Introduction
A liquidated damages (LD) clause drafted may be perfectly enforceable in Hong Kong for a construction project — but unenforceable in other jurisdictions as a penalty. As cross-border infrastructure investment across Asia accelerates, understanding these differences is crucial to allow parties to enter into commercial relations with clarity and confidence.
LD clauses remain a key risk allocation tool in construction and infrastructure projects, yet their enforceability and practical application vary significantly across Asia. This blog is the eighth in our series: “Shaping Asia’s Infrastructure”, which explores legal developments shaping infrastructure across Asia. It compares the approaches to LD clauses taken in Singapore, Hong Kong, Vietnam, and Japan, highlighting key legal principles and practical considerations for parties operating across multiple jurisdictions.
At a glance
| Singapore | Hong Kong | Vietnam | Japan | |
| Governing test of when a clause is for LD | Genuine pre-estimate of loss (applies UK House of Lords approach in Dunlop, as affirmed in Singapore Court of Appeal decision in Denka) | Legitimate interest / proportionality (applies UK Supreme court approach in Cavendish) | Emerging statutory framework (Construction Law 2025) | Freedom of contract + general principles review |
| Penalty clauses | Void and unenforceable | Void and unenforceable | Enforceable but capped at 8% of breached obligation – can be additional to LD | Enforceable – can be additional to LD |
| Judicial adjustment of LD clause | No – either enforceable (if not a penalty) or unenforceable (if it is) | No – either enforceable (if not a penalty) or unenforceable (if it is) | Uncertain | Courts may reduce or invalidate |
Key takeaways
Several themes emerge from this comparison:
- The penalty doctrine is evolving — but not uniformly. Even among common law jurisdictions, Singapore has held firm on the traditional Dunlop “genuine pre-estimate” test, while Hong Kong has embraced the broader Cavendish proportionality framework. Parties should not assume that an LD clause tested in one jurisdiction will survive scrutiny in the other.
- Civil law jurisdictions bring distinct challenges. Vietnam’s historic conflation of LD with penalties and unliquidated damages created significant uncertainty. The Construction Law 2025 is a welcome step, but much depends on the implementing decree. Japan’s approach is more settled, but the absence of a bright-line proportionality test means that enforceability turns on general principles applied on a case-by-case basis.
- Proportionality is the common thread. Across all four jurisdictions — regardless of the doctrinal framework — courts are sceptical of LD clauses that bear no reasonable relationship to the loss or interest they purport to address. Excessiveness is the surest path to unenforceability.
- Drafting matters. Tailoring LD clauses to the specific breach, building in severability mechanisms, articulating the legitimate interests at stake, and supporting stipulated amounts with evidence or objective formulas are practices that improve enforceability across all four jurisdictions.
For parties negotiating construction and infrastructure contracts across Asia, a jurisdiction-specific approach to LD clauses is not optional — it is essential.
Singapore
In assessing whether a clause is enforceable as an LD clause, Singapore courts have traditionally applied the UK House of Lords’ test in Dunlop Pneumatic Tyre Co Ltd v New Garage & Motor Co Ltd [1915] AC 79 (Dunlop), which requires the clause to constitute a “genuine pre-estimate of loss”.
When the UK Supreme Court reformulated that test in Cavendish Square Holding BV v El Makdessi [2015] UKSC 67 (Cavendish), there was a question whether Singapore would follow suit. To recap, the Cavendish test is whether the pertinent clause protects a “legitimate interest”, which can be applied as follows:
- first, consider whether an LD clause is a secondary obligation that arises from a breach of contract;
- second, identify the legitimate interest protected by the clause; and
- third, consider whether the LD clause “imposes a detriment on the contract-breaker out of all proportion to any legitimate interest of the innocent party in the enforcement of the primary obligation” by reference to the circumstances at the time of contracting.
