Evidence from the regulated electric utility industry

Dividend policy and corporate monitoring: evidence from the regulated electric utility industry

Robert S. Hansen

Miller |13^ has argued that providing a rational explanation for the widely practiced policy of paying dividends is among the central tasks of modern corporate finance theory. Corporate finance textbooks often rationalize dividends as a signal, a payment to clienteles, or an irrelevant residual payout of earnings. However, in this paper we focus on the role that dividends play in the process of facilitating primary capital market monitoring that reduces the corporation’s equity agency costs. Easterbrook |5^ points out that, by raising the dividend payout, the board of directors can increase the likelihood that the finn will have to sell common stock. As a consequence, there will be an investigation of management by investment banks, the securities exchanges, and capital suppliers (see Hansen and Torregrosa |8^). Or in Donaldson’s |4, p. 54^ phraseology, the external equity financing will be accompanied by the “glare of publicity and shareholder attention which accompanies the decisions and actions of management, …, attention which can be particularly disconcerting if the ultimate terms of financing are not as favorable as expected.” As a result of this capital market monitoring, there will be reduced agency costs, and therefore, an appreciation in the market value of the firm’s common stock.

We test the relevance of the monitoring theory for explaining the dividend policies of regulated electric utilities. For convenience, hereafter we refer to regulated electric utilities as utilities. We focus on this industry partly because its dividend policies are not easily reconciled by current textbook explanations. Moreover, relative to industrial firms, utilities are arguably somewhat more insulated from the discipline of other monitoring mechanisms for controlling agency costs. Furthermore, utility stockholders have the added need to obtain monitoring of the regulators. This suggests that if an important potential role for dividends is to promote primary market monitoring, evidence of this is most likely to be found in the case of utilities. Such evidence will suggest to practitioners and academics that monitoring can provide a rational basis for dividends.

It seems unlikely that the extraordinary dividend policies of utilities can be reconciled with either a clientele story or a signaling story or that their dividends are merely a residual distribution of earnings. Over the years, utilities’ payout ratios have greatly exceeded those of industrials, with historical average payout ratios exceeding 60% of net income. Yet utilities are occasionally among the largest sellers of common stock. To explain this behavior with a clientele theory, whereby “corporations provide payout ratios that correspond to investors’ preferences for payouts” (Miller and Modigliani |14^), would require that clientele members’ tax penalties plus the firms’ flotation costs be sufficiently below the net transaction cost of receiving the income as capital gains. Existing evidence is ambiguous concerning whether the necessary low personal tax rates are present (see Chaplinsky and Seyhun |2^).(1) The signaling story (see Bhattacharya |1^, John and Williams |12^, Miller and Rock |15^, and Ross |19^) seems unlikely because it is just too implausible that utilities, whose investment opportunities are rather well-known, have more valuable hidden information to signal than do industrial firms.

However, utilities’ extraordinary dividends also do not appear to be reconciled as a residual, since they have often paid out increasing dividends during periods of heavy equity financing. Given the magnitude of utility dividends, it is thus not surprising that Miller |13^ has called for more evidence supporting a rational basis for dividends.

The idea that the high dividends paid by utilities are, at least in part, a mechanism to produce monitoring of regulators has been expressed by Miller |13^, Myers |16^ and Smith |22^. For example,

Public utility managements have found a policy of high dividends combined with frequent external equity financing to be a useful strategy for forcing their regulators to keep utility rates high enough to continue attracting new funds from investors. (Miller |13^, n. 17a.)

and

By paying high dividends, the regulated firm subjects both its regulatory body as well as itself to capital market discipline more frequently. Stockholders are less likely to receive lower-than-normal levels of compensation due to lower allowed product prices when the regulatory authority is more frequently and effectively monitored by capital markets. (Smith |22^, p. 10.)

