Factors affecting the value of the stock voting right: evidence from the Swiss equity market.
Previous studies have dealt only with the separate effects of different factors that affect relative prices. Instead, we examine all known influential factors together. In particular, we find that not only voting rights, but also differences in liquidity, ownership structure, and restrictions on transferability, significantly affect the market value of shares with different voting rights.
Using the present value of future cash flows model, the value of a stock is the sum of the discounted cash flows until eternity. In a perfect world with neither agency nor information costs, only cash flows matter, and the value of a stock voting right must therefore be zero. Lease, McConnell, and Mikkelson (1983, p. 441) summarize this concept as follows: "A common stock that grants voting rights will be valued differently from one that does not, only if the future consumption opportunities provided by the two securities are different."
In the real world, control over a company is valuable, as shown by numerous takeover battles and proxy fights. There are different sources of this value. First, agency problems allow managers with a controlling interest to extract part of the value of the firm from the other owners and creditors, by using tactics such as high management salaries or excessive perquisites. There can also be nonpecuniary benefits to control, as mentioned by Barclay and Holderness (1989): Maintaining control of a family firm, for example, can be important for reasons of personal pride or family relations. If management does not own a controlling interest, the market for corporate control limits these agency costs, to some extent, by the threat of changes in control.
In addition to agency considerations, the voting right attached to a share has value because it enables shareholders to make decisions that increase the value of the cash flows. Although this increase must be shared between holders of voting and nonvoting shares, it is only the holders of voting rights who can change a firm's management or operations.
Voting shares can also be more valuable if there is uncertainty whether cash flows to the nonvoting shares are actually equal to the voting shares' cash flows. Nonvoting shareholders may be concerned that voting shareholders could find a way to profit unexpectedly at the expense of the nonvoting shareholders. Ang and Megginson (1989) report that 45 out of 49 UK firms that retired restricted voting rights stock by giving full voting rights to restricted shareholders simultaneously paid special dividends to superior voting rights stocks. Presumably, they did this to compensate these shareholders for the dilution of their voting rights.
Furthermore, the value of voting rights is affected by the probability that they will become valuable should there be a takeover in the future. In some takeovers, superior voting shares have received a higher price than have nonvoting or inferior voting shares.(1)
Also important to understanding the valuation of multiple share classes is the fact that minority shareholders in Switzerland are not as well-protected as they are in the US.(2) For example, there are no "coattail" provisions in Swiss law that require in a tender offer that minority shareholders be offered the same price as other shareholders. In the event of an acquisition, dissenting shareholders have little hope of successfully disputing the acquisition price in court. Controlling blocks often change hands without an offer to minority shareholders; if an offer is made to minority shareholders, it may be at a price lower than that paid for the controlling block. One well-known example is the acquisition of Jacobs-Suchard by Philip Morris in 1990. In this instance, the majority vote owner, Klaus Jacobs, received more than twice as much for his shares as did the public.
Even outside of an acquisition, minority shareholders have fewer rights than in the United States. Management can extract substantial private benefits of control through excessive salaries, sweetheart deals, and other perquisites, and can do so with little interference by minority shareholders. Banks traditionally cast both their own and their customers' votes in favor of management, unless they receive specific instructions to do otherwise. Proxy contests and American-style shareholder resolutions are virtually unheard of in Switzerland. Zimmermann, Bill, and Dubacher (1989) document how Swiss firms can successfully oppose unfriendly takeover attempts.
The Union Bank of Switzerland (UBS) saga highlights the plight of shareholders in Switzerland. UBS originally issued registered shares in 1975 as a defense against takeovers by foreigners, who were not allowed to own the registered shares that controlled the majority of votes. In 1991, BK Vision, a Swiss investment firm, began acquiring registered shares. By 1994, BK Vision owned more than 18% of the registered shares. UBS responded with an attempt to split the bearer shares, thus increasing the total voting power of the bearer-share class. UBS then would exchange the existing registered shares for the new bearer shares, diluting the voting rights of the registered shares by almost two-thirds according to Kunz (1994). This dilution is even more surprising given that the registered shareholders held the majority of the votes. However, the fears of the other shareholders over possible expropriations in a takeover, combined with the restriction that no shareholder could vote over 5% of the shares, allowed management to win shareholder approval of the action, although litigation is still pending. In the US, it would be exceedingly difficult for a large firm to take such an action that severely reduces the voting rights of existing shareholders. Similarly, US firms cannot put limits on the percentage of votes that any one shareholder may exercise. Within a month of the announcement, the price of UBS registered shares dropped by 15.7%, losing a large part of their premium over the bearer shares, which dropped by only 0.4% during this period.
The following section reviews the literature. Section II describes the institutional details of the classes of equity issued by Swiss firms. Section III contains information on data selection criteria and sources. Section IV describes how the premia on voting shares are decomposed into their component pieces based on voting, liquidity, and transferability, and then presents the empirical results. Shares with voting rights trade at substantial premia to the nonvoting shares. However, these premia are affected by other factors. Voting shares trade at higher premia for firms with severe inequalities in voting rights between classes. On the other hand, if more than 50% of the votes are controlled by a single entity, then the premia attached to voting rights are lower. The voting rights premia are also reduced if the voting shares are less liquid than the non-voting shares as measured by the bid-ask spread. Furthermore, transferability restrictions, such as refusing registration to foreign shareholders, reduce the voting rights premia.
I. Previous Work
Previous work on the value of corporate voting rights tends to fall into two major categories.
