DOES INDEX INCLUSION IMPROVE FIRM VISIBILITY AND TRANSPARENCY? * John R. Becker-Blease Whittemore School of Business and Economics University of New Hampshire 15 College Road Durham, NH 03824-3593 jblease@cisunix.unh.edu 603.862.3301 (voice) 603.862.3383 (fax) Donna Paul Babson College Babson Park, MA 02457-0310 dpaul@babson.edu 781.239.5345 (voice) 781.239.5004 (fax) January 2003 * We thank seminar participants at Babson College for their helpful comments and suggestions.
DOES INDEX INCLUSION IMPROVE FIRM VISIBILITY AND TRANSPARENCY? Abstract We examine changes in various measures of visibility and transparency for a sample of firms added to the S&P 500 index. We find significant increases in visibility measures such as the number of analysts making earnings estimates and the number of news stories reported in the Wall Street Journal following addition. However, there are no changes in traditional proxies for information transparency such as analysts forecast errors and standard deviation of forecasts. We explore whether the general increase in investor interest is associated with changes in capital market activity, and find significant increases in debt issues. Overall, the evidence indicates that firm visibility increases following index addition, and suggests that this increased investor interest provides a favorable environment for external financing.
It is commonly accepted that firm visibility and transparency are beneficial to shareholders. Asset transparency reduces the likelihood of distorted investment and financing decisions that arise when managers pursue their own objectives or attempt to transfer wealth from bondholders to stockholders (Morck, Shleifer and Vishny, 1990; Myers and Majluf, 1984). In addition, visible stocks are more widely held, and therefore likely to have lower liquidity premiums and lower discounts demanded by incompletely diversified investors (Amihud and Mendelson, 1986; Merton, 1987). Most of the empirical studies linking visibility to firm value examine changes associated with exchange listings. For example, Baker, Nofsinger, and Weaver (2002) find significant increases in number of analysts and news stories for stocks cross-listed between the New York and London Stock Exchanges. They also find that the increase in these visibility proxies is associated with a decrease in cost of equity capital, which they interpret as consistent with the investor recognition hypothesis in Merton (1987). Kadlec and McConnell (1994) also test Merton s investor recognition hypothesis for firms listing on the New York Stock Exchange using total number of shareholders and number of institutional shareholders as proxies for visibility or investor recognition. They find significant increases in these measures following listing, and that these changes are related to abnormal stock returns in a manner consistent with Merton s hypothesis. We examine the changes in visibility and asset transparency following a stock s addition to the Standard and Poor s 500 index (S&P 500). Addition to the S&P 500 is distinct from managerial attempts to improve firm visibility through actions such as exchange listings because the index addition decision is exogenous. Further, as Shleifer (1986) argues, firms added to the S&P 500 are already very large and have high investor 1
interest. Thus, it is arguable whether addition to the index significantly improves firm visibility or transparency. In spite of these arguments, there is evidence that stocks receive increased attention from investors following index addition. For example, Chen, Noronha and Singal (2002) find that the firm s investor base broadens following addition, and that this increase is not due solely to the holdings of index funds. They document a median change in number of shareholders ranging from 17% to 31% for a large sample of firms added to the S&P 500 from 1976 to 1999. They also document a significant increase in the percentage of the firm s shares held by institutions after index addition. This suggests that index funds do not displace existing institutional shareholders, whose interest in the firm persists following addition. The increase in number of individual shareholders, coupled with the apparent low turnover of non-index institutional holdings following addition, suggests increased investor interest following S&P 500 addition. The notion that firm visibility improves following addition, and that this improvement is beneficial to shareholders, is also articulated by some academics and practitioners. For example, Denis, McConnell, Ovtchinnikov and Yu (2002) postulate that index addition could lead to greater managerial effort because of increased scrutiny by investors or because the personal cost of failure increases with visibility. They use these arguments to motivate their empirical study of improvements in earnings expectations following index addition, and indeed find evidence consistent with anticipated increases in earnings. In addition, comments by the Managing Director and Chief Investment Strategist of Standard & Poor s (Blitzer, 2001), suggest that companies in the index perform better partly because of a list effect. He argues that this arises because the S&P 500 is often the starting point for identifying leading companies. Thus, 2
