The construction industry is high risk and disputes and insolvencies are common. As a result parties to a construction project typically seek to secure the counterparty’s contractual obligations by a variety of means, most frequently a bank guarantee. Principals often regard these instruments as the gold standard of security for its “cash-like” quality. This status stems from its unconditional nature which means that, if called on by the principal, the bank must unconditionally and on demand pay cash to the equivalent amount stated in that guarantee – irrespective of the terms of, or proof of breach of, the underlying construction contract.

But – what actually is a “bank guarantee” and, are they really “as good as cash”?

What is a bank guarantee?

What is frequently called a “bank guarantee” is in fact an unconditional performance bond given by a bank. The use of the term “guarantee” to describe such bonds has been deemed misleading by Courts.[i] This is because a guarantee, strictly speaking, is a contract of suretyship – where the guarantor (the “surety”) takes on a secondary obligation to which it is held only after the liability of the contractor is established.

Usually, an unconditional performance bond or undertaking from a bank (“bank guarantee”) or an insurance company (“insurance bond”) can be called upon by the principal without requiring proof of the contractor’s breach of the construction contract. In this case, the liability of the institution issuing such a bond is primary, and the bond is said to operate autonomously from the construction contract (however, as seen below, this is becoming less so with the erosion of the “autonomy principle”).

Set out below are common characteristics distinguishing an unconditional bond (either a “bank guarantee” or “insurance bond”) from a conditional undertaking or bond (“surety bond”, “contract of surety” or “guarantee”).

When can a bank guarantee be called on?

If a bond is unconditional, and intended to be cash equivalent, it can (subject to the below exceptions) be called on by the beneficiary upon written demand to the issuing institution, without regard to the underlying construction contract. This is what is known as the “autonomy principle.”

The beneficiary need only have a bona fide claim of a breach of contract.[ii]

Traditionally, Courts have been reluctant to interfere with the autonomy principle as the parties (often sophisticated commercial entities) have agreed to the unconditional nature of the security. Any such interference not only undermines the intention of the contracting parties, but also diminishes the commercial worth of these forms of security.

Here the intended purpose of the security comes into play.

It is common for security under a construction contract to serve a dual purpose, being:

  1. to secure the contractor’s performance of the contract / provide security against the contractor becoming insolvent
  2. to give the principal access to funds it claims notwithstanding a dispute with the contractor is on foot. Here, the security is called a “risk allocation device” as it is used to allocate the risk between the parties as to who will be out of pocket during a dispute under the contract.

Courts are more likely to uphold the autonomy principle where the parties have agreed that the security is also a risk allocation device.[iii] To allow an injunction to restrain a call on a guarantee that is intended to act as a risk allocation device would defeat the purpose of that security: i.e. that the principal will have access to funds during a dispute between the parties.

In complex transactions involving multiple parties and multiple security instruments, the intended purpose of a particular bond can be critical to a beneficiary’s right to call on that bond – as seen in the recent case of Kawasaki Heavy Industries Ltd v Laing O’Rourke Australia Construction Pty Ltd [2017] NSWCA 291 (Kawasaki) (see our recent case note here).

Exceptions to the autonomy principle

There are exceptions to the rule, and the three exceptions to the autonomy principle are:

  1. fraud by the principal
  2. unconscionable conduct by the principal (under section 20 of the Australian Consumer Law)
  3. an ‘underlying contract exception’.

Due to the commercial nature of performance bonds, the bar is set high for both fraud and unconscionability. For example, an element of predatory behaviour must be proven to establish unconscionable conduct.[iv] In a commercial context, disparity in bargaining power or use of superior bargaining power to achieve a legitimate commercial interest will not necessarily amount unconscionable conduct.[v]

Not surprisingly, the underlying contract exception is the most frequently argued exception. It is also the most contentious, as it provides the greatest opportunity for the erosion of the autonomy principle.

The underlying contract exception

This exception arises where a contractor asserts that the principal is attempting to call on the bond beyond the circumstances in which a call is permitted under the underlying contract, and applies for an injunction to prevent such a call.

