Addressing unobserved endogeneity bias in accounting studies: control and sensitivity methods by variable type

Together with their associated statistical routines, this paper describes the control and sensitivity methods that can be employed by accounting researchers to address the important issue of unobserved (omitted) variable bias in regression and matching models according to the types of variables employed. As with other social science disciplines, an important and pervasive issue in observational (non-experimental) accounting research is omitted variable bias (endogeneity). Causal inferences for endogenous explanatory variables are biased. This occurs in regression models where an unobserved (confounding) variable is correlated with both the dependent (outcome) variable in a regression model and the causal explanatory (often a selection) variable of interest. The Heckman treatment effect model has been widely employed to control for hidden bias for continuous outcomes and endogenous binary selection variables. However, in accounting studies, limited (categorical) dependent variables are a common feature and endogenous explanatory variables may be other than binary in nature. The purpose of this paper is to provide an overview of contemporary control methods, together with the statistical routines to implement them, which extend the Heckman approach to binary, multinomial, ordinal, count and percentile outcomes and to where endogenous variables take various forms. These contemporary methods aim to improve causal estimates by controlling for hidden bias, though at the price of increased complexity. A simpler approach is to conduct sensitivity analysis. This paper also presents a synopsis of a number of sensitivity techniques and their associated statistical routines which accounting researchers can employ routinely to appraise the vulnerability of causal effects to potential (simulated) unobserved bias when estimated with conventional regression and propensity score matching estimators.     

把握利润转移规律——利润战略管理探索

会计利润确认与计量的短期特质严重地扭曲了利润指标的完整面目 ,削弱了利润指标在战略管理中的积极作用。必须拓展会计利润观 ,使会计确认与计量利润不只停留于短期 ,而且关注长期趋势 ,并使短期与长期利润形成一种此长彼消的关系 ,以充分发挥利润指标的积极作用。为此 ,依据价值转移规律 ,在会计工作中牢固地树立战略利润观 ,企业基于这样的利润信息可以不断地调整并完善经营策略 ,最终赢得持续创造价值的能力。文章借鉴了西方价值转移理论 ,对我国企业如何确立战略利润的观念 ,基此开展有效的价值管理提出了具体思路     

Accounting for Foucault

Michel Foucault's concept of governmentality has been central to critical accounting research for two decades, a centrality that has placed systems of calculation as the starting point of discussions of the state, the firm and the market. We begin by outlining the development of governmentality in Foucault's own work. Despite the rich, productive nature of the concept, Foucault was careful to define governmentality as broadly and loosely as possible, the better to convey its open-endedness. The second section considers the introduction of Foucault to accounting research. The combination of Foucault and accounting history is not at all obvious, but became possible because a series of important contextual studies demonstrated that accounting history had to consider both the historicity of the profession and that its practices were vital in constructing measures of organisational performance, not simply uncovering previously obscure or hidden social realities. Moreover, accounting history studies the production of targets and measures of progress towards utility and welfare, processes that are not reducible to the firm or even to economic calculation. Our third section outlines the genesis of the ‘London School of governmentality’ and the main strands of their theoretical contribution. Finally, we examine the governmentalists’ analysis of corporate restructuring and the introduction of new production organisation by Caterpillar. Our aim is to use the Caterpillar case as the vehicle for a broader consideration of governmentality, strategy and the enterprise.     

Firm characteristics and selection of international accounting software

We report the results of a survey of international U.S. companies focused on the selection of international accounting software. We investigate the relationships among firm characteristics (size and degree of internationalization), international features of accounting software (multicurrency, multireporting, and multilingual), and general selection criteria (support and security, hardware and operating platforms, and flexibility and cost). We find that firms’ reported preferences for international features differ depending on size and degree of internationalization. However, the relationship between firm characteristics and general selection criteria is not significant. The implication from our results is that firm characteristics should be an important consideration in the selection and design of international accounting software.     

