Why is the residual income a better measure for performance evaluation of an investment center manager than return on investment ROI?

• Tweet

  Nội dung chính Show

  • Survey of 620 Companies
  • Use of Centers
  • ROI vs. RI
  • Defining Profit & Investment
  • Intracompany Comparisons
  • The Best Measure
  • What is the major advantage of residual income as a performance measure?
  • Which is an advantage of using residual income RI over return on investment ROI?
  • Why is return on investment better than divisional income as a measure of performance for an investment center?
  • Which is better RI or ROI?

• Post

• Share

• Save

• Get PDF

• Buy Copies

• Print

A spate of articles and books about return on investment has appeared during the past two or three years. Improving company earnings in relation to the capital used to generate those earnings has become a matter of great concern to many top financial and general managers. Their concern is not surprising. About three years ago, the prime lending rate reached 12%, a reflection of how high the cost of capital had risen by the mid-1970s. And, of course, as capital costs increase, so must the return on capital employed, for if a company cannot earn a return on invested funds that is in excess of the cost of those funds, the company is not economically profitable. Suspecting that many companies are dealing with this problem by establishing their divisions as investment centers, the authors of this article did a survey of the Fortune “1000” companies for 1976. They found that investment centers are in wide use and that ROI is the usual measure of their performance.

In large companies, top management’s concern about any aspect of a business’s performance is usually transmitted to the line managers to whom various operating responsibilites have been delegated. In companies with divisionalized structures (companies organized with units responsible to top management for their own profitability), top managers convey their concern about inadequate corporate ROI by delegating some of the company’s responsibility for ROI performance to particular divisions. In delegating this responsibility, top management sets up the divisions that it chooses for this emphasis as either profit centers or investment centers.

A profit center is an organizational unit that is responsible to top management for some measure of its own profitability—a measure like net income, pretax income, or net contribution. Revenues measure the unit’s outputs, expenses measure its inputs, and profit measures its excess of revenues over expenses.

An investment center is an organizational unit responsible to top management for its profitability in relation to the unit’s own investment base. Revenues and expenses are measured as in profit centers, but the assets employed are also measured. Thus an investment center is an extension of the profit center idea: profit is measured for both, but only in an investment center is this profit related to the size of the investment base.

Designating a division as one of these types of centers, then, is actually deciding between two ways of measuring what the division is contributing to the company. It is the measure of profit in comparison with the amount of capital invested in a division that makes us especially interested in the investment center as an approach to increasing company returns. Because it takes more factors into account, this approach probably gives a more accurate picture of what a division is contributing to the company’s economic well-being than does the profit center approach.

To see how major corporations are measuring their divisions’ economic performance and how they are incorporating these measurements into their management systems, we sent a questionnaire to the controller of each of the Fortune ‘1000’ industrial companies that were listed in May and June 1976. Of these companies, 62% responded.

Survey of 620 Companies

Our study was intended to answer the following general questions:

1. How many of these companies use profit centers and how many use investment centers?

2. In the companies using investment centers, which formula to relate profits to investment does management use—return on investment (ROI), which is profit divided by investment, or residual income (RI), which is profit before interest expense minus a capital charge levied on investment?

3. How do such companies define profit and investment for measuring investment center performance?

4. How do companies compare the performance of their investment centers?

In many respects, this study was similar to one conducted in 1965 by John J. Mauriel and Robert N. Anthony.1 Because of the 12-year gap between the two surveys, we were interested in comparing the Mauriel-Anthony findings with our own to identify any significant changes in companies’ approaches to investment center measurement.

Use of Centers

Exhibits I, II, and III summarize our findings about companies’ use of profit centers and investment centers. Nearly three-fourths of the 620 respondents had two or more investment centers. Their use tended to increase with company size: while fewer than 50% of the companies with sales under $100 million used investment centers, 84% of the companies with sales over $500 million used them.

Exhibit I Overall results of 620 companies surveyed

Exhibit II Use of profit and investment centers by sales volume

Exhibit III Extent of use of profit and investment centers by industry

In industries with 10 or more respondents the use of investment centers was lowest in publishing and printing (45%), and highest in measuring, scientific, and photo equipment (96%).

