Practitioners will be familiar with the oft quoted phrase that ‘the court is not a rubber stamp’ (Kelley v Corston [1998] 1 FLR 986). The court must continue to exercise its discretion under s 25 MCA 1973 even when presented with an agreement between parties. The court will scrutinise the agreement to ensure fairness in all the circumstances of the case.

Where an issue arises as to: (1) the exact terms of any agreement reached and whether they should be sanctioned; and (2) whether an agreement has been reached, the guidance in Xydhias v Xydhias [1999] 1 FLR 683 comes into play. As recently summarised by Moor J in Pierburg v Pierburg (No 2) [2022] EWHC 2701 (Fam), [2023] 2 FLR 81:

‘[43] Xydhias is authority for the proposition that, in relation to agreements reached in the family law context, ordinary contractual principles do not apply. As the final award was always fixed by the court, the purpose of negotiations was to reduce the length and expense of the legal process. The court has a discretion in determining whether an accord has been reached. Moreover, even where an overall settlement had been agreed, there might well be issues remaining, for example as to the drafting or exact terms of the order, that the court would be able to determine without undermining the overall agreement.’

The facts

The factual background to Xydhias is well-known. In the run up to a final hearing, there were lengthy pre-trial negotiations. As the parties’ positions grew closer, proposals shifted to consent order form, with five versions passing between the parties. By the latter drafts, the focus was on drafting, correction of errors, and proposals for security of the lump sum payment.

During the back-and-forth of draft consent orders, the wife’s solicitors informed the court that the final hearing would not be needed, where ‘heads of terms have been agreed’, subject to terms of security and the duration of certain continuing obligations. The court was also told that a short appointment should be preserved for approval of a consent order or for negotiation of outstanding points.

At that hearing, the husband withdrew from all negotiations. The wife sought to enforce the agreement that she asserted had been reached.

The district judge found ‘the essential building blocks of an agreement were in place’ and excised from the final draft order those matters which the husband had not agreed. The husband’s first appeal was dismissed.

The judgment

The husband’s second appeal was to the Court of Appeal. In dismissing that appeal, Thorpe LJ came to the ‘cardinal conclusion … that ordinary contractual principles do not determine the issues in this appeal’. Thorpe LJ set out the applicable principles at 961 onwards, and they bear quoting in detail:

‘My cardinal conclusion … is because of the fundamental distinction that an agreement for the compromise of an ancillary relief application does not give rise to a contract enforceable in law. The parties seeking to uphold a concluded agreement for the compromise of such an application cannot sue for specific performance. The only way of rendering the bargain enforceable, whether to ensure that the applicant obtains the agreed transfers and payments or whether to protect the respondent from future claims, is to convert the concluded agreement into an order of the court. The decision of the Privy Council in De Lasala v De Lasala … demonstrated that thereafter the rights and obligations of the parties are determined by the order and not by any agreement which preceded it. The order is absolute unless there is a statutory power to vary or unless vitiated by a fact that would vitiate an order in any other division. Additionally, as was demonstrated in Robinson v Robinson … an order in ancillary relief proceedings may be set aside if the product of a material breach of the duty of full and frank disclosure. An even more singular feature of the transition from compromise to order in ancillary relief proceedings is that the court does not either automatically or invariably grant the application to give the bargain the force of an order. The court conducts an independent assessment to enable it to discharge its statutory function to make such orders as reflect the criteria listed in s.25 of the Matrimonial Causes Act 1973. …