One category looks directly at the prices of share classes with different voting rights. Shares with superior voting rights usually trade at a premium compared to shares with inferior voting rights, although the size of the premium varies substantially across countries. Lease, McConnell, and Mikkelson (1983) find an average premium of only 5.4% in the US, while Smith and Amoako-Adu (1995) find a median premium of 6.4%. In other countries, the premia are higher: Levy (1983) finds a 45.5% premium in Israel. In Italy, Zingales (1994) finds an 82% premium for voting shares over nonvoting shares, and concludes that most of this premium is a function of the private benefits of control and the inefficiency of the Italian market for corporate control.(3)
The second category examines voting rights within the context of the agency problem between shareholders and managers, as in Jensen and Meckling (1976), and the market for corporate control, as in Jensen and Ruback (1983). Easterbrook and Fischel (1983) note that voting rights usually belong to the residual claimants on the firm, who are the shareholders unless the firm is in financial distress. Voting rights are one monitoring device that these residual claimants can use. In their view, voting facilitates takeovers. The premium of voting over nonvoting shares represents the "...opportunity of those with votes to improve the performance of the corporation," as opposed to private benefits for insiders. Zingales (1994) emphasizes that a premium for voting rights reflects expectations that a contested acquisition can occur in which the private benefits of control will be high. The actual premium in a bidding situation is closely related to the private benefits of control to the party that values them second highest. Other research in this vein includes the work of Grossman and Hart (1988), Harris and Raviv (1988, 1989), Blair, Golbe, and Gerard (1989), and Mikkelson and Partch (1994) on the optimality of one-share-one-vote and the effects of unbundling voting rights. The structure of ownership is also important as seen in the work of Brous and Kini (1994), Bathala, Moon, and Rao (1994), and Raad and Ryan (1995).
Although we examine voting rights rather than segmented capital markets, the literature on market segmentation is also relevant because many Swiss companies issue shares whose transferability is restricted, resulting in segmented capital markets. These restrictions generally reduce the value of such shares, as seen in the work of Stulz (1981), Errunza and Losq (1985), Merton (1987), and Stulz and Wasserfallen (1995). Although Swiss corporation law allowed these restrictions to be arbitrary prior to 1992, a typical restriction policy is one that prevents foreigners from registering stock. This policy results in lower prices for such shares.
One well-known event in this context is the so-called "Nestle crash." On November 18, 1988 Nestle announced a change in its registration policy that allowed foreigners to become registered shareholders for the first time. As a consequence, the registered shares increased in value by more than 40%. The bearer shares, which were not restricted, dropped in value by more than 20%.(4) Before, Nestle bearer shares had cost about twice as much as otherwise identical registered shares. Stock prices of most other Swiss companies also were strongly affected by this decision, since the prospect of changes in registration restrictions became more likely.
Previous work on the value of voting rights in the Swiss equity market includes Horner (1988) and Haeberle and Pasquier-Dorthe (1991), who document that superior voting stock sells for higher prices. However, their work was before the Nestle crash, which dramatically changed the institutional setting of the Swiss market. Furthermore, our investigation incorporates additional variables known to affect stock value, including transferability restrictions and liquidity.(5)
II. Equity Capital Structure of Swiss Companies
Most Swiss firms issue different classes of common equity, although not every firm has every type of share discussed below. Recently, a single-equity-class capital structure has become more popular for publicly traded Swiss companies. In the United States, some companies also issue multiple classes of equity, but the practice is not as common as it is in Switzerland; such US firms tend to be the exception, rather than the rule.(6)
Swiss companies issue three major classes of equity:
* Bearer shares (Inhaberaktien), which carry voting rights, may be freely traded. There are no restrictions on who may own such shares. One advantage of bearer shares is that they are truly anonymous, because the company does not necessarily know who owns the shares. This secrecy may be valuable to investors who, for a variety of reasons (including tax evasion), wish to hide their wealth. Although one might think this secrecy could be useful to raiders who want to acquire a large stake in a firm, other antitakeover measures, such as a limitation on the number of shares that may be voted, make hostile takeovers more difficult in Switzerland than in the United States. For example, all the companies in this study are either majority-owned or have a maximum number of votes that can be exercised by any one shareholder or group of shareholders.
* Registered shares (Namenaktien) are registered with the corporations under the names of the individual owners. Thus, corporations know the identities of the registered shareholders. Until 1992, however, Swiss corporation law permitted firms to arbitrarily refuse to register a new shareholder. Corporations can still refuse to register someone who buys a so-called "vinculated" registered share, but now the allowable reasons for refusal are limited and must be explicitly stated in the articles of incorporation. For example, some firms can refuse to register a foreign shareholder. While the buyer would still receive dividends, he or she would not be allowed to vote.
* Nonvoting shares, also known as dividend right certificates (Genussscheine) or participating certificates (Partizipationsscheine or PS), carry no voting rights, but do pay dividends tied to the dividends on other shares. Like bearer shares, they can be freely traded and are completely anonymous.
Voting rights for registered and bearer shares usually follow the one-share, one-vote principle, independent of their par values.(7) A company that wishes to raise equity and at the same time limit dilution of current registered shareholders' votes can issue bearer shares with a higher par value. The newly issued bearer shares then carry a high price because of the correspondingly high cash flow rights attached to them, and relatively few of them need to be issued. As a result, fewer voting rights are sold per currency unit raised, and the registered shares end up with relatively higher voting rights. Typically, if both types of voting shares are issued, registered shares have more than 50% of the voting rights.
Dividend payments on the different share classes are proportional to the stated par value of the stocks ? Thus, the different classes of stocks can trade at different prices, not only by reason of voting rights, but also because dividend payments differ. When comparing the prices of the voting and the nonvoting shares to investigate the value of voting rights, one must adjust for differences in cash flow rights.
