Using this average, Division B’s RI is now greater than Division A’s, although it is still significantly lower than it was in 20X1. Even using the average, Division B’s ROI is still lower than Division A’s in 20X2.
This illustrates one of the major potential problems with ROI and RI as performance measures: encouraging short-termism.
Investing in new capital has seemingly made Division B’s performance worse. Therefore, a manager might choose not to invest in new assets, to avoid the apparent negative impact on performance. However, not investing in new assets in the short term could be damaging to a division’s competitive performance in the longer term. As such, using ROI and RI as performance measures could encourage dysfunctional decision-making, rather than promoting goal congruence (one of the key characteristics of effective performance measures).
More generally, this example also illustrates a wider potential issue in performance measurement (regardless of the specific measures), being the validity of comparisons between different entities.
Comparisons of performance are most useful when the divisions (or companies) being compared are similar. In this case, the issue is that one division has had a significant increase in capital employed while the other hasn’t.
However, there could be a range of other issues which could reduce the validity of comparisons: for example, relative age of assets (and therefore the amount of accumulated depreciation reducing net book value); assets held at cost compared to assets which have been revalued; differences between the industries in which divisions operate (and differences in their potential for growth and profit).
Evaluation of ROI and RI as performance measures
As mentioned earlier one of the key characteristics of effective divisional performance measures it that they encourage goal congruence. However, using ROI to evaluate division performance can lead to sub-optimal (or dysfunctional) decision-making.
The company in Example 2 has a cost of capital of 10%. Therefore, projects which generate a return of greater than 10% will have a positive net present value and should be undertaken, because they will increase the overall value of the company in the future.
However, consider a situation where Division A from Example 2 had an opportunity in 20X2-3 to invest in a project with an expected return of 12%. If ROI is used as the main performance measure, the manager of Division A would be likely to reject the project because the return is lower than the division’s current ROI (15.9%), so undertaking will reduce divisional ROI.
This is one of the key disadvantages of ROI: that divisional managers may decide not to undertake a project with a return in excess of the cost of capital, because it has a lower ROI than the division’s current ROI.
The conventional wisdom is that, to help overcome this problem, companies should use RI as their measure of performance rather than ROI.
In the hypothetical situation facing the Manager of Division A, the choice of performance measure being used could affect the manager’s decision:
ROI: Won’t invest
Current ROI = 15.9%. Expected rate of return on new project: 12%. Manager won’t invest because expected return is below current ROI, so investing will reduce the division’s ROI.
RI: Will invest
Target rate of return (WACC) = 10%. Expected rate of return on new project: 12%. Manager will invest, because RI is positive (ie expected rate > target rate).
In this respect, RI is a better measure to use because it encourages decision-making which is consistent with the logic of net present value (NPV), which is considered the best method of investment appraisal because it looks at the value which will be generated for shareholders if a project is undertaken.
Another potential advantage of using RI instead of ROI is that it is more flexible: different costs of capital can be applied to different divisions or investments to reflect differing levels of risk.
However, there are also problems with using RI:
Difficulty in estimating cost of capital – it can be difficult to estimate the cost of capital (or to calculate the required return from a project). However, the imputed interest charge is vital to the RI calculation.
The cost of capital should include cost of equity as well as cost of debt. In practice, investment centres are often only charged the debt portion of corporate capital, which understates the true cost of the centre’s capital.
Comparing performance between divisions – to compare the performance of different divisions, a measure needs to take into account variations in size or differing levels of investment. ROI enables this, because it shows percentages, so can be used to compared returns on divisions of different sizes. By contrast, RI is an absolute measure, which makes it difficult (but not impossible) to compare performance.
Advantages and disadvantages of ROI and RI
ROI*
Advantages
|
Disadvantages
|
• Comparable – easy to compare performance between divisions (or companies of different sizes) because it provides a ratio (%) rather than an absolute value
|
• May encourage dysfunctional decision-making – eg not investing in an opportunity whose return is greater than WACC, because the projected return is less than existing ROI
• Likely to encourage short-termism
|
*: The advantages and disadvantages of using ROCE as a performance measure are the same as for ROI.
RI
Advantages
|
Disadvantages
|
• Goal congruence – If RI of an investment is positive, then the investment will be undertaken.
• Ties into the logic of NPV
• Flexibility – can use different costs of capital to reflect levels of risk
|
• Absolute values – use of absolute values rather than % makes it harder to compare performance between divisions of different sizes.
• Difficulty in estimating cost of capital
|
There are also potential problems which are common to both ROI and RI:
- Comparability: The basis of asset valuation can distort comparison. For example, if a comparison is being done of the performance of two hotel companies, one of which revalued its properties, and the other didn’t; the company with the revalued properties will have a relatively higher asset base, and therefore a relatively lower ROI or RI.
- If assets are valued at net book value, then (all other things being equal) ROI and RI will increase as assets get older, and depreciation leads to a decline in the value of the tangible asset base. This could encourage managers to retain outdated assets (so dysfunctional behaviour, rather goal congruence).
- Short-termism: More generally, financial performance measures are typically based on short-term measurement periods, which can encourage managers to become short-term oriented. For example, managers may reject investments with a positive net present value, and high payoffs in later periods, because they initially have an adverse impact (eg as a result of capital expenditure, increasing the value of capital employed.)
- Differences in accounting policies (for example for tangible assets, inventory, intangible assets) can make it difficult to compare performance between companies.
This point about reducing the impact of different accounting policies (and reducing the impact of accounting adjustments and estimates) is part of the rationale behind economic value added (EVATM) as an alternative method of performance measurement to ROI or RI.
Economic Value Added (EVATM)
EVATM is a performance evaluation tool developed, and trade-marked, by Stern Stewart & Co consultants. The rationale behind it is similar to RI, in that a finance charge is deducted from profits in order to identify value added.
The calculation of EVATM can be summarised as: