The High Court has held that an attempt by a tenant company, PRG Powerhouse Ltd (Powerhouse), to strip away landlord guarantees provided by its parent, PRG Group Ltd (PRG), through the use of a company voluntary arrangement (CVA) was unfairly prejudicial to those landlords.

Much of the press coverage of the case, Prudential Assurance Company Ltd and Luctor Limited and others v PRG Powerhouse Ltd [2007] EWHC 1002 Ch, has concentrated on the landlords’ victory, but it is worth noting that it was partial only. While the decision clarifies that a CVA may not directly remove landlords’ guaranteed rights to rental streams, Etherton J also held that a CVA can, and did, have the indirect effect of releasing their guarantees. On the facts, such treatment was unfairly prejudicial to the landlords’ interests. However, similar structures might be used in the future and will need to be analysed on a case by case basis to see if there is any unfair prejudice. This briefing examines the case and what it might mean for you

What is a CVA?

A CVA is an arrangement between a company, its shareholders and its creditors. It is a procedure whereby the directors (or a liquidator or an administrator) propose a reorganisation plan that involves delayed or reduced debt payments or a capital restructuring. A CVA may be challenged on the grounds that it unfairly prejudices the interest of a creditor, member or contributory of the company or there has been some material irregularity at or in relation to either of the meetings (section 6(1) of the Insolvency Act 1986)


Powerhouse owned a number of high street electrical stores and superstores that it had purchased in September 2003 with financial support from PRG.

A number of the landlords of those stores, including Prudential and Luctor (the Guaranteed Landlords), took parent guarantees from PRG in respect of Powerhouse’s obligations under the leases (the Guarantees). Powerhouse got into financial difficulties and closed 35 underperforming stores (the Closed Stores) while keeping open 53 other stores it hoped it could trade profitably. The directors of Powerhouse proposed a CVA. Broadly speaking, the CVA proposed that all creditors of the Closed Stores, including the Guaranteed Landlords, were to receive a dividend of 28 pence in the pound on their respective claims. All other creditors were to be unaffected by the CVA. In particular, the CVA purported:

  • to release the Guarantees in return for the dividend paid by Powerhouse to the Guaranteed Landlords ie it tried to strip away the Guarantees (clause 3.12); and
  • to oblige the Guaranteed Landlords not to claim against PRG under the Guarantees ie they were to be treated as having been released (clause 3.15).

Not surprisingly, the CVA obtained the requisite statutory majority at a meeting of all the creditors of Powerhouse, which included the creditors whose rights and obligations were not affected by the terms of the CVA. Powerhouse subsequently went into administration on 1 August 2006, although the CVA was not terminated.

The Guaranteed Landlords issued proceedings challenging the CVA. There were two preliminary issues before the High Court:

  • whether the Guarantees were released or should be treated as having been released by reason of the CVA (the first preliminary issue); and
  • if so, whether the CVA unfairly prejudiced the interests of the Guaranteed Landlords (the second preliminary issue).


First preliminary issue

Etherton J held that only Powerhouse, and not any third party (ie PRG), had the benefit of and could enforce the rights and obligations conferred by the CVA. The statutory CVA mechanism did not enable clause 3.12 to operate directly to release PRG’s liability under the Guarantees.

However, a CVA could provide that a creditor cannot take steps to enforce an obligation of a third party to the creditor that would give rise to a right of recourse by that third party against the debtor company. There was no difference between an obligation of a creditor not to enforce a contract with a third party and an obligation to deal with the third party as if the creditor’s contract with it did not exist. Accordingly, clause 3.15 was effective to impose an obligation on the Guaranteed Landlords not to claim against PRG under the Guarantees and to treat them as though they had been released.

On the first preliminary issue, therefore, the Guaranteed Landlords were only partially successful (resulting in their winning only 75 per cent of their costs).

