A sterner test: one of the weaknesses of the current body of research on how firms use EVA, the performance measurement system developed by Stern Stewart & Co, is that much of the practical evidence has come from Stern Stewart itself. Josie McLaren reports on a CIMA-sponsored study into the development of EVA in three firms in New Zealand

Financial Management (UK), July-August, 2003 by Josie McLaren

The increasingly popular philosophy of benefit-sharing is that the explicit recognition of secondary stakeholders such as employees, suppliers and customers maximises value to the primary stakeholders--ie, the shareholders. Economic value added (EVA) is one measurement system that's consistent with this belief. Managers can set a goal of EVA maximisation in order to achieve the primary objective, recognising the secondary stakeholders explicitly via value drivers that are identified throughout the business.

EVA has been vigorously promoted by its developer, Stern Stewart & Co. It claims that the system can be cascaded through an organisation to business units and even to processes and products. The rationale is that there are no conflicts of interest between the planning and control function, or between shareholders and managers. It's argued that managers think like owners and so have the incentive to achieve the primary objective.

Last year a CIMA-sponsored research project considered two state-owned enterprises in New Zealand and investigated differences in EVA usage between these firms and a third, publicly quoted, company. The organisations were of different sizes (see table 1) and operated in separate industries, but they could all be described as decentralised businesses with the overriding objective of maximising shareholder wealth.

At the sub-unit level, the balanced score-card was used to provide a direct focus on value drivers. This was seen to complement EVA. Companies 1 and 2 introduced the balanced scorecard after implementing EVA, but company 3 already had a scorecard in place. On its introduction, EVA was slotted straight into the scorecard framework as a financial measure. The interviewees identified the following factors as critical for a successful implementation:

* support from the chief executive and members of the board;

* a phased introduction;

* the treatment of EVA as a project, complete with a project leader;

* the provision of training for all staff intending to work with EVA.

None of the companies saw EVA as a complete measure of performance. They all made use of extra measures--for example:

* payback period for investment decision-making (all three companies);

* a focus on earnings before interest and tax both internally and externally (companies land 3);

* the balanced scorecard for its focus on value drivers (all three companies).

The main reason they cited for this continued use of other measures was that staff, board members and outsiders--shareholders, analysts or the media, for example--didn't really understand the measure. Also, EVA was seen as subjective because of the adjustments that needed to be made (see panel, left). Other measures were thought to lend more objectivity. The issue of comprehension shouldn't be underestimated. Internally, people need to be trained to ensure that EVA is fully understood. External parties have always focused on profit. They continue to do so and their lack of understanding persists.

The research revealed an evolutionary process in the use of EVA. The stages (see figure 1) reflect the length of time it had been used.

[FIGURE 1 OMITTED]

* Company 1 had come full circle. It had moved from using EVA only at firm level to using it at business-unit level. At the end of the study it was using EVA at corporate level for its core integrated units.

* Companies 2 and 3 were both using EVA at corporate and business-unit level, as advocated by Stern Stewart. The indications were that they would continue using EVA at this level--ie, the evolutionary process was complete at the second stage.

There were several reasons for the evolutionary process in company 1. Its use of the measure at business-unit level created certain problems that led it to pull back to corporate-level usage--clear evidence of the failure of the EVA philosophy in this case.

There was also an evolutionary process in the way that the firms calculated EVA. This related to the measure's simplification over time (see figure 2). Initially, there was a relatively complex measure of EVA involving several adjustments to the Gaap figures. But, even at implementation, none of the companies had as many as 20 adjustments--company 3 made the most, with 15. This firm currently makes two adjustments, while company 1 makes no adjustments to Gaap figures. Company 2 converts EVA information to conventional Gaap in the preparation of the standard accounts.

[FIGURE 2 OMITTED]

Again, the evolution in EVA calculation related to the issues of understanding and objectivity. The consistent message was that too many adjustments meant that:

* EVA was not well understood within any of the companies;

* it was perceived as something that arose from the arbitrary shuffling of cash flows across time periods.

All three firms had faced problems over economic dependence, which they had tried to resolve. For example, company 1 found that the introduction of EVA had actually enhanced conflicts between units because of these issues. It was viewed as a barrier to collaboration. The company therefore made changes to correct the adverse incentives that business-unit EVA had created.