A guess in the dark, or an estimate in the light? The Arbitrator’s guesstimate

Every arbitrator who has sat through final arguments faces a familiar difficulty: the breach is proved, the loss is real, but the arithmetic is nowhere to be found. Section 73 of the Contract Act, 1872 requires proof of actual loss, while commercial reality rarely permits precise calculation. In the gap between the two lies one of arbitration’s most consequential exercises: the honest estimate.

Where precise proof of loss is genuinely impossible, a tribunal may arrive at an honest estimate. In Trishala Jain v. State of Uttaranchal, the Supreme Court distinguished a ‘guess’—an estimate without specific evidence—from a ‘calculation’ based on specific evidence and a ‘guesstimate’, which combines elements of both and carries greater certainty than a bare guess. The line is narrow: too little estimation may unfairly reward the breaching party; too much may become speculation and expose the award to challenge. This article examines how common law and civil law jurisdictions navigate this boundary and what their approaches reveal about the extent of arbitral discretion in assessing damages.

India’s position has deep roots in the common law tradition. In Gemini Bay Transcription Private Limited v. Integrated Sales Service Limited, the Court traced the principle to the Calcutta High Court’s decision in Frederick Thomas Kingsley v. Secretary of State for India, which held that damages are not uncertain merely because loss cannot be proved with mathematical certainty. What is required is reasonable certainty that the loss probably followed from the wrong.

The Supreme Court has therefore permitted practical estimates of lost profit where exact proof is unavailable. In Mohd Salamatullah v. Government of Andhra Pradesh and A.T. Brij Paul Singh v. State of Gujarat, the Court awarded lost profit at fifteen per cent of contract value rather than insisting on exact proof. Construction and Design Services v. Delhi Development Authority similarly held that where the fact of loss is established but its precise amount cannot be quantified, honest guesswork may be used.

The Delhi High Court has applied the same approach in arbitration. In the Cobra Instalaciones judgment, it restored an award that apportioned loss from concurrent delays through honest estimation. In National Highways Authority of India v. ITD Cementation India Ltd, it held that practical guesswork based on reasonable material should not be disturbed merely because another figure could have been adopted. In Simplex Concrete Piles v. Union of India, it upheld an award representing roughly one-thirteenth of the sum claimed as a reasonable estimate drawn from the record.

But discretion has firm limits. In Unibros v. All India Radio, the Supreme Court set aside an award for loss of profit, holding that formulae such as Hudson’s may assist estimation but cannot, by themselves, establish loss. Evidence of lost opportunity remains necessary; the court cannot “make a guess in the dark." Batliboi Environmental Engineers Ltd. v. Hindustan Petroleum Corpn. Ltd. similarly stressed that damages must correspond to actual loss and must not create a windfall. In Radiance Infracon v. GLS Infratech, the Delhi High Court held that the law does not permit guesswork in vacuo: an estimate must have a reasoned basis and a nexus with the material on record.

The same tension appears across the common-law world. In Chaplin v. Hicks, England’s Court of Appeal awarded damages for loss of a chance despite the practical difficulty of exact calculation. Singapore’s courts have also adopted a practical approach. In CEF v. CEH, the court recognised that some losses are inherently difficult to quantify precisely, while MFM Restaurants v. Fish & Co held that difficulty in proving the exact cause or amount of loss does not justify nominal damages where substantial loss is clear. In Story Parchment Co. v. Paterson Parchment Paper Co., the US Supreme Court similarly accepted that where the wrong itself makes certainty impossible, a just and reasonable inference may suffice.

Civil law systems reach a similar result through a different route. Under French law, estimation is treated as part of the fact-finder’s ordinary function rather than an exceptional departure from proof. German law is similarly flexible, allowing estimation where the claimant has pleaded sufficient facts to support a minimum figure, without requiring exceptional circumstances or special justification.

The Gulf approach goes further by treating estimation as a normal judicial function. Under the UAE Civil Code, where compensation is not fixed by law or contract, the judge may assess it based on the harm suffered. Article 390(2) also allowed courts to revise agreed damages to reflect actual loss, even where the parties had fixed a specific amount. This differed from Dunlop Pneumatic Tyre v. New Garage, Cavendish Square Holding v. Talal El Makdessi, and Kailash Nath Associates v. Delhi Development Authority, where agreed sums are generally respected subject to applicable legal limits.

