When a 123% Tariff Becomes a Drafting Question: The India-Indonesia-Laos Determination as Contract-Architecture Stress Test

The Department of Commerce's preliminary antidumping determination on solar cell imports from India, Indonesia, and Laos arrives with a numerical asymmetry that any developer reading the order had to read twice — 123.04 percent for all Indian producers, 35.17 percent for all Indonesian producers, 22.46 percent for all Laotian producers, and a separate set of countervailing duty determinations layering on top. The Indian figure is not a tariff in any operational sense; it is a closure. At that level, no Indian-origin cell can compete on landed cost into the United States for any utility-scale procurement that has to clear a credit committee, and any supply position written against an Indian source is now, in commercial terms, a position against a price the market will not pay.

These margins did not arrive in a vacuum. They are the latest entry in a multi-year sequence in which the boundary of what counts as a dumped or subsidized cell has migrated from Chinese product, to Southeast Asian transshipped product, and now to a wider set of jurisdictions that, only twenty-four months ago, were positioned in module supply contracts as the safe harbor against precisely this kind of enforcement. Procurement teams, EPC counsel, and tax-equity diligence partners alike priced their base-case landed-cost models on the implicit assumption that the trade enforcement perimeter would remain where it had been; that assumption is no longer operative, and the contracts written under it are now the binding documents under which projects must still close.

There is a specific mechanic in antidumping procedure that is rarely modelled at term-sheet stage and almost never priced into developer base cases. The preliminary determination triggers cash deposit requirements at the rate announced; the final order, issued months later, may revise that rate up or down; and entries that cleared customs during the intervening period are subject to retroactive liquidation at the final rate, with interest. A developer who took delivery of Indonesian cells at the 35.17 percent provisional rate, comfortable that the EPC contract had a price-adjustment formula calibrated to the deposit, may find the final order imposes a different rate entirely — and the contract's adjustment clause, drafted to track duty paid at importation, does not extend to the retroactive top-up. This is the gap between the deposit number and the liquidation number, and it sits inside the EPC's books only if the contract said so explicitly. In the supply agreements we have read across the past thirty months, that explicit drafting is the exception rather than the norm.

The instinct following a tariff shock of this magnitude is to renegotiate prices. The more revealing exercise is to read the contract language already in force, because the architecture of the change-in-law clause, the indexation formula, and the termination trigger almost always predetermines the outcome of the renegotiation. Three patterns emerge. In the first, the developer carries the trade-policy risk by default — the EPC quote is structured ex-works or delivered duty unpaid, and any change in tariff lands on the project balance sheet without recourse against the supply chain. In the second, the EPC carries the risk in form but holds a force-majeure-style escape valve that converts a tariff increase above a defined threshold into a termination right at the EPC's option, leaving the developer with a half-built site, a frozen interconnection slot, and a senior debt facility whose drawdown conditions presupposed continuous construction. In the third, and rarest pattern, the contract names antidumping and countervailing duty determinations specifically as change-in-law events, with a defined sharing formula and a continuation obligation calibrated to landed cost — and only this third pattern actually allocates the risk in a way that survives a determination of this magnitude.

The headline AD margin is also not the full picture, and this is the second observation that procurement teams routinely understate when reporting up to investment committees. The countervailing duty determination layers on top of the antidumping figure, and the combined cash deposit can exceed the simple sum of the two as published when transaction value is contested or when scope rulings pull adjacent bill-of-materials components into the same calculation. For Indian product, the combined effect is structural exclusion. For Indonesian product, the 35.17 percent AD margin alone exceeds the cushion that most module supply contracts were drafted to absorb — typical price-adjustment caps in 2024 and 2025 supply agreements ran in the range of single-digit to low-double-digit percentage shifts; a 35 percent shock blows past that cap, triggers either a renegotiation right or a termination right depending on drafting, and returns the parties to the underlying balance of negotiating leverage, which in a tight-supply environment rarely favors the developer. For Laotian product, the 22.46 percent figure looks more manageable in isolation, but the layering of CVD onto AD onto Section 201 onto basic MFN rates produces a stacked landed cost that few base-case models anticipated, and the stack is not visible until the customs broker calculates the entry.

