Class Struggle
Transparent ticker symbols level the playing field for all investors.

By Najah Attig, assistant professor of finance and management science at Saint Mary’s University in Halifax.

Over the years, market regulation has evolved to provide better protection to market participants by improving information quality in the capital markets. In this context, the TSX Group decided, in 2004, to extend the ticker symbol of stocks with non-conventional voting structures to emphasize the class of shares and its attached voting rights. This rule intends to enable (minority) investors to identify non-conventional voting shares without referring to an issuer’s regulatory filings. In this paper, I examine whether this ticker symbol extension affects prices and liquidity of the involved stocks. In doing so, this paper contributes to a timely, but rare, line of studies on the information content of ticker symbol. Until recently, little attention has been given to the information content of ticker symbols. Rashes (2001) provides evidence on investors’ misperception and confusion of ticker symbols, and emphasizes the importance of understanding the information content of ticker symbol (change). More recently, Kadapakkam and Misra (2007) document significant declines in trading volume and prices of stocks subject to a voluntary change in their ticker symbols.

Arguably, small investors in multiple class shares (MCS) firms are usually ill-informed about their holdings, because the effective voting rights attached to each share (or its voting class) are not explicitly disclosed and stock symbols are not suggestive of the voting class. This lack of transparency may leave investors, in particular holders of the subordinate voting class shares, unaware of the differences between the outstanding classes of shares, and thus unaware of the quality of their holdings and the associated risks (e.g. information asymmetry, expropriation). In contrast, controlling shareholders, in control of multiple voting shares, are likely to capitalize on their informational advantage to engage in extracting private benefits of control. Dyck and Zingales (2004, p. 52) argue that the price difference (or the voting premium) between the different classes of shares can be viewed as an “indication of inadequate protections for minority shareholders and a weak corporate governance system.” Because of their non-conventional voting structure, MCS magnify the misalignment between the commensurate capital and the control rights and, thus, give controlling shareholders discretion to divert corporate resources for their private benefit.

I posit that, all else equal, the TSX rule to re-symbolize tickers of MCS stocks may increase the ability of investors (and other market participants) to have timely and accurate information about their holdings. This enhanced transparency is likely to enable investors to revise downward the assessment of their holding for better price protection against potential corporate wrongdoings. To test this hypothesis, I examine the impact of the tickers’ re-symbolizing rule on stock returns and stock liquidity. First, I find a negative and significant impact on prices of the involved stocks, suggesting that strengthening market disclosure by making tickers of public firms more informative seems to have a positive impact on investors’ ability to priceprotect themselves by revising downward the assessment of their MCS holdings. Second, consistent with the argument that liquidity reaction is affected by the price reaction and the level of public information disagreement among investors (Bailey et al., 2005), I find a significant decrease in the liquidity of the involved stocks, with the most severe decrease incurred by the lower-voting class. Last but not least, the event results of the TSX decision to discontinue the use of the ticker symbol extension point to an argument for investor irrationality during periods of confusion and potential market manipulation by major participants.

The importance of the TSX rule 1
The new TSX ticker rule aims to convey information on the voting structure of publicly traded shares. Five classes of shares with non-conventional voting structures are concerned. Non-voting shares which have no right to vote should include the suffix “NV”, multiple voting shares which have more than a voting right per share should include the suffix “MV”, subordinate voting shares which carry a right to vote, where there is another class or classes of shares outstanding that carry a greater voting right on a per-share basis should include the suffix “SV”, limited voting shares which have the right to vote only in certain limited circumstances should include the suffix “LV”, and restricted voting shares which carry a right to vote, subject to some restriction (e.g. percentage of the board that can be elected by the holders of this class or the number of shares that may be voted by the owner) should include the suffix “RV”.

