The Case of the Valid Repeat Guarantee
The case involving Adda Hotels really is the case that keeps on giving. The thrust of the litigation concerns whether a guarantee of an assignee by the same guarantor of the assignor is valid. To date the case has been fast-tracked through the Court of Appeal. Morgan J in the High Court has now determined further issues arising from that litigation. The decision highlights the inherent problems in giving effect to the Landlord and Tenant (Covenants) Act (the Act).
The law in practice
From a commercial perspective, it would seem straightforward - a tenant wants to assign to an inter-group company and its landlord is happy for the assignee to become the tenant so long as the same parent guarantees the assignee (a repeat guarantee). Every party is keen for this to happen and it makes perfect commercial sense.
Unfortunately, in respect of what should be an easy transaction, the Act lacks any form of business efficacy. By way of section 25, it has been held by the Court of Appeal that a repeat guarantee is void as the guarantor is not released from guaranteeing the assignee's obligations. The repeat guarantee amounts to an invalid guarantee because the Act provides for the release of an assignor and its guarantor (sections 5 and 24) but crucially it also provides that only an assignor can guarantee the assignee's performance of the lease covenants by way of an Authorised Guarantee Agreement (AGA) (section 16). There are no equivalent AGA provisions for a guarantor and therefore the Court of Appeal concluded that, where the original guarantor also guarantees the assignee, the subsequent guarantee was void irrespective of how much the parties wanted the transaction to happen and regardless of how much commercial sense the transaction made.
Background of Adda
In Adda, Morgan J had to consider the slightly unusual situation of an assignment that took place without consent. Consequently the original tenant (T1) and guarantor (G) were not released from their original obligations when the lease was assignment to the assignee (T2) (section 11). The landlord and tenant both wanted to unwind this assignment in order to restore the position whereby T1 became tenant again and G guaranteed T1. It should not have been difficult. The Act begged to differ. Due to the complications presented by the Act, the landlord and tenant disputed how the unwinding was to take effect.
Landlord's position - The landlord's position was the lease should simply be assigned from T2 back to T1 and G should guarantee T1. The tenant's view was that this would be void - the guarantee amounting to a repeat guarantee which the earlier case of K/S Victoria held to be void.
Tenant's position - The tenant's argument was that in order for the arrangement to be valid, T2 had to assign to a third party entity who would then onward assign to T1. G would guarantee T1. The tenant's view was that this would avoid the guarantee being a repeat guarantee and therefore the guarantee would be valid. The landlord was concerned that the third party may not assign and further that the arrangement itself amounted to a contrived arrangement to avoid the purpose of the Act (i.e. to release parties upon assignment).
Court's decision in Adda
Morgan J rejected the tenant's position.
He held that such an arrangement fell foul of section 25 as it operated to "otherwise frustrate the operation of… this Act" (section 25). The arrangement frustrated the provision in the Act that released G upon assignment (section 24(2)). Further, such arrangement in any event amounted to a sham deliberately designed to circumvent the anti-avoidance mechanism in the Act and for this reason would also be void. Morgan J's logic on this is absolutely right as it is a consistent application of K/S Victoria. It further emphasises the uncommercial nature of the Act - no matter how much the parties may have wanted the arrangement to happen and no matter how much sense the proposed arrangement made, the Act arbitrarily voids it.
Significance of Adda
What is of more interest is Morgan J's acceptance of the landlord's position.
In short, Morgan J held that the assignment by T2 to T1 and the guarantee by G of T1's tenant obligations was valid. On the face of it, the judge concluded that G could repeat its guarantee of T1 and that this would be perfectly valid.
On assignment by T2 to T1
- T1 was released from its obligations that continued following the unauthorised assignment;
- G was also released at the same time;
- T1 became liable for the tenant covenants again as assignee; and
- G became T1's valid guarantor.
Morgan J was able to thread the camel through the needle in reliance on the Court of Appeal judgment in K/S Victoria. Whilst the basic position is that a guarantor cannot guarantee the assignor's assignee, it is possible for the guarantor to guarantee the assignor's liability under an AGA. In summary, the guarantor was released "to the same extent" as the assignor (section 24(2)). Morgan J considered that it was open to him to interpret this in such a way so as to produce a "sensible commercial result".
On the facts, T1 is released on assignment and is then liable directly for the tenant covenants. G is also released and a direct guarantee would make G liable for the tenant covenants as well. By analogy with the above, Morgan J held that this arrangement amounted to a release of G "to the same extent" as T and, consequently, such release was valid. This approach is an extension of obiter comments by Lord Neuberger in K/S Victoria.
Does this mean that repeat guarantees should now be considered as valid? The answer is a firm "no".Adda should be contextualised within its own particular facts. In the case, the assignment concerned unwinding an unlawful assignment. T1 was taking liability for the tenant covenants again and G was simply mirroring T1's obligations. This is not the case where T1 is effecting a valid assignment to T2 and G is expected to guarantee T2. In this situation, G is guaranteeing T2 as opposed to T1's AGA and therefore this more typical situation will result in the guarantee being void as G is not released to the same extent as T1.
The Court did consider a rather left-field argument whereby T1 should not be considered released upon the assignment from T2 to T1 and therefore T1 and consequently G simply remain liable for the tenant covenants with no release taking effect to potentially void the guarantee. Morgan J rejected this as the Act cannot be interpreted in such a way that T1 is not released upon T2's assignment to T1.
Morgan J hinted at other alternative arrangements, although he did not comment in detail on them. These included rescinding the unlawful assignment or the parties agreeing to surrender the lease and regrant it to T1. The former arrangement would be for the Court to determine, if a suitable case presented itself. The latter surrender arrangement is unlikely to be appealing to a tenant given the likely SDLT consequences.
At the end of the judgment, Morgan J considered obiter a further alternative, namely the assignment of the lease by T2 to G and G assigning to T1 with a fresh guarantee from G. Morgan J held that the assignment by G to T1 with G guaranteeing (in the form of an AGA) would be valid. However, he remained concerned that the assignment from T2 to G would be void as a result of obiter comment by Lord Neuberger in K/S Victoria to the effect that an assignment from a tenant to a guarantor was void. That said, Morgan J noted that Lord Neuberger's comment was obiter and "somewhat tentative". He was asked the question if Lord Neuberger's statement was really correct, but declined to comment further. This is important as it casts doubt on the dicta in K/S Victoria concerning this issue.
Adda should be treated with caution and on its own peculiar facts. Adda is not authority for repeat guarantees being valid. The judge dealt with a straightforward set of facts in a complex statutory and precedent matrix and reached a practical solution that will allow the parties to effect their arrangements without risk of security invalidity. It is quite possible that the judgment would be susceptible to challenge in the Court of Appeal. However, that is highly unlikely as the parties have the result that they wanted.
Adda does continue to highlight the fundamental problems posed by the Act and the uncommercial and nonsensical situations that arise from its drafting and interpretation. Such problems are judicially recognised as well as, more importantly, by both landlords and tenants. Reform of the Act is therefore important in order to provide fairness and business efficacy in a context where parties should be free within reason to transact as they see fit. Reform of the Act is possible in such a way so as to protect the purpose behind its enactment as well as allowing business common sense to prevail where a tenant wants to assign to a third party and provide the same guarantor to that third party and the landlord commercially agrees to such assignment. In absence of reform, landlords and tenants will continue to have to find alternative ways to effect intra-group restructurings in compliance with the Act. This is an unsatisfactory state of affairs and highlights the need for swift reform of the Act.