Abstract
We develop and validate a measure of tax accrual quality. Tax accrual quality captures variation in the extent to which the income tax accrual maps into income tax-related cash flows, with lower variation indicating a higher quality tax accrual. Low tax accrual quality arises from (1) management estimation error and (2) financial reporting standards that lead to differences between income tax expense and income tax cash flows not captured by deferred tax assets and liabilities. We validate our tax accrual quality measure by showing it is associated with firm characteristics that capture both constructs and by demonstrating it predicts future tax-related restatements and internal control material weaknesses. We illustrate the importance of our measure by showing that investors view tax expense as more informative in firms with better tax accrual quality. Future researchers can use tax accrual quality to address questions related to estimation error in the income tax account.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
An example of GAAP-induced mismapping is the tax effect of discontinued operations. Discontinued operations are reflected net of tax on the income statement, such that tax expense does not reflect the tax effects of discontinued operations while cash tax payments do. This disparity leads to mismapping of the income tax accrual into income tax-related cash flows due to the proper application of GAAP. See Appendix 2 for additional examples.
We modify the equation “net income = operating cash flows + accruals” to be tax specific (i.e., tax expense = cash taxes paid + tax accrual; rearranging terms yields “tax accrual = tax expense − cash taxes paid”). Because cash taxes paid is a disclosure required by all publicly traded firms, we can directly calculate the tax accrual without referring to the change in tax-related balance sheet accounts, many of which are not separately disclosed by all firms.
Because cash taxes paid pertains only to income taxes (ASC 230-10-50-2), our measure does not suffer from noise that arises in working capital accruals quality due to cash flows from operations capturing cash flows related to both working capital and nonworking capital accruals.
In robustness tests discussed in Sect. 5.1, we replace ΔDTL_LT (ΔDTA_LT) with CAPX (ΔNOL). The variables CAPX and ΔNOL serve as proxies for the most frequent and economically significant components of deferred tax liabilities (assets) that are machine readable. We note no change in inferences.
We do not define tax accrual quality as the volatility of the tax accrual (TaxACC) because this definition would yield a measure that ignores the mismapping of an accrual into its associated cash flows and include volatility of items unrelated to management estimation error.
Valuation allowances are not considered GAAP-induced mismapping, as they impact deferred tax assets. Compustat presents DTAs net of the valuation allowance, meaning the control variable ΔDTA_LT removes changes in long-term valuation allowances from TaxAQ.
Under SFAS 109, DTAs and DTLs are classified as current or noncurrent based on the current or noncurrent classification of the underlying asset or liability to which the DTA or DTL relates, not to the timing of when the accrual will reverse. A DTA (DTL) is classified according to its expected reversal date only if the DTA (DTL) does not relate to an underlying asset (liability) (SFAS 109).
An example of the first potential source of measurement error is PPE-related DTLs being classified as long-term even though a portion of the DTL is expected to reverse next period. An example of the second potential source of measurement error is allowance for doubtful accounts-related DTAs being classified as short-term even though a portion of the DTA could be expected to reverse beyond next period.
In addition, Dechow and Dichev (2002) acknowledge there is measurement error in estimating working capital accruals quality due to using reported cash flow from operations as explanatory variables in their analysis, as a portion of operating cash flows is unrelated to changes in working capital accruals (pp. 40–41). Li et al. (2015) attempt to remove this measurement error by calculating adjusted cash flows for sales, purchases, and other current assets and liabilities but note this is difficult to do for tax-related cash flows. We acknowledge that our explanatory variable (cash taxes paid) also contains measurement error as a portion of cash taxes paid in t − 1 through t + 1 is unrelated to the tax accrual in period t.
Note that a firm’s tax year generally corresponds to its fiscal year. In addition, firms with actual income tax liabilities greater (less) than their estimated tax payments remit an additional cash tax payment when filing their tax return (reduce their third quarter estimated tax payment).
Windfalls affect CTP but not TTE, while shortfalls affect TTE but not CTP when the unrealized portion of the DTA is not offset to APIC.
Non-articulating items that do not affect net income but do affect cash tax payments could lead to GAAP-induced mismapping. An example is a change in accounting method for inventory from LIFO to FIFO. This change will not affect net income but will affect retained earnings and cash tax payments because of the LIFO conformity rule. Non-articulating items that do not affect net income or cash tax payments do not generate GAAP-induced mismapping. Examples include unrealized gains and losses or foreign currency translations and many changes in accounting principle.
Appendix 2 shows that only changes in tax reserves for uncertain tax positions related to permanent and short-term temporary BTDs affect tax accrual quality through GAAP-induced mismapping, while uncertain tax positions related to temporary BTDs can affect tax accrual quality through management estimation error. The model we use to estimate UTBs pertains to the total tax reserve (which captures uncertain tax positions related to both permanent and temporary BTDs), such that the relation we document between TaxAQ and UTB_EST captures both GAAP-induced mismapping and management estimation error.
We use an industry indicator to capture firms likely to issue stock option grants because Compustat data on grants does not begin until 2005 and does not cover all firms in our sample. We select SIC codes 30–39 and 70–89 following Lev and Nissim (2004) based on their interpretation of Table 1 in Huson et al. (2001). While Compustat includes two variables that capture the excess tax benefits from stock options from the operating and financing sections of the statement of cash flows (TXBCO and TXBCOF, respectively), during our sample period TXBCO (TXBCOF) has a nonzero and nonmissing value for only 5 (16) % of the firm-years in the Compustat universe.
We use the presence of foreign taxable income rather than the percentage of foreign sales from Compustat segment data because the latter has measurement error due to managers using discretion in determining what constitutes a geographic segment for financial reporting purposes (Dyreng and Markle 2013). For example, Apple determines its geographic segments “based on the nature and location of its customers” (Form 10-K FY2012, p. 71), which does not necessarily correspond to where income is taxable. While 61 % of Apple’s sales are US-based per its geographical segment data (p. 73), only 34 % of its pre-tax income is domestic. The correlation between indicator variables based on foreign segment presence and foreign pre-tax earnings presence is only 0.62 (p < 0.01), highlighting that geographic segment data does not necessarily capture the number of foreign tax jurisdictions (our construct of interest).
We do not distinguish between tax-related restatements driven by pre-tax accounting errors versus tax-related restatements that originate in the tax account because both types of restatements capture management estimation error within the tax account and lead to mismapping of tax expense into cash taxes paid. However, it is unlikely that our restatement and ICW predictive validity results are solely driven by flow through estimation error from pre-tax accounting errors because we control for working capital accruals quality and volatility of pre-tax book income in Eq. 3.
