I may be allowed to remark that it is difficult to understand why businessmen persist in entering upon considerable obligations in old-fashioned forms of contract which do not adequately express the true transaction.
Lord Atkin in Trade Indemnity v Workington Harbour and Dock Board [1937] A.C. 1 at 17, HL
1. Introduction
It is not uncommon, when negotiating with consultants and with contractors, whether for large or small projects, to be reminded in tired statements, heavy with condescension, of the financial strength and fine reputation of the other contracting party.
In the current environment, this counts for little. Names like Barings Bank may now be tarnished with incompetence and corruption, but prior to their collapses, they were names that commanded trust and respect. Similarly the recent collapses of Goldman Sachs, Bear Stearns and Lehman Brothers, and the apparently limitless underwrite accepted by the US Federal Government of Freddie Mac, Fannie Mae and the world’s biggest insurer (AIG) suggest that even the finest reputations can be worth little when the going gets tough.
It would be foolishly brave to suggest that we have the answers to events of such global impact, but there is a lot that can be done to improve practises and to reduce the risks in what is inevitably an uncertain industry.
This paper examines collateral contractual arrangements, which sit alongside the primary obligations of the parties under construction contracts:
(1) Parent company guarantees – which are designed to ensure that performance by project subsidiaries and specific companies is supported by their owners.
(2) Bonds – whether to secure performance by the contractor or the owner, or to protect the project against specific default.
(3) Off-site materials agreements – securing delivery of critical plant and equipment for the project.
(4) Payment of subcontractors – subcontractors typically undertaken the majority of the specialist work in the construction industry, yet the temptation is often too great for head contractors to avail themselves of the subcontractors’ balance sheets to fund the cashflow and retentions for projects.
2. Parent Company Guarantees
When the opportunity comes to be appointed for significant projects, the competitive appointment processes favoured by owners and developers (often a formal tender process) encourages contractors and consultants alike to put their best foot forward. To cast themselves in the best light, and to ensure that the client is under no doubt about their experience and suitability for the appointment.
This will tend to mean that all relevant experience and resources within the contracting company and any related companies are aggregated. Group accounts are provided, and the skills and experience from far flung related offices are called upon.
These corporate groups tend not, however, to be single legal entities. Typically, corporate divisions are established, based on geographical location, project or skills. Similarly, the shareholding tends to be tied up in trusts and holding companies, designed to maximise tax efficiency and to protect the owners, their assets and their families from creditors.
However, when tenders come in, it is also not uncommon to find that the proposed bidding party has the name of the famous company tendering for the job, but with “(Project X Development) Limited” or similar tacked onto the end. The significance of this becomes all too apparent at a time when it is far too late to do anything about it; for example, when you find that the famous company name has been changed and its 100 shares transferred to a far less well known company; or in liquidation, you discover that what looked like adequate capitalisation was secured and funded by shareholder or related party loans which exhaust the limited funds available to pay to your client and the rest of the unsecured creditors.
The difficulty with this scenario is that the contract did not secure the skills and resources which were promised with the tender. It is not surprising that most engineers and architects are reluctant to review company accounts and undertake investigations about the financial strength of tenderers before making recommendations about award.
One way around this issue is to procure a guarantee of performance directly from the holding company or major shareholder. When procuring such a guarantee, a number of issues need to be taken into account:
(1) Identity of the guarantor – this will depend largely on the nature of the guarantee, and how the corporate structure of the group is organised. A typical structure is a holding company, which is owned by a number of trusts and does little more than hold the shares in a number of subsidiaries. The skills and resources are frequently then held in other corporate divisions.
Once the various related parties have been identified, you need to ascertain where the financial value lies, and also where the skills and expertise is held. There is little point in obtaining a guarantee from a holding company, on the basis that is where the value lies, if that company does not bring with it the reputational risk and the human resources required to complete the work. Though, admittedly, the guarantee of financial support would be nice.
(2) Role of Surety – the nature of the guarantee, and the extent of the liability of the guarantor, is entirely dependent on the wording of the guarantee.[1]
A distinction is frequently made at common law between a guarantee given as a surety, who is liable for losses arising from a failure to perform, and as primary obligor, who may be required to step in and take over the obligations of the contracting party in the event of default.
In most situations, this position is clarified by avoiding the use of the term surety and ensuring the parent company is contracting as primary obligor.
(3) Corporate Control – where the guaranteeing party is genuinely a parent company or is able to exercise corporate control over the contracting party, it is frequently useful to require the parent company to covenant that until the project is complete:
(a) it will not transfer or otherwise dispose of the shares in the company in such a way that it affects the beneficial ownership and control of the company,
(b) it will not enforce the payment of any debt from the company to the parent, and
(c) it will not otherwise take any action which might impact upon or adversely affect the ability of the company to perform its obligations.
