(Court of Appeal, Moore-Bick, Kitchin, Ryder LJJ, 2 December 2014)
Financial remedies – Appeal – Compensation principle – Whether the wife’s award had adequately compensated her
[The judicially approved judgment and accompanying headnote has now published in the Family Law Reports  2 FLR 447]
The full judgment is available below.
The wife’s appeal from a financial remedies order which included an element of compensation was allowed and the matter was remitted for rehearing before a different judge.
The husband and wife divorced after 21 years of marriage. They had two adult children. The husband married the year after decree absolute was granted. He had been ordered to pay the wife periodical payments during the parties’ joint lives. The wife subsequently applied for a variation relying on the compensation principle in
Miller v Miller; McFarlane v McFarlane  3 All ER 1.
The parties reached an agreement and the wife received increased periodical payments. However, the husband then applied to terminate the periodical payments as from when he retired in order to parent his young children.
In the wife’s acquiescence new orders were made to take effect from the husband’s retirement. The judge determined on the basis of oral submissions by the husband that his employment was imminently going to cease with a dramatic reduction in his income. It was not accepted that the husband would have no or almost no earnings in the future. He would have significant earning capacity, albeit that would decrease over time. It was more than possible that he would obtain employment when the young children were settled. However, the judge found that the wife had no relevant earning capacity and that she had sacrificed her professional career for her family.
The judge approached the case on the basis that the husband had assets of £5.68m and that over the next 10 years he would receive further, significant payments. He found that there was a tangible, obvious compensation element that should be recognized and made provision to that effect. The wife’s projected income was £28,000 based on a return rate of her existing capital fund of 3.75% net. The wife contended for a rate of 3.75% gross. As her annual income needs were £80,000, based on the Duxbury formula the judge made a lump sum order of £400,000. The wife appealed on the basis that the return rate should have been calculated at a gross rate of 3.75%, therefore, resulting in a higher lump sum. She further submitted that the judge had failed to pay proper regard to the compensation principle in quantifying her needs and in considering whether she should downsize to release funds.
The appeal was allowed and the matter was remitted for rehearing by a different judge.
On the facts of the case it was not acceptable for the judge to pick a rate of 3.75% out at random which had not been identified by either party and to then apply it without having addressed the question to counsel for submissions. Without reasoning his conclusion that 3.75% net had been appropriate for one part of his calculations, the inconsistency demonstrated was fatal.
In relation to the compensation principle the judge had failed to compare both parties’ assets, income and needs positions meaning it was impossible to cross check the fairness of the overall award. A more sophisticated exercise was needed to avoid discriminating against the wife who was entitled to compensation.
Case No: B6/2014/1143
Neutral Citation Number:  EWCA Civ 1523
IN THE COURT OF APPEAL
ON APPEAL FROM THE FAMILY DIVISION OF THE HIGH COURT OF JUSTICE
Royal Courts of Justice
LORD JUSTICE MOORE-BICK
LORD JUSTICE KITCHIN
LORD JUSTICE RYDER
- - - - - - - - - - - - - - - - - - - - -
- and -
- - - - - - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - - - - - - - -
Patrick Chamberlayne QC (instructed by Stewarts Law LLP) for the Appellant
Phillip Marshall QC and Harry Oliver (instructed by Payne Hicks Beach Solicitors) for the Respondent
Hearing date: 8 October 2014
- - - - - - - - - - - - - - - - - - - - -
Lord Justice Ryder
 On 8 October 2014 the court heard an application for permission to appeal a financial remedy decision made by Coleridge J on 18 March 2014, with the appeal to follow on if permission was granted. For convenience, I shall call the appellant, the wife, and the respondent to this application, the husband. The judge ordered a lump sum payment to the wife of £400,000 to be paid by the husband upon the termination of a joint lives periodical payments order which would occur on his retirement next year.
 There is some sensitivity surrounding the identity of the adults because of personal information that the husband should be entitled to impart to his children when that is appropriate and there is accordingly an order in place prohibiting their identification, that of the children or the information concerned and this judgment is appropriately anonymised.