Thomas and Zhang (2014) note the role of tax expense as a measure of firm performance suggests β2 should be positive, while the alternative role of tax expense under the matching principle suggests β2 should be negative. The authors demonstrate that β2 is positive but flips to negative after controlling for analysts’ revisions in a firm’s future earnings growth. As we are interested in the role of the change in tax expense as an alternate measure of firm performance, we do not control for analysts’ revisions in future earnings growth in Eq. 4.
We do not use quarterly data because cash taxes paid is not a required quarterly disclosure and is not captured by Compustat (Deloitte 2013).
SFAS 109 permits firms to net their short-term DTAs/DTLs and long-term DTAs/DTLs. Accordingly, we reset 659 firm-year observations with missing net long-term deferred tax liability (TXDB) values and 22,806 firm-year observations with missing net long-term deferred tax asset (TXDBA) values to zero. As none of these firm-year observations have both TXDB and TXDBA reset to zero, we believe missing data is driven by netting such that resetting missing values is not a problem when estimating our measure.
Dechow and Dichev (2002) similarly find that the current period change in working capital in period t is not positively associated with cash flows from operations in t − 1 until they control for cash flows from operations in period t (their Panel B of Table 2). They explain the need to control for CFOt because “∆WCt is negatively correlated with CFOt, and CFOt is positively correlated with [CFOt−1], which counteracts the expected positive relation between ∆WCt and [CFOt−1]” (p. 42).
Expanding the cash flows from t − 3 to t + 3 results in significant sample loss of 9251 observations, as we lose data from the first two and final 2 years of our sample. Analysis in Table 3 can only be done using a pooled OLS regression because of lack of degrees of freedom when estimating the expanded model with t − 3, t − 2, t + 2, and t + 3 tax cash flows on a firm-specific basis. Expanding the window to 3 years will capture most long-term reversals that could affect TaxAQ. One possible exception is UTBs from permanent BTDs related to (1) nonfederal uncertain tax positions with statutes of limitation that exceed 3 years or (2) positions where firms under federal audit agree to extend the IRS’s statute of limitations beyond 3 years. An alternate way to assess the magnitude of measurement error from long run UTB reversals is to see how much variation in TaxAQ is explained by UTBs. We find that UTBs explain only 0.2 % of the variation in TaxAQ (see Sect. 4.5), suggesting that the potential for measurement error is small. Transactions that affect long-term valuation allowances, UTBs related to temporary BTDs, or permanently reinvested earnings have the potential to reverse in the long run, but do not affect TaxAQ because these transactions also affect changes in long-term DTAs and DTLs (which we control for in Eq. 1b). Thus, their lack of reversal outside of t − 1 to t + 1 is not a concern. See Appendix 2 for a detailed discussion of these three items.
To provide a benchmark for the expanded TaxAQ model, in untabulated analyses we also expand Dechow and Dichev’s working capital accruals model to include cash flows from operations from t − 3 to t + 3. While the coefficients on operating cash flows at t − 2 and t − 3 are insignificant (p > 0.10), the coefficients on operating cash flows at t + 2 and t + 3 are significant (p < 0.10) but small in magnitude (0.012 and 0.011, respectively). We note a small increase in adjusted R2 of 0.003 relative to the working capital accruals model that only includes operating cash flows from t − 1 through t + 1.
T-statistics in Panel B (C) in Table 3 are determined based on the distribution of the 18,314 (691) coefficients obtained from regressions at the firm (industry) level and ignore any cross-sectional correlation in the data, suggesting the results should be viewed as descriptive rather than definitive.
The UTB_EST coefficient could be insignificant in Panel C because the model used to estimate UTB_EST (Rego and Wilson 2012, Eq. 1) implicitly assumes the tax reserve is estimated and recorded similarly pre- and post-FIN 48. However, pre-FIN 48, it is unclear whether and at what dollar value firms were recording tax reserves. Post-FIN 48 (as of 2007) firms are required to separately disclose their UTBs. We define UTB_ACTUAL as a firm’s UTBs (TXTUBEND) scaled by total assets (AT). The correlation between UTB_EST and UTB_ACTUAL is only 0.33 (p < 0.01, untabulated). This low correlation suggests UTB_EST contains significant noise.
Ideally, we would like to use unrestated data in our empirical tests. However, many of our variables of interest are unavailable in the Compustat unrestated U.S. quarterly data dataset. A tax-related restatement is expected to yield a revised tax expense value but not a revised cash taxes paid value. As our tax accrual quality measure captures variation in the mapping of tax expense into cash taxes paid, we believe we will still be able to detect worse tax accrual quality in restatement firms using restated data for our empirical tests. Per Audit Analytics, tax-related restatements are due to “errors or irregularities in approach, understanding, or calculation associated with various forms of tax obligations or benefits. Many of these restatements relate to foreign tax, specialty taxes or tax planning issues. Some deal with failures to identify appropriate differences between tax and book adjustments” (Usvyatsky and Whalen 2014, p.25). ICWs are due to “internal control deficiencies in approach, understanding or calculation associated with various forms of tax obligations or benefits [and can] relate to foreign tax, local taxes or tax planning issues … the accounts impacted can include expense, deferral or allowances” (Cheffers et al. 2010, p.37).
There are 311 (233) firm-year observations in our sample with a tax-related restatement (ICW). We compute the 32 % increase in the predictive ability of observing a tax-related restatement by dividing the 2.25 % frequency of tax-related restatements among the 311 firm-year observations with the worst TaxAQ (untabulated) by the 1.70 % unconditional probability of observing a tax-related restatement and subtract one. We compute the 72 % increase in the predictive ability of observing a tax-related ICW by dividing the 4.29 % frequency of tax-related ICWs among the 233 firm-year observations with the worst TaxAQ (untabulated) by the 2.50 % unconditional probability of observing a tax-related ICW and subtract one (untabulated).
To further highlight that TaxAQ is distinct from AQ, in untabulated analyses, we replace TaxAQ in Eq. 3 with the component of TaxAQ orthogonal to AQ (e.g., the residual from regressing TaxAQ on AQ). We expect the estimated coefficient on the component to be negative and significant and find evidence consistent with this prediction (p < 0.05). The coefficient on AQ remains insignificant.
De Simone et al. (2014) also model current period ICWs (Table 3, Panel A). Similar to their model, we control for losses, mergers, extreme sales, Big Four, log of assets rather than log of market capitalization (correlation of 0.91), and foreign taxable profits rather than foreign sales (because foreign taxable profits better captures the underlying construct of tax complications from foreign operations). We do not include controls for total audit fees, influence, foreign currency transaction, Shumway, and auditor resignation in our tabulated analysis because they result in substantial sample loss of 22 %; in untabulated analysis, TaxAQ is still significant (p = 0.01) after the inclusion of these additional controls. In addition, including these additional controls when Y = TAX_RESTATEt+2 still yields a significant coefficient on TaxAQ (p < 0.01).
We find that the Cook’s D procedure predominantly removes outliers in ∆PTBI and ∆RET.
All findings in Table 7 are robust to including the Fama–French beta factors instead of market-to-book and log of market value of equity.
We would like to capture the difference between financial reporting (expenses) and tax reporting (deductions, which generally correspond to cash payments) for these two items. However, there are no statement of cash flow variables in Compustat that capture current period expenditures on intangible assets or defined benefit plan cash contributions.
References
Ayers, B., Jiang, X., & Laplante, S. (2009). Taxable income as a performance measure: The effects of tax planning and earnings quality. Contemporary Accounting Research, 26, 15–54.
Belsey, D., Kuh, E., & Welsch, R. (1980). Regression diagnostics: Identifying influential data and sources of collinearity. New York: Wiley.
Cazier, R., Rego, S., Tian, X., & Wilson, R. (2015). The impact of increased disclosure requirements and the standardization of accounting practices on earnings management through the reserve for income taxes. Review of Accounting Studies, 20, 436–469.
Cheffers, M., Whalen, D., & Thrun, M. (2010). SOX 404 Dashboard: Year 6 update. Sutton, MA: Audit Analytics.
Choudhary, P., Koester, A., & Pawlewicz, R. (2015). Do auditor-provided tax services affect estimation error in the income tax expense? http://ssrn.com/abstract=2320479.
De Simone, L., Ege, M., & Stomberg, B. (2014). Internal control quality: The role of auditor-provided tax services. The Accounting Review 90(4), 1469–1496.
Dechow, P. (1994). Accounting earnings and cash flows as measures of firm performance: The role of accounting accruals. Journal of Accounting and Economics, 18, 3–42.
Dechow, P., & Dichev, I. (2002). The quality of accruals and earnings: The role of accrual estimation errors. The Accounting Review, 77, 35–59.
Dechow, P., Ge, W., & Schrand, C. (2010). Understanding earnings quality: A review of the proxies, their determinants and their consequences. Journal of Accounting and Economics, 50, 344–401.
Dechow, P., & Schrand, C. (2004). Earnings quality. New York: The Research Foundation of CFA Institute.
Deloitte. (2013). Checklist for quarterly report on SEC Form 10-Q. http://www.corpgov.deloitte.com/site/caneng/self-assessments-tools-and-other-resources/financial-reporting-tools/. Last accessed December 13, 2014.
Dhaliwal, D., Gleason, C., & Mills, L. (2004). Last-chance earnings management: Using the tax expense to meet analysts’ forecasts. Contemporary Accounting Research, 21, 431–459.
Doyle, J., Ge, W., & McVay, S. (2007). Determinants of weaknesses in internal control over financial reporting. Journal of Accounting and Economics, 44, 193–223.
Dyreng, S., Hanlon, M., & Maydew, E. (2008). Long-run corporate tax avoidance. The Accounting Review, 83, 61–82.
Dyreng, S., & Markle, K. (2013). The effect of financial constraints on tax-motivated income shifting by US multinationals. http://ssrn.com/abstract=2336997.
Ernst & Young. (2014). To the point: New projects added to FASB and EITF agendas. http://www.ey.com/publication/vwluassetsdld/tothepoint_bb2799_agendaprioritization_14august2014/$file/tothepoint_bb2799_agendaprioritization_14august2014.pdf. Accessed 13 Dec 2014.
Fama, E. F., & French, K. R. (1992). The cross-section of expected stock returns. The Journal of Finance, 47(2), 427–465.
FASB. (1975). Accounting for contingencies. Statement of Financial Accounting Standards No. 5. Norwalk, CT: FASB
FASB. (1992). Accounting for income taxes. Statement of Financial Accounting Standards No. 109. Norwalk, CT: FASB.
FASB. (2004). Share-based payment. Statement of Financial Accounting Standards No. 123(R). Norwalk, CT: FASB.
FASB. (2006). Accounting for uncertainty in income taxes: An interpretation of FASB statement No. 109. FASB Interpretation No. 48. Norwalk, CT: FASB.
Francis, J., LaFond, R., Olsson, P., & Schipper, K. (2004). Costs of equity and earnings attributes. The Accounting Review, 79(4), 967–1010.
Francis, J., LaFond, R., Olsson, P., & Schipper, K. (2005). The market pricing of accruals quality. Journal of Accounting and Economics, 39, 295–327.
Francis, J., LaFond, R., Olsson, P., & Schipper, K. (2007). Information uncertainty and post-earnings-announcement-drift. Journal of Business Finance and Accounting, 34(3), 403–433.
Francis, J., & Schipper, K. (1999). Have financial statements lost their value-relevance? Journal of Accounting Research, 37(2), 319–352.
Frank, M., & Rego, S. (2006). Do managers use the valuation allowance account to manage earnings around certain earnings targets? Journal of the American Taxation Association, 28, 43–65.
Gleason, C., Pincus, M., & Rego, S. (2011). Consequences of material weaknesses in tax-related internal controls for financial reporting and earnings management. http://ssrn.com/abstract=1509765.
Goh, B., Lee, J., Lim, C. Y., & Shevlin, T. (2014). The effect of corporate tax avoidance on the cost of equity. http://ssrn.com/abstract=2237742.
Grant Thornton. (2013). Grant Thornton tax function optimization. https://www.grantthornton.com/~/media/content-page-files/tax/pdfs/white-papers-survey-reports-articles/2013/Tax-Function-Optimization-Survey-2013.ashx. Last accessed December 13, 2014.
Guenther, D., Matsunaga, S., & Williams, B. (2013). Tax avoidance and firm risk. http://ssrn.com/abstract=2153187.
Hanlon, M. (2005). The persistence and pricing of earnings, accruals, and cash flows when firms have large book–tax differences. The Accounting Review, 80, 137–166.