From an agency perspective, we interpret this idea as emphasizing that dividends promote monitoring of what we call the stockholder-regulator conflict. Thus, in effect, in the case of utilities there is this additional monitoring role for dividends that complements Easterbrook’s |5^ notion that dividends promote monitoring of the stockholder-manager conflict. We add that this monitoring rationale has the additional implication that paying particularly high dividends to disgorge discretionary cash flow can bring the primary market monitoring to bear on the expenditure of funds for new investments. Paying out large amounts of earnings, in spite of known forthcoming capital expenditure requirements for valuable investments, creates insufficient funds, thereby strengthening the link between the timing of capital investment and the receipt of the primary market monitoring. Thus, this additional monitoring benefit is somewhat akin to the benefit from reducing managers’ possible misuse of funds by disgorging “free cash flow” (see Jensen |9^ and Stulz |23^).

The monitoring explanation for utility dividends has the following testable implications. First, because utility regulators’ short tenure gives them a short-term incentive bias, utility dividends should be large enough to induce occasional equity financing. Second, a necessary empirical regularity should be that utility dividend payouts and the severity of the stockholder-regulator conflict are directly related.

Our tests are conducted in each of two recent five-year periods: the five-year period ending in 1985, which is characterized by high but declining industry-wide investment growth and financing; and the more recent five-year period ending in 1990, which is characterized by secular asset growth yet low industry-wide growth. Overall, our findings from these tests are consistent with the conclusion that the monitoring rationale for dividends is an important ingredient in the explanation of utility industry dividend policy.

The rest of the paper proceeds as follows. In section I, we discuss a more succinct view of the stockholder-regulator conflict and the nature of the dividend-induced capital market monitoring of that conflict. In section II, we develop the empirical implications of the monitoring rationale for utility dividends and report the results of the empirical tests. We conclude the paper with summary remarks in section III.

I. Utility Dividends and Monitoring

The delegation of corporate decision-making to managers by shareholders creates a stockholder-manager conflict, which is associated with equity agency costs (see Jensen and Meckling |10^). Various internal and external mechanisms exist that may control these agency costs. These include, but are not limited to, leverage, which will replace equity agency costs with debt agency costs (Jensen and Meckling |10^); compensation packages, which can better align the interests of managers with those of stockholders (Fama |7^); and the disciplining effects of the takeover market (see Jensen and Ruback |11^). Easterbrook |5^, Hansen and Torregrosa |8^, and Rozeff |20^ suggest that primary market monitoring, received when the firm raises equity capital, provides an additional mechanism for controlling equity agency costs. When they raise equity capital, finns are monitored by investment banks and suppliers of new capital. Such monitoring produces unique value because it is focused yet does not suffer from the collective choice problem that normally accompanies the monitoring incentives of individual shareholders. Stockholders receive and pay for this primary market monitoring on a pro rata basis.

In general, the mix of mechanisms actually chosen to control agency costs will vary across different types of firms, depending on the availability and relative cost-effectiveness of those costs (see Crutchley and Hansen |3^). Because our focus is on utility dividend policy, our discussion will emphasize the comparative ability of dividends in controlling equity agency costs of utilities.

A. Equity Agency Costs Under Regulation

The separation of ownership and control in the case of utilities is arguably more complete than in the case of other corporations. In the utility industry, some control over decision-making is delegated to management, but much other control resides under regulatory supervision. Regulators are either elected officials or political appointees who are charged with providing “fair” regulation that ensures “confidence in the financial integrity of the enterprise, so as to maintain its credit and to attract capital” (Supreme Court decision in Federal Power Commission vs. Hope Natural Gas Co., 320 U.S. 591, 603 (1944)). Consistent with the landmark Hope Decision, state regulatory commissions are expected to set fair and reasonable rates for electric services while just meeting operating expenses and providing a fair return to investors. However, as Evans and Garber |6^ have recently argued, as political appointees, regulators may also act as ratepayers’ agents, being responsive to ratepayers’ pressures to keep the rates “low.” The incentive to act as ratepayers’ agents can be bolstered by the typically short tenure of regulators: often just a single four-year appointment (Phillips |18^). The fact that regulators are prohibited from owning common stock in the regulated firm (Phillips |18^) also reduces their incentive to act in the stockholders’ interest. This view highlights the stockholder-regulator conflict and the need for stockholders to monitor regulators. For example, absent any counterbalancing pressure from the stockholders to increase economic profits, regulators may tend to respond primarily to the pressures from ratepayers and their representatives, like the Office of Public Counsel. In Evans and Garber’s |6^ view, the regulator maximizes his or her utility by optimally balancing the opposing pressures exerted by various groups, rather than by responding exclusively to the pressures of any one group.(2)

