nach oben

Review of Accounting Studies

Erschienen in:

01.03.2010

The pricing of conservative accounting and the measurement of conservatism at the firm-year level

verfasst von: Jeffrey L. Callen, Dan Segal, Ole-Kristian Hope

Erschienen in: Review of Accounting Studies | Ausgabe 1/2010

Einloggen

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config

KI-gestützte Suche

Aus

Abstract

This paper analyzes the relation between equity prices and conditional conservatism and introduces a new measure of conservatism at the firm-year level. We show that the asymmetric properties of conservative accounting, the existence of non-accounting sources of information, and the properties of GAAP related to special items combine to generate a nonlinear relation between unexpected equity returns and earnings news (the shock to expected current and future earnings). Based on this model, we construct a conservatism ratio (CR) defined as the ratio of the current earnings shock to earnings news. CR measures the proportion of the total shock to expected current and future earnings recognized in current year earnings. Ranking firms according to CR, we show empirically that higher CR firms have more leverage, increased volatility of returns, more incidence of losses, more negative accruals, and increased volatility of earnings and accruals, consistent with the literature on conservative accounting.

Vorheriger Artikel Regulation fair disclosure and the cost of equity capital

Nächster Artikel Disclosure of GAAP line items in earnings announcements

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

über 102.000 Bücher
über 537 Zeitschriften

aus folgenden Fachgebieten:

Automobil + Motoren
Bauwesen + Immobilien
Business IT + Informatik
Elektrotechnik + Elektronik
Energie + Nachhaltigkeit
Finance + Banking
Management + Führung
Marketing + Vertrieb
Maschinenbau + Werkstoffe
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

über 67.000 Bücher
über 340 Zeitschriften

aus folgenden Fachgebieten:

Bauwesen + Immobilien
Business IT + Informatik
Finance + Banking
Management + Führung
Marketing + Vertrieb
Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Jetzt informieren

Nur mit Berechtigung zugänglich

In other words, earnings news is the conventional current earnings surprise plus the surprise to future earnings (appropriately discounted).

Hence, CR is a meaningful measure of conservatism only if conditioned on the sign of the shock.

We use the terms “unexpected,” “revision to,” and “shock to” interchangeably in this paper.

Positive (negative) raw returns are neither necessary nor sufficient conditions for good (bad) news. Basu (1997), in a sensitivity analysis, subtracts total market returns from firm returns. However, this measure of unexpected returns is potentially misspecified because it fails to account for systematic risk.

The conventional view of the earnings-return paradigm is that the current level of earnings provides information about expected future cash flows and, this in turn (partially) determines the current level of security returns. We do not contest this. However, the standard Basu-type analysis of conservative accounting focuses correctly on the asymmetry between “good” news and “bad” news events on equity valuation. Good news and bad news refer to revisions or equivalently to shocks, not levels. Thus, it is more useful to analyze conservative accounting with a revisions approach rather than a levels approach. Specifically, the perspective of this study is that revisions to current earnings provide information about revisions to expected future cash flows which, in turn, (partially) determines revisions to equity returns.

Focusing on the conventional earnings surprise instead of earnings news, results in a potential correlated omitted variables problem.

Specifically, Basu (1997) uses a reverse regression of price-deflated earnings (EARN) on an indicator variable for negative stock returns (D), stock returns (R), and stock returns interacted with the indicator variable (subscripts omitted): \( {\text{EARN}} = {\text{a}}_{0} + {\text{a}}_{1} {\text{D}} + {\text{a}}_{2} {\text{R}} + {\text{a}}_{3} {\text{R}}\,{*}\,{\text{D}}.\) He finds the coefficient \( {\text{a}}_{ 3} \) to be significantly positive.

Some studies refer to differential timeliness as “earnings conservatism” or “conditional conservatism” as compared with “balance sheet conservatism” or “unconditional conservatism” (as reflected in the market-to-book ratio) (e.g., Beaver and Ryan 2005; Pae et al. 2005).

We perform tests that address these issues in Sect. 6.3.

In contrast, Riedl and Srinivasan (2007) do not find a significant difference in response coefficients across positive and negative special items.

In the empirical analysis below, we specifically control for discount rate news.

Scenarios 1 and 2 of the example in the Appendix and the related Fig. 1 illustrate this linearity result.

Scenarios 1 and 3 of the example in the Appendix and the related Fig. 2 illustrate this nonlinearity result.

Scenarios 1 and 3 of the example in the Appendix and the related Fig. 3 illustrate this nonlinearity result.

The return decomposition (Eq. 2) is derived from the definition of the market to book ratio and the clean surplus relation; hence, it is an identity. See Callen (2009) for a formal proof that emphasizes the tautological nature of the Vuolteenaho (2002) return decomposition model.

