Table 3 Sample firms across levels of financial reporting quality
Related Figures (15)
The initial sample consists of 710 firms listed on the New York Stock Exchange in October 1929. Returns are adjusted for dividends and stock splits. Characteristics of firms deleted in the sample selection process Fig. 1. Cumulative value-weighted market return from January 1, 1926, until December 31, 1935, based on all firms included on the CRSP database. The darker line in the graph represents the drop in cumulative market return during October 1929. October 1929 stock returns compared to returns in surrounding three-year period The sample consists of 540 firms listed on the New York Stock Exchange in October 1929, excluding railroads and financial services firms (2-digit SIC codes 40 and 60-67). Returns are based on all monthly data available on the CRSP database. The buy-and-hold return, Rgy, is the gross return from the earliest to the latest month in the holding period (e.g., October 1927 to September 1929); the compounded monthly return is (1 + Rpy)” — 1, where k is the number of months over which Rpy; is calculated (e.g., 24 in the case of a return over October 1927-September 1929). All returns are adjusted for dividends and stock splits. Heke. 7 7 a _ - The sample consists of 540 firms listed on the New York Stock Exchange in October 1929, excluding railroads and financial services firms (2-digit SIC codes 40 and 60-67). Financial reporting variables are for the most recent fiscal year ending no later than June 1929. The variables are defined as ISTRANSP=Number of following items disclosed: sales, cost of sales, depreciation expense, other operating expenses, income tax expense. BSTRANSP=Number of following items disclosed: property, plant and equipment (net of accumulated depreciation); intangible assets; depreciation reserve; reserves other than for depreciation; earned surplus. AUDITOR =? if the firm’s financial statements were audited by one of the nine largest auditors at the time, based on Merino et al. (1994); 1 if they were audited by a small auditor (i.e., all other auditors); 0 if they were not audited. CONSERV = 1 if the firm reported on the balance sheet intangibles assets valued at nominal amounts (e.g., $1), 0 otherwise. Principal factor analysis of financial reporting quality variables Pearson correlation The sample consists of 540 firms listed on the New York Stock Exchange in October 1929, excluding railroads and financial services firms (2-digit SIC codes 40 and 60-67). Financial reporting variables are for the most recent fiscal year ending no later than June 1929. The variables are defined in Table 3. “and * denote significant at the 0.01 and 0.10 levels based on two-tailed tests. 5.2. Financial reporting choice model and definition of independent variables Descriptive statistics of independent variables in Eq. (1) and Pearson correlations with financial reporting quality factor The sample consists of 540 firms listed on the New York Stock Exchange in October 1929, excluding railroads and financial services firms (2-digit SIC codes 40 and 60-67). Financial reporting variables are for the most recent fiscal year ending no later than June 1929. The variables are defined as QUALITY =Score for principal component factor in Table 4, Panel C. AGE=Number of months before October 1929 since the firm was first listed on the CRSP database. TECH = 1 if the firm is in one of the following industries: aircraft (3-digit SIC code 372), automotive (371), communications (481, 482, 489). electronics (363, 366, 369), film and entertainment (781, 783, 791), industrial machinery (351-356), office equipment (357), photography (381, 383, 384, 387), or electric utilities (491, 493); 0 otherwise. CVEARN =Coefficient of variation in net income, using all available annual data over the five-yea1 period ending no later than June 1929. BETA=Beta from firm-specific value-weighted market model. based on all available monthly returns on CRSP over December 1925—September 1929. ROE=Net income divided by shareholders’ equity. ISSUE=1 if the firm issued shares within two years before October 1929, 0 otherwise. LEVERAGE=Debt plus preferred stock, divided by common equity. INCCONF = | if the firm has at least one of the following: income bonds, noncumulative preferred stock or participating preferred stock; 0 otherwise. CONTCONF = 1 if the firm has at least one of the following: a voting trust, multiple classes of common shares, or preferred stock with voting rights; 0 otherwise. DELAWARE = 1 if the firm was incorporated in Delaware. MK TSHR = Total assets as a percentage of the sum of total assets of all sample firms in the firm’s 2-digit SIC code. SIZE = Natural logarithm of total assets. DIVIDEND=1 if the firm paid ordinary dividends before October 1929, 0 otherwise. REGULATED = 1 if the firm is in one of the following industries: communications (3-digit SIC codes 481, 482, 489), oil and gas (460), nonrail transportation (419, 422, 440, 450, 474), railroad equipment and services (471), electric utilities (491, 493), or gas utilities (492, 499); 0 otherwise. “and *” denote significant at the 0.01 and 0.05 levels based on one-tailed tests. ee a a ee ee a ae ee ee a a en, en a ea The sample consists of 540 firms listed on the New York Stock Exchange in October 1929, excluding railroads and financial services firms (2-digit SIC codes 40 and 60-67). Financial reporting variables are for the most recent fiscal year ending no later than June 1929. All variables are defined in Table 5. sek dee . and * denote significant at the 0.01, 0.05 and 0.10 levels, respectively, based on two-tailed tests. Pearson correlations between independent variables in Eq. (1) Table 6 Regression results for Eq. (1) with QUALITY as the dependent variable The sample consists of 540 firms listed on the New York Stock Exchange in October 1929, excluding railroads