The regulatory process can have a mixed impact on the stockholder-manager conflict. On one hand, regulatory oversight may reduce managers’ ability to appropriate stockholders’ wealth and consume perquisites. On the other hand, by setting corporate revenues on a cost-plus basis, regulators may set into motion a managerial incentive structure that potentially conflicts with stockholders’ interests.(3) Regulation, thus, can attenuate some aspects of the managerial agency costs while exacerbating others.

B. Why High Dividend Payouts Are Important for Utilities

Managers and shareholders of unregulated firms normally have access to several different internal and external mechanisms for controlling agency costs. Here we describe how utilities’ access to most of these mechanisms may be somewhat limited. Yet, we argue that their access to the dividend policy mechanism for monitoring is not limited, and we suggest that their costs of paying dividends may actually be below the costs paid by other types of firms.

Relying on greater leverage, utilities, like industrials, can reduce shareholders’ potential wealth losses from the agency relation (Jensen and Meckling |10^). Moreover, because using debt binds managers more firmly to a long-term commitment to disgorge cash, it reduces the agency costs of free cash flow like the overinvestment of uncommitted funds (Jensen |9^ and Stulz |23^). The historical fact that utilities maintain high debt ratios is consistent with the conclusion that they face these and possibly other equity agency costs. However, too much debt also raises the costly prospects of financial distress and increased costs of debt contracting. Moreover, for utilities, the marginal benefits of more debt are reduced because the well-known tax benefits of additional debt, rather than going to stockholders, flow through to ratepayers. In addition, debt has limited ability to invite monitoring that is directed at the stockholder-regulator conflict.

A second possible internal mechanism for disciplining decision-makers is their incentive compensation contracts and the incentive effects stemming from their common stock ownership (see Fama |7^ and Jensen and Meckling |10^). Managerial ownership of common stock is significantly lower in the utility industry relative to industrial finns and is therefore less effective as a mechanism for aligning managers’ interests with shareholders’ interests. Recent evidence suggests that compensation packages provide some incentives for aligning managers’ interests with stockholders’ interest. However, these incentive-aligning contracts are ineffective for aligning regulators’ interests with stockholders’ interests. Moreover, regulators are prohibited from owning the common stock of the utilities they regulate. Thus, the possible incentive-alignment benefits from compensation contracts and equity ownership are also of limited use for resolving stockholders’ conflicts with managers and regulators of utility firms.

In the unregulated firm, agency costs are also potentially bounded by disciplinary pressures from the takeover market. Consistent with this, the utility industry has experienced some restructuring during the past decade.(4) However, since virtually all of the restructurings are consolidations of one regulated utility with another regulated utility, their disciplinary effects on the stockholder-regulatory conflict have been blunted. Moreover, the Public Utility Holding Company Act limits the power of the takeover market to reduce equity agency costs since it mandates that any nonregulated company that owns 10% or more of a regulated utility will itself be subject to regulation. Thus, in the case of utilities, the ability of the takeover market to affect equity agency costs is significantly constrained by the regulatory process.

