Measuring investment center performance I: return on investment and residual income

An investment center manager controls both profit and the assets used to earn it. Evaluating such a manager on profit alone would ignore the capital tied up: a division earning $3,000,000 on $20,000,000 of assets is doing better than one earning the same profit on $40,000,000. Measures for investment centers therefore relate profit to capital. ACCT-102 introduced the two classic measures, return on investment and residual income, in a few paragraphs. This lesson works through them in depth: how they are calculated, how they decompose, the choices that change their values, and above all the behavior they encourage. Lesson 5 extends the analysis to economic value added and the problem of measuring assets.

Return on investment

Return on investment is operating income divided by invested capital. It is widely used because it is easy to understand, comparable across divisions of different sizes and with external benchmarks such as industry returns, and consistent with how investors think about returns.

The DuPont method, developed at the DuPont company early in the twentieth century, splits ROI into two drivers. Return on sales, or margin, measures how much operating income each dollar of sales generates. Investment turnover measures how many dollars of sales each dollar of invested capital supports. ROI is their product, so a division can improve ROI by raising margin, by raising turnover, or both.

# ROI, DuPont analysis and residual income for Cormorant's divisions, 2026 (thousands)
rate = 0.10   # required rate of return set by head office
div = {"Appliances": (40_000, 3_000, 20_000),
       "Components": (24_000, 1_800, 15_000),
       "Services":   (12_000,   900,  4_000)}
print(f"{'':<11}{'Margin':>8}{'Turnover':>9}{'ROI':>7}{'RI':>7}")
for name, (sales, oi, assets) in div.items():
    margin, turnover = oi / sales, sales / assets
    roi = margin * turnover
    ri = oi - rate * assets
    print(f"{name:<11}{margin:>8.1%}{turnover:>9.2f}{roi:>7.1%}{ri:>7,.0f}")

             Margin Turnover    ROI     RI
Appliances     7.5%     2.00  15.0%  1,000
Components     7.5%     1.60  12.0%    300
Services       7.5%     3.00  22.5%    500

All three divisions earn the same 7.5 percent margin, so the differences in ROI come entirely from turnover. Services, an asset-light business, turns its $4,000,000 of assets three times a year and earns 22.5 percent. Appliances turns its assets twice and earns 15 percent. Components, with heavy machinery and tooling, turns its assets 1.6 times and earns 12 percent. Residual income at a 10 percent required rate is $1,000,000 for Appliances, $500,000 for Services and $300,000 for Components.

Measuring income and capital

ROI looks precise, but its value depends on choices about both the numerator and the denominator. For income, most companies use operating income before interest and taxes, because division managers usually do not control financing and tax decisions, which are made centrally. Some use controllable margin (lesson 2) to evaluate the manager and segment margin to evaluate the division. Including allocated head-office costs lowers ROI but adds no information about the manager.

For capital, the choices include total assets available, total assets employed (excluding idle assets that head office has told the division to hold), and net assets, which deduct current liabilities from total assets on the grounds that suppliers' credit is a free source of funds the division manager can influence. Using net assets encourages managers to stretch payables, which may or may not be good for the company. Most companies use an average of opening and closing balances, because income is earned over the whole year; using opening balances is simpler but lets managers time investments near year end. Lesson 5 examines the further question of whether assets should be measured at gross book value, net book value or current cost.

The fundamental problem with ROI

ROI encourages managers to maximize a ratio, which is not the same as maximizing value for the company. A manager evaluated on ROI will reject any project whose return is below the division's current ROI, even if that return is above the company's cost of capital. The project would add value for the company, but it would lower the division's average return. Conversely, a manager may dispose of assets earning more than the cost of capital but less than the division's current ROI, raising the ratio while destroying value. This is the classic goal congruence problem with ROI, and it is most severe in the highest-performing divisions.