Second preliminary issue

As the court held that the Guarantees had been indirectly released, Etherton J had to consider whether the treatment of the Guarantees under the terms of the CVA was unfair. In doing so, one of the comparisons that Etherton J found helpful was with the position if, instead of a CVA, there had been a formal scheme of arrangement. In that situation, the different classes of creditors would have been required to meet and vote separately. He held that the fact that a CVA involves differential treatment of creditors is a relevant factor but will not necessarily be sufficient to establish unfair prejudice. While a particular class of creditors might have blocked a scheme, that would not automatically mean they have been unfairly prejudiced where a CVA is used.

Etherton J found that the Guarantees were of value, particularly as PRG was solvent, and would have been enforceable but for the CVA. However, under the terms of the CVA the claims (present, future and contingent) of the Guaranteed Landlords were to be discharged ‘at a fraction of their value’ so that other creditors could be paid in full.

The Guaranteed Landlords were prejudiced most by the CVA because they were to receive the same dividend as all other creditors, irrespective of their Guarantees. The CVA effectively valued the Guarantees at nil. In fact, the Guaranteed Landlords would have suffered least on an insolvent liquidation of Powerhouse where all the other unsecured creditors would have received nothing.

Such an ‘illogical and seemingly unfair result’ could not have been achieved via a scheme, as the Guaranteed Landlords would have formed their own class separate from other unsecured creditors and vetoed the scheme. In addition, the scheme would not have included creditors who were to be paid in full. By using a CVA, ‘the votes of those unsecured creditors who stood to lose nothing from the CVA, and everything to gain from it, inevitably swamped those of the Guaranteed Landlords who were significantly disadvantaged by it.’ Accordingly, in failing to put a value on the Guarantees, Etherton J held that the terms of the CVA were unfairly prejudicial to the Guaranteed Landlords.


The commercial property market had been worried about the possibility that CVAs might be used to remove landlords’ rights to guaranteed rental income. The concern was that this would have had a substantial effect on the valuation of leasehold property where a parent guarantee was used to support a tenant’s covenant strength.

While the decision clarifies that a CVA may not directly release guarantees provided by, for example, the debtor’s parent, it can have the indirect effect of doing so by obliging a creditor to treat the guarantee as though it had been released. As Etherton J made clear, the fact that a CVA involves differential treatment of creditors will not necessarily be sufficient to establish unfair prejudice. Landlords will need to consider on a case by case basis whether or not they have been unfairly prejudiced where a CVA has the effect of indirectly releasing their guarantee. The onus will be on the landlords to challenge the CVA and prove unfair prejudice.

Of course, it is not just landlords who should take note of this decision – any creditor who has a guarantee should be aware of its implications. Where a CVA purports indirectly to release a creditor’s guarantee, the key question will be whether sufficient value has been ascribed to the release such that the CVA is not unfairly prejudicial.

It is too early to predict whether the decision might lead to changes in market practice when drafting leases, rent guarantees or even general financing arrangements. For example, a lender could try to restrict a borrower from including a term in any future CVA that prevents the lender from claiming under its guarantee (although there may be a public policy issue in a lender purporting to restrict the possible terms of a borrower’s CVA). Of course, failure by the borrower to adhere to such a term would simply give rise to contractual damages – a claim that may itself be compromised under the CVA! Alternatively, a landlord might try to insist that a guarantor agrees, in the event of a CVA of the tenant, not to have recourse to the tenant should it be required to meet its guarantee obligations to the landlord.

This would seemingly remove the need for the tenant to seek to compromise the landlord’s guarantee rights. However, there would need to be compelling reasons as to why a guarantor would accept such a commercially unattractive position. It might also be possible to oblige the guarantor, in the lease, not to seek to discharge its liability to the landlord in a tenant CVA and to indemnify the landlord if a CVA does have this effect. These suggestions, however, have not been tested and it is unclear whether, ultimately, they would be effective.

It is clear from the decision that a CVA may be able to suspend a beneficiary’s rights under a guarantee provided such suspension is valued ‘fairly’. However, while fairly will not automatically mean full value, it is difficult to imagine a situation where anything less than full value would be appropriate, certainly where the guarantor is clearly solvent. Arguably, it might be commore difficult where the guarantor’s financial situation is more precarious – for example, in the case of a guarantee provided by one of the debtor’s subsidiary companies.