This issue was squarely addressed by the Singapore courts in Denka Advantech Pte Ltd v Seraya Energy Pte Ltd [2020] SGCA 119 (Denka). The Singapore Court of Appeal reaffirmed the Dunlop framework and held that an LD clause is enforceable only if it represents a “genuine pre-estimate of the likely loss”, assessed at the time of contracting. The court declined to adopt the broader “legitimate interest” test in Cavendish, reasoning in part that Singapore law recognises compensation for loss as the only legitimate interest justifying LD clauses.
Denka further offers practical lessons for parties negotiating LD clauses under Singapore law:
- Severability provisions: Where the stipulated sum in the LD clause is excessive, Singapore courts will treat the entire LD clause as an unenforceable penalty. Parties should therefore include language to sever an unenforceable part of the provision, to mitigate the risk of the entire LD clause being struck down.
- Structure payments as primary obligations: The Singapore courts take a “substance over form” approach in determining whether an obligation is primary or secondary (i.e., obligations triggered by breach). The distinction remains critical: courts will scrutinise only secondary obligations under the penalty doctrine. Primary obligations are upheld even if they involve bad commercial bargains.
- Avoid indiscriminate lump sums: A single fixed sum payable regardless of the nature or severity of the breach is likely to be presumed a penalty.
- Tailor LD clauses to the breach: Clauses calibrated to the duration of the breach — such as those “gradated according to the remaining duration of the contracts” as in Denka — are more likely to be upheld.
- Use of multipliers: Where a percentage multiplier is implemented, parties should be prepared to justify it with expert evidence. In Denka, the court accepted expert testimony showing that the 40% multiplier reasonably reflected estimated losses.
Hong Kong
While Singapore has reaffirmed the traditional Dunlop approach, Hong Kong has moved in the opposite direction.
Hong Kong previously applied the Dunlop test (see The Bank of East Asia Ltd v Yip Chi Wai and Others [2011] 5 HKLRD 761). However, in Law Ting Pong Secondary School v Chen Wai Wah [2021] HKLRD 185, the Court of Appeal changed course consistent with Cavendish. This ruling was subsequently re-affirmed in Li Chiu Wah Joseph v Hong Kong Society of Notaries [2023] HKCA 809 – a decision upheld by the Court of Final Appeal (see (2023) 26 HKCFAR 529)).
The Cavendish approach has been applied in subsequent decisions, including Top Gear Investment Ltd and Others v PACM Group (Holdings) Ltd [2024] HKCFI 2347 and Wing Lok Construction & Engineering Co., Limited v Techoy Construction Co. Ltd. [2025] HKCFI 5546.
When drafting LD clauses under Hong Kong law, parties should consider the following factors:
- What are the legitimate interests? Parties should consider and be prepared to articulate the legitimate interests they seek to protect if the LD clause is challenged. The court will require evidence to establish these interests (see, for example, Wing Lok Construction & Engineering Co., Limited).
- Ensure proportionality: Sanctions imposed must not be “out of all proportion” to the interest protected. A clause imposing a substantial late charge over a minor delay, for instance, will require substantive evidence justifying its proportionality.
Vietnam
If Hong Kong represents a common law jurisdiction modernising its penalty doctrine, Vietnam illustrates the challenges of applying LD concepts within a civil law framework still finding its footing.
While LD clauses are commonly used in Vietnamese construction contracts, their treatment by the courts has been inconsistent, leading to questions over enforceability and raising issues of contract certainty.
Historically, courts have generally treated LD as either: (i) penalties, which are capped at 8% of the value of the breached obligation; or (ii) unliquidated damages, which require proof of actual loss and are often fiercely disputed.
In response to this uncertainty, the National Assembly enacted Law No. 135/2025/QH15 on Construction (the Construction Law 2025). This is the first time Vietnamese legislation has recognised LD clauses as a standalone concept, distinct from both penalties and general damages. For foreign investors and contractors, this is a notable shift.
Article 86.2 provides that: “Compensation for damages shall be determined on the basis of actual damages [or] predetermined damages corresponding to obligations under the construction contracts that are breached [and] the extent of such breaches”.
