The Incremental IRR Decision Rule

Why Compare Incremental Returns?

When projects are mutually exclusive, only one can be selected. A project with the highest standalone IRR is not automatically the best choice: IRR measures a return relative to that project's own initial investment, while the decision may be whether the extra investment in a larger project earns enough to justify choosing it instead of a smaller one. The incremental IRR decision rule answers that question.

The method starts by ranking alternatives by their initial investment. For a pair of alternatives, subtract the lower-investment project's cash flow from the higher-investment project's cash flow at each date. The resulting incremental cash flows represent the additional investment and additional benefits associated with upgrading. Their IRR is compared with the required return, denoted by $k$.

Definition: For a higher-investment project $H$ and a lower-investment project $L$, the incremental cash flow at time $t$ is $\Delta CF_t=CF_{H,t}-CF_{L,t}$. The incremental IRR, $r_{\Delta}$, is a rate that makes the present value of these incremental cash flows equal to zero: $$ \sum_{t=0}^{n}\frac{\Delta CF_t}{(1+r_{\Delta})^t}=0. $$ When the incremental cash flows are conventional and have a unique IRR, choose the higher-investment project if $r_{\Delta}>k$; choose the lower-investment project if $r_{\Delta}<k$. At equality, the alternatives have the same NPV at $k$.

A conventional incremental cash-flow stream has an initial negative investment followed by nonnegative receipts, with at least one positive receipt. This pattern gives a single meaningful incremental IRR under the usual project assumptions. The decision rule also assumes comparable alternatives are evaluated over the same relevant period, that the cash flows are measured consistently, and that the same required return is appropriate for both.

Why the Incremental Rule Works

The rule is a comparison of NPVs expressed as a rate threshold. At a common required return $k$, the difference between the projects' NPVs is exactly the NPV of the incremental cash flows:

NPV_H(k)-NPV_L(k) =\sum_{t=0}^{n}\frac{CF_{H,t}-CF_{L,t}}{(1+k)^t} =\sum_{t=0}^{n}\frac{\Delta CF_t}{(1+k)^t}.

For a conventional increment, write its initial outlay as $-I$, where $I>0$, and its later receipts as $R_t\geq 0$. Its NPV at rate $q$ is $-I+\sum_{t=1}^{n}R_t/(1+q)^t$. Each positive receipt has a present value that declines as $q$ increases. More formally, wherever the derivative is defined for $q>-1$,

\frac{d}{dq}\left(-I+\sum_{t=1}^{n}\frac{R_t}{(1+q)^t}\right) =-\sum_{t=1}^{n}\frac{tR_t}{(1+q)^{t+1}}<0.

Thus the incremental NPV decreases with the discount rate. It is zero at $r_{\Delta}$, so it is positive at any required return below $r_{\Delta}$, and negative at any required return above it. Because that incremental NPV is $NPV_H(k)-NPV_L(k)$, an incremental IRR above $k$ means the higher-investment project has greater NPV at the hurdle rate.

Applying the Rule in Order

For several mutually exclusive alternatives, the comparison is made in increasing order of initial investment. First check the smallest investment against the no-investment alternative if doing nothing is available. Then assess whether each increase in investment is worthwhile. The comparison is about the incremental cash flows, not a ranking of standalone IRRs.

Rank by initial investment.
List the feasible alternatives from the smallest initial outlay to the largest. Do not rank them by standalone IRR.

Form the increment.
For an upgrade, subtract the lower-investment alternative's cash flow from the higher-investment alternative's cash flow at every date, including time 0.

Check the cash-flow pattern.
Confirm that the incremental stream is conventional and has a unique IRR before relying on the simple rule.

Compare with the hurdle.
Upgrade when the incremental IRR exceeds the required return; do not upgrade when it is below the required return.

Continue through the ranked alternatives.
Use the currently preferred alternative as the base for the next feasible upgrade, and confirm the base alternative itself is acceptable relative to doing nothing.