The equity capital structure of Alusuisse, shown in Table 1, is representative of the equity structure of a typical major Swiss firm in 1991. Alusuisse, a large aluminum company, had nonvoting shares with a par value of 25 Swiss francs (SF), registered shares with a par value of SF 125, and bearer shares with a par value of SF 250. The 1990 dividends paid to these different shares (SF 3.5, SF 17.5, and SF 35, respectively) were proportional to their par values. However, because the votes for the registered and the bearer shares are one per share, the registered shares have 52.5% of the total votes, even though they contribute only 32.1% of the nominal capital for the firm.
At first glance, the prices of Alusuisse shares roughly accord with the dividend rights. However, when the prices are adjusted for differences in dividends, the bearer and the registered shares usually sell for a premium compared to the nonvoting shares. The bearer shares can also sell at a premium to the superior voting registered shares, because the value of the registered shares' higher voting rights can be more than offset by their lower liquidity, restrictions on transferability, and limits on the votes that may be exercised.
Multiple classes of shares can be used as a defense against takeovers. A firm that needs to raise capital, but whose current owners wish to maintain control, can issue either nonvoting shares or shares with inferior voting rights. However, this will only work if the superior voting shares are held by a controlling group. If the higher voting shares become widely held, it can make it easier for a raider to buy cheap votes and gain control. To guard against this, many Swiss firms place explicit limits on the number of votes that any one group of shareholders can vote at any one time. In the Alusuisse example, no shareholder may exercise more than 5% of the total votes. A shareholder who controls more votes than can be exercised may still wish to acquire more votes to prevent them from being [TABULAR DATA FOR TABLE 1 OMITTED] voted against him or her. By enfranchising the other share classes, an incumbent management can also use a multiple-class structure as a defense against a raider who has purchased registered shares, as in the UBS case above.
Our sample period extends from April 1990, when trading volume data first became available for Swiss stocks, through December 1991. We intentionally choose a period that was stable in terms of major events affecting Swiss voting rights. During this period, there was neither an upward nor a downward trend in the average relative price ratios of voting to nonvoting shares. Sufficient time had passed after the Nestle crash for the market to absorb the shock. We choose a relatively short period of time, because over a longer period, too much company-specific data, such as voting rights restrictions and equity structure, would change. No proxy fights or hostile takeovers of the sample firms occur during the sample period. We intentionally look at the value of voting rights during "normal times," which allows us to control for differences in liquidity, transferability, and company-specific items.
This stable period is both an advantage and a limitation of our study. These results may not necessarily apply in circumstances different from those studied here. A longer time period would have allowed us to examine the effects of changes in the company-specific variables and perhaps the effects of takeovers, but then our results would have been affected by the Nestle crash and the July 1992 changes in the Swiss Company Act.
The new Swiss Company Act is a significant change in Swiss corporate law (Peters, 1992). Corporations can no longer arbitrarily refuse to register shareholders; reporting standards are greatly improved; and stock repurchases are now explicitly legal. The minimum required par values for voting stock have been decreased, allowing firms to split their shares and issue lower-priced stock. Prior to the new law, firms wishing to sell low-priced shares to the retail market often resorted to selling nonvoting shares, which could have lower par values. The new law also limits the fraction of nonvoting shares that firms can issue. These changes reduce the attractiveness of issuing nonvoting shares; thus, a majority of the companies in our sample retired their nonvoting shares by exchanging them for voting shares. In anticipation of possible exchange offers, the value of the remaining nonvoting stocks generally increased relative to the voting shares. Because incorporating data from this time period would have led to a downward bias in the estimated value of voting rights, we end our sample period in December 1991.
The following criteria were used to select the companies in the sample:
* We consider only Switzerland-based companies whose stocks are listed on a Swiss stock exchange.
* Companies must have at least two classes of equity outstanding, one of which is nonvoting. We need a pure nonvoting security as a benchmark for measuring the value of shares with voting rights.
* The nonvoting, and at least one of the voting, shares must have an average monthly trading volume of at least SF 8 million during the sample period as reported by the Association Tripartite Bourses (ATB).(9) This criterion avoids problems with missing or stale prices.
We eliminate for company-specific reasons two firms that meet these criteria. Union Bank of Switzerland (UBS) and Ascom offered to exchange voting for nonvoting stock during this period. Since the relative prices of the voting and nonvoting shares were clearly a function of the specific exchange offers, we exclude these two companies from the sample.
The selection criteria include 70 equity securities (24 bearer, 17 registered, and 29 nonvoting stocks) issued by 29 different firms. The sample consists of 41 pairs of one voting and one nonvoting stock. It includes seven of the top nine, and 14 of the top 20 firms in terms of market capitalization. Of the six most highly capitalized firms, only UBS is absent. The median market capitalization is SF 1.925 billion. The 70 securities include 26 of the top 30 stocks in Switzerland ranked by trading volume, as obtained from the Swiss Stock Guide (1991).
Daily closing prices from the Zurich stock exchange, as well as volume data representing trades on all Swiss stock exchanges for these securities, were provided by Telekurs AG. Bid-ask spread data were obtained from various issues of the Kursblatt der Zurcher Effektenborse. Information on shareholder restrictions was obtained from Kaufmann and Kunz (1991). Other data were obtained from publicly available information, especially the Swiss Stock Guide (1991).
IV. Empirical Evidence
In general, the value of a voting stock consists of:
* The value of a nonvoting stock (NV) of the same company (adjusted for cash flow rights).
* The value of differential liquidity (L) between the voting and nonvoting shares of the same company.
* The value of differential transferability (T) between the vinculated voting shares and the nonvoting shares of the same company.
* The value of the voting right as affected by company-specific factors (C), such as the concentration of votes or the number of votes per unit of cash flow rights.
* The value of the voting right itself (control), after the adjustment for the company-specific factors.
The value of a voting share V can therefore be decomposed as follows:
V = NV + L + T + C + [Epsilon] (1)
where [Epsilon] represents the residual.
