Question 1 Table 1 Firm characteristics upto three-years before business combination (BC) laws’ implementation Thistable reports summary statistics for firm characteristics on three years beforea new BC law is adopted. Based on theadoption of the BC law or not, two samples are considered. Column 1 representsthe summary statistics of the sample of firms that incorporated the BC law inthe following year. The 2nd column represents the statistics on thesame period but for the firms that were not part of the states in which the lawswere implemented. The mean and standard deviation (in brackets) of thevariables are reported individually in each column.
The 3rd columnrepresents the p-values of t-tests performed between the two sample means. InTable 1 the average characteristicsof the firms are presented, based on data three-years before each BC law isimplemented. The mean-coefficient and standard deviation values when comparingColumn 1 and Column 2 are close in value. However in Column 3, where thep-value of the mean differences is calculated, it can be seen that the firmshave different stock volatility, cash-flow volatility, assets anddebt-to-assets ratios.
The only similarities appear between firms’3-year-assets CAGR and a Return-on Assets (only below 10% level) when comparingthe two sample groups. Whencomparing our results with the ones of Todd A. Gormleya andDavid A. Matsa we observe similarities in the direction and sign of thevariables but also differences in significance: the value of Stock Volatilitymean is higher in Column 2 than in Column 1, however the mean difference thatthey found is not significant while ours is; the Cash-flow Volatility coefficienthas similar values between Columns with the standard deviation slightly higherin Column 2, again significance in our result while the original paper one isnon-significant; the 3rd row represents total assets. For assets,the mean coefficient is lower in Column 2. Once more, significance found in ourdata while the original paper contradicts it; the Return on Assets variable isin-line with the original paper results, Column 2 has a lower mean value and ahigher standard deviation and no significant difference between the two groups;Debt to Assets mean coefficients and standard deviations have similar valuesbetween the first two columns, however there is significance in our meandifference while the original paper proves otherwise; the last variable in our reportis the 3-year-assets CAGR which has a similar pattern to the original paper,higher mean and standard deviation in Column 2. It is also insignificant, thusfirms in both columns have similar average CAGR.
ToddA. Gormleya and David A. Matsa found in their paper similar characteristicsbetween samples for all the above variables at all significance levels. In ourtable we find similar direction and values to the original paper; however thesignificance of the variables is different. This difference can be caused bythe sample selection, number of observations or even robustness as in ourresults we did not cluster at an incorporation level. Question 2 Table 2Theeffect of BC law on Stock Volatility and Distress Risk This table reportscoefficients from firm-panel regressions of a firm’s stock volatility anddistress risk and how are these affected by the BC law, while controlling forthe firm fixed effects (FE), state-by-year fixed effects (FE) andindustry-by-year fixed (FE). The dependent variables used are volatility ofdaily stock returns which is represented in Column 1, a firm delisting codebased on liquidation, bankruptcy, or other performance-related reason shown inColumn 2, Column 3 represents volatility of yearly operating asset returns, volatilityof quarterly ratios of cash flow to assets is explained in Column 4, and logcash in Column 5. The sample contains observations on the time-period 1976 to2006.
Standard errors are reported in parentheses and are clustered at anincorporation level. Significance level is presented as follows: ??? denotessignificance at the 1% level. The effect of the BC law on stock volatility anddistress risk characteristics are shown in Table2. There are two dependent variables that are significantly affected by theBC law introduction at a 1% significance level: Stock Volatility and OperatingAsset Volatility.
Average stock volatility declines by approximately 4.6% forfirms that are affected by the BC law relative to firms that operate in thesame state and industry but are unaffected by the law. Operating AssetVolatility declines as well with approximately 1.
5% for firms affected by theBC law when compared with firms that are unaffected by the law. Thus, there isevidence that firms reduce their business risk as both stock volatility andoperating asset volatility decreased but no evidence that firms reduced theirfinancial risk or distress risk as the performance-related exit variable, cashand cash-flow volatility have insignificant values. When comparing our results with the original paperoutcomes, we notice several similarities and differences. When it comes tosimilarities, Todd A.
Gormleya and David A. Matsa proved thatStock Volatility is significant at 1% and Operating Asset Volatility issignificant at 5%. We find similar results but with Operating Asset Volatilitysignificant at 1%. Another similarity is the insignificance of Cash flowvolatility, which provides less evidence of a decline in financial risk.