Employing the control-structure device of either profit centers or investment centers was not a new practice for most of our respondents (see Exhibit IV). Only 34 companies (5.7%) using such centers for measurement purposes had begun doing so as early as the past 5 years, whereas 226 companies (38%) had had them for over 25 years.

Exhibit IV Experience with profit and/or investment centers

Not surprisingly, the companies for which those centers were new approaches tended to be the small ones in our sample: only 1 company out of 128 with sales over $1 billion had added the center concept within the past 5 years, whereas 21 companies out of 203 having sales under $200 million had begun the center approach during that 5-year period.

When compared with what Mauriel and Anthony found, our results indicate that the profit center—investment center concept has gained maturity: they found that over one-third of their respondents using investment centers had begun doing so in the 5 years preceding their 1965 study and that over one-half had begun in the previous 10 years.

ROI vs. RI

Although many managers seem to think of using an investment center approach as being synonymous with measuring return on investment, some companies relate profits and the investment base by using the residual income measure. Because each measurement has strengths and weaknesses, many companies make both calculations for their investment centers.

As shown in Exhibit V, measuring ROI only was by far the most common approach: 65% of respondents having investment centers were using it, while 28% were measuring both ROI and RI, and only 2% were measuring RI alone. (The Mauriel-Anthony results were 52% for ROI only, 41% for both, and 6% for RI only.)

Exhibit V Methods used to evaluate investment centers

ROI has several advantages that may explain its wide use:

1. ROI makes unlikes comparable. Since return is a ratio, it “normalizes” for divisions or companies differing in investment base size. For example, one can meaningfully compare a large steel company’s ROI with a small steel company’s ROI.

2. ROI, being a percentage-return measurement, is consistent with how companies measure the cost of capital. For example, one can say that a company with an 8% ROI (before capital costs) is faring poorly if its cost of capital is 10%.

3. ROI is useful for people outside the company. The ROI measure, unlike residual income, can be calculated by outside financial analysts for purposes of evaluating the economic performance of a company and for making intercompany performance comparisons. Many top executives thus want their division managers to focus on ROI performance, since outsiders (especially potential investors and their advisers) are focusing on it.

In our opinion, however, ROI has potential drawbacks as a measure of an investment center’s performance. Two of these are:

1. The same decisions that increase a center’s ROI may decrease its economic wealth. For example, in an investment center whose current ROI is 25%, its overall ROI can be increased by disinvesting in an asset whose ROI is 20%; yet, if the cost of invested funds is less than 20%, the absolute amount of the investment center’s profit after taking account of capital costs will decrease.

2. Given an investment opportunity whose ROI is above the cost of capital but below an investment center’s current ROI, the center’s manager may forgo this opportunity, since the investment, while economically sound, will lower the center’s ROI below the current level.2

The use of residual income as a measure deals with these drawbacks, both of which pertain to investments—actual or prospective—whose return falls between the cost of invested capital and the centers average ROI. If an investment center’s performance is measured by residual income (i.e., if its profit in excess of capital costs is shown in absolute dollars rather than in percentages), then investments generating a profit in excess of capital costs will increase RI and be attractive to the center’s manager.

In the first case, then, the manager will probably not be motivated to disinvest; in the second, he will probably want to make the investment.

For these reasons, many authorities agree that RI is superior to ROI as a measure of investment center financial performance. Why, then, the great preference of companies for ROI?

First, because ROI makes investment center performance measurement congruent with the way “outsiders” measure a company’s overall economic performance.

Second, because other authors, including John Dearden and David Solomons, have, in our opinion, exaggerated the practical importance of ROI’s giving a manager the “wrong signal” on investments whose return is between the cost of capital and an investment center’s current ROI.3

In the case of using disinvestments to increase overall ROI (which is a more “visible” case to an investment center manager’s superiors than is the manager’s suppressing ideas or proposals for new investments), our informal interviews with investment center managers lead us to believe that these people are unwilling to risk having to explain to a superior why they have eliminated economically sound assets.