Given the utility’s leverage and compensation policies, it may achieve further potential reductions in equity agency costs through its dividend policy. By keeping dividends high enough, the board of directors may raise the prospect of primary capital market monitoring of both managers and regulators. If regulators set rates of return below stockholders’ fair rate of return, such capital market monitoring will reveal this fact to the various constituencies through an increased difficulty in raising funds. Further, the issuance of new equity may trigger an update in the commission ranking by financial institutions, similar to the updating of bond ratings associated with new bond issues. A lowering of commission rank based on “unfair” regulation increases the cost of capital and reduces access to markets. The capital market monitoring and the increased difficulty in fund raising, as well as any further related difficulties, such as possible disruption of service, expose to the general public the regulators’ failure to meet their mandate to preserve the financial integrity of the utility. To the extent that capital market monitoring revelations reduce the value of the regulators’ reputational capital, the knowledge of an impending underwritten equity offering will exert pressure on regulators to provide stockholders with a fair rate of return on their capital.(5)

It would also appear that the costs associated with dividend-induced monitoring are significantly lower for utilities than for industrials. The direct flotation costs of issuing new equity can be at least partially passed through to the ratepayers in the case of utilities. For example, this can occur when cost-plus pricing is used, so that regulators may allow the flotation costs as an add-on to the allowed return on equity. Also, utilities face a relatively inelastic product demand. Moreover, the indirect cost in terms of adverse price reaction is much lower for utilities than for industrials (Pettway and Radcliffe |17^).

Two sets of empirical implications emerge from the above discussion. First, the utility industry is expected to have a higher dividend payout ratio than other industries. This is because the marginal benefit of utility dividends is greater, reflecting both the additional benefit of reducing the stockholder-regulator conflict (Smith |22^) and the lesser effectiveness of substitute monitoring mechanisms. Also, to the extent that flotation costs can be passed on to ratepayers through the regulatory process, the marginal cost of dividend-induced monitoring is lower. Second, within the utility industry there should be cross-sectional regularities relating dividend payout ratio to proxy measures for the severity of the stockholder-manager conflict, the stockholder-regulator conflict, and the cost of monitoring these conflicts. We test these implications in the following section.

II. Empirical Evidence

This section describes the empirical tests of the monitoring rationale for dividends and presents the test results.

A. Industry Payout Ratio Test

The monitoring rationale for dividends predicts that, because dividends are both more beneficial and less costly for utilities, utilities should pay out a greater proportion of earnings than do industrial firms. Thus, the monitoring explanation predicts an industry effect:

POR(i) |is less than^ POR(u).

To test the industry-effect hypothesis, the mean dividend payout ratio of electric utilities is compared with the mean dividend payout ratio of S&P 400 industrial firms during two recent five-year periods: 1981-1985 and 1986-1990. The mean payout ratio is the equally weighted average of each firm’s ratio of cash dividends paid (COMPUSTAT II data item #21) to stockholders’ income during the respective periods (COMPUSTAT II data item #237). For 1981-1985, the mean payout ratio of the utilities is 66.25% while that of the S&P 400 industrials is 36.16%; for 1986-1990, the respective means are 69.56% and 33.77%. The respective mean differences of 30.09% and 35.79% are statistically significant at the 1% level (t = 14.24 and 14.91, respectively). Thus, consistent with the monitoring rationale, utilities as a matter of long-term corporate policy pay out a significantly larger proportion of stockholders’ income as cash dividends than do nonregulated corporations, in spite of being more capital intensive, thereby increasing the likelihood of dividend-induced equity financing.

B. Cross-Sectional Payout Ratio Test

A utility’s incentive to resort to dividend-induced capital market monitoring increases with the severity of the shareholder-manager conflict and the shareholder-regulator conflict. A utility’s payout ratio should decrease with the cost of the dividend-induced monitoring and with the need for growth-induced external financing, which substitutes for dividend-induced external financing. Accordingly, we specify and estimate the following cross-sectional model for the years 1985 and 1990:

|POR.sub.i^ = ||Beta^.sub.0^ + ||Beta^.sub.1^ |COMMRANK.sub.i^ + ||Beta^.sub.2^ |OWNSHIP.sub.i^ + ||Beta^.sub.3^ |FLOTCOST.sub.i^ + ||Beta^.sub.4^ |TAGROW.sub.i^ + |e.sub.i^ (1)

where

|POR.sub.i^ Utility i’s target payout ratio measured as the sum of all dividends paid (COMPUSTAT II data item #21) during the five years prior to and including the ending year, over the sum of all stockholder earnings (COMPUSTAT II data item #237) over the same period.