analyst coverage and general investor interest are likely to increase following addition, which he argues, result in gains for shareholders. The relation between asset transparency, visibility, and index inclusion is therefore an empirical question, which we address in this paper. We focus on two primary areas of analysis. First, we examine whether visibility and transparency improve by documenting changes in several measures of information asymmetry and visibility following index addition. We then investigate what we consider a natural implication of increased investor interest, which is better access to capital markets. Thus, we investigate whether firms have abnormal increases in their frequency or amounts of capital raised following addition. We motivate our hypothesis that the intensity of capital market activity is linked to increased asset transparency with the Myers and Majluf (1984) model. They argue that asymmetric information between managers and investors with respect to the value of a firm s assets can harm firm value by leading to underinvestment. In their model, investors believe that the equity issue decision reveals managers private information that the firm is overvalued, and therefore discount the firm s share price upon issue announcement. This adverse selection problem can lead managers to forego profitable projects if they require external financing, and especially equity financing. Similarly, Stiglitz and Weiss (1981) demonstrate that asymmetric information can affect the cost of new debt, which in turn can result in underinvestment. Thus, we test whether financing activities following index addition suggest an improved information environment. We find significant increases in the number of analysts making earnings estimates and in the number of news stories reported in the Wall Street Journal following index 3
addition. Consistent with our hypothesis that firms with greater investor attention are likely to have easier access to the capital markets, we find that newly added firms significantly increase their amounts of external debt financing. In addition, firms have significant increases in their cash holdings, indicating that the proceeds of the financing do not merely fund current investment projects. Rather, it appears that firms take advantage of the increased visibility following index addition to issue new capital, even if there is no immediate use for the proceeds. This combined evidence suggests that investor attention increases following addition, and that firms exploit the improved visibility by increasing their intensity of capital market activity. The remainder of the paper proceeds as follows. The next section details the sample selection and study design. Section II provides evidence on information and visibility changes following index addition. We investigate the implications of such changes in Section III by documenting the changes in capital market activity. Section IV identifies areas for further research and Section V concludes. I. Study Design Our sample comprises non-financial firms that were added to the S&P 500 list from 1980 to 1997. We begin with an initial sample of 301 firms, which we identify from Lexis-Nexis news stories, Wall Street Journal announcements, and data from Standard & Poor s. We delete 40 firms for various reasons such as the firm being a division of a previously listed firm, additions that are simply name changes, and firms missing data on Compustat. We also delete 54 financial services firms because the investing activities of these firms are not directly comparable with industrial firms. After 4
these deletions, we end up with a sample of 207 index additions. We later delete 8 firms that we could not match with a control firm, yielding a final sample of 199 firms. We collect data from CRSP, Compustat, I/B/E/S, and Security Data Corporation (SDC) to perform our empirical tests. The CRSP database provides stock return data, and the SDC database provides data for debt and equity issues, and for acquisitions of U.S. and foreign targets by added firms. The I/B/E/S data provide our measures of information asymmetry, and we collect accounting data from Compustat. We first document changes in proxies for firm visibility and transparency of the information environment. We follow Beneish and Gardner (1995) and measure visibility based on the number of news stories reported for a firm. We collect the frequency of stories reported in the Wall Street Journal Index in the four years surrounding index addition. We also use the number of analysts as a measure of visibility, following Brennan and Subrahmanyam (1995). To measure changes in asset transparency, we use traditional measures of information asymmetry such as the accuracy of consensus forecasts and dispersion among forecasts, as reported by I/B/E/S. There is evidence that analysts forecasts are optimistic at the beginning of the fiscal year (Fried and Givoly, 1982; O Brien, 1988; Easterwood and Nutt, 1999). Therefore, similar to Thomas (2002), we collect forecasts made in the last month prior to actual earnings announcements. Krishnaswami and Subramaniam (1999) use the last month in the fiscal year to standardize the impact of the bias. However, given that actual earnings announcements can occur from one to several months after the end of the fiscal year, we believe that a better point of standardization is relative to the earnings announcement date. 5