By definition, the residual includes the value of the voting right after the adjustment for company-specific factors. Note that this residual, unlike standard regression residuals, does not have an expected value of zero. Dividing by NV and rearranging the terms yields the following equation, which examines the relative values as a fraction of the nonvoting share price:
V/NV - 1 = [Epsilon]/NV + C/NV + L/NV + T/NV (2)
The voting share premium (V/NV - 1) equals the sum of:
Relative residual value
+ Relative value differential of voting rights caused by company-specific factors
+ Relative value differential caused by liquidity factors
+ Relative value differential caused by differences in transferability
Therefore, models of the following general form are estimated:
V/NV - 1 = [Alpha] + [[Beta].sub.a]C/NV + [[Beta].sub.b]L/NV + [[Beta].sub.c]T/NV (3)
A. Description of Variables
To measure the premium in value of the voting over the nonvoting shares, we use a measure of the relative price of the voting shares with respect to the nonvoting shares, adjusted for cash flow rights. For each pair of stocks (a), we compute the average of the daily log ratios of the voting share price to the nonvoting share price, adjusted by the ratio of the par values. The dependent variable, LPRATIO, is then defined as:
[LPRATIO.sub.a] = 1/n [summation of] ln ([V.sub.ai]/[NV.sub.ai] x [P.sub.NV]/[P.sub.V]) where i=1 to n (4)
n = Number of days for which observations are available
[V.sub.ai] = Price of the voting share for firm a on day i
[NV.sub.ai] = Price of the corresponding nonvoting share for firm a on day i
[P.sub.NV] = Par value of nonvoting share
[P.sub.V] = Par value of voting share.
We use logarithms to achieve both a symmetric distribution for premiums and discounts and easy interpretation, because logarithms lead directly to decimal fractions. The values of LPRATIO for each firm's voting shares are listed in Table 2. They vary from -21% (Swiss Re registered share) to 67% (Pirelli bearer share), with an average of 13%. In general, the values are higher for bearer shares.(10) Only registered shares show negative values. At first glance, since the value of voting rights cannot be negative, it may appear odd that negative values persist over an extended time period. However, in some cases, the losses from reduced liquidity and transferability outweigh the value of the voting rights, as our regression results show.
1. Company-Specific Factors
Several company-specific factors affect the value of a particular company's voting shares. The value of a single vote is not just the value of control divided by the number of shares outstanding. It reflects the value of a vote to the marginal investor, which is affected by the relative holdings of the different shareholders. If more than 50% of the shares are owned by another entity, that entity controls the firm. No matter what the minority shareholders do, they cannot gain control without the assent of the majority. Even though the votes of minority shareholders in such firms cannot affect the outcomes of any votes at the present, voting shares are still more valuable than nonvoting shares: Minority shareholders can participate in some of the voting share benefits if the controlling shareholders transfer wealth from the nonvoting shareholders to the voting shareholders.
Furthermore, minority shareholders' voting rights can become valuable if the controlling majority later breaks up. For example, this happens if the shares in a family-controlled firm are dispersed among feuding heirs, or if the controlling shareholder runs into financial difficulty and must sell. The effect of shareholder concentration is captured by the dummy variable CO, which is set equal to one if more than 50% of a company's votes are controlled by a single parent company, person, pool, or family.(11) Since the existence of a controlling shareholder should reduce the immediate power of voting rights, we expect a controlling influence to reduce the value of voting rights and lead to a negative coefficient for CO.
The votes of companies without any registered shares outstanding can be acquired easily and anonymously. This increases the effectiveness of the market for corporate control and, consequently, the value of these votes. For companies without registered shares, the dummy variable NOREG takes the value of one, and zero otherwise. Because the lack of registered shares should increase the value of the votes to the voting bearer shares, we expect a positive coefficient.
Not only is there the obvious difference in voting power between the voting and nonvoting shares, but there may also be a voting inequality between registered and bearer shares. Such an inequality can make it easier for the company to take actions that would harm those not in control, thus making the voting rights even more valuable.
Again, Alusuisse demonstrates how the voting rights can differ between different classes of voting shares. The Alusuisse bearer shares have a par value of SF 250, and the registered shares have a par value of SF 125, but each share has one vote. Because cash [TABULAR DATA FOR TABLE 2 OMITTED] flow rights are proportional to par values, a registered shareholder gets twice as many voting rights for the same cash flow rights as a bearer shareholder.
To capture the effect of different par values on voting rights, we constructed the explanatory variable LPAR by taking the natural logarithm of the ratio of the par value between one voting share class and the other voting share class (if there is one). For Alusuisse, it takes the value of 0.693 (= ln 2) for the (superior) registered shares and -0.693 (= ln 1/2) for the (inferior) bearer shares. The value is zero (= ln 1) if both share classes have the same par value. It is also set to zero (neutral) if the company has only one type of voting shares outstanding.(12) Since we expect voting rights to be valuable, superior voting shares should trade at a premium, leading to a positive coefficient for LPAR.
While LPAR captures the inequality of voting rights between voting shares, it does not take into account the nonvoting shares. We therefore use the Levy (1982) measure of voting power inequality, based on a Lorenz inequality measure, that measures inequality across all types of shares. The LEVY index is a standardized range between zero and one. For example, a firm that has only one class of stock, or a firm with only registered and bearer shares with equal par values and voting rights, has a LEVY value of zero. A firm with all its voting power concentrated in a special class of stock with no cash flow rights, only votes (e.g. a "golden share"), has a LEVY value of one.
Figure 1 demonstrates the calculation of the LEVY value for Alusuisse. On the horizontal axis, the cumulative fractions of paid-in capital for the different share classes are plotted in ascending order, beginning with the one with the lowest par value per vote. On the vertical axis, the cumulative fractions of the votes are plotted correspondingly. The LEVY value represents twice the shaded area between the actual plot and the 45 degree line that represents no voting rights inequality. Mathematically, the LEVY value for Alusuisse is 0.251.