Put another way, although their companies may formally use ROI instead of RI to measure division performance, managers realize that assets generating a positive RI should not be scrapped just because ROI measurement has conceptual flaws.

In the case of suppressed attractive investments also, ROI’s wrong signal probably is not so deleterious in practice because a sizable amount of new investments that a division makes are nondiscretionary in nature. It is not uncommon today for 20% to 40% of a company’s capital budget to be earmarked for projects that are not justified on economic grounds by using discounted cash flow or payback calculations—for instance, projects related to meeting pollution control or OSHA requirements, replacing or upgrading administrative facilities, and improving employee cafeterias.

Since nondiscretionary projects require capital but do not recover capital costs, they represent a capital cost burden to be earned by those projects that are evaluated on economic grounds. For the corporation to recover all of its capital costs, then, the discretionary projects must recover their own cost of capital and the capital costs of nondiscretionary projects. Viewed in this light, investments whose return is only between the cost of capital and the current average ROI are probably not desirable after all.

Another reason the use of ROI does not, as alleged, seem to discourage investments that would pull down an investment center’s overall average ROI has to do with the nonfixed nature of a division’s ROI target from year to year. Other writers, including the ones we have mentioned, seem to assume that if an investment center’s current ROI is, say, 25%, the center’s manager will automatically discard any investment proposal with an ROI below 25% on the ground that the center’s future ROI target will necessarily be at least 25%.

We question the validity of that assumption.

We feel that if this manager proposes a 21% project, for example, and if that project is approved and implemented, then the center’s future ROI targets will be accordingly adjusted downward.

This argument, in turn, is based on how investment center ROI targets are apparently set. Other authors seem to assume that top management sets an ROI target for an investment center (at a level at least as high as the center’s current ROI) and then the center tries to develop budgeted sales, profits, and assets consistent with this target.

Our opinion is that, at least in many companies, the sequence is: (1) budget sales, (2) budget profit, (3) budget assets, and (4) on approval of steps 1 through 3, divide budgeted profit by budgeted assets to arrive at budgeted ROI. The key approvals are in steps 2 and 3, not in step 4.

Thus, in the preceding example, the approved 21% project would be reflected in steps 2 and 3 and “automatically” reflected in 4, even though the resultant budgeted investment center ROI may be lower than the center’s ROI for the current year.

At this point, we want to emphasize that we are not saying that the conceptual weaknesses of ROI that have been so articulately and persuasively pointed out by other authors do not exist. Rather, we are saying: the use of ROI has increased significantly during the very years its weaknesses have been publicized; we feel that designers of financial control systems do not intentionally design major flaws into these systems, therefore, ROI’s conceptual weaknesses must not be felt to lead in practice to poor decisions.

Defining Profit & Investment

Both profit center and investment center measurements require specific definitions of what constitutes profit and what constitutes investment.

Elements in the profit calculation: One alternative in defining profit is to calculate it according to the same generally accepted accounting principles (GAAP) as are used to calculate net income in the company’s published financial statements. But this, while an obvious alternative, is indeed only one option. Variations from GAAP for internal reporting can include any of those listed in the lower portion of Exhibit VI.

Exhibit VI Determination of profit

The upper portion of this exhibit shows that two out of five companies do define profit to be the same as reported net income. Many companies evidently feel that a division manager should easily be able to relate the division’s profitability to the total net income that the corporation reports to its shareholders and other interested outside parties.

Exhibit VI reflects the fact that, in the companies which do define profit differently from net income, the variations fall almost entirely in the category of eliminating expenses over which a division manager has no direct control—taxes, interest on corporate debt, and allocated headquarters expenses.

This finding is consistent with the notion espoused by Solomons and others that for purposes of evaluating a division manager’s performance (as opposed to the performance of the division as an economic entity) the measure used should be exclusive of costs not significantly influenced by the manager.4

A significant finding in Exhibit VI is that most companies attribute interest expense to individual investment centers. Charging interest expense to investment centers serves to remind managers that invested funds are not a free resource. The reader should note, however, that an investment center’s fair share of corporate interest expense understates the true cost of the center’s capital, because interest is a charge for only the debt portion of capital.