In consequence, it is clear that the award to an applicant for ancillary relief is always fixed by the court. The payer’s liability cannot be ultimately fixed by compromise as can be done in the settlement of claims in other divisions. Therefore the purpose of negotiation is not to finally determine the liability (that can only be done by the court) but to reduce the length and expense of the process by which the court carries out its function. If there is a dispute as to whether the negotiations led to an accord that the process should be abbreviated, the court has a discretion in determining whether an accord was reached. In exercising that discretion the court should be astute to discern the antics of a litigant who, having consistently pressed for abbreviation, is seeking to resile and to justify his shift by reliance on some point of detail that was open for determination by the court at its abbreviated hearing. If the court concludes that the parties agreed to settle on terms then it may have to consider whether the terms were vitiated by a factor such as material non-disclosure or tainted by a factor within the parameters set in Edgar v Edgar. Finally in every case the court must exercise its independent discretionary review applying the s.25 criteria to the circumstances of the case and to the terms of the accord. This approach particularly applies to accords intended to obviate delivery of briefs for trial. Different considerations may apply to agreements not negotiated in the shadow of an impending fixture. …

Litigants in ancillary relief proceedings are subjected to great emotional and psychological stresses, particularly as the date of trial approaches. In my opinion there are sound policy reasons supporting the conclusion that the judge is entitled to exercise a broad discretion to determine whether the parties have agreed to settle. … The court has a clear interest in curbing excessive adversariality and in excluding from trial lists unnecessary litigation. A more legalistic approach, as this case illustrates, only allows the inconsistent or manipulative litigant to repudiate an agreement on the ground that some point of drafting, detail, or implementation had not been clearly resolved. Ordinarily heads of agreement signed by the parties or a clear exchange of solicitors’ letters will establish the consensus. Hopefully a case such as this requiring the exercise of the judge’s discretion will be a rarity.’

On the evidence, Thorpe LJ held that ‘the parties had concluded a compromise during the week before the hearing. Throughout that week it was the husband who was pressing for a settlement and plainly there came a point at which the wife agreed his terms. All that remained unresolved was either mechanics or trivial.’

As to the principles, headline points can be drawn out as follows:

• Contractual principles do not apply within financial remedy proceedings to create an enforceable contract on agreement.
• An agreement is not enforceable until converted into an order of the Family Court.
• The court will not automatically or invariably convert an agreement into an order, and must instead undertake the s 25 exercise.
• When viewed in that context, negotiations between the parties, and the reaching of an agreement, is designed to reduce the time and resources required by the court in exercising its discretion under s 25 MCA 1973.
• The court can also, by way of shortened procedure, determine whether an agreement has in fact been reached, and if so whether there are any vitiating factors, before then coming to the s 25 exercise.
• Where core issues are agreed, residual matters (e.g. drafting, timings/implementation) can be determined by the court and are thus out of the parties’ hands at that stage.
• By retaining the ability to determine whether an agreement has been reached, the court aims to reduce further litigation arising from attempts by any party to resile from agreement on peripheral grounds.
• Signed heads of agreement or clear exchange of solicitors’ letters are recommended to show a final agreement.

Is Xydhias untouched 25 years on?

It is often overlooked that in the later case of Soulsbury v Soulsbury [2007] EWCA Civ 969, [2008] 1 FLR 90, Thorpe LJ’s conclusions in Xydhias came under scrutiny. Ward LJ said at [40] that:

‘[o]ne has to say that there are some who are critical of the “cardinal conclusion” that “ordinary contractual principles” do not apply to determine whether or not the parties had reached a concluded agreement.’

Ward LJ found no controversy in the court having discretion as to whether or not an agreement reached should be sanctioned by the court and made into an order. He reviewed the authorities confirming that the court has a duty to inquire into proposed settlements before making an order, and noted the statement of Ormrod LJ in Thwaite v Thwaite [1982] Fam 1 that this does ‘represent a significant departure from the general principle frequently stated in cases arising in other divisions of the High Court, that the force and effect of consent orders derives from the contract between the parties leading to, or evidenced by, or incorporated in, the consent order’. This difference in treatment was accepted.

The concern raised by Ward LJ was that the court pursuant to Xydhias could now exercise its discretion in determining whether or not any such agreement was reached. At [40], Ward LJ highlighted that Xydhias had given the Family Division a different and unique test for establishing the formation of the underlying agreement itself.