Worked Comparisons

Worked Example: The Larger Project Has a Lower Standalone IRR

A firm can install either a compact or an expanded sorting line. The compact line has cash flows of $-\$40{,}000$ at time 0 and $+\$60{,}000$ at the end of year 1. The expanded line has cash flows of $-\$70{,}000$ at time 0 and $+\$96{,}000$ at the end of year 1. The required annual return is $k=0.15$. Assume all cash flows occur as stated and the projects are mutually exclusive. Which line should the firm select?

Method. Rank by initial investment: the compact line is the lower-investment alternative. Form the expanded-minus-compact increment:

\Delta CF_0=-70{,}000-(-40{,}000)=-30{,}000,\qquad \Delta CF_1=96{,}000-60{,}000=36{,}000.

The incremental IRR solves $-30{,}000+36{,}000/(1+r_{\Delta})=0$. Therefore $1+r_{\Delta}=36{,}000/30{,}000=1.2$, so $r_{\Delta}=0.20$. Since $0.20>0.15$, the additional investment earns more than the required return; the rule favors the expanded line.

Check at the required return. The incremental NPV is $-30{,}000+36{,}000/1.15=-30{,}000+31{,}304.35=\$1{,}304.35$, rounded to the nearest cent. It is positive, as expected. For context, the compact line's standalone IRR is $60{,}000/40{,}000-1=0.50$, while the expanded line's is $96{,}000/70{,}000-1\approx0.3714$. Those standalone rates do not decide the upgrade: the incremental comparison does. A common mistake is to select the compact line just because its standalone IRR is higher. The incremental NPV check confirms that this would miss value at the stated hurdle.

Worked Example: Evaluating Successive Increases in Scale

A hypothetical logistics company is choosing one of three mutually exclusive depot designs. Their time-0 investments and year-1 receipts, in thousands of dollars, are: Design A, $-30$ and $+41$; Design B, $-60$ and $+90$; Design C, $-90$ and $+134$. The annual required return is $k=0.30$. Assume no other cash flows and conventional project cash flows. Select the preferred design, provided the smallest design is acceptable relative to doing nothing.

Method. The investment amounts are $30$, $60$, and $90$, so the order is A, B, C. Design A's IRR is $41/30-1=0.3667$ to four decimal places. Since $0.3667>0.30$, A is acceptable relative to no investment.

Now compare B with A. The incremental cash flows are $-60-(-30)=-30$ at time 0 and $90-41=49$ at year 1. The incremental IRR is $49/30-1=0.6333$ to four decimal places. Since $0.6333>0.30$, B is preferred to A.

Next compare C with the currently preferred design, B. The incremental cash flows are $-90-(-60)=-30$ at time 0 and $134-90=44$ at year 1. Thus $r_{\Delta}=44/30-1=0.4667$ to four decimal places. Since $0.4667>0.30$, C is preferred to B. The resulting choice is Design C.

Check. At the hurdle rate, the B-over-A incremental NPV is $-30+49/1.30=-30+37.6923=7.6923$ thousand dollars. The C-over-B incremental NPV is $-30+44/1.30=-30+33.8462=3.8462$ thousand dollars. Both are positive, consistent with both upgrades passing the hurdle. The common mistake is to stop after finding B's incremental IRR over A, or to compare C directly with A while ignoring the ordered upgrade process. The calculations show that each successive increase is worthwhile at the specified return.

Worked Example: Incremental Cash Flows Arrive After Two Years

A utility is considering a standard pump station costing $\$100{,}000$ now and paying $\$144{,}000$ at the end of year 2, or an advanced station costing $\$140{,}000$ now and paying $\$200{,}000$ at the end of year 2. There are no year-1 cash flows. The annual required return is $k=0.15$. Assume the stated year-2 receipts are certain for purposes of this comparison. Should the utility upgrade?

