Abstract
We find that analysts are more likely to provide cash flow forecasts in countries with weak investor protection. This finding is consistent with our hypothesis that market participants demand (and analysts supply) cash flow information when weak investor protection results in earnings that are less likely to reflect underlying economic performance. Our results suggest that information intermediaries respond to market-based incentives to attenuate the adverse effects of country-level institutional factors on earnings’ usefulness. These findings contribute to the literature by shedding light on the institutional determinants of analysts’ research activities, and on the nature of the financial information they generate.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Specifically, our control variables consist of a country-level disclosure index, country-level foreign investment, whether the firm is audited by a Big Five auditor, the number of analysts following, firm size, whether the firm is cross-listed on U.S. stock exchanges, and industry and year dummies. In addition, we perform a test that includes the inverse Mills ratio to control for potential self-selection bias (Heckman, 1979), because our sample firms are followed by analysts and thus are not randomly chosen.
Following DeFond and Hung (2004) (and many others), we capture investor protection by the extent of the laws that protect investors’ rights and the strength of the legal institutions that facilitate law enforcement. While our investor protection variables are correlated with legal origin, we do not use legal origin in our primary analyses because Ball et al. (2003) document that some “common law origin” countries have civil law-like institutions. Nonetheless, in robustness tests we repeat our hypothesis test after replacing our investor protection variables with legal origin variables and find results that are qualitatively the same as those that obtain using our investor protection variables.
We note that cash flows can still be influenced by management’s discretion (Mulford & Comiskey, 2002). For example, managers can manipulate cash flows by timing payable and receivable decisions. However, because manipulating cash flows involves altering real business activities (such as deferring or accelerating expenditures), we expect them to be more costly than manipulating accounting accruals.
This inference is consistent with a recent Special Report in The Economist that comments on problems of comparing accounting information across countries (The Economist, 2002):
“Standard-setters admit that no country has adequate rules on the recognition of revenues. A solution in the meantime may be to look at cash, which is far harder to disguise or invent. Comroad duped its auditor about its revenues, but it could not conceal the fact that its cash flow was negative.”
This prediction is consistent with the views of a sell-side analyst specializing in Latin American companies. In our conversations, this analyst indicated that there is widespread lack of trust in the reported earnings of Latin American companies and hence that there is high demand for additional analyses, including analyses pertaining to cash flow information.
Since there are arguments both for and against analysts’ propensity to forecast cash flows in countries with poorer investor protection institutions, an alternative approach would be to make our hypothesis two-sided. However, we believe that the most compelling arguments favor an increased propensity to forecast cash flows in countries with weak investor protection environments, and hence we make our hypothesis one-sided. We note, however, that our analyses use two-sided p-values throughout the paper.
This is consistent with our conversation with an analyst at Bunting Warburg. Specifically, the analyst indicates that it is more difficult to compute cash flow forecasts for companies in countries with inadequate disclosure. Since forecasting cash flows is a fairly costly and difficult process that involves predicting items such as working capital and deferred taxes, increased disclosure of financial statement data is likely to reduce the cost of compiling these forecasts.
For example, Patrick O’Donnell, chief of global equity research at Putman Investments, states that the most difficult task in cross-border investments is to achieve “true comparability” between, for example, U.S. and Argentinean oil companies. Although Putman prides itself on having analysts who understand different accounting methods, O’Donnell notes “there will always be quirks and twists.” (Meisler, 1997).
For example, the global telecommunication team at Morgan Stanley Dean Witter states: “Wireless companies are most commonly valued on a discounted cash flow basis... Due to the different accounting treatment for goodwill,... amortization expense among operators can vary significantly. For this reason, it is difficult to compare wireless operators on an operating income basis” (Morgan Stanley Dean Witter, 1999).