A high LEVY measure indicates that votes are concentrated in a relatively small fraction of paid-in capital, making it more likely that the superior-voting-rights shareholders will attempt to exploit the other shareholders class(es). As opposed to CO, which indicates that no outsider can obtain control and its private benefits, the LEVY value captures the potential for agency costs. Such agency costs could be due to a divergence of interests among the holders of different stock classes, despite the fact that a company is or is not controlled. The LEVY measure indicates that someone can control the company with a relatively low fraction of the equity. This may lead to a reduction in the total value of the company, but it also implies that the voting rights are more valuable, since they could be useful not only to take over a firm, but also to expropriate other equity classes. Therefore, a positive coefficient is expected.
2. Liquidity Differences
We must also control for differences in the liquidity of the shares. Because more liquid securities usually command a premium, the differences in the prices of the share classes also reflect liquidity differences. The usual concepts of liquidity contain both a transaction cost component (including costs such as the bid-ask spread, market impact, commissions, and taxes) and a time dimension for finding the counterpart to a trade. We use the average bid-ask spread as a proxy for liquidity.(13) Because a lower spread implies lower transactions costs, ceteris paribus, a lower-spread stock is considered more liquid than a higher-spread stock.
To measure the average spread, the closing bid and ask prices for the 70 stocks in the sample were taken from 21 randomly selected copies of the Kursblatt der Zurcher Effektenborse (one for each month during the sample period). The average of the 21 logarithmed ratios of ask to bid prices for each stock is multiplied by 100 to determine the percentage value, which we use as a liquidity measure. Because relative stock prices will be considered, we need a relative measure of liquidity - the difference between the average logarithmed spreads of the voting and nonvoting stock (LSPRDIF). A positive value indicates that the voting stock has a higher average bid-ask spread and is thus less liquid than the nonvoting stock, and vice versa. Because liquidity is valuable, we expect LSPRDIF to have a negative coefficient.
Finally, a dummy variable (MOSHACL) is set to one for companies with more than one class of voting shares, and zero otherwise. This variable can be interpreted as an unused potential for a simpler equity capital structure. The reduction of share classes would lead to enhanced liquidity of the remaining share class. Given this explanation, we expect a negative coefficient.
3. Restrictions on Transferability
The transferability of some registered shares was severely restricted during our sample period. As we have noted, a company could arbitrarily refuse voting rights to an investor. This uncertainty and the potential disregard of shareholder interests reduces the value of the voting right of such a share. Therefore, we use a dummy variable (VREG) that is set to one for shares with restrictions, and zero otherwise. VREG is split into two subgroups: 1) severely vinculated registered shares (VREGS), where companies can arbitrarily refuse to register shareholders, or reject foreign shareholders, or both; 2) less severely vinculated registered shares (VREGL), where the policy of acceptance is clearly declared in the articles of incorporation, and is limited to a maximum of total or registered votes that can be exercised at the shareholders' meeting.
B. Regression Results
To estimate the impact of the different factors on the relative prices of voting and nonvoting shares, we regress the logarithmed price ratios (LPRATIO) on the above explanatory variables. The results, presented in Table 3, are consistent with our expectations that liquidity is valuable, as are voting rights, and that restrictions on transferability reduce value. All of the coefficients have the expected signs, most of them are significant, and the model fit is good.
Differences in voting power between the voting shares, as measured by LPAR, indicate that superior voting shares are worth more, as demonstrated in the results for the first regression (Model A). For example, if the share class with superior voting rights has half the par value of the share class with lower voting rights, then there is an overall difference of over 10% between the voting and nonvoting shares.(14) This is evidence that voting rights are valuable.
The value of voting rights for bearer shares in companies without registered shares increases by more than 10% over firms with registered shares. This is shown in the coefficient for NOREG in Model A. However, this result is not significant at conventional significance levels. A closer examination of the data reveals that this insignificance is associated with splitting the transferability restriction measure VREG into the less severely vinculated VREGL and the more severely vinculated VREGS. In the second regression, Model B, NOREG is significant at the 5% level when vinculated shares are not split up. All of the other variables have similar and significant coefficients in [TABULAR DATA FOR TABLE 3 OMITTED] the two models.
The CO variable also confirms our ex-ante expectations: If a company has a stockholder with a stable majority, the premium for the voting shares is reduced by over 13%. This significant negative effect can be attributed to the strongly reduced probability that voting rights traded in the market will become decisive.
The coefficient for the LEVY value indicates that the more unequal the distribution of the voting rights, the higher the premium for votes. This result strongly supports the "expropriation hypothesis." The more the distribution of voting rights deviates from the distribution of cash flow rights, the easier and more likely it is that someone can reach a majority of votes, and thus capture benefits at the expense of minority shareholders. These potential private benefits and conflicts of interest between shareholder groups increase the value of the voting right by approximately 44% of the LEVY inequality index.
The significant impact of the relative bid-ask spread (LSPRDIF) demonstrates the importance of the transaction-cost component of liquidity in security pricing. Previous studies that did not control for these effects should be considered with caution. An improvement in the relative liquidity of the nonvoting shares by 1% will reduce the premium of the voting over the nonvoting stock by over 12%, ceteris paribus.
The value of voting rights in companies with more than one class of voting stock outstanding (MOSHACL) is significantly reduced by the market, almost 15% in Model B. This reduction can be attributed to the shares' reduced liquidity, which is caused by the existence of several voting share classes. LSPRDIF quantifies the liquidity differences between the voting and nonvoting share classes. MOSHACL measures the effect that the adoption of a single-voting share class has on further increasing the liquidity of the voting relative to the nonvoting shares.
