Promises, promises

An indemnity is a contractual promise to be responsible for another’s loss.

It creates what lawyers call a “primary obligation”: the obligations of the person giving the indemnity are independent of the underlying contract to which the indemnity relates. If that underlying contract is set aside for any reason, the indemnity will remain valid.

What might a decent indemnity look like? Here’s one based on a 1997 agreement between two companies in the same group - Rust and PB - as part of a transfer of Rust’s assets to PB.

The agreement read: “PB assumes responsibility for the satisfaction, fulfilment and discharge of all the liabilities and the contracts of Rust outstanding at close of business on 31 December 1996 and PB hereby indemnifies and covenants to keep PB indemnified against all proceedings, claims and demands in respect thereof.”

Covering all liabilities.

So far, so robust. Or so you might think if you were Rust and looking to rely on the indemnity. In simple terms, the indemnity appears to cover all liabilities and contracts as at a certain date. And when Rust (now in liquidation but acting on legal advice) accepted liability for negligent geotechnical engineering work carried out before the 31 December 1996 date, it looked to PB to indemnify it for the just-over £8m of damages it owed by way of the judgment debt.

PB refused to pay, arguing that the indemnity didn’t apply. Mr Justice Akenhead agreed with PB. His judgment in Rust Consulting vs PB was the start of Rust’s problems. The judge held that the definition of contracts in the 1997 agreement, which effectively limited the indemnity to obligations yet to be performed, didn’t help
Rust because, by the time of the agreement, the geotechnical work had been carried out. The fact that the work was subsequently alleged to be defective - the breach of contract was latent and not patent - didn’t alter that conclusion.

The most obvious point to emphasise is the need for indemnities to be as widely worded as possible

Why Rust lost

More helpfully for Rust, the judge did accept that the definition of liabilities was wide enough to cover contingent or latent defects. However, Rust lost the case because the indemnity only related to “proceedings, claims and demands” in respect of those liabilities and the consent judgment wasn’t caught by that definition. In a conclusion that some may find surprising, the judge held that Rust had to establish that it was actually liable for the amount, notwithstanding the fact that it had agreed to the judgment debt.

Rust also raised an “estoppel argument”: because PB had been given notice of the claim, it should be forbidden (or estopped) from challenging the judgment as a result of the indemnity. Rust lost this argument as well.

Has the party giving the indemnity agreed to pay whatever sum the court may have awarded? The judge held that clear words would be needed to show this. PB hadn’t indemnified Rust in relation to, for example, “any relevant judgement sum”. Because the indemnity only covered actual liabilities, Rust had to show that the liability was properly due and owing.

Wording is key

So how might you avoid Rust’s fate? The most obvious point to emphasise is the need for indemnities to be as widely worded as possible. The indemnity in this case seems widely drafted: it wasn’t. Indemnities will be interpreted against the party trying to rely on them; this is a well-established principle, but the case is a helpful reminder of the consequences of that principle being applied in practice.

A party looking to enforce an indemnity is well advised to consider the following:

• Check the indemnity to see if the settlement will be binding
• Ensure compliance with any preconditions required by the indemnity
• Invite the indemnifying party to approve the proposed settlement beforehand
• Consider whether you can join the indemnifying party to any underlying litigation so that they are bound by any judgment.

Stuart Pemble is a partner, and James Richards a solicitor, at Mills & Reeve