|COMMRANK.sub.i^ Utility i’s regulatory commission rank as of year-end 1985 and 1990, obtained from the Salomon Brothers Regulatory Ranking as reported in their semiannual publication, Electric Utilities. These commission rankings are based on Salomon Brothers Inc’s judgement concerning the rate of return the cognizant regulatory commission permits the utility to earn. Factors that enter into this judgement include (but are not limited to) the allowed returns on equity and the rate base, the commission’s treatment of construction work in progress in computing the rate base, the type of rate base, the method of accounting permitted, and the speed with which new rates are granted. Salomon’s rankings run from a high of A to a low of E-, which we convert to a numerical scale ranging from a high of 13 to a low of 1, and denote as COMMRANK. A relatively high value of COMMRANK, implying a more favorable ranking, suggests a lower degree of stockholder-regulator conflict, lesser need to obtain capital market monitoring, and consequently, a lower payout rate. The coefficient ||Beta^.sub.1^ is hypothesized to be negative.(6)

|OWNSHIP.sub.i^ Utility i’s stockholder-ownership concentration, as measured by an approximation of the Herfindahl Index of the firm’s common stock ownership as of year-ending 1985 and 1990. OWNSHIP is defined as

|Mathematical Expression Omitted^,

where |s.sub.j^ is the average percentage of stockheld by a stockholder in ownership group j, |n.sub.j^ is the number of stockholders in group j, j = m for managers, j = i for institutional stockholders, and j = o for remaining stockholders.(7) The firm’s managerial stockholdings are obtained from the semiannual publication, Spectrum VI (CDA Investment Technologies, Inc., Maryland), institutional holdings are obtained from the quarterly publication, Moody’s Handbook of Common Stocks (Moody’s Investor Services, Inc. New York), and the number of stockholders is obtained from COMPUSTAT II (data item #100). As the stock ownership concentration OWNSHIP increases, there is greater per-owner benefit from monitoring managers; hence, there is more intensive monitoring in place and less need for dividend-induced capital market monitoring. Thus, the coefficient ||Beta^.sub.2^ is expected to be negative.

|FLOTCOST.sub.i^ Utility i’s expected cost of the monitoring process, as measured by the firm’s historical average flotation cost incurred in selling common stock, expressed as a percentage of the gross proceeds. Using the historical cost captures most idiosyncracies in each firm’s expected flotation cost and avoids additional errors from using, say, the firm’s most recent cost. Because the equity financing flotation cost data are obtained from the Security and Exchange Commission’s Registered Offering Statistics(ROS) tape, each firm’s historical average FLOTCOST is estimated from equity offerings during the period 1971-1986.(8) A higher FLOTCOST implies a greater cost of using the dividend mechanism for monitoring, hence, a lower payout ratio, ceteris paribus. Accordingly, the coefficient ||Beta^.sup.3^ is hypothesized to be negative.

|TAGROW.sub.i^ Utility i’s growth rate in total assets, measured over the five-year periods ending in 1985 and 1990, respectively. Total assets data are obtained from COMPUSTAT II (data item #6). This variable proxies for the influence that increased use of funds for long-term investment has on the demand for monitoring. Ceteris paribus, as we have argued, when capital investment is high there is added demand for capital; thus, dividends do not need to be raised as high to induce the same level of primary market monitoring. Therefore, a higher TAGROW is hypothesized to be inversely associated with payout ratio, and thus, the coefficient ||Beta^.sub.4^ is expected to be negative.

The sample of firms used to estimate the model consists of all utilities that paid out cash dividends and had positive earnings over the respective five-year periods, 1981-1985 and 1986-1990, and for which all other relevant data are complete.