Firm ratings, momentum strategies, and crises: evidence from the US and Taiwanese stock markets

This paper investigates whether there is a link between momentum profitability and firm ratings. We follow traditional and practical (non-) investment-grade classifications to divide into three rating groups, high, median, and non investment-grade group (HIG, MIG, and NIG) since firm ratings express risk in relative rank order to contain valuable information. This study considers the US and Taiwanese stock markets. We find that firm ratings momentum strategies can even earn positive profits, larger than na?ve momentum, supporting that firm ratings can be used to strengthen naive momentum effects. By comparisons, the US firm ratings momentum with NIG produces larger profits than HIG but opposite in direction and V-shaped pattern in Taiwan. With an examination of crises on firm ratings momentum, we find that firm ratings momentum indeed helps increase the payoff during (non-)crises although firm ratings momentum profits should be strong following non-crises states and weak following crises states. However, firm ratings momentum profits partially result from the predictability of business cycle, calendar months, and information asymmetries. Our results highlight the critical importance of using firm ratings screens in empirical momentum studies.     

FINANCIAL STATEMENT USERS‘ PERCEPTIONS OF FACTORS AFFECTING THE ABILITY OF AUDITORS TO RESIST CLIENT PRESSURE IN A CONFLICT SITUATION

This study investigates the perceived effects of four factors on the likelihood that an auditor will acquiesce to the demands of a client when a dispute arises over a significant financial reporting issue. A repeated measures experiment was conducted using 49 security analysts and 69 bankers as subjects. Analysis of Variance results indicate that an auditor is perceived as being more likely to acquiesce to his client's demands when the accounting issue in dispute is not dealt with clearly by an accounting standard. Auditors also are viewed as being more likely to acquiesce when: (1) competition in the audit environment is aggressive; (2) the audit firm is small; and (3) the audit firm provides management consulting services to its audit client.     

Tax competition and the international distribution of firm ownership: an invariance result

How does the international distribution of firm ownership affect the outcomes of tax/subsidy competition for mobile plants? As corporate ownership becomes increasingly globalised, this question becomes increasingly important for policy. We prove a strong invariance result in the context of the tax/subsidy competition between two host countries for a monopoly firm’s plant. Both the equilibrium plant location and the equilibrium tax/subsidy offers are independent of the international distribution of the firm’s ownership. The reason is that the tax/subsidy competition equalises the firm’s post-tax profits across countries, making owners of capital indifferent towards the location of production.     

Commentary: Research Productivity Measurement and the Field of Academic Accounting*

The previous issue of Canadian Accounting Perspectives contained three studies that examined various aspects of performance measurement in universities. In this issue, we step back from the acts of performance measurement and disclosure to consider the context within which the demand for increased performance measurement has arisen. The current commentary considers the use of research productivity measures in the field of academic accounting and why the use of these measures is drawing such an equivocal response from academics in the field. Starting from three theoretical perspectives, we consider the use, response, and consequences associated with research productivity measurement.     

The effect of measurement timing on the information content of customer satisfaction measures

We use two sets of customer satisfaction measures obtained from a homebuilding company to examine the effect of measurement timing on the association between customer satisfaction measures and future financial performance. The research site employs two separate consulting firms that measure customer satisfaction at different times from the same homebuyer population. A national consulting firm captures customer satisfaction at a fixed time in the year following purchase (the “NF” measure), whereas an industry-focused, boutique consulting firm captures customer satisfaction at three specific points in time (30 days, 5 months, 11 months) after purchase (the “BF” measures). We analyze data for the period 2002–2004 and have the following findings: first, customers’ satisfaction varies over a homebuyer's consumption period. Comparing across the three BF measures, we find that on average a homebuyer is most satisfied 30 days after purchase and least satisfied 11 months after purchase. Second, we compare the NF measure with the BF measures and find significant differences in their predictive abilities for future financial performance. The BF measures are significant leading indicators of future financial performance, as measured by higher revenues and profits and lower warranty costs, but the NF measure is not. Additional analyses indicate that the relatively higher predictive ability of the BF measures is due to the more precise timing of those measures, rather than differences in measurement content. Finally, we find that the point of diminishing returns to improvements in customer satisfaction varies across customer satisfaction measures obtained at different points in the consumption period. We conclude that timing has a significant impact on the information content of customer satisfaction measures, at least for goods and services that are consumed over extended period of time.     

The Relation Between Patent Citations and Tobin's Q in the Semiconductor Industry

The market value of a firm is largely determined by the expected returns to the firm's tangible and intangible assets. However, accounting data generally excludes intangible assets. Financial variables which are constructed in part from accounting data, such as Tobin's Q, are thus biased. If measures of intangible capital are successful in explaining variation in Q, then a case can be made for incorporating such measures into future research. In high technology industries, such as the semiconductor industry, valuing a firm's intangible assets requires the valuation of its technological capital. Past studies have relied heavily on simple patent counts and research and development expenditures to quantify the technological component of a firm's intangible assets. This paper examines the ability of measures of intangible capital to explain variation in Q and considers an additional data source, patent citations. We find that stock variables created from citation data contain relevant information about the market's valuation of intangible assets.     