The effect of the transferability restrictions is seen in the strongly negative impact (-15.5%) of the restrictions captured by VREG. Model A indicates that the severity of the restrictions does matter, as indicated by the lower magnitude of the coefficient for the less severely vinculated shares (VREGL).
The model fit is good, with an adjusted [R.sup.2] near 70% for both models. Our model is also robust to different subperiods. Running the Model A regressions separately over seven non-overlapping three-month intervals (not shown here) provided similar parameter estimates for each subperiod, with an average adjusted [R.sup.2] of 62.9%.
It is possible to use our results to estimate the "pure" value of a voting right, and to estimate the average fractions of voting share premia that are attributable to liquidity factors and to transferability restrictions. The intercept [Alpha] is influenced by any factors that are not captured by our explanatory variables. These factors include the value of the pure voting right relative to a nonvoting share, after adjusting for company-specific factors. The intercept of 0.18 in Model B can be interpreted as the value of the pure voting right. However, misspecification of the functional form of the model, and the omission of important variables, can bias the intercept. We attempt to control for as many important factors as possible, and the high explanatory power of the model indicates that the risk of omitted variables is low. Misspecification of the functional form is also unlikely, since other specifications (including quadratic terms and interaction effects) do not result in substantial differences in the intercept. (In the interest of brevity these results are not reported here.(15)) The 18% therefore represents a reasonable estimate of the value of the pure stock voting right in Switzerland over the price of an otherwise identical nonvoting share, if we control for the influence of all the variables discussed above, including the company-specific factors.
V. Summary and Conclusions
Liquidity, transferability, ownership structure, and voting rights affect the relative prices for the 41 pairs of voting and nonvoting shares of 29 Swiss corporations that we investigate. Voting rights have a substantial economic value. In our sample, the pure stock voting right carries a premium of 18%. This number fits between the approximately 5% found for the United States by Lease, McConnell, and Mikkelson (1983) and the 82% that Zingales (1994) found for Italy. This premium could indicate that a nontrivial level of private benefits can be achieved by controlling a Swiss company: Switzerland's lower level of legal protection for minority shareholders (relative to the United States) makes it more likely that the voting shareholders can expropriate value from the other shareholders. Thus, voting rights are more valuable as a monitoring device in the sense of Easterbrook and Fischel (1983). The more the distribution of voting rights deviates from the one-share, one-vote principle, the higher the value of the voting right. This can lead capital markets to expect significant value added through changes in the control of these firms. This is consistent with Zingales' (1994) findings for Italy, where poor legal protection of minority shareholders is associated with high private benefits of control and high value for voting rights.
Among the many other factors influencing the relative prices of voting and nonvoting stock, restrictions on transferability and liquidity differences appear to be very important: Fewer restrictions on transferability and higher liquidity lead to a higher relative price. Another significant factor is the pattern of share ownership: As we might expect, companies in which the majority of votes are held by a single entity have lower voting rights premia, because remaining shares have less voting power. The likelihood that the market for corporate control can play its role is strongly reduced. On the other hand, the voting rights of companies that issue only bearer shares appear to be priced higher by the market. Superior voting stocks are also significantly higher priced than otherwise identical securities.
These findings raise interesting questions for further research. It would be useful to obtain results from additional countries and compare them with those reported here. Different legal and market structures in different countries would be expected to lead to different values for voting rights. Furthermore, the recent retirement of nonvoting shares by many Swiss firms provides a natural test for the conjecture that the total value of a finn may be reduced by having multiple classes of equity. Research that examines this question is underway.
The authors wish to thank Claudio Loderer, Tobias Studer, participants at the Georgetown University Finance Seminar, and two anonymous referees for helpful comments. All errors, of course, remain ours. The authors also wish to thank Telekurs AG for providing the data used in this study. Partial financial support for this research by Georgetown University, the Georgetown Center for Business-Government Research, and the Swiss National Science and Max Geldner Foundations is gratefully acknowledged.
1 DeAngelo and DeAngelo (1985) list four examples of takeovers in which holders of voting shares received explicit premiums over the amounts paid to otherwise identical nonvoting shares. Bergstrom and Rydqvist (1992) report that such premiums are common in Sweden. Zingales (1995) provides several more examples. The authors are aware of four such situations in Switzerland.
2 See Kaufmann and Kunz (1991) and Meier-Schatz (1993) for more information on the rights of Swiss shareholders.
3 Other investigations of the value of voting rights include Lease, McConnell, and Mikkelson (1984), Horner (1988), Bergstrom and Rydqvist (1990a,b), Haeberle and Pasquier-Dorthe (1991), and Ang and Hodges (1993). Studies of the effects of changes in voting rights include Bhagat and Brickley (1984), Partch (1987), Jarrell and Poulsen (1988), Cornett and Vetsuypens (1989), Mikkelson and Partch (1994), and Maynes (1992).
4 See Herman and Santoni (1989), Stulz and Wasserfallen (1995), and Loderer and Jacobs (1995) for more on the events surrounding the "Nestle crash."
5 For example, it is well documented that more liquid assets are more valuable than otherwise identical, but less liquid, assets. See Amihud and Mendelson (1986, 1987, 1988, 1991), Diamond and Verrecchia (1991), Silber (1991), Dares and Ehrhardt (1993), and Angel (1994) for work demonstrating the value of liquidity.
6 The NYSE refused to list inferior-voting-rights securities from 1926 to 1986. In 1989, the NYSE adopted the language of SEC Rule 19(c)4, which generally prohibits actions that disenfranchise shareholders. Although the SEC rule was later struck down by the courts, it remains the NYSE policy. Thus, multiple-class equity structures are rarely found among large firms in the US.
7 This is true for all companies in our sample.
8 Some nonvoting stocks have no par value. In such cases, the dividend is declared directly proportional to one of the voting classes in the company's articles of incorporation.
