Sector policy and PLI

Every few months the government announces a policy meant to build up a sector — a scheme to make more electronics at home, more solar cells, more speciality chemicals, more medical devices. The most common shape it takes now is the production-linked incentive, or PLI: the government offers to pay a company a percentage of its extra output, but only if the company first invests a set amount and produces above a set threshold. The moment such a scheme is announced, the qualifying stocks tend to jump.

The beginner reads it simply: the government is going to pay these companies to grow — that has to be good. And here is the discipline this module builds: a policy tailwind is worth exactly as much as the delivery behind it, and not a rupee more. A PLI is not a grant handed over on announcement day. It is a conditional promise — pay-on-performance — that a company collects only after it has built the plant, produced the output, cleared the thresholds and had its claim verified. The gap between the announcement and the first cheque is measured in years, and it is littered with schemes that were celebrated and then quietly under-delivered.

This is the promise ledger — a habit this part of the shelf keeps returning to — applied to policy. When you hear "the government has announced a scheme," you write it in the promise column, not the delivered column, and you wait to see what actually gets built.

A production-linked incentive is a government payment tied to how much a company produces above a baseline, available only if the company also invests a committed amount within a set window. Think of it as the government saying: build the factory, hit these output numbers, and we will pay you a small percentage of the extra you make, for a few years.

Two features define it, and both matter for how you read it:

It is conditional and back-ended. The company spends first — capital to build capacity, working capital to run it — and is paid later, only if it clears the investment and output bars the scheme sets. Miss the bars and you get little or nothing. This makes PLI fundamentally different from an old-style subsidy that flows on every unit sold. It is pay-for-performance: the reward follows delivery, not eligibility.

It is time-limited. A PLI scheme runs for a defined window — often around five years — and then ends. So any profit it adds is temporary by design. A company leaning heavily on PLI has a slice of earnings with a built-in expiry date.

Why does the government prefer this shape? Because it wants factories and jobs on the ground, not just announcements, and pay-on-delivery is harder to game than an upfront grant. That is genuinely smarter policy design. But the same feature that makes it good policy — you only get paid if you deliver — is exactly what makes it dangerous to over-read as an investor. The announcement is free and instant; the delivery is expensive and slow; and the market keeps mistaking the first for the second. As this shelf keeps warning, how much of an announced scheme will actually be delivered, and when, is not forecastable — you can read the design; you cannot bank the outcome in advance.

Lay a PLI scheme out as a ledger with two columns — promised and delivered — and trace how an item moves from left to right. It has to pass several gates, and at each one it can stall.

Announced. The scheme is notified: sectors, incentive rates, investment thresholds, output baselines, the window. This is pure promise. Nothing has been earned by anyone.

Eligible. A company applies and is approved as a participant. This is a door, not a payment. Eligibility says you may earn the incentive if you deliver — it guarantees nothing.

Invested. The company actually spends the committed capital and builds the capacity. This is the expensive, slow, risky gate. Land, clearances, equipment, commissioning — a plant is not built in a quarter.

Produced. The plant runs and output crosses the scheme's threshold. Only above the baseline does incentive accrue, so a plant that runs below target earns proportionately little.

Claimed and verified. The company files a claim; the government verifies the numbers against the scheme's conditions.

Paid. The incentive is finally disbursed — and, as with any government payment, this can lag, becoming a receivable in the meantime, exactly as the previous module described for subsidies.

The reader's move is to ask, for any PLI story, which column is this in? Announced and eligible are promises the whole sector can share. Invested, produced and paid are delivery, which only some companies will actually achieve. The value is in the second column; the excitement is almost always in the first.

Watch one PLI scheme land on two composite companies in the same sector. illustrative

Company A treats the scheme as a reason to do what it was half-planning anyway. It commits the capital, builds the plant on schedule, ramps output past the threshold, files clean claims, and after a couple of years begins receiving the incentive. Its profit gets a genuine, if time-limited, lift. When the reader checks, the PLI shows up as real cash in the cash-flow statement and as a receivable that is actually being collected.

Company B issues a stirring press release the week the scheme is announced, promising a huge new facility. Its stock jumps hardest of all. But land and clearances drag, the plant slips two years, output never comfortably clears the threshold, and the incentive it eventually collects is a fraction of what the announcement implied. The profit boost the market priced in on day one never fully arrives.

Same scheme, same eligibility, opposite outcomes — and on announcement day, the market could not tell them apart, because on announcement day both had only a promise. The difference was delivery, which is invisible until it happens. That is why the promise ledger matters: it stops you paying delivered-column prices for promised-column news.

Sector policy is the purest example on this shelf of a true thing being over-priced. Hold the limits.

It cannot tell you who will deliver. Eligibility is shared; delivery is not. The announcement cannot distinguish the company that will actually build and produce from the one that will only announce. That is settled over years of execution you cannot see on day one.

It cannot tell you how much will actually be earned. PLI pays on output above a threshold, for a limited window. The headline scheme size is a ceiling for the whole sector, not a figure any single company will collect. Real receipts are usually a fraction of the launch-day imagination.

It cannot tell you the timing. Plants take years; claims lag; payments can become receivables. The distance from announcement to received cash is long and, as always on this shelf, unforecastable in its precise timing.

It cannot tell you the profit is durable. A PLI-boosted profit is part operating, part subsidy-with-an-expiry-date. Valuing the whole boosted figure as if it were permanent means paying for earnings that partly vanish when the scheme ends. And policies can be altered mid-stream — thresholds tightened, windows changed — because a scheme is policy, not a contract.

The errors around policy schemes are remarkably consistent.

1. Pricing a promise as a payment. An announced scheme is in the promise column. Treating it as banked profit ignores the years of expensive delivery between the notification and the first rupee received.

2. Confusing eligibility with the reward. Qualifying for a scheme is a door, not a cheque. The incentive flows to measured investment and output, so eligible companies can end up worlds apart.

3. Rewarding the loudest announcer. The company with the grandest press release is often the one most tempted to over-promise. Announcements collect nothing; production does.

4. Valuing time-limited profit as permanent. A PLI boost expires with the scheme. Paying a high multiple on incentive-inflated earnings means paying for profit that will partly disappear.

5. Forgetting the scheme can change. Thresholds, rates and windows are policy, and policy can be redesigned mid-stream. What was promised is not guaranteed.