 The key issue in the appeal is whether the judge erred in his calculation of the lump sum award by applying a rate of return on investments of 3.75% net when the applicant says that it should have been 3.75% gross. There is also a dispute between the parties as to the application of the principle of ‘compensation,’ which arguably would have had a significant effect on the computation of the award. The husband opposes the grant of permission or in the alternative submits that the appeal should be dismissed. At the conclusion of the hearing we reserved judgment. For the reasons which I now explain, I would grant permission and allow this appeal.
The background circumstances
 The parties married on 30 July 1983. The husband was 54 years of age and the wife 55 at the date of the hearing. The parties separated in August 2004 after 21 years together and the wife’s petition for divorce was presented on 22 December 2004. Decree nisi was pronounced on 17 June 2005 and decree absolute on 12 August 2005. There are two adult children of the marriage. In May 2006 the husband remarried. He has two young children from his new relationship. It is a relevant fact that because of a serious personal issue he wishes to retire in the Summer of next year to allow him to be the full time parent of those children.
 The parties are both qualified accountants. They met when they were trainees in the same firm. The husband has been with a major international financial services firm since 1984 and was Chief Operating Officer of a tax practice of the firm until June 2010. The wife worked until 1990 eventually as a finance director in a retail organisation. Although she then took redundancy, the cessation of her career coincided with the birth of their first son who is now 24 years of age and a daughter who is now 21. None of the children of either party were concerned with the application before the court below.
 On 3 August 2005 Deputy District Judge Solomons made an order consequent upon the divorce that the husband pay to the wife periodical payments of £90,000 each year during the parties’ joint lives. That order was agreed. The papers reveal that at that time there was a division of the assets such that the wife received £1.37m and the husband received approximately £1.2m. I note that Coleridge J made an error in relation to that asset division but it does not appear that the error was material to his subsequent deliberations.
 Following the decision of the House of Lords in Miller v Miller; McFarlane v McFarlane  UKHL 24,  2 AC 618, the wife applied for a variation. It is apparent that in making that application she relied upon the principle of compensation, which is to be derived from the opinions of their Lordships in that decision. There was no agreement between the parties then or now about the applicability or effect of the application of that principle to their circumstances with the consequence that without privilege being waived this court cannot make any assumption about the basis for the agreement subsequently reached between the parties. That has not occurred and there are no recitals setting out the understandings or bases for the order. The agreement that was reached was approved by the late Mrs Justice Baron on 20 June 2007 and the order records that the periodical payments sum was increased to £150,000 each year.
 On 5 November 2012 the husband made an application to terminate the periodical payments that he was making to the wife with effect from his retirement. The wife’s response until questioned about it by the judge during the hearing was that the application was premature. She changed her position during the hearing before Coleridge J and acquiesced in the court making new orders to take effect on the husband’s retirement with the consequence that the quantification of any terminating payment to her from the husband was the subject of oral submissions rather than detailed written submissions.
 It is common ground that when considering an application under section 31(7)(a) of the Matrimonial Causes Act 1973, as amended, the court has the power to vary the underlying periodical payments provision, if it is not to be terminated, and also to capitalise the same under section 31(7B). The judge was required to have regard “to all the circumstances of the case” in his consideration of an application to which section 31(7) of the 1973 Act applied. That involved consideration of any change in any of the matters to which the court was required to have regard when making the original order(s) under sections 25(1) and (2) of the Act. The relevant provisions are as follows:
s 31(7)(a): “in the case of a periodical payments order made on or after the grant of a decree of divorce … the court shall consider whether in all the circumstances and after having regard to any such change it would be appropriate to vary the order so that payments under the order are required to be made or secured only for such further period as will in the opinion of the court be sufficient (in the light of any proposed exercise by the court where the marriage has been dissolved of its powers under section (7B) below) to enable the party in whose favour the order was made to adjust without undue hardship to the termination of those payments.”
s 31(7B): “the court has power, in addition to any power it has apart from this subsection, to make supplemental provision consisting of any of-
(a) an order for the payment of a lump sum in favour of a party to a marriage; […]”
 The wife submits that although she regarded the husband’s application as premature she was prepared to capitalise her existing periodical payments order. She submitted to Coleridge J that her existing capital fund of £1m savings would produce an annual income of £28,000 assuming a rate of return on that fund of 3.75% gross which she submitted was the rate of return that was to be regarded as ‘standard’. Her income shortfall given her needs as quantified in the existing annual sum of £150,000 would be £122,000 which would require a Duxbury fund of £2.6m, which was her claim.