For the exception to be established by a Court, there must be “clear words” in the underlying contract that limit the principal’s rights to call on the performance bond.[vi] However, the necessary “clear words” are often difficult to distinguish.[vii]

If established, the Court will then examine the terms of the underlying construction contract to determine whether the principal rightfully called on the bond in accordance with those terms.

An obligation on the principal to provide notice to the contractor prior to a call on security is an example of a term that may exist in the underlying contract, which if found to be expressed in “clear words”, could restrict a principal’s unconditional right to call on that security. This can be difficult to reconcile with the autonomous nature of an unconditional bond, especially if the terms of the unconditional bond itself expressly exclude any obligation on the principal to provide notice to the contractor prior to a call on the bond.

Fighting back with a no injunction clause – fruitful or futile?

One way principals have attempted to thwart a contractor from successfully obtaining an injunction restraining a call on a security is by including a “no injunction clause” in the underlying contract.

If Courts are more willing to look behind an unconditional bond to the terms of the underlying contract to determine whether an underlying contract exception exists, then it is thought that more weight will also be given to an express intention of the parties from precluding the contractor seeking an injunction from restraining a call?

However, the case law on “no injunction clauses” appears unsettled. Some Courts have held that a no injunction clause is invalid as it is an ouster of the Court’s jurisdiction.[viii] Indeed, it is a long standing principle that parties cannot contract out of the Court’s jurisdiction, or preclude a Court from granting a remedy at law.[ix]

However, some recent decisions have cast doubt on this position. In Anaconda Operations Pty Ltd v Fluor Daniel Pty Ltd [1999] VSCA 214, the Court held “It matters not that it may be argued that the latter part of this provision is bad as ousting the jurisdiction of the Courts. The important thing is that both parts of it show an intention that the owner is to be at liberty to call on the security at any time”.[x] Further, the Western Australian Court of Appeal held that if a no injunction clause is unenforceable as an ouster of the jurisdiction of the Court, it is nevertheless permissible to take it into account in construing whether the parties intended the principal to have unfettered recourse to the security.

Therefore, where the parties have expressly agreed that the contractor will not injunct a principal’s right to have recourse to security, Courts appear to struggle to reconcile that intention with an award of an injunction on the basis of a finding that the “no injunction” clause is invalid.

Practical take outs

The underlying contract

  • Whether you are the principal or contractor, always carefully consider the wording of any clause relating to security
  • Principal’s should:
  1. reflect the intended purpose of the security is in the contract. If it is intended to act as a risk allocation device, clearly state that the principal may have recourse to it, pending the resolution of a dispute between the parties
  2. consider including a “no injunction / obstruction clause” – however these may be held to be invalid as ousting the Court’s jurisdiction. It may survive if:
  3. the intention of the clause is to preclude a contractor from restraining a call where the principal has a bona fide claim to the security and the purpose of the security is to act as a “risk allocation device”. A “no injunction clause” that seeks to completely defeat access to the Courts, or preclude a party from seeking other kinds of remedy (such as damages) is likely to be invalid.[xii]
  4. the terms of the underlying contract, and the bond itself, are consistent and unequivocal with regards to the security being an autonomous, unconditional form of security which the principal may have recourse to for a bona fide claim
  5. the clause also contains less prohibitive restrictions on the contractor which may survive if the “no injunction” part of the clause is severed from the clause e.g. the contractor must not hinder or obstruct the principal’s right to recourse of the security.[xiii]
  6. ensure that there are no contradictory conditions or restrictions on the principal’s recourse to security in the underlying contract e.g. notice provisions
  7. always go for an unconditional performance bond by a bank or reputable insurance company over any other kind of security
  8. include a clause that the form of security must be on terms approved by the principal, or include an approved pro forma bond as an attachment to the contract.

Reviewing the terms of an unconditional performance bond If you are a principal:

  1. make sure it includes the words: unconditional and on demand
  2. make sure it is from a bank or a reputable insurance company
  3. include an assignability clause
  4. resist an expiry date, but if one is needed ensure it is suitably in the future (well past the likely date of final completion)
  5. carefully check all details are correct and are consistent with the terms of the underlying contract (party names, ACN’s, the amount of the bond and any notice provisions)
  6. execute as a deed to avoid any issues of consideration.