Method. Subtract the standard station's cash flows from the advanced station's cash flows. The incremental cash flows are $-\$40{,}000$ at time 0, $0$ at year 1, and $+\$56{,}000$ at year 2. The IRR equation is $-40{,}000+56{,}000/(1+r_{\Delta})^2=0$. Therefore $(1+r_{\Delta})^2=56/40=1.4$, and $r_{\Delta}=\sqrt{1.4}-1\approx0.1832$, to four decimal places.

Since $0.1832>0.15$, the incremental IRR rule favors the advanced station. Check this against the incremental NPV at the required return: $-40{,}000+56{,}000/(1.15)^2=-40{,}000+56{,}000/1.3225\approx-\$40{,}000+\$42{,}344.05=\$2{,}344.05$, rounded to the nearest cent. This is positive, agreeing with the decision. The common mistake is to treat the receipt as if it arrived in year 1 and compute $56/40-1=0.40$. The receipt arrives at the end of year 2, so the rate must be compounded for two periods.

Limits and Common Exam Traps

The incremental IRR rule is not a substitute for checking the cash-flow pattern or the acceptance of the base alternative. In particular, a difference between two conventional projects need not itself be conventional. If incremental cash flows change sign more than once, the increment may have multiple IRRs or no nonnegative IRR. In that case, comparing a selected IRR with the hurdle can give an ambiguous or misleading answer. Use the projects' NPVs at the required return to establish which has greater value, and interpret any IRR roots with care.

Other frequent errors include subtracting in the wrong direction, omitting a time period, or treating an incremental IRR as the standalone IRR of the larger project. Keep the direction explicit: higher-investment project minus lower-investment project. For an upgrade, the time-0 increment should ordinarily be negative; the later increments show what additional cash flows are received for that additional outlay. If the upgrade changes cash flows at several dates, subtract at every date rather than comparing only total receipts.

Key takeaway: For mutually exclusive projects, standalone IRRs do not establish the preferred scale. Rank projects by investment, form the higher-minus-lower incremental cash flows, and compare the incremental IRR with the required return only when the increment has a conventional pattern and a unique IRR. At the hurdle rate, the sign of incremental NPV verifies the decision.

Check Your Understanding

For each problem, assume the stated cash flows are complete, annual periods are used, and the required return applies to both alternatives unless stated otherwise.

Project A costs $\$25{,}000$ now and returns $\$35{,}000$ in one year. Project B costs $\$40{,}000$ now and returns $\$56{,}000$ in one year. At $k=0.25$, which project is preferred by the incremental IRR rule? (Answer: B; the B-over-A incremental IRR is $0.4000$, above the hurdle.)
Project A costs $\$50{,}000$ and returns $\$70{,}000$ in one year. Project B costs $\$80{,}000$ and returns $\$104{,}000$ in one year. At $k=0.15$, which is preferred? (Answer: A; the B-over-A incremental IRR is $34/30-1=0.1333$, below the hurdle.)
Project A has cash flows $-\$40{,}000$ now and $+\$54{,}000$ in one year; Project B has $-\$65{,}000$ now and $+\$88{,}000$ in one year. At $k=0.30$, which is preferred? (Answer: B; the incremental IRR is $34/25-1=0.3600$, above the hurdle.)
Three alternatives, in thousands of dollars, have cash flows: A, $(-20,+29)$; B, $(-50,+68)$; C, $(-80,+104)$, with the second entry at year 1. At $k=0.25$, which should be selected if A is acceptable relative to no investment? (Answer: B; B over A has incremental IRR $39/30-1=0.3000$, while C over B has $36/30-1=0.2000$.)
Project A costs $\$60{,}000$ now and returns $\$90{,}000$ at the end of year 2. Project B costs $\$100{,}000$ now and returns $\$145{,}000$ at the end of year 2. There are no year-1 cash flows. At $k=0.16$, should the firm upgrade from A to B? (Answer: Yes; the incremental IRR is $\sqrt{55/40}-1\approx0.1726$, above the hurdle.)

What You'll Learn