An incentive is a conditional receipt, not a business case

Over the past few years Indian manufacturers have been handed a new kind of capital-allocation question. Central and state governments have introduced incentive schemes, production-linked and otherwise, that pay companies for building capacity in particular products and selling from it. The schemes differ, and I will not describe any one scheme's clauses here, because the detail changes and matters only to the companies that apply. Their broad shape is familiar: a payment, often tied to incremental sales or to capacity actually sold, spread over several years, available only if the company commits a minimum investment, reaches sales or value-addition thresholds on time and keeps its records in the form the scheme requires.

I support the purpose of such schemes, and some Indian companies have used them well. My concern is how they enter boardrooms. In the proposals I have seen, at Ilvara and in conversations with other CFOs, the incentive tends to arrive inside the business case rather than beside it. Returns are calculated with the incentive included, the incentive turns a marginal plant into an attractive one, and the board approves a capital commitment whose economics depend on conditions management has not yet met.

Start with the plant that would exist anyway

My first rule is simple. Every capital proposal at Ilvara shows its return without any incentive, against our 15 per cent post-tax hurdle, before anything else is discussed. If the plant clears the hurdle on its own, the incentive is welcome upside. If it does not, the board is being asked a different question: whether the incentive, with all its conditions, is worth the risk of owning a plant that would otherwise not be built.

Accounting offers a useful discipline here. Under Ind AS 20, a government grant is recognised only when there is reasonable assurance that the company will comply with the conditions attached to it and that the grant will be received. I think a board should apply the same test before it approves the spending. A business case that counts an incentive the company could not yet recognise in its own accounts is counting money on assumptions its auditors would reject.

There is also the question of what the plant does after the incentive ends. A scheme that pays for five or six years leaves the company owning an asset for twenty or more. If the product's margins are thin without support, the incentive period becomes a subsidised entry into a business that will be uneconomic for most of the asset's life. That may still be a sensible strategic bet, but it should be presented to the board as a bet.

Thresholds are the real terms

My second rule is that the thresholds are the terms of the deal, and they deserve as much analysis as the payment. Most schemes pay only if sales or production reach set levels within set years, and many reduce or withhold payment when they do not. That turns the incentive into an option whose value depends on demand for a product the company has not yet sold, in a market the scheme itself is encouraging rivals to enter.

We therefore value incentives as a range. For each proposal we model the incentive under low, middle and high demand cases, with the threshold tests applied year by year, and we show the board the probability-weighted value beside the headline figure. We also show what happens to the plant's return if the company meets each threshold a year late, which in our experience is the likeliest way to fall short. In some cases the headline incentive survives this treatment almost intact. In others, most of it disappears.

What we chose instead at Dahej

In 2023 Ilvara's board considered a ₹460 crore greenfield plant to make equipment for green-hydrogen production, a category for which a central-government incentive was available to manufacturers that met its investment and sales conditions. Our executive chairman saw it, with good reason, as an entry into a market that could grow for decades, and one of our independent directors agreed with him.

Our analysis was less encouraging. Without the incentive, the plant's expected return was about 8.5 per cent. With the full incentive, if every threshold was met on time, it rose to about 14 per cent, still short of our hurdle. Weighted across the demand cases, it was nearer 11 per cent. The technology depended on a licence from a foreign partner whose own commercial position was uncertain, and the sales thresholds assumed orders from projects that had not yet reached financial closure.

At the same time our Dahej yard had twenty-six months of backlog in heavy reactors, our most profitable product, and a ₹171 crore extension there showed a return of about 21 per cent without any incentive at all. I recommended the Dahej extension and, for hydrogen, a smaller route: fabricating balance-of-plant equipment for other manufacturers in existing bays, with no capacity commitment, until the market showed its depth. The board approved Dahej and deferred the greenfield plant, with an agreed set of market signals that would bring it back for decision.

The trade-off was visible. We gave up the chance to apply for the incentive in that round, a competitor did apply, and three of our process engineers left to join it. The chairman has been entitled to remind me of that, and I am not certain the deferral will look right in ten years. The Dahej bay was commissioned in October 2024 and earned about ₹230 crore of revenue in its first twelve months. Our hydrogen fabrication for others is small but profitable, and it is teaching us the product without asking us to stake the balance sheet on it.

The broader point is not that incentives are bad or that greenfield plants are rash. It is that an incentive is a conditional receipt from a counterparty, and deserves the analysis we would give any other: what must we do to earn it, how likely are we to do it on time, and what are we left holding if we do not. A board that asks those questions will still approve some incentive-linked plants. It will approve them knowing what it has bought.