Suppose that in 2002, a cartel caused EUR 1,000 in damages to a victim. Is the victim 20 years later entitled to only EUR 1,000? Of course not. The amount must be adjusted for interest. EU law is clear on this point, along with the requirement that interest must accrue as of the time the damage occurred. Other modalities, however, are left to national laws. When working out interest in any concrete case, several common issues should be considered. The following attempts to foster systematic treatment of this interesting (sic!) topic.

Some introductory notes and caveats:

- The following discussion is about pre-judgement interest only, not post-judgement.
- The focus is on damages caused by cartels, but some portions are also pertinent to other infringements, such as abuse of dominance cases.
- The intricacies of national laws are not considered. The working paper by the European University Institute on interest on damages for infringements of competition law¹ (initiated and supported by CDC), with its country chapters, is well suited for those wishing to take a deep dive.

1. Interaction of quantification method and interest

Assume that the cartel victim is a company selling a single product, which is produced using a cartelised input. For this input, the victim was overcharged and is hence entitled to damages. Under the principle of full compensation (restitutio in integrum), the victim is to be placed in the same financial position that it would be in, had the competition law infringement not occurred.

Now distinguish two idealised approaches to quantification of the claim:

a) Data about the development of the value of a large cohort of comparable companies unaffected by the cartel suggest that the victim’s value today would be 80% higher than it is. The damage claim is hence for 80% of the victim’s current value.²

b) According to a multiple regression analysis that seeks to explain the contemporaneous prices of the cartelised product at the time of each transaction, the cartel caused the victim to overpay the amount O₁ at time t₁, O₂ at time t₂, … O_n at time t_n.The victim might have partially passed on those overcharges to its customers via higher prices (“pass on”), which in turn might have led to lower sales volume (“volume effect”). After these effects have been suitably netted, one obtains a series of contemporaneous, nominal damages D₁, D₂, … D_n.

If approach a) is used, no interest calculation is needed because the approach directly yields the financial situation that the company would be in today. With approach b) however, the victim must be awarded interest on the damage amounts to be fully compensated. This is because approach b) does not yield the company’s financial situation today.

In the following, we assume that we are in a situation where approach b) has been used. Note that categories like damnum emergens, lucrum cessans, and interest are simply part of the overarching principle of full restitution.³

2. Large impact of interest

Because it usually takes a while for compensation to be paid in cartel damages cases, interest is very important. Using the example from the introduction (a damage of EUR 1000, occurring 20 years ago) and assuming a simplistic interest regime under which the rate stays constant if the interest rate were 4%, the interest amounts to 0.04*20*D = EURO 800 with simple interest. If interest were compound, it would amount to (1.04²⁰-1)*D = EUR 1191.12.

3. Interest as of the occurrence of the harm

European Law, from the case-law of Marshall⁴, Manfredi⁵, Irimie⁶, to the Damages Directive⁷, (which merely summarises the acquis in this regard), is clear on two points: To achieve full compensation, the national law must also grant interest, and interest must accrue as from the date the damage occurred. Other than that, it leaves the modalities of interest calculation to the Member States’ civil laws, subject only to the principle of effectiveness.

The national civil laws all have provisions for the accrual of interest on debts. Some national regimes require the debtor be notified of the debt, sometimes by a specific formal act, before interest starts accruing; alternatively, the time from which interest starts accruing must be agreed in the contract.⁸ Such rules were designed by the historic legislators with contractual debts in mind. Still, these rules are also applied to tort claims in many national laws.

In a typical cartel case, this raises issues. Cartels are clandestine and often run for a long time, sometimes decades. In addition, years pass between dawn raids and the publication of a competition authority decision. By the time a victim knows of the existence of the damage and who the debtors are, several decades might have passed since the first overcharge was paid. If a claimant were then only reimbursed the nominal amount of the overcharge, one would systematically fall severely short of achieving “full compensation”.