Exhibit 1 reports relevant statistics for the variables used in the regression for each subperiod.(9) The sample size declines from 81 companies in 1985 to 70 companies in 1990. This decline reflects the consolidation that took place in the utility industry. Difference of means tests reveal that the payout rate increased significantly from 69.16% to 76.43% and that the average ranking of utilities improved significantly from 6.48 to 7.07 between 1985 and 1990. There is also a significant 37% decline in the average asset growth rate from the earlier period to the later period. Taken together, these changes are indicative of a better average performance, from stockholders’ perspective, and a general slowdown in the average rate of growth of the firms.(10)

Exhibit 1. Summary Statistics for Model Variables

1985 1990

(obs=81) (obs=70) Difference

Variable Mean Std. Mean Std. Between

Dev. Dev. Means

POR 69.16% 12.00 76.43% 14.02 7.27

(3.43)(***)

COMMRANK 6.48 1.86 7.07 1.54 0.59

(2.06)(**)

OWNSHIP 6.17 4.58 8.59 12.59 2.42

(1.61)(*)

FLOTCOST 5.24% 3.37 4.51% 2.24 -0.73

(-1.54)

TAGROW 7.88% 5.00 5.05% 6.47 -2.83

(-3.02)(***)

Notes: t-statistics are reported in parentheses.

*** Significant at the 0.01 level.

** Significant at the 0.05 level.

* Significant at the 0.10 level.

Exhibit 2 reports the regression results for both the 1985 and 1990 periods. The coefficient of COMMRANK is negative and statistically significant, supporting the conclusion that a more severe stockholder-regulator conflict, as reflected in a lower regulatory ranking, is associated with a higher payout.(11) This is consistent with the hypothesis that dividends serve as a mechanism for monitoring the stockholder-regulator conflict. The coefficient of OWNSHIP is negative and statistically significant, consistent with the argument that firms with higher managerial agency costs, as reflected in lower ownership concentration, will rely more on higher payouts to control the agency conflicts. The coefficient of FLOTCOST is negative and significant. This finding is consistent with the argument that the flow through of flotation cost to ratepayers is, at best, partial and varies across utilities and is dependent, perhaps, on the severity of regulation. Lastly, the coefficient of TAGROW is also negative, as anticipated. This is consistent with the conclusion that, at higher rates of investment, dividends can be reduced without reducing the demands for external equity capital. These findings show that utilities faced with higher regulatory and managerial conflicts, lower flotation costs, and lower asset growth pay proportionally greater dividends. These findings are consistent with the conclusion that dividends are used to increase the probability of primary market monitoring. Overall, these results are consistent with the hypothesis that paying higher dividends increases the monitoring of the stockholder-manager and stockholder-regulator conflicts.(12)

TABULAR DATA OMITTED

III. Conclusion

Corporate dividend policy has long been regarded as an unresolved economic puzzle, which Miller |13^, among others, suggests requires a rational resolution if the prevailing economic paradigm of corporate finance is to continue. This paper recognizes that regulated electric utility dividend policy provides a unique opportunity to apply the monitoring rationale for dividends and to conduct empirical tests of that rationale. The monitoring threat produces counterbalancing capital market pressure to increase economic profits and increase the frequency of regulator accountability to shareholders. We present an analysis suggesting the monitoring rationale as a viable explanation for regulated electric utility dividend policy and the extraordinary practice of issuing large amounts of new equity while, over the same period, paying proportionally large cash dividends. Our findings are consistent with the monitoring hypothesis that these utilities use dividend-induced equity financing to control equity agency costs that arise out of the stockholder-regulator and stockholder-manager conflicts.

1 Moreover, it requires a particular life-cycle twist: that often much of the per-period dividend income of an existing clientele be provided by an entirely new clientele. Otherwise, the existing clients are voluntarily recycling billions dollars on personal account, a habit that should be maintainable at a much lower cost by some other means.

2 It should be theoretically possible to model the regulators’ optimal response by appropriately specifying his or her utility function. We have chosen to test the empirical implications of such a response.

3 The cost-plus basis for revenue setting will not, in itself, solve the regulator-stockholder conflict since the source of the conflict itself lies in differences in the perceptions of what constitutes “fair” cost-plus. Regulators aligned with ratepayers may have incentives to understate the utility’s required return on capital, for instance.