The correctness of that part of the decision did not need to be decided in Soulsbury, and Ward LJ expressly stated at [42] that it was not for the Court of Appeal ‘to pronounce upon the correctness of the “cardinal conclusion”’. He did, though, consider Thorpe LJ’s reason for arriving at that conclusion as material and requiring consideration.

Thorpe LJ’s reasoning was, he said, that ‘the fundamental distinction that an agreement for the compromise of an ancillary relief application does not give rise to a contract enforceable in law’. It followed, said Ward LJ, that pursuant to Xydhias, ‘the only way of rendering a bargain to make payment of money enforceable would be to convert the concluded agreement into an order of the court’. He considered that ‘stated in those terms, it cannot be correct’, and that in expressing this view Thorpe LJ had ignored the dictum of Butler-Sloss LJ (as she then was) in Kelley v Corston [1998] 1 FLR 986 and the opinion of Lord Diplock in De Lasala v De Lasala [1980] AC 546, which Ward LJ had considered at [35] as confirmation that an agreement can be enforced.

Ward LJ also noted that Thorpe LJ was of the view that ‘[t]he parties seeking to uphold a concluded agreement for the compromise of such an application cannot sue for specific performance’ but in Merritt v Merritt [1970] 1 WLR 1211 the Court of Appeal allowed specific performance.

Ward LJ therefore concluded at [45] that:

‘the cardinal conclusions expressed by Thorpe LJ are stated in terms which are too wide. I accept that if there are negotiations to compromise a claim for ancillary relief, then there is a duty to seek the court’s approval as is stated in Smallman v Smallman [1972] Fam 25. But as Smallman states, and I do not see how that authority of this court can be ignored by me, even an agreement subject to the approval of the court is binding on the parties to the extent that neither can resile from it.’

It would seem, therefore, that Soulsbury is authority for the proposition that an agreement between parties to settle their respective claims upon divorce can give rise to an enforceable contract and, to that extent, the dicta in Xydhias that the only way to make an agreement binding is to convert it into a court order was disapproved.

Two further points are of interest when considering the Soulsbury ‘challenge’:

(1) At the outset of Xydhias, neither party advanced the argument that ordinary contractual principles did not apply when determining whether or not there was an agreement. The husband argued that the absence of agreement as to security, and thus no signed heads of agreement, meant that all material terms were not agreed; accordingly, it was said, contractual principles dictated there would be no enforceable contact. The wife countered, but also by reference to contractual principles. It was therefore common ground between the parties that those principles applied (as it had been in the courts below). This was noted by Thorpe LJ in his judgment, and again by Ward LJ in Soulsbury at [37]. It was only during the hearing (and at the instigation of the court) that the wife’s position changed to include the alternative formulation that formed the basis of Thorpe LJ’s judgment.

(2) The husband obtained, on paper, permission to appeal to the House of Lords. However, prior to any hearing taking place before the House of Lords, the parties reconciled and remarried, bringing an abrupt end to the proceedings. The husband’s challenge, which again placed reliance on contractual principles, was therefore never considered by the House of Lords – despite the court’s clear appetite to consider the same, given the granting of permission.

Despite this, no advance of the Soulsbury challenge in opposition to Xydhias has come through in judgments thereafter. Indeed, the issue has been somewhat sidestepped. For example:

(1) In S v S (Ancillary Relief) [2008] EWHC 2038 (Fam), [2009] 1 FLR 254 Eleanor King J (as she then was) drew together at [23] a number of propositions, including that ‘[t]he existence of a concluded agreement is a matter of great weight’. She then considered Ward LJ’s dictum in Soulsbury (cited above) and considered that:

‘It is not necessary for the purposes of this judgment to consider how Ward’s LJ recent observation fits with the body of case-law. Its significance for the purposes of the case management decision I have to make is that it is a further example of the importance of agreements in the eyes of the Court of Appeal.’

