How to get out of a guarantee

There are not many legal cases which are claimed to have a potential financial impact of £38bn across the property industry, or to represent ‘Armageddon’, but both these claims were made in relation to Prudential Assurance Company Ltd v PRG Powerhouse Limited [2007]. While that may have been a little over the top, it is not hard to see the reasons for alarm.


If you accept as a tenant a company which is not particularly strong financially, you may require that its obligations be guaranteed by its parent company. If the tenant company becomes insolvent, you have the guarantee to call upon. The significance of Powerhouse is that it represented an attempt to use an insolvency procedure to relieve the parent company of its guarantee liabilities – at the very time when landlords needed to call on them. The implications would have wiped billions off property valuations for institutional landlords.
The actual decision in the case favoured the landlords, and so the dire outcomes predicted have been avoided. Nevertheless, the judge’s reasoning has not closed the door to Armageddon as firmly as landlords might have wished.

#  PRG Powerhouse Limited (“Powerhouse”) was the third largest electrical retailer in the UK.
#  It was owned by PRG Group Limited (“PRG”), a company listed on the New Zealand Stock Exchange.
#  Powerhouse traded from 24 high street stores and 110 superstores in the UK.
#  A number of the landlords of these premises had taken guarantees from PRG in relation to Powerhouse’s obligations under the leases.
#  Powerhouse found itself in financial difficulties, and made proposals for a company voluntary arrangement, which was approved at a meeting of creditors in February 2006.


Company voluntary arrangements, or CVAs, were introduced by the Insolvency Act 1986. In essence, the procedure provides opportunity for a company which might otherwise be wound up to continue trading, if it can come to an arrangement with a sufficient majority of its creditors.

#  An insolvency practitioner called a ‘nominee’ assists the company to draw up the proposal, reports to the court on it, and calls meetings of the company and creditors to vote on it.
#  If it is approved by the required majority (75% by value of those voting), even those creditors who do not agree are bound by it as if they did.
#  Usually the same insolvency practitioner (now known as the ‘supervisor’) puts the proposal into effect.

Typically, the proposal involves some restructuring of the company’s business, together with wiping away historic debt by means of paying creditors only a proportion of what they are owed.


The principle of the CVA proposal made to (and approved by) Powerhouse’s creditors was a simple one. Some 35 under-performing stores needed to be closed and it was thought that Powerhouse would then be able to trade profitably from the remaining stores. PRG (as well as funding the insolvency practitioners’ fees of the CVA) would provide full parental support to the reconstructed business of Powerhouse.


As regards the closed stores, PRG would pay employees in full up to the termination of their employment, plus redundancy pay. Trade suppliers would also be paid in full. The landlords of those stores would also be paid in full up to a cut-off date.
As to rent falling due after the cut-off date, the landlords’ claims would be calculated as the rent for a specifed number of months (varying according to length of term remaining) and the landlords would be paid a proportion of those claims, anticipated to be in the order of 28p in the £. PRG proposed to make available a fund of £1.5 million for this purpose.


In a number of instances, PRG had given the landlords guarantees, the potential liability on which has been estimated in total at £75 million. If called upon to pay, PRG would in turn have a ‘back-door’ claim against Powerhouse under general guarantee law, so in order to draw a line under Powerhouse’s liabilities, this issue had to be dealt with as part of the CVA.

The solution adopted by the nominee was to propose, as part of the CVA, that PRG should be released from its guarantees. The landlords, understandably, were unhappy, and a group of them launched a challenge to the CVA.

The outcome is something of a do’s and don’t’s list for insolvency practitioners – and certainly doesn’t amount to saying that parent company guarantees can never be released by a CVA.


#  Don’t try to achieve a release with direct effect between the parent company and the landlords. The CVA gives rise to an arrangement between the debtor company and its creditors, and it is the debtor company which can enforce it, not any third party such as the parent company.
#  Do confine the terms of the CVA to arrangements between the debtor company and its creditors – even though this may have the effect of relieving a third party of liability. For example, had the CVA provided that the landlords should enter into agreements to surrender the leases, which in turn would have terminated PRG’s liability, there would have been no technical problem with that.
#  Do provide for the landlords not to enforce their rights against the parent company. This is an obligation owed to and designed to benefit the debtor company by preventing ‘back-door’ claims, even though it benefits the parent company as well. In Powerhouse, the court was prepared to find that this obligation arose by necessary and obvious implication from an express provision that the guarantees were to be “treated as having been released”. Again, as an arrangement between the debtor company and the landlords, there was no technical problem with that.
#  Do consider whether the proposal is “unfairly prejudicial” to the landlords. This is the rock on which the Powerhouse CVA foundered, in the end, and the court’s approach to the question will be closely studied by insolvency practitioners.