Conceptually, an investment center should also be charged for equity capital. However, because the cost of equity capital cannot be objectively measured, both shareholder income statements and most companies’ divisional income statements overstate “true” profitability. (In the 225 companies not charging divisions with interest expense, the overstatement of profit is, of course, greater.)

It is not likely that many companies will begin imputing a total cost of their divisions’ capital to investment centers until the cost of equity is recognized as an expense by generally accepted (external) accounting principles.

Components of the investment base: Companies using either ROI or RI must decide how to define investment. The three concepts outsiders use most when calculating ROI are total assets, invested capital (total assets less current liabilites, which is equivalent to long-term liabilities plus owners’ equity), and owners’ equity (equivalent to assets minus all liabilities).

In our experience, return on owners’ equity is used primarily by outside analysts in their advisory role to potential or present shareholders. Return on invested capital is of more concern to people in top management—especially the chief financial officer—who are monitoring the return earned on all funds committed long-term to the corporation, whether by creditors (including bondholders) or shareholders.

Return on total assets employed seems to be the most relevant of the three common ROI calculations for investment center evaluation because an investment center manager’s overriding responsibility is to use the center’s assets as efficiently as possible. From this manager’s viewpoint, the liabilities and equity side of the balance sheet may be of little interest, for the manager does not know—or care—what mix of current liabilites, long-term debt, and owners’ equity was used to finance his division’s assets.

However, if an investment center is relatively large, the division’s manager may in fact have some impact on how his center is financed. For example, the division may have the authority to decide when individual accounts payable shall be paid or may have its own line of short-term credit, especially if it is a wholly owned but legally separate corporation.

In our opinion, subtracting the controllable current liabilities from the division’s assets has merit when one is defining the investment base. Of course, such a subtraction alters the investment definition from that of total assets to one more akin to invested capital.

There are also investment-base definitional questions on the asset side. For example, if an investment center’s cash is centrally controlled, or if headquarters determines credit policy and has collection responsibility, then an argument can be made for omitting cash or accounts receivable from the defined investment base, or for including those accounts at arbitrary amounts (e.g., cash equal to 10 days’ expenses, receivables equal to 45 days’ sales).

Also at issue here is an analogy to the problem of allocated headquarters expense when one is defining profit: the question of how to treat shared assets when calculating a division’s investment base.

Exhibits VII and VIII show how 459 respondents defined the asset base (investment base) for their investment centers. Nearly all companies included receivables, inventories, and fixed assets used solely by the investment center in calculating the centers investment base. Fewer than half felt it useful to include allocations of shared facilities in their calculations, and about half deducted external payables and other current liabilities.

Exhibit VII Assets included in the investment base

Exhibit VIII Treatment of liabilities

The findings in Exhibits VII and VIII are quite similar to those of Mauriel and Anthony, except ours show significantly fewer companies allocating headquarters assets and more deducting external payables.

The results in Exhibits VII and VIII also show that in measuring the financial performance of investment centers, most companies include in the investment base only those items significantly influenceable by the center’s manager. This finding in turn indicates that the ROI and the RI measures are being used primarily to evaluate the division’s managerial performance rather than the division’s economic performance, since the latter purpose requires full allocation of all balance sheet items and expenses to profit-producing subunits of the organization.

Since managers are usually evaluated more frequently than is the economic health of their investment centers, and since including non-controllable items in an investment center’s asset base may impair motivation, omitting non-controllable items from the investment base is a sound practice.

Asset valuation: What asset categories to include and what liabilities to deduct when one is defining investment is a question quite apart from how to value the assets that are already included. Asset valuation is an issue especially with respect to plant and equipment, where the obvious valuation alternatives are gross book value (undepreciated historical costs), net book value, and replacement costs.

We found that most of the companies surveyed value plant and equipment at net book value (85%) and use the same depreciation method they do for shareholder reporting (92%), as shown in Exhibit IX. In their earlier study Mauriel and Anthony found 73% of their respondents using net book value, so its use has increased over the years.