Hanlon, M., Laplante, S., & Shevlin, T. (2005). Evidence on the possible information loss of conforming book income and taxable income. Journal of Law and Economics, 48, 407–442.
Holthausen, R., & Verrecchia, R. (1988). The effect of sequential information releases on the variance of price changes in an intertemporal multi-asset market. Journal of Accounting Research, 26(1), 82–106.
Hribar, P., & Collins, D. (2002). Errors in estimating accruals: Implications for empirical research. Journal of Accounting Research, 40, 105–134.
Huson, M. R., Scott, T. W., & Wier, H. A. (2001). Earnings dilution and the explanatory power of earnings for returns. The Accounting Review, 76(4), 589–612.
Internal Revenue Service. (2013). Instructions for Form 1120-W. http://www.irs.gov/pub/irs-pdf/i1120w.pdf. Last accessed December 13, 2014.
Jones, K., Krishnan, G., & Melendrez, K. (2008). Do models of discretionary accruals detect actual cases of fraudulent and restated earnings? An empirical analysis. Contemporary Accounting Research, 25(2), 499–531.
Kothari, S. P. (2001). Capital markets research in accounting. Journal of Accounting and Economics, 31, 105–231.
Krull, L. (2004). Permanently reinvested foreign earnings, taxes, and earnings management. The Accounting Review, 79, 745–767.
Lassila, D. R., Omer, T., Shelley, M., & Smith, L. M. (2010). Do complexity, governance, and auditor independence influence whether firms retain their auditors for tax services? Journal of the American Taxation Association, 32(1), 1–23.
Lev, B., & Nissim, D. (2004). Taxable income, future earnings, and equity values. The Accounting Review, 79, 1039–1074.
Li, V., Shevlin, T., & Shores, D. (2015). Mitigating measurement error in the AQ measure of accrual quality. Working paper, University of Washington.
Linebaugh, K., Thurm, S., & Lessin, J. (2013). Apple tax bill overstated to investors. The Wall Street Journal, May 21.
Lobo, G., & Zhao, Y. (2013). Relation between audit effort and financial report misstatements: Evidence from quarterly and annual restatements. The Accounting Review, 88(4), 1385–1412.
Plumlee, M., & Yohn, T. (2010). An analysis of the underlying causes attributed to restatements. Accounting Horizons, 24(1), 41–64.
Public Company Accounting Oversight Board. (2007). Auditing Standard No. 5: An audit of internal control over financial reporting that is integrated with an audit of financial statements, Appendix A—Definitions. Washington, DC: Government Printing Office.
Raedy, J., Seidman, J., & Shackelford, D. (2011). Is there information content in the tax footnote? Working paper. http://ssrn.com/abstract=1759266.
Rego, S. (2003). Tax avoidance activities of U.S. multinational corporations. Contemporary Accounting Research, 20, 805–833.
Rego, S., & Wilson, R. (2012). Equity risk incentives and corporate tax aggressiveness. Journal of Accounting Research, 50(3), 775–810.
Schrand, C., & Wong, F. (2003). Earnings management using the valuation allowance for deferred tax assets under SFAS No. 109. Contemporary Accounting Research, 20, 579–611.
Srinivasan, S., Sijamic, W. A., & Yu, G. (2014). Admitting mistakes: Home country effect on the reliability of restatement reporting. The Accounting Review 90(3), 1201–1240.
Thomas, J., & Zhang, F. (2014). Valuation of tax expense. Review of Accounting Studies 19, 1436–1467.
Trochim, W., & Donnelly, J. (2007). Research methods knowledge base (3rd ed.). Mason, OH: Thomson Custom Publishing.
Usvyatsky, O., & Whalen, D. (2014). Financial restatements: A thirteen year comparison. Sutton, MA: Audit Analytics.
Vuong, Q. H. (1989). Likelihood ratio tests for model selection and non-nested hypotheses. Econometrica, 57, 307–333.
Wilson, R. (2009). An examination of corporate tax shelter participants. The Accounting Review, 84(3), 969–999.
Acknowledgments
This paper has benefitted from comments by Bill Baber, Brad Blaylock, Katharine Drake, Alex Edwards, Andrew Finley, Michelle Hanlon (2013 FARS discussant), Susan Krische, Rick Laux, Sean McGuire (2013 CFEA discussant), Wayne Nesbitt (2013 ATA discussant), Mort Pincus, Katherine Schipper, tax readings groups at the University of Arizona and the University of Texas, workshop participants at American University, George Mason University, Penn State University, and conference participants at the 2012 Washington Accounting Research Symposium, the 2013 FARS mid-year meeting, the 2013 ATA mid-year meeting, and the 2013 Conference on Financial Economics and Accounting (CFEA). Shevlin acknowledges financial support from the Paul Merage School of Business at the University of California-Irvine. Choudhary and Koester acknowledge financial support from the McDonough School of Business’ Center for Financial Markets and Policy at Georgetown University. Choudhary coauthored this paper prior to joining the PCAOB. The views expressed in this paper are her personal views and do not necessarily reflect the view of the Board as a whole or members of its staff or Board. This paper was previously titled “Assessing the quality of the income tax accrual.”
