Indemnities – unlimited limitation periods?

It is well known that parties rely on contractual indemnities to allocate risk and to provide an avenue for the indemnified to recover loss.


Indemnities are attractive not only because common law principles such as mitigation and foreseeability are avoided but because they play havoc with limitation periods. This article considers indemnities in the limitation context and highlights the grey area as to how long a party has to call on an indemnity and, in turn, how long the indemnifier stays on the hook.

Types of indemnities

Indemnities are broadly either:

  • Preventing loss indemnities (“hold harmless”): the obligation is to prevent the indemnified party suffering the loss or harm covered by the indemnity (see Firma C-Trade SA v Newcastle P & I Assn (The Fanti) [1991] 2 AC 1); or
  • Reimbursement indemnities (“make good”): the obligation on the indemnifier is to reimburse the loss or harm actually suffered, but not to prevent it from being suffered in the first place (see Caledonia North Sea Limited v British Telecommunications Plc (The Piper Alpha) [2002] 1 Lloyd’s Rep 553).

Indemnities can provide some certainty as to the recovery of loss in terms of scope and quantum. However, in practice, parties often overlook including an express provision as to when the indemnity expires. This creates uncertainty over how long a party will be able to call on an indemnity and, conversely, over how long the indemnifying party will be “on the hook” for its liability because the type of indemnity affects the start of the limitation period, i.e. when the breach occurs. To illustrate, breach of a “hold harmless” indemnity occurs when the indemnified party suffers harm of the kind covered by the indemnity whereas breach of a “make good” indemnity occurs when the indemnifier refuses to honour the indemnity by failing to reimburse the indemnified party for the losses covered by the indemnity.

Limitation periods and indemnities generally

The law on limitation periods in New Zealand is primarily governed by the Limitation Act 1950 (LA1950), the Limitation Act 2010 (LA2010), and in the construction context, by the 10 year longstop period provided by the Building Act 2004 (BA2004).

The LA2010 is of greater significance to practitioners because the LA1950 only applies where an act or omission that forms the basis of a cause of action occurred on or before 31 December 2010 (LA2010, section 59). The cause of action under an indemnity is breach of the indemnity, which is a breach of contract claim. Under section 14 of the LA1950, an action accrues at the first point in time when everything has happened that would need to be proved to enable a plaintiff to obtain judgment for a sum of money (see also Mt Albert City Council v New Zealand Municipalities Co-Operative Insurance 2 NZLR [1981] 708 at 713.). From that point, an indemnified has six years to bring a claim (LA1950 section 4). Therefore, where breach of an indemnity has occurred on or before 31 December 2010, most claimants will now be out of time to bring a claim.

The LA2010 provides three limitation periods of relevance to indemnities:

  • The “primary period”, which is six years from the date on which the claim for monetary relief is based (LA2010, sections 11(1) and 12);
  • In certain circumstances, a “late knowledge period”, which is an additional three years from the end of the primary period (see LA2010 sections 11(2), 11(3), 14 (1) & 14(2)); and
  • Where a claim is based on an obligation not enforceable until a demand is made, time does not run until the date of default on the demand (LA2010 section 5(1)(a)). This is relevant to “make good” indemnities and is considered below.

It is also important to note that, under section 41, parties are able to contract out of the LA2010, which can adjust (or remove) the limitation periods above to suit the parties’ agreement. If proceedings arise from “building work” (defined in section 7 as work for, or in connection with, the construction, alteration, demolition, or removal of a building, including design work), the BA2004 provides a 10 year long stop from the date of the act or omission on which the claim is based (section 393). In the context of indemnities, time will start to run under the long stop at the point of breach of the indemnity.

On demand indemnities

“Make good” indemnities often require as a condition precedent that a demand is made on the indemnifier. The cause of action for breach of the indemnity will not arise until demand is made and the indemnifier fails to honour that demand, which may not be for some time after the losses were in fact incurred. There is limited New Zealand case law concerning on demand indemnities. However, section 5 of the LA2010 was considered in Official Assignee v Wheeler [2015] NZHC 1644. In this case, Mr Wheeler’s assets were sold to a family trust for $2 million in 2004. No money changed hands but the $2 million was listed as a debt owed to Mr Wheeler by the trust. In 2014, Mr Wheeler was declared bankrupt and the Official Assignee made demand in 2014 on the trust to repay the $2 million balance. The trustee argued that the agreement was entered into in 2004 so the statutory limitation applied and the Official Assignee’s demand was time barred.

Venning J held that on the terms of the agreement the trustee’s obligation to pay was not triggered until demand was made. As the demand was made in 2014, section 5 applied and the date from which the limitation period commenced was the date on which the trustees defaulted by failing to pay the Official Assignee’s demand (para [44]). This case confirms that where there is an “on demand” condition in an indemnity, time will not run until demand has been made. However, more concerning is that there is no apparent “expiry” of the ability to call upon the indemnity – which in practice could potentially leave the indemnified party “on the hook” indefinitely.

A similar issue may arise where loss covered by a “hold harmless” indemnity may not be ascertainable for some time, for example, where the loss protected by the indemnity is limited to the costs of hiring a replacement contractor to complete works under a building contract. In Thom v Davys Burton [2008] NZSC 65, the Supreme Court said: “Damage will be contingent, and hence not actual for limitation purposes, if the plaintiff will suffer no damage at all unless and until a contingency is fulfilled. That will be so if the damage results from the plaintiff being exposed to a liability which is contingent on the occurrence of a future uncertain event. A good example is where the liability is that of a guarantor and is contingent on a default by the principal debtor, in contrast to the undertaking … of a direct and present liability which falls due in the future.”

In such circumstances, it appears that limitation will not run until the loss is ascertained or the contingency is fulfilled, again apparently leaving the indemnifier on the hook for an indefinite time. (See further Wardley Australia Limited v The State of Western Australia (1992) 175 CLR 514, and the Canadian case of Penhold (Town) v Boulder Contracting Limited [2009] ABQB 550.) In light of the above, whilst it would be expected that a party who has suffered loss that is recoverable under an indemnity would seek immediate redress, it does highlight risks that parties should consider when agreeing to (or accepting) an indemnity.


The legal operation of an indemnity is ultimately a question of construction – what did the parties agree? However, unless parties expressly limit the period within which the indemnified is able to call on an indemnity, the indemnifier is arguably accountable for an uncertain period of time. It may be helpful to consider the approach taken in Canada to avoid this, where some standard form construction contracts now contain an express period to call on an indemnity, and contrast it with clause 7 NZS3910:2013 which does not (see A Wallace and V Merritt, “Contractual Indemnities in Construction Contracts: A Contractor’s Perspective”, (2017) J. Can. C. Construction Law 1). This issue is exacerbated when a guarantor agrees to guarantee the indemnifier’s obligations. The lack of case law on the timeframes within which the indemnified must call on the indemnity is not helpful. However, in practical terms, these may not be issues that arise frequently because parties are usually motivated to call on indemnities as soon as possible, given solvency risks for the indemnifier and the commercial desire to address issues as they arise and move forward, rather than pursue legacy claims. Nevertheless, when considering whether to offer an indemnity, the indemnifier should turn its mind to the extent to which the indemnity cuts across applicable limitation periods. 

Contact Us
Phone 09 303 5270
Fax 09 309 3726