While these provisions may seem draconian to some, their purpose is simply to ensure that the skills and resources offered during tender are applied to the project, and that the parent company does nothing to impair that performance.
(4) Release – depending on the terms of the guarantee, typically the guarantor will be released on actual completion of the work. This will not release the guarantor from breaches which predated completion, for example in the case of latent defect in the works.[2]
There are a number of situations in which, unless the issue is specifically covered in the guarantee, the owner may inadvertently release the guarantor from its obligations:
(a) Repudiation of the contract – where the owner repudiates principal contract, and the contractor accepts that repudiation, the guarantee will also be discharged. As with all such matters, there is an element of degree, in relation to the nature of the breach and the risk undertaken.[3]
Conversely, if the owner innocently accepts a repudiation by the contractor, the guarantor will not be released.[4]
(b) Fraud – normally, a fraudulent act on the part of the owner which is material to the obligations of the guarantor or permitting fraud will discharge the guarantee. However, where a clerk of works deliberately covered up defective work, which was discovered after completion and full payment, the guarantor was not released, regardless of the fact that the certificate of completion had been issued.[5]
(c) Non-disclosure of a relevant fact – there is no general obligation akin to material disclosure in the insurance industry (not being contracts of the utmost good faith). However, in some circumstances where the contract is beyond what would normally be expected between the parties, the guarantee may be discharged. This would not extend to circumstances like ground conditions where the contract allocated that particular risk to the contractor.[6]
(d) Laches – conduct between the contractor and the owner, without the knowledge or consent of the guarantor, which is prejudicial to the guarantor’s interest.[7]
(e) Material alteration to the contract – much like laches, if the alteration is prejudicial to the guarantor, the guarantor may be discharged. Where a contract contains a variation clause or allows for extensions of time, then ordering a variation or granting an extension of time will generally not be sufficient to discharge the guarantor. However, if the extension is a general waiver, for example to put time at large, or the variation is beyond the scope of work, for example new work which would not generally fall within the concept of a variation, then the guarantor may well be discharged.[8]
From a drafting perspective, these issues are simply resolved by expressly providing that the obligations of the guarantor are unaffected by, the guarantor also waives notice of, any of these matters.
3. Bonds
Form of Bond
At its simplest, a bond is a promise by a third party bank or other financial institution to pay a certain sum of money, in the event that certain conditions are met.
In practice, bonds take two forms – default or contingent bonds, and on-demand bonds.
(1) Default or conditional bonds
In the construction industry, most contractors will offer a bond issued by a bank or contractors’ bonding company and the contractor “jointly and severally”, promising to pay a sum equating to somewhere between 5% and 10% of the contract sum in the event of default by the contractor.[9]
Requiring the contractor to sign this form of bond only further confuses the issue, in my view. The Contractor has no obligations under the bond, and should have no involvement beyond procuring the bond from its bankers.
Most owners and banks, and many contractors, believe that if you make a demand under this form of bond, then the bond issuer will simply pay the amount claimed. Sadly, nothing could be further from the case.
It is incumbent on the bond issuer to enquire as to whether or not there has been default, and to satisfy itself that the amount claimed is in fact due.[10] This effectively means that a bond issued in this form is little more than a guarantee for liquidated damages, that can only be called once the contractor has accepted its default, or that matter has been determined in arbitration and on appeal.
It can also mean that the bond issuer is a guarantor, and may be called upon to perform the contractor’s obligations.
In such a circumstance, it is extremely unlikely that a bond for between 5 and 10% of the contract sum will go very far in covering the damages and costs awarded in arbitration and liquidated damages for delay (typically capped at 10% of the contract sum).
It is useful to compare this approach with the US, where conditional bonds are commonly used and the bonded sum is typically for the entire contract sum.
(2) On-demand bonds
The alternative is an on-demand form. These bonds are not preconditioned by default or by proof of loss. There is an implicit assumption in the contract, as opposed to the form of bond, that the client who calls the bond will account to the contractor for the amount paid by the bondsman. [11]
The benefit of these bonds is that the client can make demand at the point in time where the contractor is actually in default, rather than at the end of costly and uncertain litigation. Where the contractor is refusing to comply with its contractual obligations, or the client has decided to terminate the contract, the amounts typically secured by bonds in New Zealand (between 5 and 10% of the contract sum) should usually be sufficient to cover the client’s additional costs in procuring another contractor to complete the work. Calling the bond is, all too frequently, the end of the matter.