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix 1: Variable definitions
Variable | Definition |
|---|---|
Tax accrual quality variables | |
CTPjt | Cash taxes paid related to income taxes (TXPDjt), scaled by total assets (ATjt) |
∆DTA_LTjt | Change in the long-term portion of the deferred tax asset (TXDBAjt − TXDBAjt−1), scaled by total assets (ATjt). Because SFAS 109 permits firms to net their short-term DTAs/DTLs and long-term DTAs/DTLs and in practice many firms do, we reset missing values of TXDBjt equal to net DTA/DTL (TXNDBjt) less short-term DTL (TXDBCLjt) less short-term DTA (TXDBCAjt), with missing values of TXDBCLjt (TXDBCAjt) reset to zero when TXDBCAjt (TXDBCLjt) is not equal to missing. If TXDBAjt is missing and TXDBjt is not missing, TXDBAjt is reset to zero (N = 22,086) |
∆DTL_LTjt | Change in the long-term portion of the deferred tax liability (TXDBjt − TXDBjt−1), scaled by total assets (ATjt). Because SFAS 109 permits firms to net their short-term DTAs/DTLs and long-term DTAs/DTLs and in practice many firms do, we reset missing values of TXDBjt equal to net DTA/DTL (TXNDBjt) less short-term DTL (TXDBCLjt) less short-term DTA (TXDBCAjt), with missing values of TXDBCLjt (TXDBCAjt) reset to zero when TXDBCAjt (TXDBCLjt) is not equal to missing. If TXDBjt is missing and TXDBAjt is not missing, TXDBjt is reset to zero (N = 659) |
TaxACCjt | Total tax accrual, calculated as total tax expense scaled by total assets (TTEjt) less cash taxes paid scaled by total assets (CTPjt) |
TaxAQjt | Standard deviation of the residuals from firm-specific estimates of Eq. 1b (TaxACCjt = β0 + β1CTPjt−1 + β2CTPjt + β3CTPjt+1 + β4∆DTL_LTjt + β5∆DTA_LTjt + εjt), multiplied by −1, so larger values indicate better tax accrual quality. As Eq. 1b is estimated over 8-year rolling windows, TaxAQ is calculated using residuals from t − 7 to t |
TTEjt | Total tax expense (TXTjt), scaled by total assets (ATjt) |
Other tax measures | |
CASHETR3jt | Three-year cash-based effective tax rate, defined as cash taxes paid (TXPDjt) summed over periods t − 2 through t, divided by pre-tax income (PIjt) less special items (SPIjt) summed over periods t − 2 through t |
GAAPETR3jt | Three-year GAAP-based effective tax rate, defined as total tax expense (TXTjt) summed over periods t − 2 through t, divided by pre-tax income (PIjt) summed over periods t − 2 through t |
PERMBTD3jt | Three-year permanent book–tax difference, defined as pre-tax income (PIjt) summed over periods t − 2 through t minus estimated taxable income summed over periods t − 2 through t minus grossed-up deferred tax expense (TXDIjt ÷ U.S. statutory tax rate) summed over periods t − 2 through t, scaled by lagged total assets (ATjt−1). Estimated taxable income is measured as the sum of federal current tax expense (TXFEDjt) and foreign current tax expense (TXFOjt) grossed up by the top U.S. statutory tax rate of 0.35, minus the change in net operating loss carryforwards (TLCFjt). If both TXFEDjt and TXFOjt are missing, we measure current tax expense as the difference between total tax expense (TXTjt) and deferred tax expense (TXDIjt) |
TOTALBTD3jt | Three-year total book–tax difference, defined as pre-tax income (PIjt) summed over periods t − 2 through t less estimated taxable income summed over periods t − 2 through t, scaled by lagged total assets (ATjt−1). Estimated taxable income is measured as the sum of federal current tax expense (TXFEDjt) and foreign current tax expense (TXFOjt) grossed up by the top U.S. statutory tax rate of 0.35, minus the change in net operating loss carryforwards (TLCFjt). If both TXFEDjt and TXFOjt are missing, we measure current tax expense as the difference between total tax expense (TXTjt) and deferred tax expense (TXDIjt) |
ETR_STDjt | Following Guenther et al. (2013), the standard deviation of cash ETR [(TXPDjt ÷ (PIjt − SPIjt)] from t − 4 through t for observations where (PIjt − SPIjt) > 0 |
ETR4_ETR3jt | Following Dhaliwal et al. (2004), the fourth-quarter GAAP ETR less the third-quarter GAAP ETR; GAAP ETR is defined as year-to-date tax expense (Compustat quarterly TXTQj) divided by accumulated pre-tax income (Compustat quarterly PIQj) |
Firm characteristics related to tax accrual quality | |
DISC&EXTRAjt | Indicator variable set equal to one when a firm reports a large discretionary/extraordinary item [defined as discontinued and extraordinary items from the statement of cash flows (XIDOCjt) > 1 % of revenue (REVTjt)] and set equal to zero otherwise |
ESO_EXERCISEjt | Sum of the value of annual option exercises (Equilar value realized on exercise) scaled by total assets (ATjt). Data available in Equilar from 2002 to 2009 |
ESO_INDUSTRYjt | Indicator variable set equal to one if a firm operates in an industry with potentially large tax deductions from the exercise of options (defined as industry SIC codes 30–39 and 70–89) and set equal to zero otherwise |
FOREIGNjt | Indicator variable set equal to one when a firm reports nonzero and non-missing foreign tax expense (TXFOjt) and set equal to zero otherwise |
PTBI_VOLjt | Standard deviation of pre-tax income (PTBIjt) scaled by total assets (ATjt), measured from t − 7 through t |
SIZEjt | Natural log of total assets (ATjt) |
TAX_LOSSjt | Indicator variable set equal to one when current tax expense (TXCjt) is less than zero and set equal to zero otherwise |
UTB_ACTUALjt | Unrecognized tax benefit (TXTUBENDjt) scaled by total assets (ATjt). Data available post-2006 |
UTB_ESTjt | Predicted value of unrecognized tax benefits, estimated from Equation 1 in Rego and Wilson (2012) |
Predictive validity variables | |
NONTAX_ICWjt+2 | Indicator variable set equal to one if a firm reports an internal control material weakness in t + 2 and TAX_ICWjt+2 = 0 and set equal to zero otherwise |
NONTAX_RESTATEjt+2 | Indicator variable set equal to one if a firm restated its t + 2 financials and TAX_RESTATEjt+2 = 0 and set equal to zero otherwise |
TAX_ICWjt+2 | Indicator variable set equal to one if a firm reports an internal control material weakness in t + 2 due to a tax-related issue (IC_IS_EFFECTIVEjt+2 = ‘N’ and NOTEFF_ACC_REAS_KEYSjt+1 = ‘41’ which relates to “tax expense/benefit/deferral issues,” from Audit Analytics’ SOX 404 Internal Controls database) in our sample period for which data are available (2004–2010) and set equal to zero otherwise |