We note that the information asymmetry problem varies with the source of foreign investment. For example, if the foreign investment is mostly from countries with similar accounting standards and institutional background, the information asymmetry among foreign investors should not be severe. Since we are unable to find data on the sources of foreign investment across countries, we acknowledge that the foreign investment variable is measured with error. However, we do not expect the noise in this variable to bias in favor of supporting our hypothesis.
Another measure of foreign equity investment is foreign equity portfolio investment. We do not use foreign equity portfolio investment because these data are available for only 15 of our 36 sample countries. In addition, World Bank documents suggest that data on foreign equity portfolio investment often suffer from measurement error and inconsistency because periodic reporting in many developing economies lacks clarity, adequate disaggregation, and comprehensiveness (World Bank, 2001).
We acknowledge, however, that while auditor size is traditionally used in cross-country studies to control for audit quality, it is possible that Big Five auditors do not necessarily represent the set of high quality auditors in a particular country. For example, in a study of German companies, Ashbaugh and Warfield (2003) find that the two largest audit firms in Germany have the greatest local market expertise.
We note that the efficiency of the judicial system component of the law enforcement institutions variable is measured much earlier than our investigation period. We therefore rerun our analysis using the rule of law component of this variable (because it is measured over a period ending in 1995) as a proxy for law enforcement institutions (as in La Porta, Lopez-de-Silanes, Shleifer, & Vishny, 1997). The results show that the signs and significance levels of the coefficients on our hypothesized variables are consistent with the results currently reported in our primary analysis in Models 2 and 3 of Table 4.
For Hong Kong, this measure is averaged over the 1998 to 2001 period because the World Bank World Development Indicators do not disclose pre-1998 foreign investment for Hong Kong. We repeat our analyses excluding Hong Kong and find that the signs and significance levels of the all hypothesized variables are consistent with the results reported in our primary analysis (Table 4).
We also perform sensitivity tests that consider alternative methods of controlling for dependence among our error terms as reported in Sect. 4.7.
While thousands of firm-year observations in the I/B/E/S database have earnings forecasts but no cash flow forecasts, only 18 firm-year observations have cash flow forecasts but no earnings forecasts. Thus, we restrict our analysis to firm-year observations that have earnings forecasts.
We assume that analysts provide cash flow forecasts to I/B/E/S only when investors demand cash flow information; that is, we do not assume that all analysts who use cash flow forecasts for internal analysis necessarily submit these forecasts to I/B/E/S for use by equity investors. This is consistent with our correspondence with an analyst at Prudential Securities who states that “cash flow forecasting is important and we do it for all of the companies we cover. Our published estimates though are earnings, as that is what investors look at.”
While the low percentage of cash flow forecasts for U.S. firms is consistent with DeFond and Hung (2003) and Wasley and Wu (2006), we perform additional analysis aimed at assessing whether the I/B/E/S database underreports the frequency of cash flow forecasts for U.S. firms. Specifically, we randomly select a sample of 100 firm-analyst pairs that do not have cash flow forecasts and attempt to trace them back to the analysts’ report as reported in the Investext database. Of the 100 firm-analyst pairs, we are able to identify 72 analyst reports, none of which includes operating cash flow forecasts on the cover or summary page of the report. While six of the 72 reports include operating cash flow forecasts in the body of the report, they are typically included toward the end of the report, consistent with the forecasts for these six companies being peripheral to the analysts’ analysis. Thus, we conclude that the I/B/E/S database appears to include all cash flow forecasts that are likely to be demanded by investors.
While Compustat’s Global Industrial/Commercial database excludes financial service firms such as banks and insurance companies, we note that a significant portion of our sample firms still belong to the finance sector according to the I/B/E/S classification. Further investigation indicates that the majority of these companies are related to insurance/real estate agents and brokers.
While not reported in Table 1, only 4% of the I/B/E/S firms with earnings forecasts also had cash flow forecasts during 1993, the first year for which I/B/E/S began reporting cash flow forecasts. Because the sample size is small in 1993, we begin our analysis with 1994 data.
