The Law Commission



PART VI

CALCULATING COMPOUND INTEREST

6.1     In this part we consider the practicalities of calculating compound interest. We first ask whether there should be a prescribed computer programme or tables. We then consider whether any prescribed system should use annual, quarterly or monthly rests.

COMPUTER PROGRAMME OR TABLES?

6.2     In the Consultation Paper we asked if compound interest should, in practice, be calculated through a computer programme, through tables or in some other way. We provisionally proposed a computer programme that would be common to all courts, suggesting that tables may be "somewhat cumbersome". ^[1]
6.3     Most of those who responded on this point favoured the creation of a computer programme, which could carry out a wide variety of calculations with minimal difficulties. It was suggested that most lawyers were familiar with using computers, and had ready access to them. The Association of Personal Injuries Lawyers said that a large proportion of their membership already used computer programmes to calculate interest, which could easily be adapted to calculate compound interest. We were also told that the Family Law Bar Association provides its members with a computer programme that can carry out compound calculations.
6.4     Many respondents on the other hand argued that tables should be available as well as or instead of a computer programme. Tables were thought to have two main advantages. First, they were more portable. Lawyers could take them to court for last minute calculations, either in court or at the door of the court. Second, they were transparent. The parties could refer to them and discuss them in the event of a dispute, rather than having to take the sums on trust.
6.5     Those involved in personal injury work were familiar with reading sums from published tables. As one lawyer put it, "we already have the Ogden Tables… and they are brimming over with multipliers". Several judges told us that they would be happy to read appropriate figures from tables, they though would be unhappy to refer to a computer programme. They were used to dealing with tables, and any anomalies or differences of perception could be discussed openly.
6.6     We are convinced that there is a need both for a prescribed computer programme (for complex or unusual calculations) and for prescribed tables (which will be readily accessible in court and during last minute settlements). We set out how each of these would work below.

A prescribed computer programme

6.7     As discussed in Part IV, compound interest calculations may be carried out in different ways to produce different results. ^[2] It is essential that the potential for dispute is removed, by prescribing the way that calculations should be done. We see substantial advantages in the Court Service providing a simple, readily accessible computer programme. Lord Justice Dyson commented If a programme is used, it is essential that the same programme is used by and is available to, all those involved in the litigation process. One possibility might be to create an interest calculator and to place it on the public pages of the Court Service Website. We endorse this suggestion. It would not preclude commercial providers from developing their own programmes, using the method of calculation prescribed. We think, however, that the programme should also be provided on a public service basis, so that litigants are not required to pay for access to it.
6.8     We recommend that the Court Service should produce a computer programme to calculate compound interest and make it readily accessible on its website.
6.9     The computer programme will perform many calculations that the tables cannot. First, it will give interest calculations that are correct to the day, while the tables (for reasons to be explained) give multipliers on a month by month basis. Second, it will be able to cope with variations from the specified rate. For example, a judge might award compound interest at one or two percentage points above the specified rate on all or some of the damages, for all or some of the period. The programme should be able to cope with this. Third, the programme will be able to deal with complex damages, arising at different times. For example, in personal injury cases it will be able to calculate interest on past losses from the time they arose, rather than assuming that they all date from the same time. ^[3]
6.10     The Personal Injuries Bar Association noted that, as most calculations would be carried out by the parties rather than by the judge, it was not necessary for the computer programme to be available in every courtroom. They suggested that it would be useful if at least one copy of the programme were available in each courthouse, so that the parties could use it for last-minute checking.
6.11     We are aware that the computer programme will take resources to develop. It will also need to be updated each year as the specified rate changes. It is important that compound interest is not introduced until the requisite plans are in place.

Prescribed tables

6.12     The tables are designed to cover only the basic calculation. The aim is to provide a multiplier that can be applied to the principal sum to give compound interest at the specified rate.
6.13     Appendix F provides sample tables. We think that it is important to keep the tables simple and easy to use, even at the expense of some loss of accuracy. We therefore recommend that the tables should calculate interest from the month in which interest started to the month in which interest finished.
6.14     For cases that start or finish in the middle of the month the tables will provide a fairly rough and ready figure, as they will ignore interest for the remaining days. However, the accuracy should be sufficient for most purposes. As Appendix F shows, on a debt of £100,000 over five years, at an interest rate that fluctuates from 4-5%, the tables give a result that is, at most, plus or minus £250 from the "pure maths" figure. By contrast, using simple rather than compound interest would produce a figure that is £2,862 less.
6.15     We think that in many cases the parties would be willing to tolerate some loss of accuracy in order to have a clear figure at the end of the case. Where the parties wish to have a more accurate result, they will need to use the computer programme.
6.16     A set of tables would need to be around 20 pages long, to cover awards or payments made in a two-year period and started anything up to twenty years earlier. ^[4] We do not think it would not be an unduly cumbersome document for judges to keep on their shelves, or for advocates to take to court. Judges would probably benefit from some training in using the tables, but we do not think that it would need to be time consuming.

6.17 We recommend that the Court Service publishes tables to allow the calculation of compound interest at the specified rate.

ANNUAL, QUARTERLY OR MONTHLY RESTS?

