Influential Observations and Inference in Accounting Research

Leone, Andrew J.; Minutti‐Meza, Miguel; Wasley, Charles E.

doi:10.2139/ssrn.2407967

Cited by 73 publications

(76 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The procedure simply replaces some of the more extreme observations with less extreme values. Other methods exist, but this method is by far the most popular, and it seems to be applied in a majority of studies (Leone et al 2014). …”

Section: Ts Compared To Ols With Winsorizationmentioning

confidence: 99%

See 1 more Smart Citation

Linear valuation without OLS: the Theil-Sen estimation approach

Ohlson

Kim

2014

Rev Account Stud

View full text Add to dashboard Cite

OLS-based archival accounting research encounters two well-known problems. First, outliers tend to influence results excessively. Second, heteroscedastic error terms raise the specter of inefficient estimation and the need to scale variables. This paper applies a robust estimation approach due to Theil (Nederlandse Akademie Wetenchappen Ser A 53: [386][387][388][389][390][391][392] 1950) and Sen (J Am Stat Assoc 63(324):1379-1389, 1968) (TS henceforth). The TS method is easily understood, and it circumvents the two problems in an elegant, direct way. Because TS and OLS are roughly equally efficient under OLS-ideal conditions (Wilcox, Fundamentals of modern statistical methods: substantially improving power and accuracy, 2nd edn. Springer, New York 2010), one naturally hypothesizes that TS should be more efficient than OLS under non-ideal conditions. This research compares the relative efficiency of OLS versus TS in cross-sectional valuation settings. There are two dependent variables, market value and subsequent year earnings; basic accounting variables appear on the equations' right-hand side. Two criteria are used to compare the estimation methods' performance: (i) the inter-temporal stability of estimated coefficients and (ii) the goodness-of-fit as measured by the fitted values' ability to explain actual values. TS dominates OLS on both criteria, and often materially so. Differences in inter-temporal stability of estimated coefficients are particularly apparent, partially due to OLS estimates occasionally resulting in ''incorrect'' signs. Conclusions remain even if winsorization and the scaling of variables modify OLS.

show abstract

Section: Ts Compared To Ols With Winsorizationmentioning

confidence: 99%

“…SeeLeone et al (2014) for a discussion of alternative robust estimators.8 However, especially in environmental sciences, there is an extensive list of research that applies TS to estimate trends in pollution, climate change, and rainfall, where the data do not satisfy the assumptions for classical parametric methods (e.g.,Hirsch et al 1982).…”

mentioning

confidence: 99%

Linear valuation without OLS: the Theil-Sen estimation approach

Ohlson

Kim

2014

Rev Account Stud

View full text Add to dashboard Cite

show abstract

“…To cope with the impact of influential observations on the statistical inference, we use robust regression based on MM estimation (Leone et al, 2015) to estimate our models. Table 2 reports the robust regression results of equation (1).…”

Section: Sample and Resultsmentioning

confidence: 99%

“…However, a positive relationship between production and the current period sales change was shown in Roychowdhury (2006). To cope with the impact of influential observations on the statistical inference, we use robust regression based on MM estimation (Leone et al 2015) to estimate our models. The results are qualitatively the same.…”

Section: Sample and Resultsmentioning

confidence: 99%

Control-Oriented Inventory Production and Sales Activities in Chinese Manufacturing Firms

2018

AFR

View full text Add to dashboard Cite

This study uses a sample of Chinese manufacturing firms to examine the association between current period inventory production costs and sales changes in adjacent past and future periods. The results find an overall significant association between inventory production and the past period sales changes, but no association between inventory production and the future period sales changes. Further, we find that in firms with high return on assets, inventory production is only associated with the past period sales changes, but not with the future period sales changes; in contrast, in firms with low return on assets, inventory production is only associated with the future period sales changes, but not with the past period sales changes. The results, on one hand confirm our perception of control oriented management accounting practices in China, on the other hand suggest that firm performance is associated with such management emphasis in China.

show abstract

“…Specifically, the outlier tests include (i) winsorizing (truncating) all variables at the top and bottom 5 percentiles (1 percentiles), rather than 1 percentile winsorization used in Table 3; (ii) removing outliers that are defined as having an absolute studentized residual value greater than two (Belsley, Kuh, and Welsch 1980) or as having Cook's D value greater than four divided by the number of observations (Cook 1977); (iii) removing outliers where DEV_TARGET and REV_TARGET move in the opposite directions as these unusual revision behaviors could represent either measurement errors, such as human errors in data collection procedures, or the results of unusual circumstances affecting a subset of the data; (iv) using different subsamples such as S&P 500 LargeCap, 400 MidCap, and 600 SmallCap; and (v) employing the least absolute deviations (LAD) estimation, a form of robust regression also known as median regression, which has been used by many prior studies as it is believed to be less susceptible to the problem of outliers (Leone, Minutti-Meza, and Wasley 2013;Ortiz-Molina 2007;Chen, Liu, and Ryan 2008;Choi, Kim, Liu, and Simunic 2009;Indjejikian et al 2014a). asymmetry using firm-specific time-series analyses.…”

mentioning

confidence: 99%

Executive Bonus Target Ratcheting: Evidence from the New Executive Compensation Disclosure Rules

Kim

Shin

2017

Contemporary Accting Res

View full text Add to dashboard Cite

This study investigates whether firms revise executive bonus compensation targets based on past performance. Studies in this area suffer from a lack of detailed information related to executive performance targets. Using mandatory disclosures of executive compensation information under the U.S. Securities and Exchange Commission's new disclosure rules, this study provides the first large‐sample evidence of bonus target ratcheting. There are three major findings: (i) executive bonus targets ratchet and they ratchet asymmetrically; (ii) the degree of target ratcheting and ratcheting asymmetry vary with executive equity incentives and investment opportunities; and (iii) performance relative to bonus target is serially correlated.

show abstract

Influential Observations and Inference in Accounting Research

Cited by 73 publications

References 27 publications

Linear valuation without OLS: the Theil-Sen estimation approach

Linear valuation without OLS: the Theil-Sen estimation approach

Control-Oriented Inventory Production and Sales Activities in Chinese Manufacturing Firms

Executive Bonus Target Ratcheting: Evidence from the New Executive Compensation Disclosure Rules

Contact Info

Product

Resources

About