 In the alternative, the wife submits to this court that if the judge had applied a rate of return of 3.75% gross to the capital fund that he identified, that rate of return would have produced a lump sum of £746,000 not £400,000. She does not contend for that sum save in the event that she does not succeed on the other significant limb of her case which is that the judge failed to have proper regard to the compensation principle in quantifying her needs and in particular whether she should have to downsize to release funds to generate income. She submits that to be required to do so damages the compensation element of her existing award, which should not be prejudiced by the changes that are to be made for the parties’ futures.
 The wife also submits that in coming to a conclusion about whether the judge erred in applying a rate of return to her capital fund of 3.75% net, this court should have regard to the draft judgment that was handed down to the parties for typographical correction. It is said that the draft judgment, together with the email exchange that followed, demonstrates a change in the substance of the judge’s reasoning before he handed down the final judgment.
 The husband submits that Coleridge J was entitled to apply the rate of return that he did (i.e. 3.75% net). He says that there is no ‘industry standard’ rate of return to be applied to non-amortised (as opposed to Duxbury) capital funds and even less so a rate of return that is acknowledged by the judges of the Family Division of the High Court to be of general application. There is no reported case dealing with the question and this court should decline to be drawn into identifying such a rate, at least in the absence of additional evidence and submissions on the point.
 The husband also submits that there was no evidence before Coleridge J about the rate of return that he should apply. That being so, he was entitled to take notice of a rate that he (the judge) thought was appropriate and apply it to the underlying capital fund that he had identified. In the alternative it is submitted that in the exercise of his broad discretion the judge was entitled to conclude that £400,000 was fair.
 The husband submits that the exercise of discretion is incapable of challenge and that it was overtly reasonable concluding as the judge did that:
a) the wife’s future annual income requirement would be £70,000 uplifted to £80,000 having regard to her previous expenditure (that is excluding her accumulation of savings) and her future needs at the time the husband is likely to retire;
b) the wife might reasonably be expected to use not only her existing savings as a capital fund to produce income but also £500,000 that she could release from the equity of her home were she to re-house from property valued at £1.75m to something more modest and valued at £1.25m;
c) the wife would be able to produce £56,250 income from a capital fund of £1.5m which if rounded down to £55,000 would leave a shortfall of £25,000 to be met from a Duxbury fund which would then be £400,000.
 Finally, the husband submits that the judge took into account the principle of compensation in the following respects:
a) he attributed only £500,000 of the equity in the wife’s home to be part of her long term capital fund for income purposes and he assumed that the wife should have the benefit of that capital fund throughout i.e. he did not assume that the fund would be amortised;
b) although he took into account the £1m she had saved he did not subject that sum to an exercise which involved its amortisation;
c) he assumed a steady state income need for the wife after the husband’s retirement rather than any ‘step down’ thereafter;
d) he did not take into account any further savings that the wife might accrue between the date of determination and the husband’s retirement; and
e) he uplifted his estimation of the wife’s net income needs from £70,000 to £80,000 ‘to be generous’.
 The judge made findings that are relevant to the issues in the appeal. He decided that the husband’s employment was imminently and predictably going to cease with a dramatic reduction in his earnings and earning capacity. He did not accept the husband’s position that he would have no or almost no earnings in the future. He held that the husband would have a very significant earning capacity, which would be of dwindling value as time progressed. He thought it would be more than a possibility but not a probability that he would obtain employment when the young children were settled after the inevitable events that faced him and them.
18] The judge made no finding about the wife’s earning capacity save to acknowledge her claim that she had sacrificed a serious professional career for the sake of the family and that she earned a very small amount as a counsellor, which he decided to ignore. One can deduce that he was satisfied she had no relevant earning capacity.
 The judge identified a significant difference between the asset presentations of the parties and resolved them in the following terms:
“I shall approach this case on the basis that the husband has present assets worth in the region of £5.68m but in the not too distant future and over the course of the next 10 years he will be entitled to receive further and in some respects significant payments.”