There are considerable difficulties with these forms of bond for contractors, not least the impact on working capital (see below) and the risk that an unscrupulous client will call the bond without justification.
In the case of Edward Owen v Barclays Bank,[12] an English company supplied glass houses to Libya. Performance was secured by a bond for 10% of the purchase price, to be paid “on demand without proof or conditions”. The Libyan purchaser called the bond once the glass houses were delivered. The English company sought to injunct its bank from paying out on the demand.
In the absence of fraud, Lord Denning held that the bank must pay, observing that such bonds were “virtually promissory notes payable on demand”, and warning of the consequences to the entire banking industry if such demands were not met.
While this case is salutary, it speaks more of the risks involved in contracting with unscrupulous clients rather than any ills inherent in on-demand forms of bond. It should also be noted that the grounds for resisting demand (as opposed to trying to prevent payment by the bank) have widened in the years since the Edward Owen case was decided. Most courts would have little compunction in implying into most contracts a negative pledge not to call the bond without good cause, and to account for all amounts paid.[13]
There has been considerable discussion, and no shortage of heat, surrounding the issue of the form of bond, who it should be issued by and in what circumstances it may be called. This is not helped by the confusing and archaic language frequently used in traditional forms of bond, as noted by Lord Atkin in the quote at the head of this paper.
Unless the purpose and intention of bonds is clarified between the parties, the situation is unlikely to improve. In New Zealand, it is fair to say, that most banks would be unpleasantly surprised to learn that the form of bond attached to the industry standard form of construction contract (NZS3910) is actually a default bond which requires them to enquire as to the nature of the breach and the losses sustained before any payment is made.
Bond issuer
The next issue of contention is who should issue the bond – a bank or a surety/ insurance company. In general terms, provided the form of bond is agreed and the issuing party is financially sound (usually by reference to Standard & Poors or similar credit rating), this should not cause any difficulty. Like many issues in the construction industry, it is often not quite so straightforward.
Banks typically offer the best quality of covenant, and it matters little to the banks whether the bond is in on-demand or conditional form.[14] They will go through the same assessment, and charge the same amount for the bond regardless of its form (most recently in the region of 1.75% of the bonded amount).
Of more difficulty for the contractors, the banks treat the bond in the same way as any other loan facility. They will review the creditworthiness and the asset backing of the contractor to consider whether or not the contractor can repay the amount secured (if demanded) and, if it can’t, whether or not such a sum can be recovered from secured assets.
In effect, this will mean that in New Zealand where contractors are, in the main, poorly capitalised, the amount secured by the bond will be reduced from the contractor’s working overdraft or other loan arrangements. So while the amounts charged by banks are relatively modest, contractors are generally reluctant to provide such bonds. For owners, this can also be an early indication of the capital strength (or lack of it) of the contractor.
There is a further issue with the bank issued on-demand bond, and that is while the bank will have satisfied itself that there is adequate asset backing for the amount secured, calling the bond will frequently be an event of default under other banking arrangements, and will often put the contractor into receivership. All too frequently, therefore, the decision to call the bond will be accompanied by a notice of termination.
Insurance companies generally take a different view, treating bonds as contingent liabilities or insurance contracts rather than loan facilities, taking them “off balance sheet” for the contractor. The cost of such bonds is calculated on the same basis as insurance premiums (anywhere between 3.5 and 5%).
The difficulty with such bonds, aside from their expense, is that they will typically be in default or contingent form. In the context that the cost is always passed on to the client, and they can only be called at the conclusion of proceedings (either in arbitration or court), these forms of bond are of dubious worth. It is understandable why in the US such bonds cover the entire contract amount.
When used
Bonds are generally required in six distinct situations:
(a) Tender bonds – securing a sum in the region of 2.5% of the project budget, these are to ensure compliance with tender procedures. These are not commonly used in New Zealand, but they are used in jurisdictions of uncertain practices.[15]
(b) Principal’s Bond – there is allowance for a bond to be provided by owners in some standard forms.[16] While used only very rarely, in the current economic environment they should be considered more often by contractors than they typically are.
(c) Performance bonds – securing 5 to 10% of the contract sum, usually to cover the disruption and additional costs of rectifying a breach or engaging another contractor to complete the works.
(c) Advance Performance Bonds – securing an advance payment of up to 10% of the contract sum to enable the mobilisation of the contractor or the early procurement of contractor’s equipment or goods and materials. The amount secured by the bond will usually reduce by a percentage of payments certified (eg 25%), after an initial threshold has been passed (eg payment of 10% of the contract sum by instalments).