The tax-equity dimension of the determination deserves separate examination, because the structural exposure runs through a different layer of the capital stack than the supply contract. Tax-equity investors underwrite to a fixed-flip yield calculated against a base-case ITC value and a base-case generation stream; their model sensitivity to landed module cost is asymmetric, because they sit above the development equity in the loss waterfall during construction and below it in the recovery once the project flips. When a tariff shock raises module cost by twenty or thirty percent, the development equity absorbs the first dollar of loss; the tax equity absorbs nothing if the project still hits its placed-in-service deadline, but it absorbs disproportionately if the project misses — and the missed-deadline scenario is precisely what a tariff-driven supply renegotiation tends to produce. The diligence question that lender's counsel and tax-equity counsel ask after a determination of this scale is not what the new module price is; it is what the contractual mechanism is that converts a price shock into either a milestone-preserving renegotiation or a milestone-breaking termination, and the answer is found in the same change-in-law clause that the developer's procurement team is now reading carefully for the first time.

The instinct following an enforcement event of this scale is to convene the procurement team and re-source. The empirical lesson from earlier enforcement cycles — the 2018 Section 201 determination, the 2022 anti-circumvention investigation, the 2024 polysilicon traceability enforcement — is that vendor diversification is necessary but not sufficient, and that the projects which absorbed each successive shock with the least damage were not those with the most diversified vendor lists but those whose contracts had been drafted to anticipate the possibility of an enforcement event without naming any specific one. The drafting move that ages well is not if antidumping duties are imposed on Country X; it is if any cumulative trade measure increases the landed cost of the contract scope by more than Y percent, the parties will share the increment as follows. The first formulation ages badly the moment the next investigation moves to a country not named; the second survives every iteration of enforcement because it is calibrated to landed cost, not to jurisdiction. The cost of inserting the second formulation at term-sheet stage is functionally zero; the cost of not having it once the determination publishes is a renegotiation conducted from the weaker side of the table.

The vulnerability that the current determination exposes is not a vendor problem and it is not a tariff problem; it is a contract-architecture problem that has been allowed to accumulate across the entire US utility-scale solar pipeline because the conditions under which it would matter were treated as tail risk. They are no longer tail risk. They are the visible state of the trade enforcement environment, and any project whose financial close depends on landed-cost assumptions written before the preliminary determination is now operating with a disclosed gap between the number in the model and the number that will appear on the customs entry. The IFI and tax-equity diligence questions that follow — and they will follow, because lender's counsel reads these orders as carefully as developer counsel — are not how exposed is the project but where in the contract stack is the exposure absorbed, and is that absorption actually enforceable against a counterparty whose own balance sheet has been hit by the same determination.

This is the precise point at which our Procurement and Vendor Governance practice intersects with our Deal and Contract Advisory work, and the intersection is not incidental. We do not treat the procurement question as ending at vendor selection or at RFP scoring; we treat it as continuing through clause architecture, because the landed-cost certainty that a project model assumes is either built into the supply contract, the EPC contract, and the change-in-law allocation between them, or it is conceded by silence. The work we do at the clause level is the work that determines whether a determination of this scale is a manageable repricing or a structural break: drafting the trade-measure trigger so that it captures cumulative effect rather than single-jurisdiction events; calibrating the price-adjustment formula so that it tracks landed cost net of any deposit-versus-liquidation gap; sequencing the termination triggers so that a developer is not left with the worst combination of a tariff-locked supplier and an interconnection slot that cannot wait. We do this in coordination with the deal architecture across the entire counterparty stack — offtake, EPC, supply, senior debt, tax equity — because the same allocation question runs through every layer, and the answer in each must be consistent for the project to remain financeable.

The preliminary determination is not the final order, and the gap between the two will reward parties whose contracts were written to survive both. The question worth asking, before the next investigation announces its scope, is not which jurisdiction is safe; it is which clause in the supply stack — read in full, against the EPC, against the change-in-law allocation, against the price-adjustment formula — actually carries the trade-policy risk, and whether that allocation is the one the project's economics ever assumed.