To highlight the importance of the TSX re-symbolizing rule, I present one case of MCS firms. The first firm is Headline Media Group Inc (TSX: HMG). As of January 7, 2004, the proxy circular shows that the authorized equity capital of Headline consists of 82,626,200 Class A Subordinate Voting Shares and 10,000 Special Voting Shares are issued and outstanding. Each Class A Subordinate Voting Share has attached thereto one vote for each share held. Similarly, the holders of the Special Voting Shares are entitled to one vote for each share held. Although both classes seem to be equal with respect to their disclosed voting rights, holders of the special voting shares can elect the majority of the authorized number of directors of Headline. Translating this condition into effective voting rights, one can see that each special voting share has X voting rights, where X satisfies the following equation:

therefore X = 8,600 votes per share. To emphasize the subordinate voting structure of Class A shares, a suffix “SV” was added to its ticker.

Data
I collect daily market data on stock prices (i.e. closing price, closing bid, and closing ask), returns, trading volume, and monthly outstanding shares from the TSXCanadian Financial Markets Research Centre (TSXCFMRC) database. The initial sample for this study is the list of the 141 TSX-listed stocks that were subject to the TSX re-symbolizing decision. First, I exclude stocks of firms under protection or plan of arrangement or reorganization, stock of subsidiary of foreign firms, and stocks that changed ticker symbol during the estimation period (e.g. M8 Entertainment Inc.). Then, I restrict my sample to stocks with at least 100 daily observations during the estimation period and a maximum of 10 missing observations in the event window. These screenings reduce my sample to 97 involved stocks (tickers), comprising 26 non-voting stocks, 10 multiple-voting stocks, 61 subordinate voting stocks (and limited- and restrictedvoting stocks). I use firms’ management proxy circulars, published in 2004 and available at www.sedar.com, to manually collect data on the different outstanding classes of shares, the size of ultimate ownership and control stakes, the identities of their owners, and the structure of the board of directors (e.g. size and proportion of outside directors). Interestingly, I find that the publicly traded multiple-voting shares have, on average, 21 votes per share, and controlling shareholders need to own only 6.61% cash flow rights to control 20% of the voting rights of a Canadian public MCS firm. Equally important, the controlling shareholders control about 77% of the outstanding multiple-voting shares and less than 10% of the lower voting shares.

Methodology
I apply the standard event study methodology to examine the effects of the TSX ticker transparency-enhancing rule on stock prices and stock liquidity. The event day is April 27, 2004, date of public announcement of the TSX initiative to re symbolize tickers with non-conventional voting rights. The estimation period is defined from day 11 to day 210 before the announcement day. The event window is defined as 10 days around the event date. Daily abnormal returns (ARs) are calculated based on the excess return (over normal returns predicted by the market model) during the event window:

are the abnormal, actual and normal returns of stock i on day t. is the market index return. I use two measures of stock liquidity. I start by looking at the trading volume, which is usually associated with the incorporation of information changes in stock prices (e.g. Morse, 1980). To minimize the firm-size effect and obtain a normalized measure, I use the daily turnover as a proxy for volume, where daily turnover is daily trading volume divided by the number of outstanding shares. The second liquidity measure is the daily bidask spread (BASP), defined as the quoted spread divided by the bid-ask midpoint

I calculate abnormal turnover for a security as the difference between its turnover during the test period and the average turnover of the estimation period:

where and are the abnormal and actual turnover for firm i on day t.

is the average turnover of firm i over the estimation period. I use an analogous method to estimate the average abnormal bid-ask spread . Finally, average abnormal levels of the considered variables are aggregated over the event window to estimate the cumulative average abnormal levels.