Bonuses and Non-Public Information in Publicly Traded Firms

Recent research in accounting explores how firms use “individual” or “non-financial” measures of performance in executive compensation contracts. We model a firm that conditions bonus payments to executives on information that is not available to those outside the firm. This raises two issues. First, market participants may use the magnitude of such payments to infer the non-public information. Second, because information that is non-public is, by extension, non-verifiable, the firm cannot write explicit contracts based on it. Combining the relational incentive contracts and financial signaling literatures, we examine equilibria of a signaling game in which bonus payments from a firm to a manager convey non-public information regarding the firm’s future cash flows. Our main result is that increases in corporate myopia can, under some conditions, lead to increased profits. This finding is contrary to that typically found in financial signaling models.     

An Examination of the Biases in Estimating the Benefit of Debt Insurance

This paper examines the biases in previous studies of estimating the benefit of debt insurance. We identify three possible sources of estimation biases: selection bias, marketability bias, and premium bias. Our findings indicate that both the selection bias and the marketability bias cause an underestimation of the actual benefit of bond insurance, while the premium bias results in an overestimation. Future research could benefit from this study by explicitly accounting for these estimating biases in obtaining a more accurate evaluation of the role of debt insurance in the market.     

The Impact of Irreversibility,Uncertainty and Timing Options on Deprival Valuations and the Detection of Monopoly Profits

This paper has two purposes. The first is to derive rules identifying the deprival value of an asset (i) which is irreversible to one extent or another; (ii) the benefit stream of which is subject to continuing uncertainty; and (iii) for which an option to wait exists as to when to reacquire should the owner be deprived of it. The second is to consider whether accounting rates of return employing these deprival value rules can be developed to help in the detection of monopoly profits in circumstances where investment decision-making takes place in the presence of irreversibility, uncertainty and the existence of timing options. The ‘new’ deprival value rules for an asset differ from the ‘conventional’ ones in that present value less the value of the option to reinvest in the asset appears in the ‘new’ rules wherever present value appears in the ‘conventional’ rules. Examples are provided which suggest that ‘new’ and ‘conventional’ deprival value rules can differ materially. A further result is that accounting rates of return can be developed using the ‘new’ deprival value rules that are, in principle, useful in the detection of monopoly profits. Nonetheless, in practice such use requires a level of information that renders the result superfluous in the sense that the provision of replacement cost balance sheet data, combined with the level of information needed, is sufficient to reveal the presence of monopoly profits directly.     

Major shareholders’ trust and market risk: Substituting weak institutions with trust

This study examines the impact of foreign controlling shareholder trust on firm market risk using two measures of total and idiosyncratic risk. An extensive global sample of 12,496 firm-year observations from 43 countries is employed. The results show that firms controlled by foreign trusting shareholders display lower levels of risk in both market measures. Trust appears more important for firms based in countries with a less favourable institutional environment, whereby it varies with the investment horizon of foreign controlling shareholders. The results are robust after controlling for cultural measures, endogeneity, selection bias and alternate model specifications.     

A model of cost-based transfer pricing

In most decentralized organizations, goods and services are transferred between divisions. These transfers are frequently recorded in the accounting books of the divisions; the term transfer price refers to the dollar amount of the interdivisional exchange. This study considers two main issues: (i) the costs and the benefits of delegating decisions through a system of transfer pricing and divisional performance evaluation, and (ii) the performance of one common method of pricing intrafirm transactions: cost-based transfer pricing.The study analyzes a firm in which each divisional manager has better information about the divisional environment than what is known by the firm's top management. The first half of the paper demonstrates that the firm can attain the optimal level of profits with a compensation system utilizing (i) reports by divisional managers describing in complete detail each manager's private information, and (ii) divisional performance evaluation with cost-based transfer pricing. Next, a situation is considered in which divisional managers are not able to communicate their private information to the firm's top management because of complexity of divisional environments or managers' specialized expertise. In this bounded-rationality setting, a managerial-compensation system employing cost-based transfer pricing allows the firm to earn strictly higher expected profits than if all decisions are made by the firm's top management relying on divisional managers' reports.Financial support from the Unisys Corporation is gratefully acknowledged.     