 The judge dealt with the history of agreements between the parties in these terms:
“The wife in this case has been properly and fairly treated thus far by the previous orders. It is entirely wrong to try and back track and reargue or reopen the provision which has been made in the then circumstances and where it has been received and accepted and there has been no appeal;”
 The compensation question was answered in an unequivocal way:
“This case does retain a tangible, obvious compensation element which deserves recognition one way or another even at this stage. It has been factored in up to now and there is no reason why it should simply be ignored”.
The judge decided that compensation would be recognised in four ways:
a) only £500,000 of the value of the wife’s home was to be regarded as part of the long term income fund;
b) neither that sum nor the £1m of savings was to be amortised;
c) there was to be no further step down in the wife’s income needs after the husband’s retirement; and
d) the additional savings between the order and the husband’s retirement were to be the wife’s to keep without them being taken into account (estimated at £100,000).
 The wife’s actual budget i.e. her annual income needs less the potential to make savings was calculated at £70,000 to which £10,000 was added to ensure that ‘it was generous to her’.
 I do not propose to enter into the exercise of analysing the differences between the judge’s confidential draft judgment and his final judgment. The former is not his judgment and the latter is. If the judge made an error of calculation, as it is submitted he did, then all that such an analysis would achieve is to demonstrate that the judge neither identified it as an error nor corrected it. That does not make it any less or more of an error, which is a matter for this court to determine. If the judge changed his mind then there is no doubt on the authorities he was entitled to do so and the perfected text is that which should be scrutinised by this court.
 It is not necessary to indulge in an examination of the authorities on the confidentiality of the draft. That is not to undermine the existing law on that topic but simply to decline to be drawn into an argument that it is not necessary to resolve in order to decide this appeal.
 The rate of return chosen by the judge was not that submitted to him by the wife. The wife in her closing submissions submitted to the judge that her income projections were “on the basis of a standard Family Division 3.75 percent return, about £28,000 per year” (for the avoidance of doubt that is 3.75% gross). The rate and the alleged standard nature of that rate were twice relied upon by the wife. The husband challenged the proposition that there is a “Family Division rate of 3.75%” in his closing submissions. Neither party otherwise addressed the alternatives either by evidence or during submissions although there is an assumption in the parties calculations that 3.75% gross was an appropriate rate of return for the judge to apply. It should be noted that 3.75% net would be 5.4% gross, a rate of return that would require some justification.
 The judge can be forgiven for coming to a decision based upon the actual rate that the wife had contended for if that was his own view even if there is no industry standard. The husband identified no other rate and it would have been entirely reasonable for the judge, having taken notice of the economic and market conditions, to have relied upon one component of the Duxbury assumptions to come to his view. If he had done that the rate he would have chosen would have been 3.75% gross. I am very firmly of the view that there is no ‘industry standard’ even less an acknowledgement by the Family Division judges that there is or should be such a rate. For my part, I would firmly decline the implied invitation to identify such a rate or to give guidance that there should be a parallel process to that employed by the Duxbury Committee to produce such a rate.
 I am entirely neutral on the question of whether it is a cost effective exercise to embark on such a process. That must be a matter for others. What I do take on board is the wise warning given by Holman J in F v F (Duxbury Calculation)  1 FLR 833 at 849 where he rejected on the facts of that case rates of return above and below the then Duxbury assumed rate for the purposes of a Duxbury calculation. He opined that there may be particular facts that required higher or lower assumed rates of return which might justify expert evidence in support. He then cautioned against any standard practice arising of experts being instructed on the point given that the assumed rate of return was only a starting point for consideration of the section 25 factors in the Act.
 What in my judgment was not acceptable in this case was for the judge to pick a rate out of thin air, which was not the rate identified by either of the parties i.e. 3.75% net and then apply that rate to his calculations without addressing the question to the advocates for them to make submissions. If he had come to a conclusion about a rate having heard submissions on the point and having decided why he preferred one rate as against another then, absent an irrational decision, his exercise of discretion would have incapable of challenge. Furthermore, the judge applied 3.75% net to the wife’s capital funds but used the Duxbury formula, which incorporates a rate of return at 3.75% gross when calculating the lump sum that the husband would have to provide. Although the exercises being performed are different, without reasoning his conclusion that 3.75% net was appropriate for one part of his calculations, the inconsistency demonstrated is fatal.