o Powerhouse argued that the court should not assume that the guarantees were of any real value; however, there was enough material before the court to justify an inference that they were. In a suitable case, though, expert evidence on the parent company’s finances could persuade a court otherwise.

o Powerhouse argued that the landlords would have the opportunity to re-let, perhaps on more advantageous terms, but this was dismissed on the basis that there was no evidence of that. Again, in a suitable case, there might be convincing evidence on this issue.

o The Powerhouse CVA proposal did not distinguish between those landlords of the closed stores who did not have the benefit of parent company guarantees, and those who did. Each group was to receive 28p in the £, regardless of the fact that the latter were additionally being required to give up the benefit of the guarantees. This arose because the landlords’ claims were to be calculated by reference to length of remaining lease term; in future, insolvency practitioners might consider adopting a different formula to take account of this point.
o The clinching point was that, had there been an insolvent liquidation, the landlords would still have had the benefit of the guarantees, and would have been the group of unsecured creditors which would have suffered least. Under the terms of the CVA, by contrast, they would be the group which would suffer most, receiving a dividend which placed no value on the lost guarantees, and which represented a fraction of their full claims, in order that other creditors could be paid in full.

o Finally, the court also noted the position of PRG, which hoped by this arrangement to reduce their exposure from £75 million to £1.5million, even though by entering into the guarantees they had assumed the risk of Powerhouse’s insolvency.


Powerhouse leaves it open, then, for future attempts by parent companies to divest themselves of guarantee liability. It is clear that there are structures which are quite workable legally, subject to the question of whether the landlords are unfairly prejudiced by them.
When it comes to questions of fairness, like questions of what is reasonable, there are always two sides to every story. While landlords might regard it as outrageously unfair that a parent company could evade its obligations in this way, from the point of view of the tenant and its parent it may seem iniquitous that the landlord can sit back and collect the rent from the parent company without being required to make any effort to re-let the premises, particularly where the parent company is trying to put in place a rescue package to save the tenant and protect creditors as much as is reasonably possible.

The question as to unfair prejudice need not inevitably be answered in the landlords’ favour, and much will depend on the commercial realities, as well as on the detail of the CVA proposal.


#  It has not been common for landlords to pay much attention to CVA proposals made in relation to their tenants. From now on, landlords will have to take an interest, and be prepared to get involved, attend meetings and try to head off this sort of scheme.

#  Landlords will be looking to the detail of their standard guarantee clauses. An ability to require the guarantor to take up a new lease in the event of the tenant entering into a CVA may be useful, in that the guarantor would be liable as principal debtor under the new lease, and unable to pass the liability along to the tenant. There could therefore be no justification for relieving the guarantor of that liability in the CVA.

#  Although the Powerhouse decision went in landlords’ favour, the scope for a similar scheme to work in future will strengthen a landlord’s hand in requesting a rent deposit in addition to, or in substitution for, a guarantee. This will not be popular with tenants.

#  A straightforward solution, from a landlord’s point of view, is to insist on the parent taking the lease, with a clause permitting intra-group sharing. A refinement on that could be for parent and subsidiary to be co-tenants, with provision restricting the ability of the co-tenants to claim indemnity as between themselves. Either requirement might make a rent deposit look more attractive to the parent.
Of course, how these issues play out in any individual situation is going to be a matter for negotiation and will depend largely on the bargaining strength of each party. It remains to be seen whether any settled alteration in market practice will emerge.

The Problem With Side Letters

Side agreements are often personal in nature, though this begs the question of which party they are personal to. In a recent case, a tenant successfully enforced a side agreement against its landlord, where the agreement had been entered into with the landlord’s predecessor in title. The obvious lesson for investors is to look into any such arrangements closely before purchase, but the decision illustrates why they can be less than ideal from a tenant’s point of view as well.


#  The legal elements of the case were straightforward.

#  The defendant (‘Deanwater’) was negotiating to take a lease, which was to contain a tenant’s break clause.

#  They anticipated exercising the break, but did not want to be saddled with a dilapidations claim when they did so.

#  The then landlord (‘Lionbrook’), assured them that no terminal dilapidations claim would be made, and on the strength of that assurance they entered into the lease, on terms including a tenant’s repairing obligation.

#  In due course Deanwater exercised the break; shortly after it had taken effect, Lionbrook transferred their interest in the property to the claimant, and also assigned to them the cause of action for breach of repairing obligations.