To avoid the influence of outliers while conserving sample size, we winsorize these variables at the top and bottom 1% of their distributions. Our results are qualitatively the same when we truncate observations in the top and bottom 1%. We note that the sample size is smaller than our full sample as reported in Table 1 because of the additional data requirements.
The Heckman first-stage probit estimation used to compute Lambda has a pseudo R2 of 27% with the following coefficients, where two-tailed p-values are in parentheses and coefficients on industry dummies are not reported:
$$ \begin{aligned}{} Select{\text{ }} = & {\text{ }}4.10{\text{ }} + 0.01{\text{ }}French{\text{ }}Origin - 0.43{\text{ }}German{\text{ }}Origin - 0.24{\text{ }}Scandinavian{\text{ }}Origin{\text{ }} \\ & (1.00){\text{ }}\quad \quad \quad \quad \;(0.69){\text{ }}\quad \quad \quad \quad (<0.01)\quad \quad \quad \quad \quad\quad(<0.01){\text{ }} \\ & - 0.02\ Foreign{\text{ }}Investment{\text{ }} + 0.01\ Disclosure{\text{ }} + 0.00{\text{ }}Per{\text{ }}Capita{\text{ }}GDP{\text{ }} + 0.34{\text{ }}Firm{\text{ }}Size{\text{ }} \\ & (<0.01)\quad \quad \quad \quad \quad \quad(< 0.01) \quad \quad \quad (< 0.01)\quad \quad \quad \quad \quad (< 0.01) \\ & - 0.01{\text{ }}Capital{\text{ }}Intensity + {\text{ }}0.04{\text{ }}Sales{\text{ }}Growth{\text{ }} - 0.03{\text{ }}Market {\text {-}}to{\text {-}}Book + {\text{ }}0.19{\text{ }}Leverage{\text{ }} + 0.01{\text{ }}Loss.{\text{ }} \\ & (0.37)\quad \quad \quad \quad \quad \quad (<0.01)\quad \quad \quad \quad (<0.01)\quad \quad \quad \quad \quad (<0.01)\quad \quad \,(0.48) \\ \end{aligned} $$Thus, I/B/E/S analysts tend to cover firms in countries with English legal origin, less foreign investment, greater disclosure, and larger per capita GDP, and further, they tend to cover firms that are larger and that have higher sales growth, smaller market-to-book, and higher leverage. In addition, we note that the number of observation in Model 3 is smaller because of the additional data requirements for the first-stage probit regression.
We also note that the coefficient on firm size is significantly negative in Model 3 but significantly positive in Model 2. This is probably because firm size is significantly correlated with the inverse Mills ratio (in untabulated tests, the Pearson correlation coefficient between these variables is −0.84). That is, since larger firms are more likely to be covered by analysts, the effect of firm size on analysts’ propensity to issue cash flow forecasts becomes negative after correcting for the self-selection bias.
Because Table 2 finds that several correlations among our independent variables are reasonably large, we follow Allison (1999) to assess whether multi-collinearity impacts the coefficients in Models 2 and 3 of Table 4. Specifically, we rerun these models using an OLS regression after adjusting the linear combinations of our independent variables with the weight matrix used in the maximum likelihood algorithm. We then use OLS regression diagnostics to detect potential multi-collinearity (Belsley, Kuh, & Welsch, 1980). The results (not tabulated) indicate that the variance inflation statistics (the degree to which the standard error of a coefficient is increased because of the degree to which the independent variable is correlated with the other predictors) for our hypothesized variables are below the commonly used cutoff of 4.0.
We also rerun our analysis in Models 2 and 3 of Table 4 after replacing our investor protection variables with the proxies of earnings properties (the earnings management and earnings’ value relevance scores reported in Table 3). This analysis (not tabulated) shows that the coefficients on earnings management scores (earnings’ value relevance scores) are significantly positive (negative) at p < 1% (two-tailed). Thus, the analysis provides further evidence consistent with our assumption that the investor protection variables are reasonable surrogates for the ability of earnings to capture underlying economic performance.
