Liquidated Damages vs Penalty in India (Section 74 Explained)

A liquidated damages (LD) clause names, in advance, the amount one party will pay the other if a specific breach happens, usually late delivery, missed milestones, or a data breach. The one thing most people get wrong: they assume Indian law follows the English rule that a "genuine pre-estimate of loss" is enforceable while a "penalty" is void. It does not. Section 74 of the Indian Contract Act, 1872 collapses that distinction. Whether the clause is labelled liquidated damages or a penalty, the party complaining of breach can only recover reasonable compensation, capped at the named sum, and usually still has to show a loss actually happened. This guide (published by Adira, which makes contract review and CLM software, a commercial stake in you understanding contracts well, but this explainer stands on its own) walks through the statute, the case law, and what to check before you sign.

Plain meaning

An LD clause tries to do one useful thing: avoid a fight over the exact rupee value of a loss after a breach, by agreeing the number upfront. "If delivery is late, Vendor pays Client 0.5% of contract value per week of delay, capped at 10%." That is the whole idea, it saves both sides a drawn-out damages trial where you have to prove lost profit or cost of cover with evidence.

In common law jurisdictions like England and the US, this only works if the number is a genuine pre-estimate of loss made at signing. If it looks set to punish rather than compensate, courts strike it down as an unenforceable "penalty." India does not run this two-track system. Section 74 treats a named sum and a "stipulation by way of penalty" as the same thing, and caps what you can recover from either at reasonable compensation, not the full stated amount, regardless of the label used.

Who it protects and what triggers it

LD clauses protect the non-breaching party from the cost and uncertainty of proving damages from scratch. They are common wherever a delay or shortfall has a real but hard-to-pin-down cost: construction delay, IT project milestones, SLA breaches, delayed goods delivery, and confidentiality breaches with reputational fallout.

The trigger is narrow by design, a specific, defined event, not "any breach." A well-drafted clause names the exact failure (missing a delivery date, falling below an uptime percentage) and ties a calculable sum or formula to it. If the trigger is vague, or the same figure applies to breaches of wildly different severity, that usually means it was not actually pre-estimated, it was picked as a number big enough to scare the other side.

What to look for

Four things decide whether an LD clause is a fair risk-allocation tool or a trap:

1. Is there a genuine pre-estimate, or just a number? Look for language showing the parties turned their minds to actual likely loss at signing, "the parties agree this represents a genuine pre-estimate of loss likely to be suffered." Its absence does not make the clause unenforceable in India, Section 74 does not require that language, but its presence makes the clause far easier to enforce in full when the loss is hard to prove precisely.
2. Is it proportionate? Compare the LD amount to what a real loss from that breach would plausibly be. A one-day delay penalty of 5% of total contract value is a red flag; a per-week rate that adds up to something in the range of realistic loss over the likely delay period is not.
3. Is there a cap, and does LD stack with termination and indemnity for the same event? Stacking remedies for a single breach, LD plus termination plus an uncapped indemnity, is one of the most common ways these clauses get abused.
4. Does the clause claim LD is payable with no need to show any loss at all? Some clauses say the sum is due "automatically on breach, regardless of loss." That language does not override Section 74; courts still generally look at whether loss occurred, except where proving the precise loss is genuinely difficult and the sum was a reasonable pre-estimate (more below).

A quick test: does the clause state a calculation basis rather than a flat number, is it capped, and does the same breach also trigger a separate indemnity or open-ended claim elsewhere in the contract? No, no, yes to those three questions in order means treat the clause as high risk, whatever it is labelled.

The Indian position: Section 74, Contract Act 1872

Section 74 of the Indian Contract Act, 1872 reads:

"When a contract has been broken, if a sum is named in the contract as the amount to be paid in case of such breach, or if the contract contains any other stipulation by way of penalty, the party complaining of the breach is entitled, whether or not actual damage or loss is proved to have been caused thereby, to receive from the party who has broken the contract reasonable compensation not exceeding the amount so named or, as the case may be, the penalty stipulated for." Source: Section 74, Indian Contract Act, 1872

Three things follow. First, it does not matter whether the contract calls the sum "liquidated damages" or "penalty", both are treated the same way. Second, the named sum is a ceiling, not an entitlement: you can never recover more than the stated amount, but Section 74 does not automatically hand you the full amount either. Third, "whether or not actual damage or loss is proved" sounds like it dispenses with proof of loss altogether. The Supreme Court has read it more narrowly: where proving the exact figure is difficult, the court does not need precise mathematical proof, not that loss can be presumed when none occurred. (A separate Exception carves out bail bonds and statutory performance bonds, irrelevant to an ordinary commercial LD clause.)