and financial services firms (2-digit SIC codes 40 and 60-67). Financial reporting variables are for the most recent fiscal year ending no later than June 1929. The regression results are based on ordinary least squares estimation. For continuous independent variables, the standardized effect is the predicted change (in standard deviations) in QUALITY based on a one-standard-deviation increase in the independent variable; for indicator variables, it is based on a change from 0 to 1. All variables are defined in Table 5. et ea The sample consists of 540 firms listed on the New York Stock Exchange in October 1929, excluding railroads and financial services firms (2-digit SIC codes 40 and 60-67). Financial reporting variables are for the most recent fiscal year ending no later than June 1929. Regression intercepts are omitted. The results for the models with ISTRANSP, BSTRANSP and AUDITOR as dependent variables are based on ordered logit estimation, those for the model with CONSERV are based on logit estimation. In this latter model, REGULATED drops out from the regression because it predicts lack of conservatism perfectly; we also exclude from the estimation sample the 30 firms that are regulated. All variables are defined in Tables 3 and 5. and * denote significant at the 0.01, 0.05 and 0.10 levels, respectively, based on one-tailed tests for signed predictions and two-tailed tests otherwise. Test statistics are based on heteroskedasticity-consistent standard errors. Ta tan clanifnant at tha N1N lawal avr hattar hacad an tun tailed tact but nf annncite olan tr that nrvedintad Regression results for Eq. (1) with components of QUALITY as the dependent variable The sample consists of 540 firms listed on the New York Stock Exchange in October 1929, excluding railroads and financial services firms (2-digit SIC codes 40 and 60-67). December 1925 is the first month in which data are available on the CRSP database. The buy-and-hold return is the gross return from the earliest month in which the firm entered the CRSP database until September 1929. All returns are adjusted for dividends and stock splits. The financial reporting quality variables are defined in Tables 3 and 5. “and “™* denote significant at the 0.01 and 0.05 levels based on two-tailed tests. The null hypothesis is that there is no trend in the means across the five periods. Mean stock returns and financial reporting quality measures by year that firms entered the CRSP database The sample consists of 540 firms listed on the New York Stock Exchange in October 1929, excluding railroads and financial services firms (2-digit SIC codes 40 and 60-67). Financial reporting variables are for the most recent fiscal year ending no later than June 1929. The variables are defined in Table 5, except: RETURN=October 1929 stock return, adjusted for stock splits and dividends. PB= Market capitalization of common equity at the end of September 1929, divided by shareholders’ equity. SHRTO = Average share turnover (i.e., number of shares traded divided by the weighted average number of shares outstanding) over the six months before October 1929. PRCMTM = Autocorrelation of monthly stock returns, based on all available monthly returns on the CRSP database over January 1927—September 1929. GROWTH = Earnings growth over the two most recent consecutive fiscal years ending no later than June 1929. Descriptive statistics of independent variables in Eq. (2) and Pearson correlations with October 1929 stock returns Table 10 Pearson correlations between independent variables in Eq. (2) The sample consists of 540 firms listed on the New York Stock Exchange in October 1929, excluding railroads and financial services firms (2-digit SIC codes 40 and 60-67). Financial reporting variables are for the most recent fiscal year ending no later than June 1929. All variables are defined in Tables 5 and 10. ae kak . and “ denote significant at the 0.01, 0.05 and 0.10 levels, respectively, based on two-tailed tests. Table 11 The sample consists of 540 firms listed on the New York Stock Exchange in October 1929, excluding railroads and financial services firms (2-digit SIC codes 4 and 60-67). Financial reporting variables are for the most recent fiscal year ending no later than June 1929. The GMM and 2SLS regression results are base: on the independent variables in Table 7 serving as instruments for QUALITY. For continuous independent variables, the standardized effect is the predicte change in the October 1929 stock return based on a one-standard-deviation increase in the independent variable; for indicator variables, it is based on a chang from 0 to 1. All variables are defined in Tables 5 and 10. aetok ok . and “* denote significant at the 0.01, 0.05 and 0.10 levels, respectively, based on one-tailed tests for signed predictions and two-tailed tests otherwise. Tes statistics are based on heteroskedasticity-consistent standard errors. The sample consists of 540 firms listed on the New York Stock Exchange in October 1929, excluding railroads and financial services firms (2-digit SIC codes 40 and 60-67). Financial reporting variables are for the most recent fiscal year ending no later than June 1929. The regression results are based on GMM estimation using the independent variables in Table 7 as instruments for QUALITY. All variables are defined in Tables 5 and 10. “and * denote significant at the 0.01, 0.05 and 0.10 levels, respectively, based on one-tailed tests for signed predictions and two-tailed tests otherwise. Test statistics are based on heteroskedasticity-consistent standard errors. The predicted signs for BETA follow the signs of the average monthly returns for August—December 1929 reported in Table 2. Lee a a ae a ae a. ee ae ee a ee ee a” ee ee ee ae ee i a a ee ee a ee ee ae ee Regression results for Eq. (2) with stock returns in months adjacent to October 1929 as the dependent variable