We winsorize each variable at the top and bottom 1% of its distribution.
These control variables are not included in DH because, while our study includes 36 countries, DH focuses only on the U.S. and does not have these four variables in its analysis.
We use two-month CARs instead of two-day CARs as in DeFond and Hung (2003) because Compustat Global only provides monthly data on stock prices and prices are likely to fully incorporate information before public release in some international markets.
While we would like to restrict our analysis to the same individual analysts, we find that international analysts are frequently identified only by their industry name in I/B/E/S. In addition, it is unlikely one would find many analysts covering a large number of our sample countries, since analysts typically specialize in regions or cover a small number of countries.
The proportion of forecasts from global research firms by country has a mean of 18%, a median of 17%, and a standard deviation of 9%. The country with the highest proportion is Australia (36%) and that with the lowest proportion is Turkey (4%).
Throughout the paper we define “consistent with the results reported in Models 2 and 3 of Table 4” to mean that the coefficients on our investor protection variables remain significantly negative at p < 10% (two-tailed).
While not tabulated, we also repeat this analysis restricting the sample to the top three research teams (each covers an average of 35 of the 36 countries in our sample) and find results that are the same as those reported in Table 4. Thus, our results do not appear to be driven by differences in analyst research teams.
The negative coefficient on the dummy variable indicating mandatory reporting is consistent with two related explanations. First, because mandatory cash (and funds) flow statement reporting is associated with strong investor protection laws and enforcement, the dummy may be picking up some factors in the legal environment not captured by our investor protection variables. Second, analysts may be more motivated to supply cash flow information when it is not mandated because earnings quality tends to be poorer in such environments and cash flows are useful in helping market participants interpret poor quality earnings.
Bushman and Smith (2001) suggest that good communications infrastructure results in financial information being widely, quickly, and cheaply disseminated to economic agents through distribution channels such as the financial press, radio, television, and the Internet.
We also expect the correlation among residuals to be better captured by country-industry clusters than by country clusters alone, because analysts are often country-industry experts, suggesting a higher correlation across country-industries than across countries. Consistent with this expectation, we find that the intra-cluster correlation among residuals is 0.08 among country-industry clusters versus 0.05 among country clusters, suggesting country-industry clusters better capture the correlation among the residuals in our analysis.
We do not rerun Model 3 of Table 4 because this test is an alternative way to control for potential self-selection bias.
References
Ahearne, A., Griever, W., & Warnock, F. (2004). Information cost and home bias: An analysis of US holdings of foreign equities. Journal of International Economics, 62, 313–336.
Ali, A., & Hwang, L. (2000). Country-specific factors related to financial reporting and the value relevance of accounting data. Journal of Accounting Research, 38, 1–22.
Allison, P. (1999). Logistic regression using the sas system, theory and application. Cary, NC: SAS Institute Inc.
Altman, E.I. (1968). Financial ratios, discriminant analysis, and the prediction of corporate bankruptcy. The Journal of Finance, 23, 589–609.
Ashbaugh, H., & Warfield, T. (2003). Audits as a corporate governance mechanism: Evidence from the German Market. Journal of International Accounting Research, 2, 1–21.
Ball, R., Kothari, S.P., & Robin, A. (2000). The effect of international institutional factors on properties of accounting earnings. Journal of Accounting and Economics, 29, 1–51.
Ball, R., Robin, A., & Wu, J.S. (2003). Incentives versus standards: Properties of accounting income in four East Asian countries. Journal of Accounting and Economics, 36, 235–270.
Baptista, R. (1999). The diffusion of process innovations: A selective review. International Journal of the Economics of Business, 6, 107–129.