Accounting Conservatism: A Literature Review

Accounting conservatism is one of the most important properties of financial reporting. The goal of this article is to gain a better understanding of accounting conservatism. We explicate the evolution of conservatism over its long history. Accounting conservatism is indispensable because the main parties of a firm demand conservatism to mitigate agency costs. Various methods are used to measure accounting conservatism, which include balance sheet measures, income statement measures and earnings/stock return relation measures. Empirical research into accounting conservatism has flourished over the last two decades and we focus on the cross‐sectional and time‐series variations in conservatism. We conclude that accounting conservatism is important and cannot be excluded from accounting standards.     

How do different performance measures affect managerial time orientation? Empirical evidence from sales managers in the oil and gas industry

This study investigates the relative effect of performance measures on managerial time orientation. We collect survey data on the actual time allocation of sales managers for tasks that affect financial performance on the short-, medium-, and long-term horizons. In addition, we obtain survey data on the specific metrics used by an oil and gas firm and classify them into three groups: traditional accounting (gross margin and budgeted costs), nonfinancial (market share and sales volume), and accounting returns (economic value added — EVA). Based on partial least-squares analysis, our results suggest that, in our setting, both nonfinancial and accounting return measures can supplement traditional accounting metrics to mitigate potential short-term orientation by inducing sales managers to consider mainly not only sales tasks but also investing tasks, which will affect the firm results more than a quarter ahead. In addition, our results imply that accounting return metrics are not better than nonfinancial measures in inducing a longer-term orientation in our research setting.     

Predictors of revenue shifting and expense shifting: Evidence from an emerging economy

Prior literature established that managers engage in Revenue Shifting (RS) and Expense Shifting (ES) with an intent to report favourable operating performance; our paper extends such research in a new direction by investigating both forms based on the need, ease, and advantage of each form of shifting strategy. The study identifies firm-specific factors that incentivize firms to prefer RS over ES and vice-versa. We undertake a longitudinal study (2001–2019) using a sample size of 39,634 firm-years, enlisted in the Bombay Stock Exchange (BSE). Our results show that peer-performance, size, financial leverage, growth opportunities, accounting flexibility, and age of the firm are important determinants of RS and ES. Specifically, our results exhibit that large, levered, old, and high-growth firms are engaged in RS, whereas small, young, firms with lesser accounting flexibility, and firms operating below peer-performance are involved in ES. These results are robust to controlling for accruals earnings management, real earnings management, endogeneity, self-selection bias, and alternative measures of RS and ES. Our findings are helpful to auditors and investors in improving awareness of forms of classification shifting.     

CEO age and tax planning

This study investigates the association between CEO age and corporate tax planning. Using a sample of 11,537 firm‐year observations from the fiscal year 1997–2013, I find CEO age exerts an economically significant influence on firms’ tax policies, incremental to economic determinants identified in prior research. Specifically, CEO age is positively related to cash and GAAP effective tax rates, and negatively related to permanent book‐tax difference, suggesting that older CEOs are less likely to take actions to lower tax burden. The results hold across different model specifications and robustness tests to address potential bias arising from endogeneity, sample selection issue, and the confounding effect of CEO tenure.     

Reexamining the managerial ownership effect on firm value

Whether equity-based compensation and equity ownership align the interests of managers with stockholders is an important question in finance. Early studies found an inverted U-shaped relation between managerial ownership and firm value, but later studies using firm fixed effects found no relation. Managerial ownership levels change very slowly over time which may mask an ownership effect on firm value when using a fixed effect model. This is due to a much smaller within firm variation than between firm variation. We demonstrate that using pay-performance semi-elasticity, rather than pay-performance sensitivity as a measure of managerial ownership incentives, results in meaningful variation within firm over time. The greater within firm variation increases the power to detect a relation between managerial ownership and firm value with fixed effect regressions. As in the early research on this issue, we find a significant inverted U-shaped relation between managerial ownership and Tobin's Q in fixed effects regressions and after controlling for endogeneity with both two-stage and three-stage least squares regressions. Our results are consistent with incentive alignment at low levels and risk aversion at high levels of managerial ownership.