# Would each division manager accept the project? ROI versus residual income
rate = 0.10
divisions = {   # current operating income, assets; project income, investment
    "Appliances": (3_000, 20_000, 260, 2_000),
    "Services":   (  900,  4_000, 180, 1_000),
    "Components": (1_800, 15_000,  95, 1_000),
}
for name, (oi, a, pi, pa) in divisions.items():
    roi_now, roi_new = oi / a, (oi + pi) / (a + pa)
    ri_change = pi - rate * pa
    print(f"{name}: project ROI {pi/pa:.1%}; division ROI {roi_now:.1%} -> {roi_new:.2%}; "
          f"RI change {ri_change:+,.0f}")
    by_roi = "accept" if roi_new > roi_now else "reject"
    by_ri = "accept" if ri_change > 0 else "reject"
    print(f"   ROI-evaluated manager would {by_roi}; RI-evaluated manager would {by_ri}")

Appliances: project ROI 13.0%; division ROI 15.0% -> 14.82%; RI change +60
   ROI-evaluated manager would reject; RI-evaluated manager would accept
Services: project ROI 18.0%; division ROI 22.5% -> 21.60%; RI change +80
   ROI-evaluated manager would reject; RI-evaluated manager would accept
Components: project ROI 9.5%; division ROI 12.0% -> 11.84%; RI change -5
   ROI-evaluated manager would reject; RI-evaluated manager would reject

The Appliances project returns 13 percent, above the 10 percent required rate, but below the division's current 15 percent. Taking it would lower division ROI to 14.82 percent, so a manager evaluated on ROI would reject it, while residual income rises by $60,000 and an RI-evaluated manager would accept it. Services shows the same conflict even more sharply: an 18 percent project is rejected under ROI because the division already earns 22.5 percent. In Components the conflicts disappear: its 9.5 percent project is below both the division's ROI and the required rate, so both measures reject it, correctly.

Residual income

Residual income is operating income minus a capital charge equal to the required rate of return times invested capital. It measures the income a division earns above the minimum its capital must earn.

Because residual income increases whenever a project earns more than the required rate, it solves the goal congruence problem in the example above. Any project with a positive net present value at the required rate will, over its life, add to residual income in present value terms, a link that lesson 5 develops. Residual income also lets head office apply different required rates to divisions with different risks, so that a riskier division must earn more to show a positive result.

Residual income has its own weaknesses. It is an absolute amount, so larger divisions tend to show larger residual income, making comparisons between divisions of different sizes difficult; many companies compare each division's residual income with its own target rather than with other divisions. It depends on the required rate, which head office must set and justify. And it shares ROI's dependence on accounting measures of income and assets, with all the problems of depreciation and historical cost that lesson 5 discusses.

Setting targets and evaluating trends

Whichever measure is used, the performance target matters as much as the measure. Comparing Components' 12 percent ROI with Services' 22.5 percent says little, because the businesses have very different capital intensity. A fairer evaluation compares each division with its own budgeted ROI or residual income, with its own past trend, and with external benchmarks for similar businesses. Budgeted targets should reflect the assets the division is expected to hold and the investments it is expected to make, so that a manager who undertakes an approved expansion is not penalized for the temporary fall in ROI it causes.

Many companies use several measures together. A division might be evaluated on residual income against budget for the bonus, with ROI reported for comparison with industry, and with non-financial measures (lessons 9 and 10) to guard against short-term thinking. The single-measure systems examined in this lesson are useful teaching devices, but few well-run companies rely on one number.

Short-term thinking

Both ROI and residual income measure performance one period at a time. A manager can raise either in the short run by cutting discretionary spending on research, training, maintenance or advertising, which reduces current expenses but may harm future results. A manager close to retirement or expecting a transfer has a particular incentive to do so. Investments with long payback periods, which reduce current income through depreciation and start-up costs before they earn returns, look unattractive. Remedies include evaluating managers over several years, adding non-financial measures of the health of the business, reviewing discretionary spending against plans, adjusting the accounting measures (lesson 5) and paying part of compensation in deferred or long-term forms (lesson 11).

Evidence from practice

Surveys over several decades have found that most large companies use ROI or a similar return measure for divisional evaluation, and that a substantial minority use residual income or economic value added, often alongside ROI. Research on the behavioral effects has documented the underinvestment problem predicted above, particularly in high-ROI divisions, and shown that companies switching from ROI to residual-income-type measures tend to dispose of more underperforming assets and invest more selectively. ACCT-421 returns to these measures in the context of strategic performance management and capital investment decisions.