Many jurisdictions have solved this issue for similar constellations in civil litigation in which the victim can only bring a claim a long time after the harm was caused. In those jurisdictions, judges do not have to break new ground when deciding about cartel damages cases but can simply avail themselves of such exceptions from the notice requirement. Those exceptions might have the form of a dogmatically separate class of “compensatory” interest, interest on debts of “value”, rather than a certain amount, or the notice requirement simply not applying in case of a concealed tort. In jurisdictions that have seen significant private enforcement activity, such as Spain and Germany, this point is settled.⁹ Interest is regularly calculated as from the time the victim disbursed the overcharge (but see below: ne ultra petita).

In jurisdictions that have a notice requirement but have not developed suitable exceptions that could be applied (analogously) to cartel damages cases, judges must set aside the notice requirement and any other national rule that may prevent the interest from running from the time the victim suffered the harm. EU law unambiguously demands it.

4. General statutory rates vs individualised rates

In the following it is useful to distinguish between general (statutory) rates on the one hand, and individualised, case, and claimant specific rates on the other.

a) Statutory rates

EU case-law strongly suggests that Member States must allow victims of infringements of EU law to rely on standard rates for interest, without having to provide any special proof of their adequacy in the case at hand.¹⁰ It lies in their nature as general rates that such rates cannot recreate the exact but-for situation of each victim. Using such rates will not exactly lead to “full compensation” in each case, but will merely achieve that in a rough, imperfect way.¹¹ While interest must start running from the date the victim disbursed the amount, EU law leaves the details (the rate and whether the interest is simple or compound) to the national laws, subject to the principles of equivalence and effectiveness. In systematically assessing the statutory interest situation, the following questions commonly arise:

i. Which statutory rate?

There are often several national statutory rates, and the national legal system needs to be analysed to decide which rate applies to cartel damages.

ii. Compound or simple interest?

Financial reality is compound. A bank does not offer a credit with simple interest and interest earned on savings accounts earns interest itself. Inflation correction is mathematically equivalent to compound interest with the inflation rate and a sensible individualised interest calculation must use compound interest: The result of last period’s calculation is the basis for the current period.

Despite of that, in the domain of statutory interest, many national regimes still stipulate simple interest, and the EU courts mostly award simple interest. There hence seems little room to argue that, in the domain of applying the general, statutory regime, EU law demands compound interest. Instead, this is left to the laws of the individual Member States. Nonetheless, judges and practitioners seeking to align the law with market reality should consider whether the national regime allows for compounding, which is often the case for debts of “value” (dette de valeur / waardeschuld / Wertschulden / deuda de valor).

iii. One rate to rule them all?

There are significant differences among the national statutory rates. On the first of January 2006, the Belgian statutory interest rate was 7%, the Spanish 4%, and the French 2.11%. One might question whether the considerable variation in rules and rates across Member States is justified, especially in cases involving victims from various European Member States harmed by the same cartel. With capital markets highly integrated, and a subset of the Member States even using the same currency, the differences seem hard to justify.

The rates commonly awarded by the EU courts – typically the ECB fixed rates tender rate plus 3.5 percentage points¹², or plus two percentage points¹³, do not apply directly to damages claims between cartelists and harmed parties. But should they? Should there be a uniform rate, independent of whether the harm was suffered in France or Belgium?

While the thought has some appeal, there are also considerations against it.

First and most importantly, there is no clear legal basis for such a uniform rate. The European case-law merely stipulates the principles mentioned above but otherwise leaves the modalities of statutory interest to the Member States. While the Damages Directive could have introduced a uniform regime in terms of type of interest (compound or simple) and rates, this was not done.

Second, while financial reality would militate for less variation between the Member States, some variation seems defensible from an economic vantage point. At least in the short term, inflation rates between Member States vary significantly, including between different Eurozone countries. Comparing price indices of November 2020 with November 2021 shows an increase of just 2.6% for Portugal but 7.1% for Belgium. And for the weighted average cost of capital (WACC)—one candidate metric for “individualised” rates (see b), below)—studies show there are differences between countries that are likely larger than those between different industry sectors. This is because some methods for determining the inputs to calculate the WACC vary and because there are good arguments in favour of country-specific risk premiums.

iv. Does notice/filing of a claim still matter?