As a sensitivity analysis we estimate earnings news as the sum of accruals news and cash flow news (see Callen and Segal 2004). The idea behind this alternative estimation is that the breakdown of earnings into cash flows and accruals may provide a better prediction of future cash flows than earnings alone. No inferences are affected if we use this alternative approach. However, breaking down earnings to accruals and cash flows precludes us from computing our conservatism ratio so we do not pursue this approach.

The book-to-market ratio is included in the parsimonious VAR because the decomposition model is generated from this ratio. Vuolteenaho (2002) similarly includes the book-to-market ratio in his VAR specifications. It also helps to control for the firm’s growth prospects and the firm’s unconditional conservatism.

Industry subscripts are suppressed in the above equations.

Using OLS gives similar results.

Note that (I − ρA)⁻¹ is a present value operator.

Following Vuolteenaho (2002), Callen and Segal (2004), Callen et al. (2005), and Callen et al. (2006), we assume that ρ = 0.967. The results are not sensitive to this assumption.

Substituting the AR(1) dynamic of Eq. 12 into the definition of earnings news (Eq. 3) and noting that ΔE_t(roe_t+j) = (ρβ)^jε_t for j ≥ 0 yields Ne_t = ε_t/(1 − ρβ). By definition CR_t = ε_t/Ne_t = (1 − ρβ). Since ρ is very close to 1 (0.967), the result follows.

Note that if past returns have no impact on roe_t so that β₁ = 0 then roe_t is AR(1) as before and the conservatism ratio equals one minus the persistence of earnings (1 − β₂) as before.

As sensitivity tests, we repeat the analysis scaling special items by beginning of the period total assets and market value of equity. The results are similar to those reported.

Untabulated results show that if we restrict the sample to observations with non-zero SI, then the mean and median SI are −0.037 and −0.014, respectively.

Reported results are based on a parsimonious VAR with only one lag per variable. Untabulated results show that our inferences are robust to including two lags per variable in the VAR estimation.

The revision to returns is computed as the residual from the VAR return equation. See Eq. 6a.

We are not making a causality statement here but rather documenting an association.

As a sensitivity analysis, we examine the frequencies of positive and negative special items by earnings news portfolio quintiles. Going from the most negative earnings news (quintile 1) to the most positive earnings news (quintile 5), the proportion of firms with negative (positive) special items decreases (increases) monotonically with earnings news.

In Eq. 2, the revisions in returns are tautologically determined by Ne and Nr so that there are no parameters to estimate. However, we test Eq. 2 using ex post revisions in returns so that the relation has an error structure.

Note that α₂ > 0 because Nr is on the other side of the equation in a reverse regression.

Formally, Ne = CES + FNe = ROE – PROE + FNe.

We obtain similar results when estimating the regressions using the Fama–MacBeth (1973) methodology.

Although Nr_t and (r_t – E_t−1(r_t)) are highly correlated (see Table 2, Panel C), the maximum variance inflation factor of 2.18 and condition index of 2.56 are not indicative of serious multicollinearity.

We also test whether the coefficients across the Ne and SI regressions are equal by estimating both equations as a system. The coefficient on the revisions to returns in the SI regression is significantly smaller (at less than the 1% level) than the coefficient on the revisions to returns in the Ne regression (α₁) consistent with special items being a noisy measure of earnings news.

These variables obtain by decomposing Ne in Eq. 2 as per footnote 33 so that CES and ROE are the variables on the left-hand of the equation, respectively.

Equation 17 generalizes the Basu (1997) nonlinear relation. In particular, assuming inter-temporally constant discount rates so that Nr_t = 0 and assuming that shocks to expected future earnings (FNe) are identically 0 yields the Basu relation: \( {\text{ROE}}_{\text{t}} = \beta_{0} + \beta_{ 1} {\text{D}} + \beta_{ 1} \left( {{\text{r}}_{\text{t}} - {\text{E}}_{{{\text{t}} - 1}} \left( {{\text{r}}_{\text{t}} } \right)} \right) + \beta_{ 2} {\text{D}}\,{*}\,\left( {{\text{r}}_{\text{t}} - {\text{E}}_{{{\text{t}} - 1}} \left( {{\text{r}}_{\text{t}} } \right)} \right). \)

The variance inflation factors for all four regressions are less than 10, indicating that multicollinearity is not a concern.

The difference between the ratios is likely attributable to the inclusion of discount rate news, the shock to future earnings, and the predicted ROE in our regression, as required by the Vuolteenaho model.

To account for discount rate news, we subtract discount rate news from earnings news, effectively forcing the coefficient on discount rate news to take on its theoretically correct value of one. Therefore, everywhere that earnings news appears in this statement, one should read earnings news less discount rate news.