Parties can therefore rely on LD clauses so long as the compensation corresponds to: (i) obligations breached (which must be specified under the contract); and (ii) the extent of such breach.
While a promising development, the Construction Law 2025 still leaves space for ambiguity, particularly around the degree of “correspondence” required, and how the courts will distinguish between punitive penalties and LD. An implementing decree — potentially clarifying these points — is expected before or shortly after the law becomes effective in July 2026.
In practice, parties should consider the following factors under Vietnamese law:
- Avoid excessiveness: LD clauses should ensure that compensation amounts are clearly linked to specific contractual obligations and proportionate to the level of breach.
- Choice of remedy: Since the Construction Law 2025 allows LD, unliquidated damages and penalties to co-exist, parties should carefully consider which remedies to elect and how they will interact.
Japan
Japan offers yet another model — one where penalty clauses are not merely tolerated but expressly permitted, and where courts have developed a pragmatic approach to policing excess.
LD clauses are widely used in Japanese commercial contracts to facilitate efficient dispute resolution and strengthen contractual enforceability. The statutory basis lies in Article 420 of the Civil Code. Unlike in common law jurisdictions, the principle of freedom of contract permits not only LD clauses but also penalty clauses. An LD clause operates as an exclusive remedy, whereas a penalty clause allows the obligee to claim the penalty in addition to actual losses. As Article 420(3) presumes any stipulated monetary obligation for breach to constitute an LD clause (i.e., an exclusive remedy), contracts should clearly state when a monetary obligation is intended to operate as a penalty.
Before the 2020 Civil Code Reform (the Reform), Article 420(1) prohibited courts from adjusting stipulated LD amounts. In practice, however, courts limited or invalidated excessive LD clauses under general principles such as public policy, good faith, or comparative negligence. The Reform removed the prohibition, aligning the statutory text with established judicial practice, but did not create an independent statutory basis for proportionality review. Courts therefore continue to rely on general principles when reducing or denying enforcement of LD clauses that are significantly disproportionate to actual damages — either partially or, in some cases, entirely.
Key practical points for parties to consider under Japanese law include:
- Reasonableness: Courts consider the contract purpose, negotiation process, nature and duration of the breach, the actual damages, the remaining term, and the burden on the breaching party. LD amounts should be justifiable based on these factors and the intended function of the clause at contracting.
- Avoid excessiveness: While there is no fixed multiplier or threshold, courts assess excessiveness based on the totality of circumstances. Courts may reduce or invalidate an LD clause when the stipulated amount is significantly disproportionate to actual or reasonably anticipated damages.
- Bargaining power: Courts may invalidate LD clauses where there is significant disparity or abuse of a superior position, including situations that may raise competition law concerns. In balanced commercial relationships between parties of comparable sophistication and negotiating power, courts are far less likely to intervene.
- Clear calculation methods: Where damages are difficult to quantify, objective formulas — particularly those supported by relevant precedent or statutory analogies — can help demonstrate reasonableness and may increase enforceability.
- Entire agreement provisions: Although Japanese law does not adopt the parol evidence rule, an entire agreement provision may, depending on its wording, limit the scope of extrinsic evidence considered when interpreting LD clauses. However, Japanese court decisions on their effect vary, and an entire agreement clause cannot exclude the application of general principles (e.g., public policy). Careful drafting remains essential.
- Recovery beyond the stipulated amount: Under the Civil Code, an obligee generally cannot recover actual damages exceeding the stipulated LD amount. Contracts should expressly permit such recovery if intended. Conversely, obligors may include a liability cap to ensure the LD amount also operates as a ceiling, as an LD clause alone may not exclude claims under other causes of action.
This article should not be construed as legal advice.

/Passle/5832ca6d3d94760e8057a1b6/SearchServiceImages/2026-03-31-12-35-55-395-69cbbfab295ee73eda661d7f.jpg)

/Passle/5832ca6d3d94760e8057a1b6/SearchServiceImages/2026-03-27-18-47-18-856-69c6d0b63b7265709d8aad6e.jpg)