 Without doubt there are reported examples of cases where different rates of return have been applied and one can imagine circumstances where it might be necessary to hear evidence on particular market conditions or circumstances which would generate a rate specific to the facts of the case, for example, the type of investment into which the fund is placed (its nature and extent) or the place in which that investment occurs (both as to the differential benefits or detriments of the economic conditions and the taxation regime). All that does is to emphasise that the rate of return is fact specific to the individual case and like any other fact or prediction should be decided by the judge where it is not agreed.
 It is perhaps important that the reader who is not familiar with the Duxbury methodology and its assumptions should have the benefit of a summary. I take this excerpt from the narrative to Table 14 of the publication ‘2014-2015 At A Glance’, Class Legal 23rd Ed (2014):
“Duxbury calculators are based on an iterative computation, seeking the amount which if invested to achieve capital growth and income yield (both at assumed rates and after income tax on the yield and CGT on the realised gains) could theoretically be drawn down in equal inflation-proofed installments over a period (usually, but not always, the estimated actuarial life expectancy of the recipient) but would be completely exhausted at the end of the period. It is not, and never has been, an attempt to identify the sum necessary to guarantee a particular level of expenditure
The underlying ‘assumptions’ are
(1) a uniform income yield;
(2) a uniform rate of capital growth;
(3) a uniform rate of inflation;
(4) a consistent regime of taxation - with bands/allowances increasing in line with inflation;
(5) a constant level of drawdown in real terms;
(6) a consistent rate of ‘churn’ (the realisation of capital gains other than to fund expenditure); and that the recipient will
(7) survive for precisely the expected average of her (or occasionally his) contemporaries; and
(8) be or become entitled to a ‘full’ state pension; which will
(9) increase in line with prices; while
(10) the age at which a state pension is payable will not alter in the meantime.
All of the assumptions are necessary simplifications which will not materialise: as Ward LJ said in
B v B  1 FLR 20 ‘the only certainty is that it will not happen as we have predicted’. […] The assumptions include three key financial predictions - an average income yield of 3% p.a., an average capital growth of 3.75% p.a. and average inflation of 3% p.a.”
 In summary, it is not wise to assume that because the Duxbury Committee are of the opinion that in the context of their calculations 3.75% gross is achievable over the long term with a cautious investment strategy that the parties will agree that that rate is applicable to capital funds that are not to be amortised on the facts of a particular case. However, if they do agree or if the judge decides that assumption is valid on the facts of a case, I cannot for my part see how objection can be taken. If they do not, then the rate chosen by the court should be reasoned.
 Given that I have come to the conclusion that the judge was not justified on the facts of this case in selecting 3.75% net as the rate of return to apply to the wife’s capital funds and given the significant difference that the rate makes to the overall lump sum, then for that reason alone I would be minded to give permission and allow the appeal. There are, however, related questions that should be addressed in order to decide what to do as a consequence of allowing the appeal.
 The first is to scrutinise the judge’s decisions about the assets to which he applied the rate of return. If the assets presentations are agreed or his decision about them is incapable of challenge, then applying an alternative rate of return to the capital fund may be an appropriate exercise for this court to undertake.
 What did the judge decide about the assets? He first of all decided that the applications were not an opportunity to re-address decisions that had been made in the past. He must be right about that and no-one submits otherwise. He was not acting in a reviewing or appellate capacity and the parties’ agreements as approved by DDJ Solomons and Baron J must be taken as read. Furthermore, he held that this was a compensation case. It follows that whatever dissenting opinions might suggest about the applicability of the compensation principle to periodical payments and capitalisation issues, the judge decided that they applied. In my judgment, taking the two issues together, it was incumbent on the judge to explore with the parties the implications of compensation on the facts of this case as he found them to be.
 It is patent from the written submissions of the parties to this court that the asset schedules were not agreed in the court below. The wife’s asset schedule at trial amounted to £11.738m whereas the husband’s asset schedule including pensions amounted to £8.358m and the treatment of the husband’s annuity income post retirement or more particularly a capital fund representing the investment basis for the same was fundamentally in issue. The parties’ cases about the husband’s assets were significantly at variance. The wife said his assets amounted to £8.911m and the husband said £5.573. It is equally not clear how the judge determined the difference between the presentations by the parties of the value of the husband’s London and Paris homes.
