2009 PMA Conference – Professor Kenneth Merchant: On accounting measures

The opening keynote speech of Day 1 of the 2009 Performance Management Association conference was given by, Kenneth Merchant, Deloitte & Touche LLP Chair in Accountancy and Professor of Accounting at the University of Southern California Marshall School of Business. Professor Merchant is an expert in management accounting and management control systems with numerous publications in this field.
His speech addressed the challenges practitioners are faced with in selecting and using performance measures. He discussed the flaws of accounting measures and presented four alternatives. I have structured below my notes from this session.

Part 1: On accounting measures

• For management purposes accounting measures are badly flawed. The primary role of a corporation is to create and maximize shareholder value. However, accounting measures tell you little about value. Accounting measures are backwards looking. Value measures are future looking.
Evidence: A research study analysed 172 large firms that averaged an earnings per share (EPS) growth of over 15% over a 6 year period. Their correlation between annual earnings (before extraordinary items) and annual value changes was not as strong as anticipated. These correlations are discussed in several research papers: Easton et al (1992): .22, Biddle et al (1997): .25 and Erkens & Merchant (2005): Annual.18 and Quarterly .13

• Alternative financial-measure specifications are not much better. Correlations with market returns, annual (Biddle et al., 1997)
* Earnings before extraordinary items: .25
* Residual income : .16
* EVA(TM): .15
* Cash flow from operations: .14

• Flaws of accounting performance measures:
* Fail to identify real problems in a timely basis
* Lead to poor decisions, resulting in allocation of resources to “dogs” and not to “stars” and excessive short term orientation (myopia)
* Gamesmanship – ethics issues

• Why using accounting measures? Why do firms use them anyway?
* Nice to have a single, global performance indicator.
* Earnings/returns correlations are positive.
* Some nice measurement properties: timely, understandable, inexpensive, precise, objective and controllability can be tailored to the role (easy to disaggregate)
* Habit – i.e. DuPont chart has a long tradition – ROI = Sales margin * Asset turnover
* Misconceptions, such as: “Profit making firms should make profits” and ” Quarterly EPS reports drive stock prices”.
* Political expediency

• Do Accounting Measures work anywhere? What do we do if they don’t work? Answer: It should be an industry by industry analysis of annual accounting earnings / market correlations. Correlations based on information obtained from Quarterly Earnings and Market Presentations (Erkens and Merchant):
* Overall: .18
* High: Oil and gas: .58, Mining/construction: .35, Manufacturing (misc.) .32
* Low: Pharmaceuticals: .04 (annual), Chemicals – .03 (Q) and Mining and construction – 0.2 (Q)