We use analysts forecasts to develop two measures of information asymmetry. The first is the forecast error of the consensus earnings estimate, defined as the median estimate minus the actual earnings per share. We are not concerned with the direction of the error only with the magnitude and therefore, similar to Christie (1987), we measure the error as the absolute difference between the estimate and actual earnings. We deflate the error with the stock price five days before the earnings announcement date. The second measure of information asymmetry is the dispersion of analysts forecasts, measured as the standard deviation of analysts forecasts deflated by the stock price five days before the earnings announcement date. Disagreement among analysts is indicative of differences in information sets and thus greater information asymmetry. We also employ addition measures such as forecast errors based on mean forecasts, normalized forecast errors that deflate dispersion by the standard deviation in detrended quarterly earnings over the five-year period before the addition date, and the standard deviation of returns surrounding earnings announcement dates. In our empirical tests, we compare changes in the variables of interest for index additions with changes made in a control sample of firms. We match sample firms on industry, size, and market-to-book value of assets in the following manner. For each firm, we identify a pool of potential matches with the same 4-digit SIC code. From this pool, we first delete firms that were added to the S&P 500 within three years of the sample firm, and then identify all firms that are within 20% of the book value of assets of the sample firm. 1 Finally, from this group of potential industry and size matches, we select the firm with the closest market-to-book value, provided that the market-to-book is 1 The pool of potential matches includes firms that are already in the S&P 500. However, because our purpose is to pick up changes in corporate activity for newly added firms, we do not match on firms that were added to the index in the 6-year period (t-3, t+3) around the sample firm s addition year. 6
within 20% of the sample firm s measure. We then repeat the matching algorithm for the remaining sample firms without matches, but begin with a pool of potential matches in the same 3-digit SIC industry. We repeat the procedure, identifying potential matches in the same 2-digit SIC industry. For the remaining firms without matches, we increase the size band and the market-to-book band until we find a matched firm. We were unable to match 8 firms, and our final sample consists of 199 index additions. II. Changes in Information and Visibility Proxies Table 1 reports the results of tests of changes in information asymmetry associated with index addition. Prior to addition, sample firms have particularly low levels of information asymmetry of.05%, as measured by both median forecast errors and standard deviation of forecasts. In comparison, Doukas, Kim, and Pantzalis (2002) report that the median firm between 1976 and 1997 had a forecast error of 0.52%. The typically large size of index additions may account for some of the disparity, as Barry and Brown (1984) find a positive correlation between firm size and information availability. Doukas, et al. break forecast errors into size quintiles based on market value of equity. The two quintiles of the largest firms ($0.63 - $2.7 billion and greater than $2.7 billion) have median errors of 0.26% and 0.13%, respectively. The median firm in our sample of additions has total market value of equity of $1.2 billion, placing it in the second to largest quartile, and yet its error is substantially below the median firms error in that group both before and after addition. Overall, there is no evidence from these information proxies that sample firms experience a decline in information asymmetry subsequent to index addition compared to control firms. The median sample and control firms experience no change between pre- 7
and post-addition periods for forecast errors. Standard deviation actually increases significantly for sample firms by 0.02%, though this change is not statistically different from the change documented for control firms. The evidence suggests no support for the asymmetric information hypothesis. We also test for changes in visibility measures such as number of news stories and number of analysts. The results, not reported in a table, indicate that on average, 16.4 news stories per year appear about firms added to the index (excluding the addition announcement) in the two following addition compared to 14.0 per year in the two years prior and the difference is significant at the 1% level. We find no significant change in the number of news stories for control firms over the same period. In addition, the average number of analysts making earnings estimates as reported by I/B/E/S increases from 14.2 in the two years before addition to 17.3 in the two years following addition, and the difference is significant at the 1% level. This compares with a much smaller and insignificant increase in analyst coverage for control firms from 11.6 to 12.8. III. Changes in Capital Market Activity In this section, we perform further tests of the hypothesis that index addition attenuates information disparities between shareholders and managers. Although the previous section indicates that the traditional proxies for information asymmetry do not change, the significant changes in the visibility measures suggest increased investor interest following index addition. It is likely that these improvements in firm visibility are associated with increased information production that our proxies do not capture. If this is the case, then access to new capital should improve for newly added firms. We 8