 The material differences are such that it is difficult to rationalise how the judge treated a significant part of the assets for the purposes of his discretionary exercise when he did not set out the decisions he had made about the differences between the parties with any specificity. He need not have constructed his own summary balance sheet although that is a commonplace in a judgment of this kind, a narrative would have sufficed or if he accepted the presentation of one or other party (as both parties here appear to erroneously assume) he could have said so. The finding that I have set out in  above was an insufficient basis to compare the positions of the parties in what the judge decided was a compensation case. The lack of any conclusion about the valuations of all of the homes and the elements of the asset schedule presentations that were in dispute presents a barrier to the comparative exercise that was necessary.
 In the absence of an explanation there is simply too much unknown about what were the available assets and how the judge decided to treat them by reference to any reasons he had for preferring one presentation over the other. To say, as he did that “I shall approach this case on the basis that the husband has present assets worth in the region of £5.68m but in the not too distant future and over the course of the next 10 years he will be entitled to receive further and in some respects significant payments” was simply too opaque to allow of sufficient scrutiny.
 Turning then to the compensation issue. I have alluded more than once to the debate among informed commentators about the nature and extent of the principle. Given that I accept that it gives rise to the need to compare the positions of the parties as part of the judge’s exercise of discretion and that a comparison of the effect of the re-distributive options was not adequately undertaken in this case, it is necessary to set out the principle. The high point of the dictum is to be found at  in McFarlane where Baroness Hale said:
“the main family asset is the husband’s very substantial earning power, generated over a lengthy marriage in which the couple deliberately chose that the wife devote herself to the home and family and the husband to work and career. The wife is undoubtedly entitled to a generous income provision for herself and for the sake of the children, including sums which will enable her to provide for her own old age and insure her husband’s life. She is also entitled to a share in the very large surplus, on the principles both of sharing the fruits of he matrimonial partnership and of compensation for the comparable position which she might have been in had she not compromised her own career for the sake of them all.”
 The high point of the restricted interpretation of the compensation principle can be found in the obiter dicta of Mostyn J in SA v PA (Pre Marital Agreement: Compensation)  2 FLR 1028 EWHC 392 (Fam) at  (and see also B v S (Rev 2)  EWHC 265 (Fam) at  to ):
“I think that the principles concerning a compensation claim can properly be expressed as follows:-
i) It will only be in a very rare and exceptional case where the principles will be capable of being successfully invoked;
ii) such a case will be one where the court can say without any speculation, i.e. with almost near certainty, that the claimant gave up a very high earning career which had it not been foregone would have led to earnings at least equivalent to that presently enjoyed by the respondent;
iii) such a high earning career will have been practised by the claimant over an appreciable period during the marriage. Proof of this track-record is key.
iv) once these findings have been made compensation will be reflected by fixing the periodical payments award (or the multiplicand if this aspect is being capitalised by
Duxbury) towards the top end of the discretionary bracket applicable for a needs assessment on the facts of the case. Compensation ought not to be reflected by a premium or additional element on top of needs based award.”
 There is a reported case that is persuasive on how the principle is to be applied and
that is the decision of Sir Mark Potter
P in VB v JP  1 FLR 742. Although I take one small part of the exposition in that case, the full text of the judgment deserves careful consideration if and when the issue of compensation comes to be considered by this court. At , the President said this:
“Second, on exit from the marriage, the partnership ends and in ordinary circumstances a wife has no right or expectation of continuing economic parity (‘sharing’) unless and to the extent that consideration of her needs, or compensation for relationship-generated disadvantage so require”
“where it is necessary to provide ongoing periodical payments for the wife after the division of capital assets insufficient to cover her future maintenance needs, any element of compensation is best dealt with by a generous assessment of her continuing needs unrestricted by purely budgetary considerations, in light of the contribution of the wife to the marriage and the broad effect of the sacrifice of her own earning capacity upon her ability to provide for her own needs following the end of the matrimonial partnership”.