The switching regressions (Eqs. 18a and b) are of the exogenous known sample separation type first analyzed by Goldfeld and Quandt (1973). A potentially better approach is to estimate the degree of conservatism endogenously. Unfortunately, since the firm’s degree of conservatism is unobservable to the researcher, the switching regression would be of the unknown sample separation type. The latter raises difficult technical estimations issues and is the subject of ongoing research. For surveys of switching regression models, see especially Maddala (1983, 1986).

Note that SI is negative for write-offs.

Earnings news is a weighted average of the CES and revisions to returns. Hence, CES and earnings news may be of opposite sign if CES and revisions to returns are also of opposite sign.

Although the degree conservatism is expected to be stable over time, it should not be too stable if only because CR is a function of the relative shock of earnings to unexpected returns (see Eq. 15), which varies substantially over time even at the firm level.

The assumption that free cash flows remain invested in the firm and are not paid out as dividends adds additional illustrative complexity to the analysis.

Net income (NI) equals cash flow (CF) minus depreciation expense plus investment income, computed as beginning-of-period balance of marketable securities (MS) multiplied by the cost of capital (COC). PPE-MV is computed as the present value of future cash flows. Finally, market value (MV) is computed as the sum of PPE-MV and MS.

The small difference in the two numbers is due to the fact that earnings news is computed here somewhat inexactly but more simply as the sum of the changes in ROE multiplied by the beginning of period book value of equity over the life of the firm. Earnings news is defined more exactly in Sect. 4. Note that earnings news in Scenario 1 is computed based upon the period 2 report. This computation does not presuppose knowledge of the period 3 shock beyond the period 2 special items. In particular, given the asset writedown (special items) in period 2, plant cash flows in the next period equal the product of the asset value and COC. Moreover, the period 3 return on marketable securities (the beginning of the period balance of MS times COC) and depreciation expense are also known. Ultimately, special items are what link the period 3 shock to the period 2 report.

Of course, if the firm adopts a depreciation policy such that book depreciation is less than economic depreciation, the special item could potentially overstate earnings news and the revision in equity values, although this is unlikely. It is even more unlikely if free cash flows are not paid out as dividends. More often than not, the historical cost basis of asset value understates market value, so that special items are likely to understate earnings news and the revision to equity values.

Scenario 2 assumes that no positive special items are recorded. However, as noted above, under GAAP, positive special items are sometimes recognized when earnings news is positive so that empirically one should expect a weak correlation between special items and positive earnings news and special items and the revision in returns.

We convert dollar earnings news to a percentage by dividing the dollar earnings news by beginning of the period book value of equity. For example, in Scenario 2 the percentage earnings news equals 4% (=1,314/33,139). The percentage earnings news equals (by construction) unexpected ROE. The expected ROE is 15.4% (=5,110/33,139, see Benchmark Scenario), and the actual ROE is 19.4% (=6,424/33,139).

The nonlinearity result is robust to allowing non-accounting information to also inform about negative shocks. But, in that case, accounting might be irrelevant (or dominated) to the extent that external sources provide similar information about negative shocks and better information about positive shocks. More realistically, the more conservative the accounting system, the more likely is management to provide more timely information about negative shocks and the non-accounting system to provide better information about positive shocks. In that case, nonlinearity obtains and accounting matters.

Regarding an analytical proof, see footnote 17.

The numbers in Fig. 3 are based on Scenarios 1 and 3. Scenarios 1 and 2 yield the same qualitative result except that unexpected returns will be a bit more than 4% instead of 12% in the positive quadrant.

Alciatore, M., Callaway, C., Easton, P., & Spear, N. (1998). Asset write downs: A decade of research. Journal of Accounting Literature, 17, 1–39.

Antle, R., Demski, J. S., & Ryan, S. G. (1994). Multiple sources of information, valuation, and accounting earnings. Journal of Accounting, Auditing and Finance, 9, 675–696.

Ball, R., & Shivakumar, L. (2006). The role of accruals in asymmetrically timely gain and loss recognition. Journal of Accounting Research, 44, 207–242.CrossRef

Basu, S. (1997). The conservatism principle and asymmetric timeliness of earnings. Journal of Accounting and Economics, 24, 3–37.CrossRef

Beaver, W. H., & Ryan, S. G. (2005). Conditional and unconditional conservatism: Concepts and modeling. Review of Accounting Studies, 51, 269–309.CrossRef

Callen, J. L. (2009). Shocks to shocks: A theoretical foundation for the information content of earnings. Contemporary Accounting Research, 26, 1–33.CrossRef

Callen, J. L., Hope, O.-K., & Segal, D. (2005). Domestic and foreign earnings, stock return variability, and the impact of investor sophistication. Journal of Accounting Research, 43, 377–412.CrossRef