Results 2
Results for the TSX re-symbolizing rule on stock returns (RETN), liquidity turnover (TURN) and bid-ask spread (BASP) are reported in Table 1. Panel A shows a statistically significant (at 5%) average abnormal return of -0.38% on day 0. The price drop appears to start nine days before the announcement, and continued to drop significantly up to the ninth day after the announcement (-0.79%). Examining a 10-day window surrounding the announcement, I find a negative and significant cumulative abnormal return of -3.26%. Interestingly, I find significant cumulative abnormal returns of -1.31% and -1.95% over the event windows (-10,-1) and (0,10), respectively. This evidence suggests that there was information leakage regarding this event. This might be expected as many market participants (i.e. institutional investors) had been “lobbying the TSX Group” to implement the ticker symbol extension.3 Overall, the economically significant drop in the price of the involved stocks suggests that the event price-pressure relates to the information content of the ticker symbol change, plausibly because investors revised downward the assessment of their MCS holdings against potential corporate wrongdoings.

An important feature of the involved stocks is the existence of two broad classes of shares: one with superior voting rights (i.e. multiple voting) and the other with lower voting rights (i.e. subordinate, limited, restricted or non-voting rights). To test how the previous results vary across the different voting classes of shares I apply the event study on the different voting classes. Although all voting classes displayed significant abnormal returns (in mean or in median) on the announcement day, only the lower-voting class had negative and significant cumulative abnormal returns. The most distinguishable result comes from the large and consistently significant drop in the price of the non-voting shares (e.g. -5.45% over the event window (-10,10)). This evidence lends further support to the argument that holders of a lower voting class of shares (e.g. non-voting) revised downward the assessment of their holdings for further price-protection against potential corporate expropriation.

Based on the argument of Bailey et al. (2005), the liquidity reaction is affected by the price reaction and the level of public information disagreement among investors. The ticker enhanced transparency is likely to reduce disagreement about MCS tickers’ informativeness, thus leading to a decrease in trading volume (i.e. liquidity turnover). Results of the abnormal turnover support this argument. For instance, the cumulative average abnormal turnover was about -79.85% over the event window (-10,10), and seems to be evenly split on the pre- and post-announcement event periods.4 The most significant abnormal turnover was reported for the lower-voting class. However, the decrease in the turnover of the multiplevoting class is not significant. This result may also explain the non-significance of the cumulative abnormal return of this class. To some extent, the volume effect is consistent with the evidence of Kadapakkam and Misra (2007), who show that a ticker change adversely affects trading volume. A similar pattern is observed for the cumulative abnormal levels of the bid-ask spreads, with the exception of the multiple-voting class, which had a significant decrease in its average bid-ask spread. Although this result seems surprising, it may suggest that the enhanced transparency of the MCS ticker decreased the likelihood of informed trading by holders of multiple-voting shares. Importantly, I find that non-voting shares had the largest increase in the bid-ask spread (1.37%). Overall, the evidence of increased (cost of) illiquidity of MCS stocks due to enhanced transparency is consistent with the information content of market trading in response to public disclosure (Kim and Verrecchia, 1997) and with the findings of Harris (1997) who conclude that changes in transparency result in increased transaction costs or reduced liquidity.

As a conclusive postulate, the negative and significant effect of the tickers’ re-symbolizing rule on the stock returns and the two liquidity measures seems to be very related to changes in the information content of the involved stocks. Enhancing market transparency through an increase in the ticker’s informativeness seems to reduce the ability of controlling shareholders to utilize or generate private information.

Additional Results
It is important to point out that, recently, the TSX decided to discontinue the use of the new symbol rule for MCS. In its press release of December 21, 2005, the TSX stated, “These changes are the result of an extensive consultation process with market participants—many of whom had been lobbying the TSX Group for changes to the 2004 symbol modifications.” It is understandable that those market participants might have incurred some costs in their portfolio due to the enhanced informativeness of MCS tickers. I run a similar event study to test the effect of the discontinuation of the TSX symbol extension on the return, turnover, and bid-ask spread of the involved stocks. The sample used for this analysis decreased to 82 stocks, the event window was also limited to five days after the announcement day because data for 2006 is not available. The event results are reported in Table 2.