(d) Offshore Manufacturing Bonds – where a significant element of the works is to be fabricated or manufactured overseas, the client has very limited if any security over the goods and materials. The bond typically secures all amounts paid by the client for such goods and materials, and is released once the goods and materials are delivered to the site. Typically, providing a bond is more cost effective than withholding payment for the off-site goods and materials until delivery.
(e) Retention Bonds – in the course of the works, most forms of contract provide for retentions of up to 5% of all progress payments. Traditionally, the retentions are held for the rectification of defects which are discovered following completion, and are only released on the expiry of the defects period and entire performance by the contractor. While there have been constructive trust arguments in the past, such retentions are typically treated as a fund held by the client against breach by the contractor.
Like all bonding arrangements, the cashflow of the project is of more value to the contractor, and justifies the cost of an on-demand bond to secure the release of retention moneys.[17]
For the client, careful consideration needs to be given to releasing funds against a bond; funds can be released in a controlled way and they generally command more attention from the contractor than a threat to call the bond. Despite their “promissory note” nature, demands are rarely made against bonds. Conversely, to make a deduction from an amount otherwise due to the contractor to cover the cost of defects which the contractor has lost interest in rectifying is relatively standard practice. This approach is not at all so easy when the owner holds a bond, rather than a pool of funds to be released to the contractor.
This issue also brings the conflict between default bonds and on-demand bonds into stark relief. There is really no excuse for any owner releasing retentions against the default form of bond in the Fifth Schedule of NZS3910:2003.
There are varying justifications for bonds, the amounts secured by them, who they are issued by, when they are released and when demand can be made under them. The important point is to be clear why the bond is required and to look critically at the form proposed. Conversely, simply following the form of bond attached to the published standard contract may have some surprising and unintended consequences.
4. Off-Site Materials Agreements
As the saying goes, possession is 9/10 of the law.
In most standard forms of construction contract, title to work in progress vests at a minimum on payment,[18] more reasonably on the earlier of delivery to site and entitlement to payment,[19] and alternatively, whatever is on site, and anything off site which belongs to the contractor and is marked as being for the contract.[20]
The difficulty from a security perspective is that the contract price is paid by instalments, which will include payment for goods and materials procured and stored or fabricated off the site or outside the country. Clearly, the most conservative course of action is not to pay for any such goods and materials until they are delivered to the site where the owner can see them, and there is a reduced chance of the contractor or any one else removing them. That is, unfortunately unrealistic in most cases.
Ownership of goods and materials paid for should never be contentious as this should be manageable by the contractor. On that basis, while it may be helpful from a notice perspective, there is little legal benefit in trying to register a security interest under the Personal Property Securities Act.
From a practical perspective, the most that can be done is to:
(a) include a title provision in the construction contract, and require all such goods and materials which are off-site to be identified as belonging to the client and kept separate from other goods and materials on that site,
(b) procure an acknowledgement in similar terms from the owner of the site where the goods and materials are stored (and any financiers holding a general security agreement in terms of the PPSA), and
(c) require the contractor to provide a bond to the value of such goods and materials securing their delivery to the site (see above).
These scenarios can become complex.
Consider, for example, a situation where the owner has paid for the fabrication of certain essential goods. The contractor has, unfortunately, used the payments made by the client to ease the cashflow on another project. The goods are therefore either owned by the fabricator or subject to a common law lien. Without a security arrangement agreed with the supplier, if the main contractor becomes insolvent, the client is left with little alternative but to pay for the goods again.
Other graphic examples of this situation are the moulds for a vessel or for the segmented lining for an underground tunnel. While the design, or a licence to use it, may be held by the client, the moulds themselves rarely belong to the owner. In the event of default by the contractor, or the subcontractor supplying the moulds, the client will need to secure the use of the moulds to complete the works. In that scenario, it is possible to include a right to use the moulds which will meet the requirements of the PPSA (in NZ) and the equivalent legislation in other jurisdictions.
This will also create an insurable interest for insurance purposes.
5. Payment of Subcontractors
Increasingly, practitioners are extolling the virtues of a single point of responsibility in contracting, whether using novated design-build structures, or more purist design and construct forms of contract. Whichever contracting structure is adopted, the bulk of the work is inevitably undertaken by subcontractors.
With notable recent exceptions, subcontractors have become the construction equivalent of the canary down the mine. When the head contractor’s cashflow gets squeezed, the first thing they will do is to pass that on to the subcontractors, or to use the cashflow from one project to finance work on another.
It was for these reasons that the Construction Contracts Act 2002 outlawed conditional payment clauses,[21] provided for rights to be paid amounts claimed and to suspend the works if they’re not, and for the prompt adjudication of disputes. While laudable, and usually effective, the Act does rely on subcontractors understanding and exercising their rights. In practice are also inclined to be the ambulance at the bottom of the cliff.