TAX_RESTATEjt+2 | Indicator variable set equal to one if a firm restated its t + 2 financials due to a tax-related issue (RES_ACC_RES_FKEY_LISTjt+2 = ‘|18|’ which relates to “tax expense/benefit/deferral/other (SFAS109) issues,” from the Audit Analytics’ Non-Reliance Restatements database) in our sample period for which data are available (2000–2010) and set equal to zero otherwise |
Market-based tests | |
RETjt | Buy-and-hold size-decile-adjusted return to firm j over the 16-month return window (starting in the 1st month of fiscal year t and ending in the 4th month after the end of fiscal year t) |
∆PTBIjt | Change in pre-tax income (PIjt), scaled by lagged market value of equity (CSHOjt−1 × PRCC_Fjt−1) |
∆TTEjt | Change in total tax expense (TXTjt), scaled by lagged market value of equity (CSHOjt−1 × PRCC_Fjt−1) |
MBjt | Market-to-book ratio [(CSHOjt × PRCC_Fjt) ÷ CEQjt] |
LogMVEjt | Log of the market value of equity [log(CSHOjt × PRCC_Fjt)] |
Other variables | |
AQjt | Standard deviation of the residuals from firm-specific estimates of ∆WCjt = α + β1CFOjt−1 + β2CFOjt + β3CFOjt+1 + β4∆REVjt + β5PPEjt + εjt, multiplied by −1, so larger values indicate better working capital accruals quality. Following Francis et al. (2005), ∆WCjt is the change in working capital accruals (Δcurrent assets (ACTjt − ACTjt−1) − Δcurrent liabilities (LCTjt − LCTjt−1) − Δcash (CHEjt − CHEjt−1) + Δcurrent portion of long-term debt (DLCjt − DLCjt−1). CFOjt is cash flows from operations (OANCFjt), ∆REVjt is change in revenue (REVTjt − REVTjt−1), and PPEjt is gross plant, property, and equipment (PPEGTjt). All variables are scaled by average total assets [(ATjt + ATjt−1) ÷ 2]. A minimum of eight residuals per firm is required to estimate AQjt |
MVEjt | Market value of equity, defined as common shares outstanding (CSHOjt) × end of fiscal year stock price per share (PRCC_Fjt) |
PTBIjt | Pre-tax income (PIjt) |
REVjt | Total revenue (REVTjt) |
TAjt | Total assets (ATjt) |
Restatement control variables | |
BIG4jt | Indicator variable set equal to one if a firm-year is audited by a Big Four auditor (AUjt = 1 through 8) and set equal to zero otherwise |
TIER2jt | Indicator variable set equal to one if a firm-year is audited by a Tier Two auditor (AUjt = 11, 16, 17, 20, or 21) and set equal to zero otherwise |
MERGERjt | Following Lobo and Zhao (2013), indicator variable set equal to one if the amount of acquisition sales divided by revenue (AQSjt ÷ REVTjt) > 0 and set equal to zero otherwise |
FINANCINGjt | Following Lobo and Zhao (2013), indicator variable set equal to one if the sum of new long-term debt plus new equity exceeds 2 % of lagged total assets [((DLTISjt + SSTKjt) ÷ (ATjt−1)) > 0.02] and set equal to zero otherwise |
LOSSjt | Indicator variable set equal to one if net income is negative (IBjt < 0) and set equal to zero otherwise |
INV_INT_COVjt | Following Lobo and Zhao (2013), inverse interest expense coverage = interest expense divided by operating income before depreciation (XINTjt ÷ OIBDPjt). The ratio is capped at 2 and assigned a value of 2 if OIBDPjt < 0 |
SALESGRjt | Following Lobo and Zhao (2013), percentage change in sales from the prior year to the current year [(SALEjt − SALEjt−1) ÷ SALEjt−1] |
EXT_FINANCINGjt | Following Lobo and Zhao (2013), indicator variable set equal to one if FREECASHjt < −0.5 and set equal to zero otherwise. FREECASH is cash flow from operations minus lagged average capital expenditures scaled by lagged current assets [(OANCFjt − average CAPXjt−1) ÷ ACTjt−1]. Average CAPXjt−1 is CAPXj averaged over t − 3 to t − 1, where available |
TENUREjt | Natural log of the number of years a company is audited by the same audit firm |
NEG_EQUITYjt | Following Lobo and Zhao (2013), indicator variable set equal to one if total liabilities are greater than total assets (LTjt > ATjt) and set equal to zero otherwise |
CURR_ACCRUALjt | Following Lobo and Zhao (2013), change in noncash current assets from year t − 1 to t scaled by average total assets [(∆ACTjt − ∆CHEjt) − (∆LCTjt − ∆DLCjt − ∆TXPjt)] ÷ [(ATjt + ATjt−1) ÷ 2] |
LEVERAGEjt | Following Lobo and Zhao (2013), total debt divided by total assets [(DLTTjt + DLCjt) ÷ ATjt]. |
BTMjt | Book-to-market ratio [CEQjt ÷ (PRCC_Fjt × CSHOjt)] |
ICW control variables | |
EXT_SGROWTHjt | Following Doyle et al. (2007), indicator variable equal to one if year-over-year industry-adjusted sales growth [(REVTjt − REVTjt−1) ÷ (REVTjt−1)] is in the top quintile and set equal to zero otherwise |
RESTRUCTUREjt | Following Doyle et al. (2007), restructuring charges multiplied by negative one, scaled by market cap [(RCPjt x − 1) ÷ (PRCC_Fjt × CSHOjt)] |
MERGER_INDjt | Indicator variable set equal to one if the absolute value of AQCjt or AQSjt or AQIjt > 0 and set equal to zero otherwise |
Appendix 2: What is in the tax accrual (TaxACC)
The purpose the this appendix is to explain what is in the tax accrual (TaxACCjt) by illustrating common items that affect total tax expense (TTEjt) or cash taxes paid (CTPjt). Part A discusses items not in the tax accrual, either because the item (i) affects both TTE and CTP in the same direction and magnitude or (ii) does not affect either TTE or CTP. Items not in the tax accrual cannot affect TaxAQ. Part B discusses items in the tax accrual, either because the item (i) affects TTE but not CTP or (ii) affects CTP but not TTE. Part C illustrates that items that affect the tax accrual and are not controlled for in Eq. 1b affect our measure of tax accrual quality (TaxAQ).
2.1 Part A: Items not included in the tax accrual
Items are excluded from our definition of the tax accrual (TTEjt − CTPjt) for two possible reasons: the item affects both TTE and CTP in the same direction and magnitude or the item does not affect TTE or CTP. As our measure of TaxAQ is the standard deviation from regressing TaxACC on five control variables, an item must affect the dependent variable TaxACC to affect the residual.