Differencesarise from the Performance-Related Exit variable and Cash variable as well asthe percentage of data explained by the regressions, given by the magnitude of. Even if in ourresults Cash is affected positive and Performance-Related Exit negative, whichis in-line with the original paper, we find no significance in these variableswhile ToddA. Gormleya and David A. Matsa found significance at 5% level for Cash and 1%level for Performance-Related Exit. Our results show higher values for in all columns; however these values could be affectedby the observations number used in our analysis compared to the original paper.
Whileone would expect Performance-Related Exit variable and Cash to suffersignificant changed upon the implementation of the laws as more managers arewilling to play it safe in order to reduce the firms’ risks and also protecttheir own benefits and maximize their utility, we cannot verify this statementwith our current outcome. Question 3Table 3Stock Volatility and Performance-Related Exit overtime This table report coefficients of thepanel-regression of stock volatility and performance-related exit on 14 dummyvariables that take value 0 or 1 based on when the BC law passed in a specificstate. For example, if state s passedthe BC law in 1987, for a company functioning in that specific state, thevariable Beta-2 takes value 1 in year 1985 and 0 in all otheryears. The first rowdummy in the table accounts for if a firm passed the law 5 years after thatyear while the last row accounts for if a firm passed the law 8 years beforethat year. Fig.1.
Table 3 beta coefficientsof the 14 dummies for Stock volatility and Performance-Related Exit plottedagainst time. The x-axis 0 value indicates the BC-law implementation event, 8indicates 8 periods after the event while -5 indicates 5 periods before theevent. The figures show the relative changes of Stock Volatility andPerformance-Related Exit through time.
Thecoefficients presented in Table 3 arethe point estimates from the regression described in Table 3, in which we allowed the effect of BC laws to fluctuateannually around the event time on a 14 years period window. Stockvolatility side of Fig.1 plots thetiming of the Stock Volatility change within the BC law adoption. As ourfindings suggest in Table 2, when a BClaw is adopted, firms that are affected by the law suffer a decrease in StockVolatility when compared with firms that operate in the same state and industrybut are not affected by the laws. The above figure validates this statement,however, from the figure it can be observed that the law does not have animmediate effect, but, the steep decrease in stock volatility appears 1 periodafter the law is implemented in the state. This suggests that firms require 1year period to adapt to the law and also that the financial risks cannot be decreasedon short notice, visible from the graph a further steady decrease in the stockvolatility up to 8 years after the law is implemented can be observed.
The right side of Fig.1 plots the timing of the Performance-Related Exit changewithin the BC law adoption. Our previous finding shows that the BC lawintroduction has no significant effect on the Performance-Related Exit of thefirms that adopted the law. On the other hand, the Performance-Related Exitfigure plots an increase in Performance-Related Exit that coincides with the BClaw’s adoption followed by a low fluctuation of the variable over the followingyears. The findings are not in line with the paper of ToddA. Gormleya and David A. Matsa as their results show a decrease in thePerformance-Related Exit variable as an outcome of the BC law for the firmsaffected by it. Question 4Table 4The effect of BC law onfirm Leverage levels This table reportscoefficients from firm-panel regressions of a firm’s different levels ofleverage and how are these affected by the BC law, while controlling for thefirm fixed effects (FE), state-by-year fixed effects (FE) and industry-by-yearfixed (FE).
The dependent variables used are constructed with Compustat yearlydata and are the following: book leverage (Column 1) computed as (dltt+ dlc)/at, market leverage (Column 2) computed as (dlc+ dltt)/ (at-seq+ (prcc_f* csho)),short-term leverage computed as dlc/ at (Column 3), and net leverage computedas ( dltt+ dlc-che)/ at (Column 4). The sample contains observations on thetime-period 1976 to 2006. Standard errors are reported in parentheses and areclustered at an incorporation level. Significance level is presented asfollows: *** denotes significance at the 1% level; ** denotes significance atthe 5% level.
Theresults presented in Table 4quantify the effects of BC laws on various measures of firms leverage. From theresults presented in Column 3 and Column 4 it can be observed that theShort-term Leverage and the Net Leverage of the firms that are being part ofthe states in which the law was implemented are not significantly affected. Onthe other hand, the Market Leverage (Column 2) of the firms affected by the BClaw decreased by 0.65% when compared with firms that are unaffected by the lawand is significant at 5% level.