6.18     In the Consultation Paper we provisionally proposed that compound interest should be computed with annual rests unless there were good reasons for ordering rests at more frequent intervals. We thought that this would simplify calculations in short-running cases by effectively removing all cases of a year or less from the scheme.
6.19     Most respondents agreed with our proposal, without providing any independent supporting reasoning. However, both the Law Society and Bar Council argued for more frequent rests on the grounds that this would reflect existing commercial practice. The Law Society, for example, proposed monthly compounding as "this is the approach taken by most lenders on recovery, mortgage lenders and banks investing money, etc". The Law Reform Commission of Hong Kong took the same approach. Their own report recommended monthly rests to reflect the practice of the major banks in Hong Kong and they suggested that we should follow "general UK banking practice, whatever that may be". Other respondents pointed out that arbitrators almost always use quarterly rests.
6.20     The main argument given for annual rests is that they would be simpler to calculate. This is undoubtedly true if one is trying to work out compound interest from first principles using a pocket calculator. In our view, however, no-one should attempt to do this. It would be a recipe for confusion and dispute. As outlined above, compound interest should only be calculated using a prescribed computer programme or the official tables.
6.21     For debts of over a year, a computer programme or tables using monthly rests would be no more complicated to use than those using annual rests. In fact, the balance of convenience lies with monthly rests. The most confusing aspect of the system proposed above is the potential discrepancy between the computer programme (that calculates interest to an exact date) and the tables (that round to the relevant month). It is easier to add a few extra days of interest to a figure based on monthly rests than to one based on annual rests. ^[5]
6.22     The main reason why annual rests might be seen as simpler is that a compounding system based on annual rests would automatically exclude all awards or payments of less than a year. If one were to use quarterly or monthly rests, it would be necessary to exclude short debts through a specific provision to that effect.
6.23     The simplicity arguments are not conclusive either way, though on balance the monthly rest tables are slightly easier to use. In the absence of other considerations, our main aim is to reflect commercial reality: as the Hong Kong Law Commission put it, the question should be resolved by following normal banking practice. For borrowers, interest is usually compounded at least monthly.

6.24 We recommend that compound interest should be calculated using monthly rests.

A discretion to vary the compounding intervals?

6.25     The next question is whether the courts should have discretion to vary the compounding intervals, as we proposed in the Consultation Paper. The few respondents who commented on this proposal agreed with it, on the grounds that there may be cases in which other rests may be appropriate. If, for example, a judge wished to mirror the interest rate available on a building society investment account, it may be more accurate to use quarterly rests.
6.26     On the other hand, allowing judges a discretion over compounding intervals will add to the complexity of the system. The computer system would have to offer litigants the facility to vary the rests. If the facility is there, some parties will use it without understanding the implications, while others may be encouraged to make unwarranted applications for different rests.
6.27     On balance, we think that the added complication involved in allowing judges a further element of discretion to vary the compounding interval outweighs the increased accuracy it would provide. When one is using investment rates, the difference between annual and monthly rests is small (£800 on £100,000 at a rate of 5% over five years), ^[6] and does not justify adding another level of complexity to the system.

6.28 We recommend that the compounding interval should be set by rules of court. All interest under the new statutory regime would then be calculated in accordance with the prescribed interval.

6.29     This would not affect the courts' existing powers to award contractual interest or "interest as damages". For example, compound interest is available in accordance with a contractual term or trade usage; ^[7] as special damages under the second limb of Hadley v Baxendale; ^[8] or in equity in claims against those in a fiduciary position who are accountable for profits made from their position. ^[9] As discussed in Part V, in these cases judges would continue to be able to award the actual interest lost – at whatever compounding interval that might be. For example, if a contract specified quarterly rests, the debt would be calculated on the basis of quarterly rests.

"MIXED" AWARDS

6.30     We have considered how far a court should be able to mix simple and compound interest payments in a single judgment. Clearly, it should be open to a court to award simple interest on some heads of damages, while awarding compound interest on others. For example, in a personal injury claim, the court could award compound interest on past pecuniary damages, and only simple interest on the non-pecuniary damages. A court could also award compound interest on a longstanding debt, and only simple interest on a further debt that had been outstanding for less than a year.
6.31     A more difficult question is whether the court should have discretion to grant compound interest for part of the period and simple interest for the rest of the period. Our view is that mixing compound interest and simple interest in this way would substantially increase the complexity of the arrangements, without making much difference to the total sum awarded.
6.32     We recommend that the courts should not be allowed to award compound interest for part of the period of the debt, and simple interest for the rest.

SUMMARY OF RECOMMENDATIONS

6.33     We recommend that details of how compound interest should be calculated should be set out in rules of court. Calculations should be carried out in one of three ways: through use of an approved computer programme; through approved tables; or by following a detailed prescribed formula. Parties should not be allowed to carry out their own calculation in their own way.
6.34     The Court Service should make an approved computer programme freely available on its website, and should ensure the publication of suitable tables.
6.35     On balance we consider that monthly compounding would best reflect commercial reality. However, it is not our intention to include short debts within the scheme: rules of court should exclude compound interest on any debts or damages that have been outstanding for less than a year, unless the claimant can show exceptional reasons why interest should be compounded.
6.36     We recommend that the compounding interval should be set in rules of court, so that if commercial practice changes, the intervals can also be changed. All interest under the new statutory regime would then be calculated in accordance with the prescribed interval. The courts would not have discretion to alter the interval. Nor would courts be permitted to award compound interest for part of the period and simple interest for the rest.