Exhibit IX Valuation of plant and equipment

The use of net book value and the shareholder reporting depreciation method reflects companies’ desire to make internal management reports congruent with external reports. However, the use of net book value does have conceptual flaws.5

The most prominent of the flaws is that, other things being equal, ROI or RI will increase solely with the passage of time as depreciation reduces the investment base. The concern about this phenomenon is that it may discourage economically sound divisional investments in new fixed assets. Replacement cost valuation and the annuity method of depreciation may be used jointly to overcome this net book value flaw, but companies probably view this combination approach as too complicated to be practical.

As evidence of this view, note that in Exhibit IX only 2% of the respondents use replacement costs to value plant and equipment. (Mauriel and Anthony found only 3% of their respondents using replacement costs in 1965.) Most of the managers we have spoken with do not know what annuity depreciation is.

Despite its flaws, net book value is superior to gross book value. The latter creates too great a motivation to scrap older assets that, while still productive, are often idle because of the presence of newer, more efficient assets. Moreover, there is no reason why the phenomenon of ROI or RI increasing with the passage of time cannot be allowed for when one is budgeting an investment center’s future performance. Thus, of the currently practicable fixed asset valuation alternatives, most companies are using the best one.

Lease valuation: One final investment-base definition issue we researched was the treatment of leased equipment or facilities. (Our questionnaire was mailed in February 1977, only about three months after issuance of FASB Statement No. 13 on accounting for leases in shareholder reports.) As the reader can see in Exhibit X, only 34% of our respondents were capitalizing leases when they calculated the investment base for a division. (This number is up from 13% in the Mauriel-Anthony study.)

Exhibit X Treatment of leases

Apparently, then, most companies are not concerned about investment center managers trying to boost ROI by leasing assets that, seen from a purely economic standpoint, should be purchased. Companies may not worry about this because they know that thorough justification procedures and analyses are required before assets can be leased or because they have an overall policy of preparing investment center financial statements in accord with the procedures used for the published corporate statements.

Target ROIs: Exhibit XI shows how our respondents determined ROI budgets for their investment centers. Almost two-thirds set these targets by analyzing the profit potential of the division. This procedure is appropriate since, although the company as a whole has a single profit potential-risk profile, each division considered as a separate entity has its own profile, related to its line of business, the condition of its fixed assets, and so on.

Exhibit XI Methods of establishing a target ROI

Giving every investment center the same ROI target is an unsound procedure, unless the company is in a single line of business and its investment centers are geographical divisions having similarly aged assets. Only 7% of the respondents were using this method. This small percentage reflects both (1) companies’ increased sophistication in the use of ROI controls and (2) increased diversity in large corporations. (Of the Mauriel-Anthony respondents, 38% were using this single-target method in 1965.)

We were surprised to find that almost one-fourth of the respondents using ROI do not set ROI targets for their investment centers. There are two possible reasons for this. First, not setting a target may be a technique to avoid overemphasizing the ROI percentage at the expense of economic profit. Second, these companies’ top executives may believe that better performance is attained when a division manager is told “do your best” than when a specific ROI target is negotiated for that division.

RI capital charges: One of the advantages of using residual income is that the format of the calculation enables a company to apply a different capital charge rate to each asset category. These differing rates can reflect the fact that some categories of assets are riskier than others (e.g., investment in a joint venture versus investment in accounts receivable) or that some assets can be financed at lower rates than others (e.g., materials inventories can be thought of as being financed primarily by interest-free vendor payables).

Despite this conceptual advantage, only 26 (19%) of the 137 companies using RI in fact had used different capital charge rates for different types of assets (see Exhibit XII). Of these 26, the majority were using the same capital charge for a given asset type in all investment centers.

Exhibit XII Capital charge rates for residual income

Evidently, these companies believe (at least implicitly) that an investment in, say, inventory is equally risky and/or equally expensive to finance in all divisions (a questionable assumption in many diversified companies). Differences in divisions’ capital charges in these companies are therefore solely the result of the divisions’ having different asset mixes.

Intracompany Comparisons

Our final research area dealt with how investment centers’ relative performances are compared or ranked. As shown in Exhibit XIII, no one approach dominated, and many companies used more than one method.