Similarly, the Book Leverage (Column 1) of thesame firms increased by 1% when compared with firms from the same state andindustry but are unaffected by the law, significant at 1% level. Thus, we getan ambiguous effect when it comes to the change of leverage levels for thefirms affected by the laws As we have two leverage measuresthat suffer significant changes but with changes that follow oppositedirections, questions arise on how firms leverage is actually affected by theintroduction of the BC laws. When analyzing how the variables are computed,Market Leverage contains the difference between book value of equity and marketvalue of equity while the Book Leverage does not. The idea of managers playingit safe researched in the paper of Todd A.
Gormleya and David A. Matsa is howeverunclear in the above results. It seems like the total debt has indeed increased(by looking at Book Leverage) but also it may be the case of a simultaneous decreasein the market value of equity.
Nonetheless, a decrease in leverage may be asign of managers playing it safe. As managers benefit most directly fromreducing distress risk and high debt levels are a sign of distress, managersmay decrease leverage in order to reduce the distress risk and play it safe.Thus, our results show that managers are not playing it safe, as the Book Debtis actually increasing. In line with Todd A.
Gormleya andDavid A. Matsa hypothesis, we find that managers do not play it safe when itcomes to leverage, they are not reducing their personal exposure to distressrisk and it appears that the hostile takeover threat reduction effect of the BClaws have not increased managers’ incentives to take risk-reduction actions. Question 5a) Table 5Theeffect of BC law on Stock Volatility and Distress Risk on a shorter even periodThis table issimilar to Table 2.
However there isone difference in the sample period. The sample contains observations on thetime-period 1976 to 1995. Standard errors are reported in parentheses and areclustered at an incorporation level. Significance level is presented asfollows: ???denotes significance at the 1% level; ** denotes significance at the 5% level;* denotes significance at the 10% level Table 5 isa replica of Table 2 done on ashorter period of time, from 1976 to 1995.
By comparing the results from Table 2 with the above results,similarities in terms of significance can be observed: Stock Volatility issignificantly affected by the BC law at 1% level and Operating Asset Volatilityat 5% level. Average stockvolatility declines by approximately 3.9% for firms that are affected by the BClaw compared to the firms that activate in the same state and industry but areunaffected by the law. Operating Asset Volatility declines as well withapproximately 1.6% for firms affected by the BC law when compared with firmsthat are unaffected by the law. One important difference is the significance ofthe Performance-Related Exit dummy variable.
The variable is significant at 10%level on a shorter event period. We observe that firms are 0.44 percentagepoints less likely to undergo a performance-related exit after a BC law is implemented.
Thus, on a shorter time period, there is enough evidence to conclude that bothdistress risk and business risk declined to some extent after the introductionof BC laws.b) Table 6Theeffect of BC law on Stock Volatility and Distress Risk on non-finance firmsThis table issimilar to Table 2. Neverthelessthere is one difference in sample selection. Table 2 contains observations from all firm industries while thistable excludes observations linked to Finance firms.
The sample containsobservations on the time-period 1976 to 2006. Standard errors are reported inparentheses and are clustered at an incorporation level. Significance level ispresented as follows: ???denotes significance at the 1% level. In Table 6 we have represented the same dependentvariables as in Table 2 and Table 5. The above results are obtainedwith a similar method adopted in Table 2,on the full time-period observations but with financial firms excluded from thesample. A firm is considered financial if its “SIC” code has a value between6000 and 6799.
The same variables are significantly affected by the BC lawadoption: Stock Volatility and Operating Asset Volatility. Once again we havethe decline in average Stock Volatility of 4.5% for firms that have beenaffected by the BC law when compared with firms that were not affected by thelaw. Furthermore, a decrease in Operating Asset Volatility by 1.6% for firmsthat adopted the law when compared with the ones that did not undertake the lawcan be noticed.
Removing financial firms from thesample did not help with obtaining more precise/significant results. After thisrobustness check, we end up with the same significance values and magnitudelevels on the chosen dependent variables when compared with the resultsobtained on the whole sample analysis.