The UAE’s broad discretion has been narrowed by the new Civil Code, effective June 1, 2026. Article 340 replaces Article 390 and retains the power to adjust agreed compensation where it is excessive, the contract has been partly performed, or the creditor contributed to the loss. An award above the agreed amount now requires proof of fraud or gross fault. The reform therefore introduces greater structure while preserving the court’s ability to prevent unfair outcomes. Article 340 continues to preserve this power against contrary contractual arrangements.

Saudi Arabia has taken a different path. Its Civil Transactions Law 2023 recognises loss of profit as compensable, reflecting a more flexible approach to uncertain future loss. The DIFC and ADGM, by contrast, continue to follow common law principles. For tribunals seated onshore in the Gulf, estimation remains an important judicial function, although the UAE now subjects it to clearer safeguards.

Article 7.4.3(3) of the UNIDROIT Principles of International Commercial Contracts similarly leaves the assessment of damages to the tribunal where the amount cannot be established with certainty. The provision reflects the broader principle that lack of mathematical precision does not necessarily defeat a damages claim, provided the tribunal has a reasonable basis for its assessment.

Investment tribunals follow the Factory at Chorzów principle that compensation should, as far as possible, restore the position that would have existed without the wrongful act. Where future loss is uncertain, tribunals may estimate it but must avoid speculation. In Bear Creek Mining v. Peru, the tribunal rejected a discounted cash flow valuation for an early-stage project because of its uncertainty and awarded sunk costs instead. By contrast, Tethyan Copper v. Pakistan awarded nearly USD 6 billion for a mine that was never built. The lesson is that methodology alone does not create certainty; what matters is whether the reasoning connects the evidence to the amount awarded.

The guesstimate doctrine cannot be understood apart from the statutory limits on an arbitral tribunal’s authority. Section 28(1)(a) of the Arbitration and Conciliation Act, 1996 requires a tribunal seated in India, other than in an international commercial arbitration, to decide according to Indian substantive law. The power to guesstimate is therefore derivative: it exists because Section 73 of the Contract Act, as judicially interpreted, permits estimation where loss is proved but its precise amount cannot be established.

That power is fenced on two sides. First, Section 28(3) requires the tribunal to take the contract and applicable trade usages into account. In Associate Builders v. Delhi Development Authority, the Supreme Court held that an arbitrator who goes beyond the contract acts without jurisdiction, a principle reflected in Ssangyong Engineering v. NHAI as a form of patent illegality under Section 34(2A). An arbitrator who estimates loss contrary to an agreed measure of damages, or awards what the contract excludes, therefore exceeds the parties’ mandate. Garg Builders and Union of India v. Manraj Enterprises reinforce this contractual limit.

The second fence is Section 31(3), which requires the award to be reasoned. Ssangyong held that mere contravention of Indian substantive law is not, by itself, enough to set aside an award. However, an unreasoned award, or one whose reasons show no connection with the material on record, remains vulnerable to challenge. The reviewing court’s inquiry is therefore narrow as it asks whether the arbitrator gave reasons and remained within the contract, rather than whether the figure was correct. This distinction matters because an appellate court is not entitled to substitute its own assessment merely because it would have preferred a different methodology or figure. The focus remains on legality, contractual boundaries and the adequacy of the reasoning. That is what keeps estimation from becoming speculation while preserving the tribunal’s limited role in quantifying proven loss.

That is the quiet consequence of the doctrine. An arbitrator has no free-ranging judicial discretion, but a bounded power to estimate within statutory and contractual limits. Where the breach is proved, the loss is real, and precise arithmetic is absent, the estimate may stand—not necessarily because it was right, but because it was reasoned.

About the authors: Shantanu Agarwal is the Managing Partner of Lexster Law LLP. Ananya Garg is a Senior Associate and Tazeen Ahmed is an Associate.

Disclaimer: The opinions expressed in this article are those of the author(s). The opinions presented do not necessarily reflect the views of Bar & Bench.

If you would like your Deals, Columns, Press Releases to be published on Bar & Bench, please fill in the form available here.