4 Acquiring and acquired regulated utilities in some of the major utility restructurings include: American Electric Power and Columbus & Southern Ohio Electric Co. (1980); Northern States Power and Lake Superior District Power Co. (1982); Southern Co. and Savannah Electric & Power Co. (1988); and PacifiCorp and Utah Power and Light (1988). In addition, Virginia Electric Power Co. reformed into a unit of the Dominion Resources Inc. Holding Co. (1983); Cleveland Electric Illuminating Co. and Toledo Edison Co. reformed into Centerior Energy Corp. (1986); Iowa Resources Inc. and Midwest Energy Co. reformed into the holding company of Midwest Resources Inc. (1990); Northeast Utilities acquired Public Service Co. of New Hampshire out of bankruptcy (1991); and IE industries acquired Iowa Southern Inc. (1991).

5 Institutional arrangements, such as adversarial rate hearings, are assumed not to provide complete resolution of the regulator-shareholder conflict, thus giving rise to the need for market mechanisms, such as dividend-induced monitoring.

6 In addition to the monitoring rationale for dividends, one may argue that managers simultaneously pay high dividends and raise equity capital to increase their bargaining power over regulators (see Schelling |21^ for a general discourse on bargaining and Myers |16, fn. 14^ for a specific reference to the bargaining power argument in the case of electric utilities). We note, however, that so long as the bargaining power mechanism is triggered through high dividend payments, our model specificafion and results do not change. The greater the stockholder-regulator conflict is, the lower the commission rank, the greater the benefits of increased bargaining power, and the higher the dividend payout.

7 If |S.sub.k^ is the fraction of outstanding shares held by stockholder k, and there are N stockholders, then the firm’s Herfmdahl Index is

|Mathematical Expression Omitted^.

However, we cannot observe the |S.sub.k^, so we use the following approximation to obtain estimates for the |S.sub.k^. If there are |n.sub.m^ managers, who as a group hold the fraction |S.sub.m^ of the firm’s equity, then an estimate for the fraction held by each manager is |s.sub.m^ = |S.sub.m^/|n.sub.m^. Similarly, the typical institutional stockholder’s estimated fractional holding is |s.sub.i^ = |S.sub.i^/|n.sub.i^, and the average holding of each remaining stockholder is |s.sub.o^ = |1-|S.sub.m^ – |S.sub.i^^/|N – |n.sub.m^ – |n.sub.i^^. Using these estimates for the |s.sub.k^ of the management, institutional and remaining stockholder groups yields the variable OWNSHIP in the text. Our proxy for the Herfindahl Index, therefore, uses the entire ownership structure of the firm.

8 We measure flotation costs directly rather than use proxy measures, such as firm size or unsystematic risk discussed by Hansen and Torregrosa |8^.

9 Correlation matrices, not reported in the text, show insignificant correlation among the regression variables.

10 These observations are generally consistent with the monitoring argument that during periods of relatively low asset growth, firms raise their dividend payouts to increase the likelihood of external financing, thus obtaining greater monitoring. During the high-growth period, 1981-1985, the industry raised an annual average of $61.6 million of new equity per company, while paying out $91.6 million in dividends. The corresponding numbers during the low-growth period, 1986-90, are $30.6 million and $120.2 million, respectively. The data were obtained from COMPUSTAT II.

11 A possible alternative interpretation of the inverse relationship is that a higher payout ratio not only fails to achieve better treatment by regulators, but results in more “unfair” treatment. This interpretation is refuted by noting that in addition to the negative cross-sectional association, higher payout ratios are associated with higher rankings over time (see Exhibit I).

12 An additional empirical finding consistent with the monitoring hypothesis is the positive association between financing frequency and dividend payout ratios. If firms pay dividends to induce equity financing, then it should be the case that firms with higher payout ratios have higher financing frequency, ceteris paribus. We regressed the number of new equity offerings on the payout ratio (POR) with total asset growth (TAGROW) serving as a control variable. For the periods studied, with low growth and low external financing in spite of high dividends, the data do not support the hypothesis of the financing frequency test. However, over a longer period, 1971-1990, which may cover several growth cycles, the results are consistent with the hypothesis.