therefore test in this section whether firms increase the frequency and amounts of external debt and equity issues following index addition. In so doing, we are effectively testing the empirical implications of Myers and Majluf (1984) pecking order theory that firms will likely change their pattern of external financing following an event that increases investor awareness of the firm s future prospects. Table 2 presents average amounts of total external financing for the two years following index addition compared with the two years pre-addition for sample and control firms. The table indicates that sample firms significantly increase their amounts of external financing (debt and equity) in the two years following addition compared to the two years before addition. Following index addition, sample firms raise an average of $277 million in external financing compared to $113 million raised in the two years preceding addition. The difference of $164 million is significant at the 1% level. Control firms have virtually the same amounts of external financing as sample firms before index addition, and increase their amounts of external financing by only $54 million over the corresponding post-addition period for sample firms. The difference of $110 million in the change in financing is significant at the 5% level. The rest of Table 2 provides a breakdown of external capital raised by debt and equity issues and provides further insight into the changes occurring in sample firms post-addition compared to control firms. It indicates that the difference in external capital raised is driven by increased debt financing by sample firms relative to control firms, and decreased equity financing by control firms relative to sample firms. Although this table does not indicate the intended use of the proceeds from the offering, it does lend support to the hypothesized relation between index addition and investor interest. It 9
suggests that index addition provides firms with a more favorable environment to procure external financing. This is consistent with the increased visibility we document above, suggesting an improvement in the climate for raising external equity capital following index addition. However, although the evidence in Table 2 indicates increased amounts of financing overall, it does not indicate that newly added firms favor equity issues over debt issues, which is not strictly consistent with pecking order predictions. The evidence presented so far in this section indicates that firms increase their amounts of external financing following index addition. We now investigate the changes in capital structure that result from this external financing. We provide in Table 3 the changes in cash and leverage for the year following addition compared to the pre-addition year for sample firms and control firms. Panel A indicates that sample and control firms have approximately the same level of cash holdings in the pre-addition year. In the postaddition year, both groups increase cash holdings, but the increase of $14.7 million by sample firms is significantly greater than the $1.1 million increase for control firms. The increase in cash holdings suggests that the financing proceeds are not used entirely to fund current capital investments. This is consistent with the evidence presented in Table 2, which suggests that newly added firms exploit a favorable investor climate to increase their amounts of external financing, whether or not there is immediate use for the proceeds. Consistent with the evidence in Table 2 that sample firms source external financing primarily via the debt market, Table 3 indicates that these firms have significant increases in leverage following index addition. For newly added firms, book leverage (long-term debt scaled by pre-addition total assets) increases by 3.28% from one 10
year pre- to one year post-addition, compared to an increase of just under 1% for control firms. However, the difference in these changes between the two groups of firms is not significant. IV. Further Work Additional empirical tests will enhance our understanding of the relation between index inclusion, firm visibility, and securities issuance. We document a significant increase in firm visibility, as measured by news coverage and analyst following, following addition to the S&P 500. Firms also have significant increases in debt financing following addition. In further empirical analysis, we will first measure the abnormal stock returns associated with the announcement of index inclusion, which several other studies have shown to be significantly positive. 