The critical issue which the Act does not address is that clients and developers are ultimately the beneficiaries of the work, regardless of whether or not the subcontractor is paid. Most clients will advise subcontractors that the head contractor has been paid for the work (or not) and it is for the contractor to pay the subcontractors. While technically correct, such responses are unhelpful.
Increasingly, construction contracts include provisions requiring contractors to identify their subcontractors and the work they are doing and the amounts due to them in all invoices. In subsequent invoices, the contractor must then either provide a receipt showing that the subcontractor has been paid from funds previously provided, or justify any withholding from the subcontractor in terms of the contract. When coupled with the right for the client to pay the subcontractor directly, the difficulties of contractors using the cashflow from one project to finance another are considerably alleviated.
It is fair to say that this approach does not meet with universal approval. Typically this is on the dubious ground that “what the contractor pays his subcontractors is the contractors business”.
With all due respect to the proponents of this view, if head contractors had been honest and reasonable in the payment of their subcontractors, the Construction Contracts Act would not have been necessary. Clients are entitled to be satisfied that the payments they make are being used for their projects, and for clients with large capital expenditure programmes, they have a direct interest in ensuring that the subcontractors and suppliers are paid as they will be using them on future projects. Some would say they have a responsibility to ensure they get paid, though I’m not sure that has been successfully tested in a court of law.
The court of public opinion seems reasonably clear on this point. The recent collapse of the Kensington Park development, and the losses left in the hands of subcontractors and suppliers, is reminiscent of other high profile collapses in the late 1990s and early 2000s. In each case, the damage to the wider construction industry would have been considerably reduced had some one ensured that the subcontractors and suppliers were paid from progress payments.
6. Conclusion
It can be very easy to simply rely on what has been done in previous projects, or to follow what is printed in the standard forms. Sadly, it is only when the project has got into difficulty that you realise that your contract does not quite achieve what was intended.
Whatever contractual approach is adopted, the provision of collateral or secondary contracts to secure performance of primary obligations under construction contracts serve an important purpose; particularly if they are given by and appropriate party and they are drafted carefully to meet the particular purpose.
[1] Lewis v Hoare (1881) 44 L.T. 66, HL
[2] National House-Building Council v Fraser (1982) 22 BLR 143
[3] National Westminster Bank v Riley (1986) B.C.L.C. 268, CA.
[4] Lep Air Services v Rolloswin [1973] A.C. 349
[5] Kingston-upon-Hull Corporation v Harding [1892] 2 Q.B. 494
[6] See discussion by Lord Atkin in Trade Indemnity Co v Workington Harbour & Dock Board [1937] A.C.
[7] See Kingston-upon-Hull v Harding above.
[8] See Kingston-upon-Hull Corp v Harding above.
[9] The form of bond in the Third Schedule of NZS3910:2003 provides that the condition for payment “shall be null and void” if the contractor performs, which is effectively the same as conditioning the right to demand by default.
[10] Trafalgar House Construction v General Surety [1996] A.C. 199
[11] This was not the case in Edward Owen v Barclays Bank (discussed above), where Lord Denning described on-demand bonds bore the colour of a discount which should be taken into account in the price. This is generally not accepted as being applicable in the construction industry where the courts will imply an obligation to account – see Cargill SA v Bangladesh Sugar Corp [1998] 1 W.L.R. 461, CA. For a thorough discussion of the limits on unconscionable demands, see Lurie On-Demand Performance Bonds: is Fraud the Only Ground for Restraining Unfair Calls? [2008] ICLR 403-538.
[12] [1978] Q.B. 159, CA
[13] See John Lurie article in footnote 11 above.
[14] Under the ICC interbank rules, when dealing with demands under bonds and interbank transfers, banks do not enquire behind the reasonableness of a demand, or whether or not it has been properly made (see discussion of Edward Owen v Barclays Bank above).
[15] Typically Hong Kong and other jurisdictions with particular concerns with corruption.
[16] See clause 3.2 (and 4th Schedule) of NZS3910:2003.
[17] Somewhat curiously, the form of bond in lieu of retentions in the Fifth Schedule of NZS3910:2003 is a default or conditional bond. Why a client would release cash in the hand against a default bond, for which the client would need to establish loss by obtaining judgment, is difficult to understand.
[18] See NZS3910:2003 clause 5.9.3.
[19] FIDIC construction contract (Red Book) clause 7.7.
[20] NEC3 clause 70.
[21] Defined in section 13 of the Construction Contracts Act as being a clause which makes payment of the subcontractor conditional on receipt by the contractor of payment by the client.