9 If the registered shares are very thinly traded, but the nonvoting and bearer shares meet this criterion, only the nonvoting and bearer shares are included.
10 For three companies (BBC, Surveillance, and Von Roll), the values of LPRATIO are higher for the registered shares. For these companies, the registered shares have five times the voting rights of the bearer shares for equal cash flow rights.
11 A referee correctly pointed out that a company can effectively be controlled without a majority of the votes. The CO variable would, however, remain unchanged in our model if the definition of a controlling majority were set at 1/3 instead if 1/2, as there are no holdings in the sample between these values.
12 Eight companies in the sample have 1:5 inferior voting rights for bearer shares, two companies 1:4 and another three companies 1:2. Five companies have only registered, and four companies have only bearer shares outstanding.
13 We also investigated using a proxy for liquidity of the average log ratios of trading volume over the 21 months of the sample period, calculated similarly to LPRATIO. It too was significant and had the expected sign. Including both variables led to a loss of significance for the volume measure, so we chose to report only the results for LSPRDIF.
14 This is calculated as follows. Because the inferior voting shares have twice the par value of the superior voting shares, LPAR = ln2, or 0.693. Multiplying the coefficient of LPAR, 0.78, by the LPAR value of 0.693 yields an increase in the premium of 5.4% for the superior voting shares. The premium for the inferior voting shares is reduced by the same amount, leading to an overall difference of over 10% between the voting and nonvoting shares.
15 Different specifications, such as using a non-logarithmed price ratio for the dependent variable or using other explanatory variables such as trading volume and log size, did not affect the results. Running the regressions using only one class of voting stock from each firm, which resulted in only 29 observations, also yielded very similar results. Interestingly enough, using the Shapley (1953) value of the major block holdings instead of these dummies did not add to the model. In the spirit of Occam's razor, we thus chose to retain the simpler dummy variable. The authors will gladly provide the results of these additional specifications upon request.
Amihud, Y. and H. Mendelson, 1986, "Asset Pricing and the Bid-Ask Spread," Journal of Financial Economics (December), 223-249.
Amihud, Y. and H. Mendelson, 1987, "Trading Mechanisms and Stock Returns: An Empirical Investigation," Journal of Finance (July), 533-553.
Amihud, Y. and H. Mendelson, 1988, "Liquidity and Asset Prices: Financial Management Implications," Financial Management (Spring), 5-15.
Amihud, Y. and H. Mendelson, 1991, "Liquidity, Maturity, and the Yields on US Treasury Securities," Journal of Finance (September), 1411-1426.
Ang, J.S. and C.W. Hodges, 1993, "Time-Series Tests of the Marginal Value of Voting Rights in Proxy Contests for Control," Florida State University Working Paper.
Ang, J.S. and W.L. Megginson, 1989, "Restricted Voting Shares, Ownership Structure, and the Market Value of Dual-Class Firms," Journal of Financial Research (Winter), 301-318.
Angel, J.J., 1994, "The Municipal Bond Insurance Riddle," Financier: Analyses of Capital and Money Market Transactions (February), 48-63.
Barclay, M.J. and C.H. Holderness, 1989, "Private Benefits from Control of Public Corporations," Journal of Financial Economics (December), 371-396.
Bathala, T.B., K.P. Moon, and R.P. Rao, 1994, "Managerial Ownership, Debt Policy, and the Impact of Institutional Holdings: An Agency Perspective," Financial Management (Autumn), 38-50.
Bergstrom, C. and K. Rydqvist, 1990a, "The Determinants of Corporate Ownership: An Empirical Study on Swedish Data," Journal of Banking and Finance (August), 237-253.
Bergstrom, C. and K. Rydqvist, 1990b, "Ownership of Equity in Dual-Class Firms," Journal of Banking and Finance (August), 255-69.
Bergstrom, C. and K. Rydqvist, 1992, "Differentiated Bids for Voting and Restricted Voting Shares in Public Tender Offers," Journal of Banking and Finance (February), 97-114.
Bhagat, S. and J.A. Brickley, 1984, "Cumulative Voting: The Value of Minority Shareholder Voting Rights," Journal of Law and Economics (October), 339-365.
Blair, D.H., D.L. Golbe, and J.M. Gerard, 1989, "Unbundling the Voting Rights and Profit Claims of Common Shares," Journal of Political Economy (April), 420-443.
Brous, P.A. and O. Kini, 1994, "The Valuation Effects of Equity Issues and the Level of Institutional Ownership: Evidence from Analysts Earnings Forecasts," Financial Management (Spring), 33-46.
Cornett, M,M. and M.R. Vetsuypens, 1989, "Voting Rights and Shareholder Wealth: The Issuance of Limited Voting Common Stock," Managerial and Decision Economics (September), 175-188.
Daves, P.R. and M.C. Ehrhardt, 1993, "Liquidity, Reconstruction, and the Value of US Treasury Strips," Journal of Finance (March), 315-330.
DeAngelo, H. and L. DeAngelo, 1985, "Managerial Ownership of Voting Rights: A Study of Public Corporations with Dual Classes of Common Stock," Journal of Financial Economics (March), 33-69.
Diamond, D.W. and R.E. Verrecchia, 1991, "Disclosure, Liquidity, and the Cost of Capital," Journal of Finance (September), 1325-1360.
Easterbrook, F.H. and D.R. Fischel, 1983, "Voting in Corporate Law," Journal of Law and Economics (June), 395-427.
Errunza, V. and E. Losq, 1985, "International Asset Pricing Under Mild Segmentation: Theory and Test," Journal of Finance (March), 105-124.
Grossman, S.J. and O.D. Hart, 1980, "Takeover Bids, the Free Rider Problem, and the Theory of the Corporation," The Bell Journal of Economics (Spring), 42-64.