2 We will then investigate whether investors value the anticipated increase in visibility by testing for a statistical relation between the announcement returns and subsequent changes in proxies for visibility and transparency. Our interpretation of the results presented in this paper is that following index inclusion, firms receive increased investor attention, and that this attention provides a favorable environment for external financing. We argue that the increase in cash holdings we document suggests that the increased capital market activity following addition is not driven solely by the need to fund current investment projects, but by an improvement in the climate for capital issues. We plan additional tests to further investigate the relation between firm visibility and securities issuance. Specifically, we intend to test for lead-lag relations in visibility changes and securities issues in a 2 See, for example, Hegde and McDermott (2002), Beneish and Whaley (1996), Lynch and Mendenhall (1997), Shleifer (1986), and Harris and Gurel (1986). 11
multivariate setting, controlling for the cash levels required to fund current capital investment projects. V. Conclusion We examine whether firm visibility and transparency improve following addition to the S&P 500. Our empirical investigation is motivated by assertions by academics and practitioners that investor interest increases following index inclusion, and that this attention benefits shareholders (e.g. Denis et al., 2002; Blitzer, 2001). We match each newly added firm with a control firm from the same industry that is similar in size and market-to-book value of assets. We then test whether firms added to the index exhibit unusual changes in various proxies for visibility and information availability compared to their control firms. We find significant increases in analyst following and counts of news stories reported in the Wall Street Journal. This suggests that, even though index addition signals transfers of ownership to presumably passive index funds, these firms do attract increased attention from the general investing community. We do not find any significant changes in traditional proxies for asset transparency such as forecast errors and standard deviation of analyst forecasts. We also investigate a natural consequence of the increased investor attention, which is easier access to capital markets. Newly added firms have significant increases in external financing. Firms also significantly increase their cash holdings compared to a matched control sample, indicating that the capital raised does not merely fund current capital investments, but rather, is associated with exploiting opportunities to issue new capital. This combined evidence suggests that firms exploit the increased visibility by 12
increasing their intensity of capital market activity, and provides additional evidence on the sources of value to shareholders from addition to the S&P 500 index. 13
References Amihud, Y., Mendelson, H., 1986. Asset pricing and the bid-ask spread. Journal of Financial Economics 17, 223-249. Baker, H.K., Nofsinger, J. R., Weaver, D. G., 2002. International Cross-Listing and Visibility. Journal of Financial and Quantitative Analysis 37, 495-521. Barry, C. B., S. J. Brown. The neglected and small firm effects. Journal of Financial Economics 13, 283-294. Beneish, M. D., Gardner, J. C., 1995. Information costs and liquidity effects from changes in the Dow Jones Industrial Average list. Journal of Financial and Quantitative Analysis 30, 135-157. Beneish, M. A., Whaley, R. E., 1996. An anatomy of the S&P game : The effects of changing the rules. Journal of Finance 51, 1909-1930. Blitzer, D. M., 2001. The S&P 500. Speech presented at the Indexing and ETF s Summit, Broomfield, Colorado, March 26 th, 2001. Brennan, M. J., Subrahmanyam, A., 1995. Investment analysis and price formation in securities markets. Journal of Financial Economics 38, 361-381. Brennan, M. J., Subrahmanyam, A., 1996. Market microstructure and asset pricing: On the compensation for illiquidity in stock returns. Journal of Financial Economics 41, 441-464. Chen, H., Noronha, G., Singal, V., 2002. Investor recognition and market segmentation: Evidence from S&P 500 index changes. Unpublished working paper, Virginia Polytechnic Institute and State University. Christie, A., 1987. On cross-sectional analysis in accounting research. Journal of Accounting and Economics 9, 231-258. Denis, D. K., McConnell, J. J., Ovtchinnikov, A. V., Yu, Y., 2002. S&P 500 index additions and earnings expectations. Journal of Finance, forthcoming. Doukas, J. A., Kim, C. F., Pantzalis C., 2002. A test of the errors-in-expectations explanation of the value/glamour stock returns performance: Evidence from analysts forecasts. Journal of Finance, forthcoming. Harris, L., Gurel, E., 1986. Price and volume effects associated with changes in the S&P 500 list: new evidence for the existence of price pressures. Journal of Finance 41, 815-829. Hegde, S. P., McDermott, J. B., 2002. The liquidity effects of revisions to the S&P 500 index: An empirical analysis. Journal of Financial Markets, forthcoming. 14