#  The claimant then served a schedule of dilapidations, and issued these proceedings.

#  The question of whether Deanwater was entitled to rely on the assurance given by Lionbrook was tried as a preliminary issue.

Both parties were agreed that:

#  The lease contained nothing which would prevent the landlord from seeking damages for breach of repairing covenant; but

#  If the assurance was effective to prevent Lionbrook from doing so, then they had had no cause of action to assign to the claimant, and the damages claim must fail.

The question was therefore a narrow one of whether the assurance was effective, and fell to be determined in a rather unsatisfactory manner: no-one from the claimant could give evidence as to the negotiations and assurance, and Deanwater’s director gave evidence that while the correspondence and other documents before the court were authentic, he could not remember anything further. The court therefore proceeded on the basis of the documentary evidence alone. Since the assurance was not given in writing (although there was a letter which clearly referred to the assurance having been given), the door was open to argument about what the parties had actually meant.


The applicable legal principles were not in doubt. If the assurance was to be effective, it had to be either as a collateral agreement, or by way of promissory estoppel.

A collateral contract may arise where party A says that he will only enter into a written contract (in this case the lease) if the other party, party B, agrees not to enforce some provision of it against him. This gives rise to a separate but linked contract, the consideration given by A being the signing of the main contract.

A promissory estoppel arises where party B represents to party A that he will not reply upon his strict contractual rights, and in reliance upon that representation party A alters his position significantly.

As the judge observed, where the alteration of position consists of party A entering into the contract itself, the two legal doctrines become all but indistinguishable.

On the basis of the evidence, the judge was satisfied that (a) there was an assurance by Lionbrook; (b) it was to the effect that there would be no terminal dilapidations claim; (c) it remained an effective assurance despite subsequent amendment to the lease; and (d) Deanwater relied on that assurance in entering into the lease.

It followed that Lionbrook had had no claim for damages for breach of repairing covenant against Deanwater, and the purported assignment of such a claim to the claimant was ineffective.


Does this amount to a general rule that successor landlords will be bound by side agreements entered into by their predecessors? Hardly.

#  In this case, the court was only concerned with assignment of a cause of action, since the lease had already come to an end before the claimant acquired its interest in the property. If the claimant had become landlord before lease termination, then s.3 of the Landlord and Tenant (Covenants) Act 1995 would have operated to transfer the benefit of Deanwater’s repairing covenant to the claimant.

#  That said, the burden of Lionbrook’s agreement not to enforce the repairing obligations would also have transferred to the claimant, as a ‘landlord’s covenant’ within the meaning of s.3, because it was given in a collateral contract. As mentioned at the outset, though, side agreements are often personal in nature, and if personal to the original landlord, they will not bind successor landlords (BHP Petroleum Great Britain Ltd v Chesterfield Properties Ltd [2001]).

#  In that case, the tenant is then left to fall back on the promissory estoppel doctrine, under which a successor landlord might still be affected. If the landlord’s interest happens to be registered, they might be bound by an overriding interest: this would probably come down to whether the tenant was in actual occupation at the time the landlord’s interest was transferred. If it is not registered, the tenant must try to establish that the successor had notice of the estoppel.

#  In this particular case, the landlord had raised pre-contract enquiries about the dilapidations position, and had been notified that Deanwater considered the landlord had no right to pursue a dilapidations claim, but they had not inquired further. This would probably have been enough to affect them with notice of the estoppel, but that is quite fortuitous; such matters were outside Deanwater’s control.

It has to be the conclusion that fortune was smiling on Deanwater in this instance. Lionbrook was ultimately not unwilling to make amendments to the form of lease, but Deanwater did not press them to ensure that the lease correctly reflected the agreed position on dilapidations. Nor, having failed to do that, did they insist upon an unequivocal written side agreement. Plus, because of the pure chance that the lease terminated before the transfer of the landlord’s interest, they were not concerned with the ins and outs of the Landlord and Tenant (Covenants) Act 1995, overriding interests or proving that the claimant had notice of the estoppel.

Fortunate they may have been, but the failure to document the agreement correctly has cost them a trip to the High Court nevertheless – and since leave to appeal was granted, it may not be the end of the matter. The claimant, one imagines, will now be looking into the circumstances of its purchase, so there is potential for fresh litigation, too.

The moral? Ideally, agreed terms should be incorporated into the lease. If a side agreement is really necessary, it should be in writing, should expressly bind successor landlords, and should expressly state that it is a collateral contract for the purposes of the Landlord and Tenant (Covenants) Act 1995. Anything else is really inviting litigation.