Basu, S., Hwang, L., & Jan, C. (1998). International variation in accounting measurement rules and analysts’ earnings forecast errors. Journal of Business Finance and Accounting, 26, 1207–1247.
Belsley, D.A., Kuh, E., & Welsch, R.E. (1980). Regression diagnostics. New York, NY: Wiley.
Bhattacharya, U., & Daouk, H. (2002). The world price of insider trading. The Journal of Finance, 57, 75–108.
Bowen, R., Burgstahler, D., & Daley, L. (1987). The incremental information content of accruals versus cash flows. The Accounting Review, 62, 723–747.
Bushman, R.M., Piotroski, J.D., & Smith, A.J. (2004). What determines corporate transparency? Journal of Accounting Research, 42, 207–252.
Bushman, R.M., & Smith, A.J. (2001). Financial accounting information and corporate Governance. Journal of Accounting and Economics, 32, 237–333.
Bhushan, R. (1989). Firm characteristics and analyst following. Journal of Accounting and Economics, 11, 255–274.
Center for International Financial Analysis & Research. (1995). International accounting and auditing trends. Princeton, NJ: CIFAR Publications, Inc.
Chang, J.J., Khanna, T., & Palepu, K.G. (2000). Analyst activity around the world. Working paper, Harvard University.
Coopers & Lybrand (1993). International accounting summaries: A guide for interpretation and comparison (2nd ed.). New York, NY: Wiley.
DeFond, M.L., & Hung, M. (2003). An empirical analysis of analysts’ cash flow forecasts. Journal of Accounting and Economics, 35, 73–100.
DeFond, M.L., & Hung, M. (2004). Investor protection and corporate governance, Evidence from worldwide CEO turnover. Journal of Accounting Research, 42, 269–312.
DeFond, M.L., & Jiambalvo, J. (1993). Factors related to auditor-client disagreements over income-increasing accounting methods. Contemporary Accounting Research, 9, 415–431.
Degeorge, F., Ding, Y., Jeanjean T., & Stolowy, H. (2005). Des analysts following curb earnings management? International evidence. Working paper. University of Lugano and HEC School of Management.
Dyck, A., & Zingales, L. (2004). Private benefits of control, An international comparison. The Journal of Finance, 59, 537–600.
Fama, E.F., & MacBeth, J.D. (1973). Risk, return, and equilibrium: Empirical tests. The Journal of Political Economy, 81, 607–636.
Garrett, T. (2003). Aggregated versus disaggregated data in regression analysis: Implications for inference. Economic Letters, 81, 61–65.
Greene, W.E. (1993). Econometric analysis (2nd edn). New York, NY: McMillan.
Harris, T., Lang, M., & Moller, H.P. (1994). The value relevance of German accounting measures: An empirical analysis. Journal of Accounting Research, 32, 187–209.
Healy, P.M., Hutton, A.P., & Palepu, K.G. (1999). Stock performance and intermediation changes surrounding sustained increases in disclosure. Contemporary Accounting Research, 16, 485–520.
Heckman, J. (1979). Sample selection bias as a specification error. Econometrica, 47, 153–161.
Hung, M. (2000). Accounting standards and value relevance of earnings: An international analysis. Journal of Accounting and Economics, 30, 401–420.
I/B/E/S. (2002). Accounting for Estimates, How I/B/E/S Handles Global Accounting Diversity. New York, NY: I/B/E/S International.
Karshenas, M., & Stoneman, P. (1993). Rank, stock, order, and epidemic effects on the diffusion of new process technologies: An empirical model. The RAND Journal of Economics, 24, 503–527.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (1997). Legal determinants of external finance. The Journal of Finance, 52, 1131–1150.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (1998). Law and finance. Journal of Political Economy, 106, 1113–1155.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny, R. (2000). Investor protection and corporate governance. Journal of Financial Economics, 58, 3–37.