Grossman, S.J. and O.D. Hart, 1988, "One Share-One Vote and the Market for Corporate Control," Journal of Financial Economics (January/March), 175-202.
Haeberle, R. and J. Pasquier-Dorthe, 1991, "La Valeur Relative des Actions Nominatives, des Actions au Porteur et des Bons de Participation," Finanzmarkt und Portfolio Management 5(4), 280-287.
Harris, M. and A. Raviv, 1988, "Corporate Governance: Voting Rights and Majority Rules," Journal of Financial Economics (January/March), 203-235.
Harris, M. and A. Raviv, 1989, "The Design of Securities," Journal of Financial Economics (October), 255-287.
Herman, W. and G.J. Santoni, 1989, "The Cost of Restricting Takeovers: A Lesson from Switzerland," Federal Reserve Bank of St. Louis Review (November/December), 3-11.
Horner, M.R., 1988, "The Value of the Corporate Voting Right," Journal of Banking and Finance (March), 69-83.
Jarrell, G.A. and A.B. Poulsen, 1988, "Dual Class Recapitalizations as Antitakeover Mechanisms: The Recent Evidence," Journal of Financial Economics (January/March), 129-152.
Jensen, M.C. and W. H. Meckling, 1976, "Theory of the Firm: Managerial Behavior, Agency Costs, and Ownership Structure," Journal of Financial Economics (October), 305-360.
Jensen, M.C. and R.S. Ruback, 1983, "The Market for Corporate Control: The Scientific Evidence," Journal of Financial Economics (April), 5-50.
Kaufmann, H. and B. Kunz, 1991, Shareholder Restrictions in Switzerland, Swiss Research Bank Julius Baer, Zurich.
Kunz, R.M., 1994, "Fakten und Uberlegungen zur Einheitsaktie und zum Fall Schweizerische Bankgesellschaft," Finanz und Wirtschaft (November 16), 23.
Kursblatt der Zurcher Effektenborse, 1990/91, various issues.
Lease, R.C., J.J. McConnell, and W.H. Mikkelson, 1983, "The Market Value of Control in Publicly Traded Corporations," Journal of Financial Economics (April), 439-472.
Lease, R.C., J.J. McConnell, and W.H. Mikkelson, 1984, "The Market Value of Differential Voting Rights in Closely Held Corporations," Journal of Business (October), 443-467.
Levy, H., 1983, "Economic Evaluation of Voting Power of Common Stock," Journal of Finance (March), 79-93.
Loderer, C. and A. Jacobs, 1995, "The Nestle Crash," Journal of Financial Economics (March), 315-340.
Maynes, E., 1992, "Reallocation of Voting Rights and Shareholders' Wealth," Canadian Journal of Economics (August), 538-563.
Meier-Schatz, C., 1993, "Legal Aspects and Institutional Realities of Corporate Governance in Switzerland," Finanzmarkt und Portfolio Management (7), 309-321.
Merton, R., 1987, "A Simple Model of Capital Market Equilibrium with Incomplete Information," Journal of Finance (July), 483-510.
Mikkelson, W.H. and M.M. Partch, 1994, "The Consequences of Unbundling Managers' Voting Rights and Equity Claims," Journal of Corporate Finance (August), 175-200.
Partch, M.M., 1987, "The Creation of a Class of Limited Voting Common Stock and Shareholder Wealth," Journal of Financial Economics (December), 313-340.
Peters, R., 1992, "Better Deal for Swiss Shareholders," International Financial Law Review (January), 11-12.
Raad, E. and R. Ryan, 1995, "Capital Structure and Ownership Distribution of Tender Offer Targets: An Empirical Study," Financial Management (Spring), 46-56.
Shapley, L.S., 1953, "A Value for N-Person Games," in Contributions to the Theory of Games, Volume 2, H. Kuhn and A.W. Tucker, Eds., Princeton, N J, Princeton University Press, 307-317.
Silber, W.L., 1991, "Discounts on Restricted Stock: The Impact of Illiquidity on Stock Prices," Financial Analysts Journal (July/August), 60-64.
Smith, B.F. and B. Amoako-Adu, 1995, "Relative Prices of Dual Class Shares," Journal of Financial and Quantitative Analysis (June), 223-239.
Stulz, R.M., 1981, "On the Effects of Barriers to International Investment," Journal of Finance (September), 923-934.
Stulz, R.M., 1988, "Managerial Control of Voting Rights: Financing Policies and the Market for Corporate Control," Journal of Financial Economics (September), 25-54.
Stulz, R.M. and W. Wasserfallen, 1995, "Foreign Equity Investment Restrictions, Capital Flight, and Shareholder Wealth Maximization: Theory and Evidence," Review of Financial Studies (Winter), 1019-1057.
Swiss Stock Guide, 1991, Xeric, Verlag Finanz und Wirtschaft.
Zimmermann, H., M. Bill, and R. Dubacher, 1989, Finanzmarkt Schweiz: Strukturen im Wandel, Zurich, Zurcher Kantonalbank (Hrsg.).
Zingales, L., 1994, "The Value of the Voting Right: A Study of the Milan Stock Exchange Experience," Review of Financial Studies (Spring), 125-148.
Zingales, L., 1995, "What Determines the Value of Corporate Votes?" Quarterly Journal of Economics (November), 1047-1074.
Roger M. Kunz is in the Department of Corporate Finance, Universitat Basel, Switzerland. James J. Angel is Assistant Professor of Finance at Georgetown University, Washington, DC.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Special Issue: European Corporate Finance|
|Author:||Kunz, Roger M.; Angel, James J.|
|Date:||Sep 22, 1996|
|Previous Article:||Why do firms issue convertible debt?|
|Next Article:||Shareholder wealth effects of equity issues in emerging markets: evidence from rights offerings in Greece.|