Callen, J. L., Livnat, J., & Segal, D. (2006). The information content of sec filings and investor sophistication: A variance decomposition analysis. The Accounting Review, 81, 1017–1043.CrossRef

Callen, J. L., & Segal, D. (2004). Do accruals drive firm-level stock returns? A variance decomposition analysis. Journal of Accounting Research, 42, 527–560.CrossRef

Campbell, J. Y. (1991). A variance decomposition for stock returns. Economic Journal, 101, 157–179.CrossRef

Campbell, J. Y., & Ammer, J. (1993). What moves the stock and bond markets? A variance decomposition for long-term asset returns. Journal of Finance, 48, 3–38.CrossRef

Campbell, J. Y., Lo, A. W., & MacKinlay, A. C. (1997). The econometrics of financial markets. Princeton: Princeton University Press.

Dietrich, J. R., Muller, K. A., & Riedl, E. J. (2007). Asymmetric timeliness tests of accounting conservatism. Review of Accounting Studies, 12, 95–124.CrossRef

Fama, E. F., & French, K. R. (1997). Industry costs of equity. Journal of Financial Economics, 43, 153–193.CrossRef

Fama, E. F., & MacBeth, J. D. (1973). Risk, return, and equilibrium: Empirical tests. Journal of Political Economy, 81, 607–636.CrossRef

Frankel, R. M. & Roychowdhury, S. (2008). Are all special items equally special? The predictive role of conservatism. Working paper, Washington University and MIT. www.ssrn.com.

Givoly, D., Hayn, C., & Natarajan, A. (2007). Measuring reporting conservatism. The Accounting Review, 82, 65–106.CrossRef

Goldfeld, S. M., & Quandt, R. E. (1973). The estimation of structural shifts by switching regressions. Annals of Economic and Social Measurement, 2, 475–485.

Gonedes, N. J. (1978). Corporate signaling, external accounting, and capital market equilibrium: Evidence on dividends, income, and extraordinary items. Journal of Accounting Research, 16, 26–79.CrossRef

Maddala, G. S. (1983). Limited-dependent and qualitative variables in Econometrics. Cambridge: Cambridge University Press.

Maddala, G. S. (1986). Disequilibrium, self-selection and switching models. In Z. Griliches & M. D. Intrilligator (Eds.), Handbook of econometrics. Vol. III. Amsterdam: North-Holland.

Pae, J., Thornton, D., & Welker, M. (2005). The link between conservatism and the price-to-book ratio. Contemporary Accounting Research, 22, 693–717.CrossRef

Riedl, E. J. (2004). An examination of long-lived asset impairments. The Accounting Review, 79, 823–852.CrossRef

Riedl, E. J. & Srinivasan, S. (2007). Signaling firm performance through financial statement presentation: an analysis using special items. Working paper, Harvard Business School and University of Chicago.

Roychowdhury, S., & Watts, R. L. (2007). Asymmetric timeliness of earnings, market-to-book and conservatism in financial reporting. Journal of Accounting and Economics, 44, 2–31.CrossRef

Ryan, S. G. (2006). Identifying conditional conservatism. European Accounting Review, 15, 511–525.CrossRef

Shroff, P. K., Venkataraman R., & Zhang S. (2004). The conservatism principle and the asymmetric timeliness of earnings: An event-based approach. Working paper (www.ssrn.com), University of Minnesota.

Vuolteenaho, T. (2002). What drives firm-level stock returns. Journal of Finance, 57, 233–264.CrossRef

Watts, R. L. (2003). Conservatism in accounting part I: Explanations and implications. Accounting Horizons, 17, 207–221.CrossRef

Titel: The pricing of conservative accounting and the measurement of conservatism at the firm-year level
verfasst von: Jeffrey L. Callen
Dan Segal
Ole-Kristian Hope
Publikationsdatum: 01.03.2010
Verlag: Springer US
Erschienen in: Review of Accounting Studies / Ausgabe 1/2010
Print ISSN: 1380-6653
Elektronische ISSN: 1573-7136
DOI: https://doi.org/10.1007/s11142-009-9087-6

Springer Professional

Abstract

Bitte loggen Sie sich ein, um Zugang zu Ihrer Lizenz zu erhalten.

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Springer Professional "Wirtschaft"

Weitere Artikel der Ausgabe 1/2010

Do pennies matter? Investor relations consequences of small negative earnings surprises

Do analysts and investors fully understand the persistence of the items excluded from Street earnings?

Propping through related party transactions

Regulation fair disclosure and the cost of equity capital

The stock price effects of changes in dispersion of investor beliefs during earnings announcements

Disclosure of GAAP line items in earnings announcements