Legally Bound

Caution needed with on-demand bonds

Jordan ... a bond will not be “read together” with the underlying contract.

Two recent court cases have reminded us of the importance of ensuring that parties think carefully before agreeing to provide performance security in the form of on-demand bonds.

We have looked before at the popularity of on-demand bonds in the Middle East. These bonds are an instrument which will pay out on the beneficiary’s first call for payment without delay and without condition, save for meeting the required formalities of making that call. Crucially, the bond provider does not need to see any evidence that there has been any breach in contractual performance, still less any loss arising from that breach. We do see on-demand bonds which require the beneficiary to include, in the demand notice, a statement that there has been a breach of contract; but this is pure formality: a true on-demand bond is a “no questions asked” instrument.

And that takes us, again, to the interesting question about on-demand bonds: the fact that they are attached to construction contracts which go to great length to allocate responsibilities, to identify breaches and to calibrate liability for those breaches. Having agreed this complex contractual balance, by what reasoning do contractors then agree to give a bond which will pay out regardless of any of that? Nobody intends a bond to be free money, so what protections do contractors believe are out there when there are none in the bond itself?

They might rely on the law. In all jurisdictions, there is recognition that a fraudulent call on a bond will not be enforced. Some jurisdictions will also not allow an “unconscionable” demand – that is, a demand where the beneficiary’s motivation is not fraudulent but there is no honest belief in the beneficiary’s right to make a demand on the bond. We have seen before that Singapore law is relatively more prepared to intervene in such situations to prevent a payment; English Law less so.

Or contractors might rely on the underlying contract. The thinking is: the bond might be on-demand but surely the purpose of the bond (to provide security in the event of breach and loss arising) is clear in the contract, so they should be read together? Let’s see if that reasoning works.

In South Africa, a contractor joint venture (between Aveng and Strabag) was appointed by the South African National Roads Agency (“the Agency”) to build a bridge. The contract required two on-demand bonds. Construction was heavily delayed and disrupted by protests from local people and eventually the works stopped. The contractor cited violence by protesters as making the works impossible to undertake safely, and eventually the contractor gave notice to terminate the contract for prolonged force majeure.

The Agency disputed the force majeure and itself gave notice of termination. While the dispute on force majeure was going through adjudication, the Agency refused the contractor’s request to confirm that the bonds would not be called, so the contractor applied for an injunction to prevent this, on the basis that the underlying contract contained restrictions on calling the bond, including requiring the Agency to show that it had lawfully terminated the contract.

That question eventually came to the South African Supreme Court of Appeal, which decided that this contract did not contain restrictions preventing the Agency’s call. The court agreed that an underlying contract can impose conditions preventing a call on such bonds but those conditions need to be entirely clear.

The second case is also in Africa. Here, Shapoorji Pallonji and Co Private Limited was engaged by Yumm Limited as contractor to build a power station in Rwanda. Again, an on-demand performance bond was given. In the midst of delay disputes, Yumm called the bond. Among several applications arising from that step, Shapoorji asked the English High Court to order the demand to be withdrawn.

The court went through similar reasoning as in the above case and found that Yumm’s demand on the bond met the formalities required, and that there were neither clear restrictions on making a demand (the English court terming these “conditions precedent”) nor any evidence of fraud, despite fraud having been alleged.

Both courts emphasised the autonomy of bonds, that is, they acknowledged that the parties expressly agreed the arrangement of an instrument which will pay out without the need to establish default.

So, returning to the above reasoning: a bond will not be “read together” with the underlying contract to work out whether the parties truly intended to agree it. Contractual restrictions on calling a bond can work if they are entirely clear. Simply stating that the bond is given as a guarantee of due performance, or that it may be called in the event of breach, might not be enough, depending on the governing law.  


* Stuart Jordan is a partner in the Global Projects group of Baker Botts, a leading international law firm. Jordan’s practice focuses on the oil, gas, power, transport, petrochemical, nuclear and construction industries. He has extensive experience in the Middle East, Russia and the UK.