Lang, M.H., Lins, K.V., & Miller, D.P. (2004). Concentrated control, analyst following and valuation: Do analysts matter most when investors are protected least? Journal of Accounting Research, 42, 589–623.
Lang, M.H., & Lundholm, R.J. (1996). Corporate disclosure policy and analyst behavior. The Accounting Review, 71, 467–492.
Leuz, C., Nanda, D., & Wysocki, P. (2003). Earnings management and investor protection: An international comparison. Journal of Financial Economics, 69, 505–527.
Levitt, A. (2002). Take on the Street, What Wall Street and Corporate America don’t Want You to Know. New York, NY: Pantheon Books.
Meisler, L. (1997). The Complex World of Global Research, Institutional Investor, December 1997.
Morgan Stanley Dean Witter (1999). Global Telecommunications Primer, A Guide to the Information Superhighway. Morgan Stanley Dean Witter, June 1999.
Mulford, C., & Comiskey, E. (2002). The financial numbers game. New York, NY: John Wiley & Sons, Inc.
North, D. (1990). Institutions, institutional change and economic performance. Cambridge: Cambridge University Press.
Palmrose, Z. (1988). An analysis of auditor litigation and audit service quality. The Accounting Review, 6, 55–73.
Penman, S. (2001). Financial statement analysis & security valuation. New York, NY: McGraw-Hill Irwin.
Petersen, M. (2004). Estimating standard errors in finance panel data sets: Comparing approaches. Working paper, Northwestern University and NBER.
Schipper, K. (1991). Commentary on analysts’ forecasts. Accounting Horizons, 5, 105–121.
Schumpeter, J. (1934). The theory of economic development. Cambridge, MA: Harvard University Press.
The Economist (2002). Badly in Need of Repair – Special Report, Company Accounts. The Economist, May 4, 2002.
Wasley, C., & Wu, J. (2006). Why do managers voluntarily issue cash flow forecasts? Journal of Accounting Research, 44, 389–429.
Wild, J., Bernstein, L., & Subramanyam, K.R. (2001). Financial statement analysis. New York, NY: McGraw-Hill Irwin.
Wooldridge, J. (2002). Econometric analysis of cross section and panel data. Cambridge, MA: MIT Press.
Wooldridge, J. (2003). Cluster-sample methods in applied econometrics. American Economic Review, 93, 133–138.
World Bank. (2001). World development indicator. Washington, DC: World Bank.
Young, D., & Guenther, D.A. (2003). Financial reporting environments and international capital mobility. Journal of Accounting Research, 41, 553–579.
Acknowledgments
We thank Maureen McNichols (the editor), Luzi Hail (the discussant), two anonymous referees, and workshop participants at University of Arizona, University of Colorado, Dartmouth College, Duke University, Massachusetts Institute of Technology, University of North Carolina, Northwestern University, University of Notre Dame, Ohio State University, University of Oregon, University of Southern California, Stanford University, the 2003 American Accounting Association annual meetings, and the 2006 Review of Accounting Studies conference for their helpful and constructive comments. We also gratefully acknowledge the contribution of I/B/E/S International Inc. for providing earnings and cash flow forecast data from their Institutional Brokers Estimate System, and in particular to Steven Sommers of I/B/E/S for his help. These data have been provided as part of a broad academic program to encourage earnings expectations research. In addition, the paper greatly benefited from the input we received during informal conversations and correspondence with Trevor Harris and several analysts, including Laurence Madsen of Warburg Dillon Read and Fadi Chamoon of Bunting Warburg. This project was completed while Mingyi Hung was visiting The Chinese University of Hong Kong. Previous versions of this paper were titled “International Institutional Factors and Analysts’ Cash Flow Forecasts.”
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
DeFond, M.L., Hung, M. Investor protection and analysts’ cash flow forecasts around the world. Rev Acc Stud 12, 377–419 (2007). https://doi.org/10.1007/s11142-007-9030-7
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11142-007-9030-7