Exhibit XIII Methods used to evaluate relative investment center performance

In our opinion, the most equitable way to determine such rankings is to compare actual with budgeted performance. Such a budgeted-versus-actual comparison tends to reflect the differing profitability potentials of divisions and does not unduly penalize a manager whose division is in an industry with low profitability.

As that comment suggests, this approach focuses more on an investment center manager’s performance than on the relative economic attractiveness of the center as a business entity, especially if factors outside the manager’s control are allowed for when the ROI or RI budget is set.

Note that this actual-versus-budget approach was not the most-used method for financial performance comparisons in either ROI or residual income companies. We attribute this finding in part to the inadvertently ambiguous wording of the research question: a ranking of investment centers’ relative economic performance (as opposed to managerial performance) should be based on achieved ROI percentage or earned RI dollars.

However, we have also had many investment center managers tell us that they feel that their superiors at corporate headquarters do not adequately make this economic-managerial performance distinction when they evaluate these investment center managers’ performance.

Also of note in Exhibit XIII is the use as ranking criteria of (1) RI divided by investment, (2) RI as a percentage of sales, and (3) profit as a percentage of sales. Each of these has potential weaknesses. Calculating RI as a percentage of investment reflects an apparent desire not only to gain the benefits of RI but also to have the interdivisional comparison property of a percentage (which normalizes for size differences).

That calculation is acceptable for ranking purposes; but if an RI percentage is used as a target for an investment center’s manager, then all of the potential disadvantages of the ROI percentage appear, and little is gained by using RI instead of ROI.

The use of RI as a percentage of sales for interdivisional comparisons has no conceptual basis unless the divisions have equal investment turnover ratios. The isolated statement, “Division A earned 5% on sales while B earned 7%,” tells one nothing about the relative economic performance of these two divisions. If one adds that, “A’s investment turnover was two times and B’s was only one time,” then one can conclude that A’s economic performance was superior to B’s.

We were also interested that 29% of the respondents said they do not rank their investment centers’ financial performance. While such a ranking is not necessary or even desirable for evaluating division managers, a company viewing its investment centers collectively as a portfolio might be expected to mark these investments’ performance at least annually, particularly for purposes of identifying possible candidates for disinvestment (i.e., getting out of a low-profitability business).

The Best Measure

Despite the pitfalls of implementing investment center financial measures by using ROI and shareholder-report accounting principles, most companies with investment centers seem to feel that the best approach is to use ROI and to make their profit and investment definitions and valuations quite similar to those in their published financial statements.

It is our premise that most designers of financial control systems are aware of the conceptual flaws in the GAAP-ROI approach, and that these financial managers do not believe that these flaws are more than hypothetical. This belief is why larger companies’ use of ROI and GAAP for investment center financial control has increased over the past 12 years.

However, a note of caution is in order. The potential does exist for ROI as commonly implemented to motivate some investment center managerial decisions which improve the measured divisional ROI yet which are not in the best interests of the company as a whole. Both a company’s financial managers and its top corporate executives should be familiar with these potential ROI pitfalls and should be convinced that these pitfalls are indeed hypothetical and not real in their company.

1. See John J. Mauriel and Robert N. Anthony, “Misevaluation of Investment Center Performance,” HBR March–April 1966, p. 10.

2. For a more complete treatment of the conceptual flaws of ROI and their possible adverse consequences, see John Dearden’s article, “The Case Against ROI Control,” HBR May–June 1969, p. 124.

3. Ibid., pp. 127, 130; see also David Solomons, Divisional Performance: Measurement and Control (Homewood, Illinois: Richard D. Irwin, 1968), p. 64.

4. Solomons, Divisional Performance, pp. 53, 60.

5. For a full treatment of these flaws, see Dearden, “Case Against ROI,” Mauriel and Anthony, “Misevaluation,” and John Dearden, “Problem in Decentralized Profit Responsibility,” HBR May–June 1960, p. 79.

A version of this article appeared in the May 1978 issue of Harvard Business Review.

What is the major advantage of residual income as a performance measure?

Which is an advantage of using residual income RI over return on investment ROI?

Why is return on investment better than divisional income as a measure of performance for an investment center?

Which is better RI or ROI?