2.1.1 Part A, Section 1
Examples of items that affect both TTE jt and CTP jt in the same direction and magnitude (and thus do not affect the tax accrual)
Item | Definition/explanation |
|---|---|
PermBTDs | Revenues/expenses that appear in one set of books (financial or tax) but not the other. Examples include municipal bond interest income, qualifying incentive stock options, 50 % of meals and entertainment expenditures, and the domestic qualified production activities deduction (Section 199) |
TaxCredits | Reduces a firm’s tax expense and cash taxes paid dollar for dollar |
Tax True-upperm | Adjustment to tax expense made in t when estimation related to a permanent book–tax difference from t − 1 differed from the t − 1 tax return realization. This is an example of ex post management estimation error (because the tax expense amount originally accrued does not reflect the ex post tax obligation) that our model will not capture because the transaction does not affect the tax accrual |
Classifying PRE | Classifying foreign earnings earned in t as permanently reinvested abroad (PRE). This classification designates that the foreign earnings will not be repatriated and thus are not subject to the accrual or payment of US taxes |
Unclassifying and Remitting PRE | Unclassifying foreign earnings designated as PRE prior to t (which increases tax expense) and repatriating to US Parent in t (which increases cash taxes paid). This is an example of ex post management estimation error (because tax expense was not accrued in the period the income was earned) that our model will not capture because the transaction does not affect the tax accrual |
Generate and Carryback NOL | Net operating losses generated in t and carried back to a prior period tax return |
Under/over reserved for UTBsa | Unrecognized tax benefits represent financial reporting reserves for tax positions that the firm has taken where there is uncertainty about the ultimate tax treatment of the transaction. Evaluation and recognition determination of uncertain tax positions are required under SFAS 5, with revised recognition and measurement standards and separate disclosure requirements mandated under FIN 48. When a firm is “under-reserved” (i.e., settlement > UTB reserve) or “over-reserved” (i.e., settlement < UTB reserve), the portion of the under- or over-reserve will affect TTE and CTP in the same direction and magnitude and therefore will not be part of the tax accrual. This is an example of ex post management estimation error (because tax expense was not accrued in the period the income was earned) that our model will not capture because the transaction does not affect the tax accrual |
2.1.2 Part A, Section 2
Example of an item that does not affect either TTE jt or CTP jt (and thus does not affect the tax accrual)
Item | Definition/explanation |
|---|---|
Reclassification of long-term DTAs (DTLs) | DTAs (DTLs) established prior to t and reclassified from long-term to short-term in t because the DTA (DTL) will reverse in t + 1 |
2.2 Part B: Items that are included in the tax accrual
Because we calculate the tax accrual (TaxACCjt) as the difference between total tax expense (TTEjt) and cash taxes paid (CTPjt), items are included in TaxACC for one of two reasons: an item affects TTE but not CTP or an item affects CTP but not TTE. For example, temporary book-tax differences affect the tax accrual due to timing differences of when items are recognized for financial versus tax reporting, leading to deferred tax assets and liabilities. GAAP-induced mismapping occurs when the proper application of financial reporting standards results in a difference between tax expense and cash taxes paid that does not give rise to deferred tax assets or liabilities. Some of these differences are resolved over time (e.g., changes in UTBs), while others are not (e.g., tax effect of ESO shortfalls, tax effect of ESO windfalls, and the tax effects of non-articulating items). While these differences are not a reflection of managers’ inability to accurately estimate the tax accrual, GAAP-induced mismapping still obfuscates the ability of tax expense to reflect a firm’s underlying tax obligation. The tax accrual is comprised of the following items:
\({}^\wedge\)Temporary Book-Tax Differencesjt = ∆DTA_STjt + ∆DTL_STjt + ∆DTA_LTjt + ∆DTA_LTjt
2.2.1 Part B, Section 1
Examples of items that affect TTE jt but not CTP jt (and thus affect the tax accrual)
Item | Definition/explanation |
|---|---|
Temporary Book-Tax Differences (∆DTA&∆DTL) | Items that affect financial reporting and tax reporting by the same amount but in different periods. Deferred tax liabilities (assets) represent the deferral (acceleration) of tax payments relative to the tax expense reported on a firm’s income statement Common examples include the tax effects of accelerated depreciation/amortization, deferred compensation, ESOs, net operating loss carryforwards, accruing taxes on unremitted foreign earnings, etc. Changes in statutory tax rates also affect temporary book-tax differences. Management estimation error in these accounts arises when management’s estimate of tax expense (or benefit) differs from the ex post amount of cash taxes paid (refunded), regardless of when the cash flow occurs |
ΔVA | A valuation allowance (VA) is a contra-asset account that offsets DTAs not expected to be realized due to insufficient future taxable income required to utilize the DTAs Management estimation error in this account arises when there is a change in the valuation allowance that differs from the unrealized portion of DTAs. As a manager must assess the likelihood of realizing net operating loss carryforwards, management estimation error arises when management’s estimate of when a valuation allowance is needed (or no longer needed) differs from the outcome |
ΔIncome Taxes Payable | Income taxes payable represents income taxes accrued but not yet paid. They are affected by tax-related adjusting journal entries (i.e., tax return true-ups). These occur when a firm realizes prior to filing its period t financial statements that its estimate of tax expense is too high/low. While the tax expense adjustment is reflected in t, the cash payment related to this adjustment is made in t + 1 Management estimation error in this account arises from incorrect estimations regarding whether, when, where, and at what rate income is taxable. These errors are realized when a firm finalizes its period t tax returns in t + 1 |
Unclassifying and not Remitting PRE | Classifying foreign earnings earned in t as permanently reinvested abroad (PRE). This classification designates that the foreign earnings will not be repatriated and thus are not subject to the accrual or payment of US taxes Classifying foreign earnings earned prior to t as PRE in t but not remitting leads to a reversal of previously accrued tax expense, which affects TTEt but not CTPt. This item represents the correction of management estimation error from a prior period, as the estimate made prior to t (i.e., the estimate of foreign earnings being subject to US taxation) differs from the outcome expected in t (i.e., the current estimate of foreign earnings no longer being subject to US taxation). However, our model will not capture this management estimation error because we include ∆DTL_LT as a control variable |