(2) In Independent Trustee Services Ltd v GP Noble Trustees Ltd & Ors [2012] EWCA Civ 195, [2012] 3 All ER 210 Patten LJ observed as follows at [36]:

‘The statement by Thorpe LJ (at 394) that the only way of making an agreement to pay money enforceable between husband and wife was to convert it into an order of the court has been criticised in a subsequent decision of the Court of Appeal as too wide: see Soulsbury v Soulsbury … But nothing in the later judgment detracts from the proposition that the making of the order has to be a proper and fully informed exercise of the powers contained in the 1973 Act and that, once made, it is the order which therefore governs the rights and obligations of the parties.’

Practical considerations

Leaving to one side the disagreement between Thorpe LJ and Ward LJ as to legal principles, Xydhias presents practical problems.

In theory, distinguishing between core/substantive terms (i.e. upon which agreement is necessary for a party to be held to their bargain) and procedural/residual terms (i.e. upon which agreement is not necessary), should not be difficult. That is particularly so for practitioners who can step back and take a more objective assessment. In practice, subjectivity of the parties so often creeps in, where that which is very important to one person may to another be trivial.

Difficulties also arise in defining where the line is to be drawn between quantification and principle on the one hand and mechanics and mode of performance on the other. Using security of payments as an example, the nature of the security might be critical to one party who had to give it, but purely a matter of mechanics to the other. As Moor J observed in Pierburg v Pierburg (No 2) at [43] (above), if there is an overall settlement the court can determine issues that remain as to (say) drafting or exact terms of the order. The court can do so without undermining the overall agreement, thus preventing the parties being able to resile from the same. Therefore, for the party that still attaches significant importance to the matter of security, that key issue is taken from their control.

Considerable importance therefore lies on whether or not an agreement has been reached at all, and, as a consequence, conflict can arise where one party states a concluded agreement has been reached relying upon Xydhias, whilst the other refuses to accept any such agreement exists.

This difficulty may be avoided if the parties follow the guidance of Thorpe LJ. The easy solution to evidencing a concluded agreement is a signed heads of agreement. Time spent setting out core terms in a heads of agreement is invaluable in reducing potential for conflict; that is all the more so late in the day at an FDR where, although so often stretched for time, leaving with a signed heads of agreement (even without the court’s approval on a Rose v Rose basis) should always be attempted.

Negotiations are, though, often not that formal and take place through a chain of correspondence and conversations, with new matters emerging as negotiations develop. Email rather than formal inter-party correspondence has only increased this informality, and it is increasingly common for one side to declare in writing that they consider negotiations have reached such a stage that the parties are ‘Xydhias-bound’. To avoid dispute over such declarations, it is advisable to clearly define at an early stage the core/substantive issues (or, if later, as soon as they arise), to make it clear that no final agreement can be reached until those issues are resolved.

Lastly, practitioners should heed the fundamental difference between a Xydhias agreement and a Rose order (arising from Rose v Rose [2002] EWCA Civ 208) – the latter being where an agreement receives the court’s approval – and advise clients accordingly. As was relatively recently highlighted by Lieven J in Kicinski v Pardi [2021] EWHC 499 (Fam), [2022] 1 FLR 474:

‘[17] … [the judge] referred in a footnote to a useful summary in the FDR Best Practice Guidance of the difference between a Rose order and a Xydhias agreement. This has some relevance to the approach the Court should take to what was agreed in the Rose order, so I shall set it out:

“Where heads of agreement are signed rather than a consent order submitted, clients should be advised that the heads of agreement are evidence of consensus that may be subject to a ‘show cause’ application if one party attempts to resile from the agreement but such heads of agreement do not have the same status as an order (whether perfected or unperfected). Practitioners should be careful to explain to clients (and record on the face of the agreement where appropriate) whether any signed agreement is understood and agreed to be Xydhias-compliant (ie a binding agreement), Rose compliant (ie an approved agreement which amounts to a court order), or otherwise).”’