25 the roundtrip maintain dividend firms was -1.5%. In contrast, the dividend-to-zero roundtrip SEO firms had a positive 3.7% abnormal announcement return during the same interval, consistent with the market viewing the investments of dividend-to-zero roundtrip SEO firms favorably. Unfortunately, the paucity of events does not provide enough power to distinguish these groups. 12 The evidence does suggest that, for some SEO firms, acquisitions appear to be part of larger investment programs. We find no evidence consistent with behavior that we might associate with firms attempting to engage in an explicit wealth transfer by using stock to buy real assets, as predicted by the timing motive. 4. Discussion and Conclusions DeAngelo, DeAngelo, and Stulz (2010) show that a majority, but not all equity issues, follow a financing lifecycle pattern. Young firms that have high valuations and low operating cash flows raise equity capital in order to fund investment. As firms age and reap cash flows from their earlier investments, equity flows shift back to investors in the form of dividends and repurchases. Yet, as DeAngelo, DeAngelo, and Stulz note, many firms do not follow this classic lifecycle pattern as 41% of the SEO firms in their sample have paid dividends at some point. Our aim is to add to this literature by investigating firms that do not follow the classic equity financing pattern. These firms have made recent payouts, but reverse course and issue equity. These dividend-paying SEO firms also have an entangled choice of whether to maintain or cease dividend payments. If firms that cut dividend payments divert resources to valuable investments, then we might conclude that firms are pursuing an investment motive and dividends provide 12 As we note, these results should be interpreted with caution given the lack of events. We observe a total of 185 acquisitions, 21 of which are completed by the roundtrip SEO firms. 23

26 financial slack. In contrast, if these firms that cut dividends do poorly, this finding would be consistent with a viability motive, that maintaining dividends are a valuable signal of future cash flows. By observing ex post behavior and investment, we attempt to infer motivations for this equity roundtrip. We show that a significant minority of SEOs are executed by firms that have had recent payouts to equity holders and many of these firms maintain positive payouts post-seo. This roundtrip of equity capital would appear costly in terms of transaction costs. Dividends have been routinely viewed as financial slack in economic theory. How costly is accessing additional equity capital through eliminating dividends as compared to raising additional funds in the capital market? Are these choices driven by a market timing motive or are they driven by positive investment shocks? Our evidence broadly suggests a rational basis for how firms respond to market conditions for financing and investment. Following an investment motive, roundtrip firms issue equity in order to increase their investment programs. In addition, many of these firms utilize suspended dividends as financial slack for their investment programs. These roundtrip firms that end dividend payments do not suffer from declines in operating cash flows, inconsistent with a viability motive, and their investments are aligned with market valuations. We find that increased acquisition activity is also an important part of the investment programs. We do not find any systematic movement towards using stock as a means of payment, which would be more consistent with market timing motivations. The evidence collectively suggests that dividends are a valuable source of financial slack for firms with strong growth prospects. We have chosen to study SEOs rather than other forms of equity issues. Fama and French (2005) demonstrate that firms commonly issue stock and repurchase stock (and possibly pay dividends) in the same year. However, most of these equity issues are a function of ESOPs and 24

27 the firm chooses the extent to which they will neutralize ESOP redemptions via equity repurchases or dividends. SEOs are fundamentally different. SEOs are much less common, economically large, and involve significant asymmetric information, which are all quite different from ESOP redemptions. SEOs represent a fundamental change in firm financing and we know little about these firms that raise new equity following payouts to shareholders. Our findings suggest that the lifecycle pattern discussed by DeAngelo, DeAngelo, and Stulz (2010) can be interrupted by positive shocks. We started with a sample of firms that were in the payout portion of their lifecycle, but chose to change course. While not the norm, our evidence suggests that when some dividend paying firms receive a new sufficiently large and valuable project without sufficient cash flows or debt capacity, they turn to new equity capital in order to finance the investments. These firms generally are not experiencing poor operating performance and are not issuing equity in order to ensure the viability of the firm. We cannot eliminate the possibility that the firm might perceive an abnormally low cost of equity, possibly due to temporary overvaluation of the firm, and are timing the equity market. However, the positive relation between relative valuations for the firms that stop dividends and their capital investments suggest that the firm s management and the market s view of the firm s prospects are generally aligned. 25