Payback, and the rest

Surveys of what companies actually use put payback near the top, decades after every textbook declared it inadequate. That gap is worth taking seriously rather than dismissing.

Payback period

How long until the cumulative cash flows recover the initial outlay.

Accept if payback is shorter than some threshold — three years, five years, whatever the firm has settled on.

What is wrong with it

It ignores everything after the cut-off. A project paying back in two years and then collapsing scores better than one paying back in four and running profitably for twenty.

It ignores the time value of money. A rupee in year three counts the same as one in year one.

The threshold is arbitrary. Nothing derives "three years" from anything.

Work the problem:

	Payback	NPV at 10%
A: ₹5 cr × 3 years	2.0 years	₹2.43 cr
B: ₹2 cr × 10 years	5.0 years	₹2.29 cr

Here they are close — but extend B to 15 years and its NPV rises to ₹5.2 crore while its payback does not move at all. Payback is blind to everything beyond the moment the money comes back, which is where most of a long project's value lives.

What is right about it

It measures something NPV does not: how long the firm's capital is exposed.

That is a real concern, and three situations make it the binding one:

Liquidity constraints. A firm that cannot survive four years of outflow needs to know when the money returns, regardless of what the project is worth in year twelve.

Forecast decay. Cash flows five years out are guesses. A short payback means the decision rests on the part of the forecast that is least speculative — which is a genuine epistemic argument, not a failure of sophistication.

Political and regulatory risk. Where rules can change, a long payback is exposure to something no discount rate properly captures.

So: payback is a bad appraisal rule and a reasonable risk screen. Used as a constraint alongside NPV — "positive NPV and payback under five years" — it encodes a real preference. Used alone, it systematically rejects the long-lived projects that build a business.

Discounted payback

The same, but with discounted cash flows. It fixes the time-value objection and leaves the bigger one — ignoring everything after the cut-off — exactly where it was. A modest improvement, not a solution.

Profitability index

$PI = \frac{\text{PV of future cash flows}}{\text{Initial investment}}$

Accept if PI > 1, which is the same test as NPV > 0.

Its use is specific and real: capital rationing. When the firm has more positive-NPV projects than capital, ranking by NPV alone can select one large project over several small ones that together create more value. Ranking by PI selects the most value per rupee of scarce capital.

The limits are worth stating. It is a ratio, so it has IRR's scale blindness when capital is not rationed. And it breaks down with multiple constraints or when projects are not divisible.

Accounting rate of return

$ARR = \frac{\text{Average accounting profit}}{\text{Average investment}}$

Uses accounting profit rather than cash and ignores timing entirely — both of the errors this subject has spent four chapters removing. It survives because it is the number a manager is measured on, which is an agency problem (chapter 1) rather than an appraisal method.

When a firm appraises in NPV and rewards on ARR, the reward wins.

What companies actually do

Most large firms use several measures together, and that is defensible. NPV for the decision, IRR for communication, payback as a risk constraint, PI when capital is rationed.

The failure mode is not using more than one method. It is letting a secondary measure override NPV when they disagree — and the usual reason that happens is that the secondary measure is the one somebody is paid on.

The point

Payback measures how long capital is exposed, ignores everything after that moment, and ignores the time value of money — so it is a bad appraisal rule. It is a reasonable risk screen, because distant forecasts are weak and liquidity constraints are real, and it is defensible as a constraint alongside NPV rather than a substitute for it. The profitability index earns its place specifically under capital rationing, and accounting rate of return earns none.