Named Indian cases: what "reasonable compensation" actually means

Two Supreme Court judgments, read together, are what practitioners actually rely on.

Fateh Chand v Balkishan Das (AIR 1963 SC 1405) is the foundational case. The buyer paid Rs 1,000 as earnest money and a further Rs 24,000 towards a sale of leasehold rights and construction, then defaulted. The seller tried to forfeit the entire Rs 25,000 under a forfeiture clause in the agreement. The Supreme Court held that Section 74 covers not just a sum payable on breach but also a sum already paid and liable to forfeiture, and that a party can keep or recover only what a court finds to be reasonable compensation, not the full named or paid amount automatically. It allowed forfeiture of the earnest money proper but refused to let the seller keep the rest as a windfall with no shown loss.

Kailash Nath Associates v DDA ((2015) 4 SCC 136) is the modern statement of the rule and the case most LD disputes turn on. The Delhi Development Authority tried to forfeit earnest money after a land allotment fell through, arguing Section 74 let it keep the sum regardless of whether it had actually lost anything. The Supreme Court disagreed. A liquidated sum is recoverable as reasonable compensation only if it is a genuine pre-estimate of damages agreed by the parties, and damage or loss is a precondition ("sine qua non") for compensation under Section 74 in most cases. The phrase "whether or not actual damage or loss is proved" was read to mean strict, precise proof is not required where damage is difficult to assess, not that loss can be presumed with no evidence at all. On the facts, DDA had suffered no loss, so it recovered nothing, even though the forfeiture clause said the sum was payable automatically.

Sitting alongside these is ONGC v Saw Pipes Ltd ((2003) 5 SCC 705), often cited for the opposite-sounding point. ONGC deducted liquidated damages for delayed delivery of casing pipes under a clause both sides had agreed was a genuine pre-estimate. An arbitral tribunal denied the claim because ONGC had not separately proved actual loss with hard evidence. The Supreme Court set that aside: where loss is of a kind that is difficult to prove precisely, delay to an oil exploration project, in that case, and the figure was a genuine pre-estimate agreed in advance, the clause itself is sufficient basis to award the stated sum, no independent proof of exact quantum needed. Read together, the rule is: proof of loss is generally required, but where loss is real in kind and hard to quantify precisely, courts enforce a genuine pre-estimate without demanding exact arithmetic proof.

When courts reduce the amount, and earnest money

A court asks two linked questions before enforcing an LD figure: was it a genuine attempt to estimate loss at signing, and is it still, on the facts, roughly proportionate to what happened? If either answer is no, the court can award a lesser "reasonable compensation" instead, capped at, but potentially well below, the contract figure. Flat, large, round numbers with no visible calculation ("Rs 50 lakh for any breach") are risky. Figures built from a formula, a daily or weekly rate tied to a real cost driver like the cost of cover, survive scrutiny far better.

Earnest money follows a related, distinct rule. Fateh Chand treats a reasonable earnest deposit, a modest percentage of contract value paid to secure a deal, as forfeitable without separately proving loss, because the deposit itself is compensation agreed in advance for a collapsed deal. That does not stretch to instalments towards the price itself, or to unreasonably large deposits: at 25% to 50% of contract value, expect a court to treat the excess above a reasonable security amount as an unenforceable penalty.

Red flags

Normal	Red flag	Why it matters
LD amount calculated from a rate and time period (e.g. 0.5% per week, capped)	A single flat sum with no visible calculation behind it	Courts look for a genuine pre-estimate; an arbitrary round number invites reduction to "reasonable compensation" under Kailash Nath
LD payable only once loss-causing breach is established	Clause says LD is payable "automatically on breach, whether or not any loss is suffered"	Contradicts the reasoning in Kailash Nath; this language does not override Section 74, and courts can still ask whether loss occurred
LD capped, usually 5 to 15% of contract value	No cap, or LD that can exceed contract value	An uncapped or excessive figure is the clearest signal of a penalty in substance, whatever the clause calls itself
LD is the sole remedy for the specific breach it covers	LD, termination, and a separate indemnity all apply to the same single breach	Stacking remedies for one event is how a "reasonable compensation" clause turns into a multiple-recovery windfall, which courts resist
LD scales roughly with severity (daily/weekly rate)	Same LD figure applies whether the delay is one day or one year	Disproportion between trivial and serious breaches undermines the genuine pre-estimate argument
Earnest money set at a modest, customary percentage (commonly 5 to 10%)	Earnest money or advance set at 25%+ of contract value, all forfeitable	Fateh Chand allows forfeiture of a reasonable deposit, not the whole advance

Bad clause → better clause

Bad: "In the event of any delay in delivery, Vendor shall pay Client liquidated damages of Rs 10,00,000, which the parties agree is due immediately on breach regardless of any loss suffered by Client, in addition to Client's right to terminate this Agreement and claim indemnity for all losses arising from the delay."