The distinguishable result in Table 2 is the positive and significant cumulative abnormal return across the different event windows (e.g. pre- and postannouncement). The pre-event run-up suggests again that market participants anticipated such announcements (i.e. information leakage). The cumulative abnormal turnover effect lends support to this argument, as I find a significant turnover increase in the pre-event window followed by a significant turnover decrease in the post- event window. The event had a neutral effect on the bid-ask spread, suggesting that no significant information change was associated with the TSX decision to discontinue the use of the ticker symbol extension. Yet, the evidence that this (off) event affects prices and disrupts trading of the involved stocks leans toward the argument of investor irrationality in periods of confusion and potential market manipulation by major participants.

Conclusion
In this paper I explore the economic consequences of increasing tickers’ informativeness of stocks with nonconventional voting structure. The results show that the TSX re-symbolizing rule has a negative and significant impact on securities prices of both multiple and lower voting classes. The results also show a significant decrease in the liquidity of the involved stocks, with the most severe decrease incurred by the lower voting class. The evidence in this paper is of particular interest as it suggests that strengthening market disclosure by making tickers of public firms more informative seems to have a positive impact on investors’ ability to price-protect themselves. The results lend support to the argument that holders of lower voting classes revised downward the assessment of their holdings for further priceprotection against potential corporate expropriation. To some extent, the evidence in this paper stresses the importance of enhancing market transparency in curbing private benefits. Finally, the event results of the TSX decision to discontinue the use of the ticker symbol extension lean toward the argument of investor irrationality in periods of confusion and potential market manipulation by major participants.

References
Bailey W., Karolyi G. A., and C. Salva, 2005, “The Economic Consequences of Increased Disclosure: Evidence from International Cross Listings,” Journal of Financial Economics, forthcoming.
Dyck, A. and L. Zingales, 2004, “Control Premiums and the Effectiveness of Corporate Governance Systems,” Journal of Applied Corporate Finance 16, 51-73.
Harris L., 1997, “Order Exposure and Parasitic Traders. Equity Market Structure for Large and Mid-Cap Stocks,” Deutsche Börse, 31-50.
Kaul, A., Mehrotra, V. and R. Morck (2000) Demand Curves for Stocks Do Slope Down: New Evidence from an Index Weights Adjustment, Journal of Finance, 55, 893–912.
Kim O., and R. Verrecchia., 1997, “Pre-announcement and Event Period Private Information,” Journal of Accounting and Economics, 24, 395-419.
Morse, D., 1980, Asymmetrical information in securities markets and trading volume, Journal of Financial and Quantitative Analysis 15, 1129-1148.
Rashes, M., 2001, “Massively Confused Investors Making Conspicuously Ignorant Choices,” The Journal of Finance 56, 1911-1927.

Acknowledgments
I am grateful for very constructive comments from Paul Halpern and two anonymous reviewers at Canadian Investment Review. I thank participants at the EFMA meeting (Vienna, 2007) for their insightful discussions. I acknowledge financial support from the Schulich School of Business National Research Program in Financial Services and Public Policy, and the Social Sciences and Humanities Research Council of Canada. The usual disclaimer applies.

Endnotes
1. In Canada, securities market regulation is fragmented and is a matter of provincial (or territorial) jurisdiction. Currently, there is no legal prohibition or restriction on MCS structures for companies issuing stock, neither by the provincial securities commissions nor by the Toronto Stock Exchange. However, under certain circumstances, holders of restricted securities may vote on a one-for-one basis with the multiple voting security holders.
2. In this section I discuss average abnormal and cumulative average abnormal levels of the considered variables; however, similar patterns are found for median abnormal and cumulative median abnormal levels, suggesting the reported results do not seem to be driven by outliers.
3. See for instance the TSX press release (December 21, 2005).
4. This can be plausibly explained by the desire of the market participants to “spread their trades over an extended period in order to reduce the price impact” (Kaul et al. 2000).

For a pdf version of this story, click here.