 One of the purposes of the President’s careful use of language was to alert courts to the danger of double recovery in their consideration of the concepts of ‘sharing’ and ‘compensation’ - a hazard that was also in the mind of Baroness Hale. His suggestion (at ) that where the principle applies it should not be separated out from the overall exercise of discretion may be sound pragmatic advice but one must always bear in mind that the purpose of the principle is to ensure that the courts adopt a non discriminatory approach in the exercise of the discretion.
 It is tempting to enter into the debate and express a clear view about it and the application of the principle to cases of this kind. To do so would be wrong in a case where the judge at first instance accepted without reservation that this is a compensation case and that decision is not appealed to this court with the consequence that the debate to which I have referred has not been argued before this court, it has merely been flagged as a background issue of some significance.
 What is of relevance to this court is whether the judge having decided that this is a compensation case marked that fact by following through his decision. Complaint is made that the judge’s treatment of the wife’s home and savings did precisely what he said he would not do. If this is a compensation case, requiring the wife to use part of her capital fund that arguably represents the benefit she had obtained from the application of the compensation principle by downsizing is to re-create the detriment that compensation is meant to provide for unless there is an adequate comparison of the position of the parties such that on the husband’s retirement their income and asset positions remain fairly distributed. To do otherwise is to discriminate against the wife because she is not the income earner.
 One need only consider that part of the opinion of Lord Nicholls in McFarlane to see the problem that is created where the compensation issue is not adequately addressed in a case where it applies:
“ The Court of Appeal, however, seems not to have had the distinction between needs and compensation in mind when considering the way ahead. The court appears to have treated the surplus of income over expenditure as simply a means whereby the wife could accumulate a capital reserve. But that would be to mistake the purpose of this part of the district judge’s award”
and at :
“But the wife’s claim for compensation stands differently. Her compensation claim is not needs related; it is loss related. So the compensation element of her claim is not directly affected by the use she makes of her resources”
 I have no problem with the judge saying that he applied the compensation principle in the way that he did i.e. by the four ways he identified, the key element of which was the important aspect of non amortisation of the capital fund. What I question is whether that was adequate given the overall asset distribution between the parties (the element that regrettably is opaque so that I cannot express a view) and the arguably discriminatory nature of the requirement on the wife to downsize. What was necessary in this case was a comparison of both parties’ assets, income and needs positions. That is after all what these cases are about. That did not occur, with the consequence that one cannot cross check the fairness of the overall award.
 I emphasise that I do not intend to suggest that application of the compensation principle involves reconsideration of the lump sum award originally made: it does not or that downsizing is incompatible with the application of the compensation principle. As an issue downsizing is a commonplace in applications of this kind - whether or not the compensation principle applies and the court has been taken to eight decisions to demonstrate that point.1 My point is that the court needs to undertake a more sophisticated exercise than was undertaken here in order to avoid discriminating against a wife who is entitled to compensation.
 I remind myself that the purpose of this exegesis was not to decide the compensation question. That has been decided by the judge. It is to decide whether this court can embark on the necessary exercise of redistribution given the miscalculation, which has been identified. I regret that I do not believe that it is possible for this court to do that which is necessary. That would involve exercising the functions of the first instance court to identify an asset presentation that is not clear and then to apply the compensation principle to those assets by comparing the assets, income and needs positions of the parties. That is the task of the first instance court and with regret I have come to the conclusion that this case should be remitted to be re-tried by that court.
 For the reasons I have explained, I would grant the wife permission to appeal and would allow her appeal. I would remit the matter to be re-heard by a different judge of the Family Division of the High Court sitting as a judge of the Family Court.
Lord Justice Kitchin:
 I agree.
Lord Justice Moore-Bick:
 I agree
1 W v W (Financial Remedies)  2 FLR 359
J v J (Financial Orders: Wife’s Long Term Needs)  2 FLR 280
Robson v Robson  1 FLR 751
N v N (Ancillary Relief)  2 FLR 1093
Vaughan v Vaughan  2 FLR 242
McFarlane (No 2)  2 FLR 1322
B v B (Ancillary Relief)  2 FLR 1627
Lauder v Lauder  2 FLR 802