While some interest must start to accrue from the time the damage occurred without any notice given, giving notice of the claim to the cartelists, especially when done by filing a legal claim, might still trigger some kind of “switch” in the interest calculation in some constellations. From that day onward, the cartelist is definitely aware of the claim. As of notice, a different “judicial” rate might apply, or the sum of nominal amount and the respective pre-filing / pre-notice interest forms the new principal and earns interest henceforth (a single “compounding event”). Several national reports for the study on interest shows such a switch.¹⁴

v. Inflation correction as a floor?

As elucidated above, the variation in statutory interest regimes is acceptable, and so is the fact that a generalised regime will almost never hit the bullseye in terms of restitutio in integrum for a given specific victim.

From an effective deterrence and a full compensation perspective, it seems sub-optimal that most statutory rates tend to under-compensate, because companies’ profitability and capital costs are on average higher than the statutory rates of most Member States. However this likely does not, in and by itself, mean that the regime is not compliant with EU law.

But when, according to the EU case law, the principle of full compensation requires that firstly, there must be some standard, statutory interest regime that claimants can avail themselves of, and secondly, that interest must start to accrue from the time that the damage arose, this also implies that these mechanisms must at least result in a certain useful effect. Certainly, an interest regime that fails to compensate even inflation cannot be compliant with the principle of full compensation. A transient (a few months long) violation of that floor might be acceptable, but a sustained violation is at odds with imperatives of EU law.

The graph below shows the inflation rate calculated from the harmonised index of consumer prices for the EURO area (dark blue line) and a rolling, twelve-month average of the same (light blue line). The green line shows the ECB’s fixed-rate tenders fixed rate for main refinancing operations, increased by three and a half percentage points, the red line the same but increased by only two percentage points.

The rates derived from the ECB base rate that EU courts commonly use are at least above the one-year rolling average of inflation in the EURO area at all times, except for the “+2 percentage point” variant concerning the most recent period. National interest regimes that over longer periods do not respect that inflation compensation floor cannot be squared with the effet utile principle.

From this consideration, one may draw another conclusion, concerning compound interest: While EU law does not require compound interest per se, a statutory regime with rates that are just barely higher or equal to the inflation rates will have to at least grant compound interest. Otherwise, the regime would again fail to even counter inflation.

b) Individualised rates

To truly create the financial situation that would result but-for the infringement, one needs to use compound interest with individualised rates.¹⁵ For each damage amount, one would seek to answer the question: what were the consequences for this victim of not having the amount over a given period? On average, such rates are higher than the statutory rates in the various Member States, and naturally, they show a very high degree of variation over time and across companies. The chart below shows the yearly averages of return on equity (ROE) by country, as reported in the BACH database¹⁶, for five EURO area countries. The average ROE over these five countries over the entire time is 8.5%.

However, to rely on such rates, the claimant needs to prove them, and, depending on which metric is used, questions of remoteness and certainty might come into play. Various metrics could be considered (ROI, ROE, ROCE, RONA, ROIC, WACC …). This post cannot discuss their relative merit. On a higher level of abstraction, one may ask: What is the more pertinent perspective?

a) using the rates of return the claimant earned, thus implicitly assuming that the claimant would have earned the same rate of return on the disbursed overcharge (~ “lucrum cessans perspective”),

b) the rates at which the claimant had to finance itself (~ “damnum emergens perspective”)?

The Paris Commercial Court recently allowed the use of individualised rates by the claimant. Even so, while the claimant argued for using the (presumably higher) rates of return, the court ruled for using the cost of debt instead, which the claimants had argued for in the alternative.¹⁷

5. Practicalities

a) Software

If a cartelised product has been bought repeatedly over years, for example, every month, and/or payment was affected in instalments, the calculation of interest becomes burdensome. Applying the rate in force for the period for which it was in force, compound or simple, is a task best done by computers. At CDC, we use our interest module.

b) Ne ultra petita

In continental European jurisdictions the ne ultra petita principle seems to apply (nearly) universally.¹⁸ The European Courts also apply it.¹⁹ If the claimant does not ask for interest in those jurisdictions, or could ask for interest with a higher rate but does not do so, the court will not award it. This could even apply to the kind of interest: If the national law might allow awarding compound interest, the claimant should request it.²⁰

By Ben Bornemann

Interest on cartel damages