What is wrong: a flat, uncalculated sum, a "regardless of loss" assertion Section 74 does not actually permit, and LD stacked with termination and an uncapped indemnity for one event.

Better: "If Vendor fails to deliver by the Delivery Date, Vendor shall pay Client liquidated damages of 0.5% of the Contract Value per week of delay, up to a maximum of 10% of the Contract Value ('LD Cap'). The parties agree this rate represents a genuine pre-estimate of the loss Client is likely to suffer from delay, based on the cost of alternative sourcing during the delay period. LD paid under this Clause shall be Client's sole and exclusive remedy for delay in delivery, except that Client may terminate this Agreement if the delay exceeds 12 weeks or the LD Cap is reached, whichever is earlier."

What changed: the figure is a calculable rate with a stated rationale, it is capped, it is the exclusive remedy for that breach (removing indemnity and termination stacking), and termination triggers on an objective threshold rather than sitting alongside LD for the exact same event.

How it interacts with related clauses

• Termination. If the contract lets the other side terminate and claim LD for the same delay, check whether that is intentional or accidental double recovery for one continuous breach.
• Indemnity. LD and indemnity should cover different risks. LD for delay plus a separate, uncapped indemnity for the same delay event is the stacking problem courts look at critically.
• Limitation of liability. Decide upfront whether LD sits inside your general liability cap or a separate bucket. Silence here creates the same ambiguity as with indemnity.

You can flag how LD, termination, and indemnity interact directly in a document, for free, using Weave, before you send a contract back for negotiation.

US and global contrast

English and US law still run the two-track system India abandoned in 1872: a clause is enforceable "liquidated damages" only if it is a genuine pre-estimate of loss at signing; if it looks designed to deter rather than compensate, it is void as a "penalty," full stop, however reasonable it later turns out to be. The leading English case, Dunlop Pneumatic Tyre Co v New Garage & Motor Co (1915), set that test; the UK Supreme Court refined it in Cavendish Square Holding v Talal El Makdessi (2015), asking instead whether the innocent party has a legitimate interest justified by the sum. US law, state by state, applies a broadly similar penalty doctrine.

India has no void/enforceable binary. A clause is neither automatically enforceable nor automatically void for being called a penalty, every LD or penalty clause is capped at reasonable compensation and tested the same way. A court can award a reduced amount rather than throwing the clause out entirely, friendlier than the common law's all-or-nothing doctrine, but the named figure is never a guaranteed floor the way it can be once a US court accepts a clause as valid LD.

FAQ

Is a penalty clause automatically unenforceable in India, like in the US or UK? No, and this is the most common misconception. Section 74 treats "penalty" and "liquidated damages" the same way: neither is automatically valid or void. Both are capped at reasonable compensation, whatever the label.

Do I have to prove actual loss to recover liquidated damages in India? Generally yes, per Kailash Nath Associates v DDA. The exception, from ONGC v Saw Pipes, is where the loss is real in kind but genuinely hard to quantify precisely and the sum was a genuine pre-estimate; there, courts do not demand exact proof of quantum.

Is it safe to forfeit an advance or earnest money on breach? A modest, customary earnest deposit, commonly 5 to 10% of contract value, is generally forfeitable under Fateh Chand without separately proving loss. A larger advance or instalment towards the price is different; a court can order the excess above a reasonable security amount returned.

Should I stack liquidated damages with termination and indemnity for the same breach? Be careful. Courts view multiple remedies firing off one breach with suspicion. State clearly whether LD is the exclusive remedy for the trigger it covers, and reserve indemnity and termination for genuinely separate risks or a clear escalation threshold.

Does an LD clause need to say "genuine pre-estimate of loss" to be enforceable? No, that phrase is not a legal requirement under Section 74. But including it, with a stated rationale for the figure, makes it considerably easier to argue the clause deserves full enforcement rather than a reduced, court-assessed amount.

This guide gets you to understanding what Section 74 does and how Indian courts have applied it to LD and penalty clauses. It does not tell you whether a specific LD figure in your contract would survive challenge, that depends on the facts, the industry, and how the number was actually arrived at, and is not legal advice. Talk to a lawyer before you rely on, negotiate, or enforce a liquidated damages clause in a live deal.