Hertzel, M., Smith, R. L., 1993. Market discounts and shareholder gains for placing equity privately. Journal of Finance 48, 459-485. Jung, K., Kim, Y., Stulz, R. M., 1996. Timing, investment opportunities, managerial discretion, and the security issue decision. Journal of Financial Economics 42, 159-185. Kadlec, G. B., McConnell, J. J., 1994. The effect of market segmentation and illiquidity on asset prices: Evidence from exchange listings. Journal of Finance 49, 611-636. Kaul, A., Mehrotra, V., Morck, R., 2000. Demand curves for stocks do slope down: New evidence from and index weights adjustment. Journal of Finance 55, 893-912. Krishnaswami, S., Subramaniam, V., 1999. Information asymmetry, valuation, and the corporate spin-off decision. Journal of Financial Economics 53, 73-112. Lynch, A. W., Mendenhall, R. R., 1997. New evidence on stock price effects associated with changes in the S&P 500 index. Journal of Business 70, 351-383. Merton, Robert C., 1987. A simple model of capital market equilibrium with incomplete information. Journal of Finance 42, 483-510. Mikkelson, W. H., Partch, M. M., 1986. Valuation effects of security offerings and the issuance process. Journal of Financial Economics 15, 31-60. Morck, R., Shleifer, A., Vishny, R. W., 1990. Do managerial objectives drive bad acquisitions? Journal of Finance 45, 31-48. Myers, S. C., Majluf, N. S., 1984. Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics 13, 187-221. Shleifer, A., 1986. Do demand curves for stocks slope down? Journal of Finance 41, 579-590. Stiglitz, J.E., Weiss, A., 1981. Credit rationing in markets with imperfect information. American Economic Review, 71, 393-410. 15
Table 1 Changes in information asymmetry variables following index addition for a sample of 207 firms added to the S&P 500 index in the period 1980-1997 and for a sample of 196 size-, industry-, and market-to-book-matched control firms. Forecast errors are the ratio of the absolute value of the difference between median (and mean) analyst forecasts and actual earnings scaled by stock price 5 days prior to the last analyst forecast. Scaled standard deviation of forecasts is the standard deviation of last analyst forecasts before actuals are reported scaled by the stock price 5 days prior to the last forecast. The values within the brackets are the number of observations for which adequate data exist and the values within the parentheses are the Wilcoxon Signed-Rank statistics for equality of medians. *,**,*** denote significant at 10%, 5%, and 1% levels. Forecast Errors from Median Analyst Forecast Median Pre-Addition Median Post-Addition Sample 0.05% 0.07% Control 0.08% 0.14% -0.03% *** [165] (-1766) -0.03% *** [158] (-1803) Post-Pre 0.00% [186] (661) -0.01% [166] (954) -0.03% [148] (-474) Scaled Standard Deviation of Analyst Forecasts Median Pre-Addition Median Post-Addition Post-Pre Sample 0.05% 0.11% Control 0.13% 0.15% 0.02% *** [191] (3405) 0.01% ** [163] (1244) -1.00% *** [173] (-2270) 0.01% ** [165] (-1039.5) -0.01% [149] (-129) 16
Table 2 Average capital issues by sample and control firms during the four years surrounding index addition. All amounts are in millions of dollars. Average Pre-Addition (t statistic) Average Post-Addition (t statistic) (Post-Pre) (t statistic) All Issues (Debt and Equity) Sample $113.25 $276.82 $163.57 *** (t = 2.872) Control $125.64 $179.58 $53.94 * (t = 1.930) {Sample Control} -$12.39 (t = 0.470) $97.24 * (t = 1.724) $109.64 ** (t = 1.973) Debt Issues Sample $65.17 $222.05 $156.88 *** (t = 2.697) Control $97.10 $164.08 $66.98 ** (t = 2.457) {Sample Control} -$31.93 (t = -1.386) $57.97 (t = 1.117) $89.90 (t = 1.602) Equity Issues Sample $48.08 $54.77 $6.69 (t = 0.327) Control $28.54 $15.50 -$13.04 ** (t = -2.397) {Sample Control} $19.54 (t = 1.508) $39.27 ** (t = 2.283) $19.73 (t = 1.234) 17
Table 3 Changes in cash, leverage and market-to-book for sample and controls between the one and two years before and one and two years after addition to the S&P 500 Index for sample and control firms. Levels and differences are based on median values. Accounting variables are as defined in Table 5. Dollar amounts are in millions. Total assets as at the fiscal year-end before the index addition year. Panel A: Changes for one fiscal year before addition to one fiscal year following addition Sample firms Control firms s Fiscal Year Pre- Addition Fiscal Year Post- Addition Post - Pre [n] (Test Statistic) Fiscal Year Pre- Addition Fiscal Year Post- Addition Post - Pre [n] (Test Statistic) Sample - Control Post-Addition Diff. in s Sample-Control [n] (Test Statistic) Cash (millions) $48.28 $61.94 $14.72 *** [198] (4721.5) $36.55 $38.89 $1.10 ** [192] (1599) $16.77 ** [184] (1920) $16.55 *** [183] (2146) Cash/Total Assets 7.80% 11.69% 1.99% *** [198] (5327.5) 8.12% 9.45% 0.42% ** [192] (1526) 0.94% [184] (1103) 3.23% *** [183] (2943) Long Term Debt /Total Assets 15.14% 21.70% 3.28% *** [199] (5196) 20.71% 25.66% 0.80% *** [194] (2637) -6.20% ** [186] (-1763.5) 0.26% [186] (1058.5) 18
Table 3, continued Panel B: Average changes for two fiscal years before to two fiscal years following index addition Cash (millions) $45.47 $77.79 $22.73 *** [195] (5664) $36.24 $47.92 $3.89 *** [190] (2903.5) $20.19 ** [181] (1682.5) $23.25 *** [179] (2080) Cash/Total Assets 8.14% 12.84% 3.36% *** [195[ (5939) 9.41% 10.53% 1.42% *** [190] (3294.5) -0.06% [180] (731) 3.53% *** [179] (2140) Long Term Debt /Total Assets 17.00% 26.93% 7.23% *** [194] (6045) 20.61% 28.06% 3.80% *** [192] (4158.5) -6.90% ** [181] (-1546.5) 1.83% [180] (461) 19