References

1. S. Bhattacharya, “Imperfect Information, Dividend Policy, and the ‘Bird in the Hand’ Fallacy,” Bell Journal of Economics and Management Science (Spring 1979), pp. 259-270.

2. S. Chaplinsky and N. Seyhun, “Dividends and Taxes: Evidence on Tax Reduction Strategies,” Journal of Business (April 1990), pp. 239-260.

3. C. Crutchley and R. S. Hansen, “A Test of the Agency Theory of Managerial Ownership, Corporate Leverage, and Corporate Dividends,” Financial Management (Winter 1989), pp. 36-46.

4. G. Donaldson, Corporate Debt Capacity: A Study of Corporate Debt Capacity and the Determinants of Corporate Debt Capacity, Boston, Harvard University Press, 1961.

5. F. Easterbrook, “Two Agency-Cost Explanations of Dividends,” American Economic Review (September 1984), pp. 650-659.

6. L. Evans and S. Garber, “Public Utility Regulators Are Only Human: A Positive Theory of Rational Constraints,” American Economic Review (June 1988), pp. 444-462.

7. E. F. Fama, “Agency Problems and the Theory of the Firm,” Journal of Political Economy (April 1980), pp. 288-307.

8. R. S. Hansen and P. Torregrosa, “Underwriter Compensation and Corporate Monitoring,” Journal of Finance (September 1992), pp. 1537-1556.

9. M. C. Jensen, “The Agency Costs of Free Cash Flow, Corporate Finance and Takeovers,” American Economic Review (May 1986), pp. 323-329.

10. M. C. Jensen and W. H. Meckling, “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure,” Journal of Financial Economics (October 1976), pp. 306-360.

11. M. C. Jensen and R. Ruback, “The Market for Corporate Control: The Scientific Evidence,” Journal of Financial Economics (April 1983), pp. 5-50.

12. K. John and J. Williams, “Dividends, Dilution and Taxes: A Signalling Equilibrium,” Journal of Finance (September 1985), pp. 1053-1070.

13. M. H. Miller, “Behavior Rationality in Finance: The Case of Dividends,” Journal of Business (October 1986), pp. 5451-5468.

14. M. H. Miller and F. Medigliani, “Dividend Policy, Growth and the Valuation of Shares.” Journal of Business (October 1961), pp. 411-433.

15. M. H. Miller and K. Rock, “Dividend Policy Under Asymmetric Information,” Journal of Finance (September 1985), pp. 1031-1051.

16. S. C. Myers, “The Capital Structure Puzzle,” Journal of Finance (July 1984), pp. 575-592.

17. R. H. Pettway and R. Radcliffe. “Impacts of New Equity Sales Upon Electric Utility Share Prices,” Financial Management (Spring 1985), pp. 16-25.

18. C. F. Phillips, The Regulation of Public Utilities, Arlington, VA, Public Utility Reports, 1985.

19. S. A. Ross, “The Determination of Financial Structure: The Incentive Signalling Approach,” Bell Journal of Economics (Spring 1977), pp. 23-40.

20. M. S. Rozeff, “Growth, Beta and Agency Costs as Determinants of Dividend Payout Ratios,” Journal of Financial Research (Fall 1982), pp. 249-259.

21. T. C. Schelling, The Strategy of Conflict, Cambridge, MA, Harvard University Press, 1979.

22. C. W. Smith, Jr., “Investment Banking and the Capital Acquisition Process,” Journal of Financial Economics (January/February 1986), pp. 3-29.

23. R. M. Stulz, “Managerial Discretion and Optimal Financing Policies,” Journal of Financial Economics (July 1990), pp. 13-27.

Robert S. Hansen is a Professor of Finance and Raman Kumar and Dilip K. Shome are both Associate Professors of Finance at Virginia Tech, Blacksburg, Virginia.

COPYRIGHT 1994 Financial Management Association

COPYRIGHT 2004 Gale Group