ΔUTBt Unrelated to Settlements | Unrecognized tax benefits (UTBs) represent financial reporting reserves for tax positions that the firm has taken where there is uncertainty about the ultimate tax treatment of the transaction. Evaluation and recognition determination of uncertain tax positions are required under SFAS 5, with revised recognition and measurement standards and separate disclosure requirements mandated under FIN 48. This item captures (1) increases in UTBs for new uncertainties related to current period transactions and (2) changes in uncertainty regarding prior period transactions. The former increases TTEt but will not affect CTP until the position is disallowed by the taxing authority in a future period (if ever). The latter increases TTEt due to new uncertainties (with no effect on CTPt) or decreases TTEt due to reductions in uncertainty unrelated to settlements (i.e., a position is no longer deemed uncertain or a statute of limitation has expired) This is an example of GAAP-induced mismapping, as the tax expense of a firm that correctly applies FIN 48 will not reflect its ex post tax obligation In addition, management estimation error in this account arises due to difficulty in assessing which tax positions are uncertain and the amount that should be reserved for or released |
Tax Effect of ESO Exercise Shortfallpost-123(R) | Post-FAS 123(R), ESOs are recognized for financial reporting at their fair value at issuance (greater than zero) with no change in tax reporting. The timing of the financial reporting expense precedes the tax return deduction, giving rise to a DTA. If the intrinsic value at exercise is less than the fair value financial reporting expense, the unrealized portion of the deferred tax asset (“shortfall”) is offset to APIC, and any remainder increases TTEt but will not affect CTPt This is an example of GAAP-induced mismapping, as the tax expense of a firm that correctly applies FAS 123(R) will not reflect its ex post tax obligation |
2.2.2 Part B, Section 2
Examples of items that affect CTP jt but not TTE jt (and thus affect the tax accrual)
Item | Definition/explanation |
|---|---|
Tax Effect of Non-articulating Items | These transactions include the tax effect of non-articulating items (e.g., discontinued operations, extraordinary items, and other items reported net of tax on the income statement). These transactions are reflected net of tax on the income statement, such that tax expense on the income statement does not reflect the tax effects of discontinued operations (SFAS 109, pp. 35–36), but cash tax payments do reflect the tax effect of these items. These items affect CTPt but will not affect TTEt This is an example of GAAP-induced mismapping, as the tax expense of a firm that correctly applies SFAS 109 will not reflect its ex post tax obligation Because tax payments for the last three quarters of firms’ fiscal years are remitted 2 weeks prior to each quarter-end based on estimated taxable income (IRS 2013), management estimation error arises when the estimated tax payment related to non-articulating items differs from their realization on the tax return |
ΔUTBt related to settlementsa | Unrecognized tax benefits represent financial reporting reserves for tax positions that the firm has taken where there is uncertainty about the ultimate tax treatment of the transaction. If a tax position that has already been reserved for is ultimately settled for the same amount, CTPt increases but TTEt is unaffected This is an example of GAAP-induced mismapping, as the tax-related cash outflow in t relates to current tax expense accrued many periods in the past Management estimation error in this account arises due to difficult in assessing which tax positions are uncertain and the amount that should be reserved for or released |
Tax Effect of ESOspre-FAS123(R) | Pre-FAS 123(R) (June 15, 2006) ESOs were recognized for financial reporting at their intrinsic value (typically zero) at issuance and deducted for tax purposes at their intrinsic value at exercise (greater than zero). The difference in ESO values was similar to a permanent book–tax difference but treated as a credit to APIC instead of a reduction in tax expense, which reduces CTPt but has no effect on TTEt This is an example of GAAP-induced mismapping, as the tax expense of a firm that correctly applies APB 25 will not reflect its ex post tax obligation Because tax payments for the last three quarters are remitted 2 weeks prior to each quarter-end based on estimated taxable income (IRS 2013), management estimation error arises when the estimated tax payment related to ESOs pre FAS-123(R) differs from their realization on the tax return |
Tax Effect of ESO Windfallpost-FAS123(R) | Post-FAS 123(R), ESOs are recognized for financial reporting at their fair value at issuance (greater than zero) with no change in tax reporting. The timing of the financial reporting expense precedes the tax return deduction, giving rise to a DTA. If the intrinsic value at exercise exceeds the fair value financial reporting expense, the excess tax benefit (“windfall”) is credited to APIC at the exercise date. Thus the windfall will not affect TTEt but will decrease CTPt This is an example of GAAP-induced mismapping, as a firm that correctly applies FAS 123(R) will have worse tax accrual quality Because tax payments for the last three quarters are remitted 1 weeks prior to each quarter-end based on estimated taxable income (IRS 2013), management estimation error arises when the estimated tax payment related to ESOs post FAS-123(R) differs from their realization on the tax return |
2.3 Part C: Items that affect our measure of tax accrual quality
Recall that the tax accrual (TaxACCjt) includes the following:
\({}^{^{\wedge}}\)Temporary Book-Tax Differencesjt = ∆DTA_STjt + ∆DTL_STjt + ∆DTA_LTjt + ∆DTA_LTjt.
Note that Tax AJEjt, Tax Effect of Non-articulating Itemsjt, and Some Tax Aspects of ESOsjt affect taxes payable on the balance sheet. In addition, note that our model is based on the assumption that long-term classification of deferred tax assets and liabilities refers to their reversal date.
Unlike working capital accruals, several components of the tax accrual (e.g., ∆DTA_LTjt, ∆DTL_LTjt, and the long-term component of the ∆ in the valuation allowancejt) are not expected to reverse in t − 1 through t + 1. We address this in Eq. 1b by including the control variables ∆DTA_LTjt and ∆DTL_LTjt. Because Compustat reports DTAs net of the valuation allowance, ∆DTA_LTjt also controls for the long-term component of the change in the valuation allowance.
UTBs can relate to either temporary or permanent book–tax differences. Because changes in UTBs that relate to temporary book-tax differences are included in deferred tax assets and liabilities (ASC 740-10-45-12), ∆DTA_LTjt and ∆DTL_LTjt controls for UTB reversals that relate to long-term temporary book-tax differences. GAAP-induced mismapping in TaxAQ captures only UTBs related to permanent and short-term temporary book-tax differences.
In sum, TaxAQ captures both GAAP-induced mismapping and management estimation error in the tax effects of non-articulating items (e.g., discontinued operations, extraordinary items, and other items reported net of tax on the income statement), the tax effect of ESO exercises pre-SFAS 123-R, the tax effect of ESO exercise shortfalls/windfalls post-SFAS 123-R, and the nonsettlement portion of changes in UTBs related to permanent and short-term temporary book-tax differences. TaxAQ also captures management estimation error in current tax expense, short-term deferred tax assets and liabilities, and short-term valuation allowances. Estimation error in current tax expense includes tax adjusting journal entries related to permanent book–tax differences or tax credits. A limitation of TaxAQ is that it does not capture estimation error associated with long-term temporary book-tax differences, changes in permanently reinvested earnings designations, tax return true-ups, and under- or over-reserving for uncertain tax positions.
Rights and permissions
About this article
Cite this article
Choudhary, P., Koester, A. & Shevlin, T. Measuring income tax accrual quality. Rev Account Stud 21, 89–139 (2016). https://doi.org/10.1007/s11142-015-9336-9
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11142-015-9336-9