Semiconductors are in short supply. Companies from virtually all sectors, including most notably electronics and automotive, are struggling with meeting customer demand. In response to the crisis (which economists say will last until at least Q2 2022), many have taken a pro rata approach to allocating short supplies, i.e. the remaining quantities are allocated proportionally amongst all customers, based on the customers' forecasts. While sharing the pain seems intuitive in times of crisis, the competition law risks, as well as opportunities, involved seem underexposed in the current debate: Is pro rata the only legally permissible option available when addressing the current crisis? Can suppliers choose a different allocation method without running afoul of applicable competition law rules? And, last but not least, can a customer successfully run a competition law argument saying it is entitled to a larger piece of the cake than would be allocated to it under a (strict) pro rata approach? This article provides answers to (some of) the most urging competition law questions surrounding the current chip crisis and advocates a more flexible approach to dealing with it.

Short on chips – high in risk?

Shortages may lead to economic dependency. This is true particularly in the semiconductor industry where certain (e.g. automotive) chipsets are tailormade to meet very specific cus-tomer demands and production capacities are planned long ahead based on (rolling, as the case may be) forecasts. If customers cannot then easily switch demand to another supplier (who may be likewise struggling with performing even its existing supply obligations) they may be dependent.
Under German law a company that is economically dependent must not be discriminated against. Similar rules exist in France, Spain, Italy and Belgium. The legal concept is similar to that of abuse of dominance, but the threshold is much lower. The law does not require (abso-lute) market power for the non-discrimination rule to kick-in. A supplier can be found rela-tively dominant – relative in relation to one customer, but not necessarily another – simply because that customer may be dependent on being supplied by it.

Relevant case law and practice suggests that shortages are a practical example of these de-pendency rules (others include e.g. dependency due to product range or company-specific investments). Although this case law dates back to the oil crisis of the 1970s (see e.g. Higher Regional Court Berlin, decision of 04.07.1974, Kart 27/74 – AGIP II), there is no reason to consider it outdated. This is true in particular with a view to market-wide shortages as we currently see with semiconductor supplies.

Under the assumption that, in (and for the duration of) a crisis like this, chip suppliers and/or suppliers of modules containing a complete chipset may be considered (relatively) dominant and may thus be subject to specific non-discrimination rules vis-à-vis their cus-tomers. The question is then how to avoid the abuse.

Of apples and oranges

Non-discrimination does not mean treating everyone equally. It means treating apples as apples and oranges as oranges. Put differently: a company must provide objective justifica-tion as to why one (group of) customer(s) is treated differently from another. Allocating short supplies pro rata amongst all customers will likely stand this test. Each customer is given its fair share (‘quota') in the total cumulative volume of all orders. No customer is treated differently compared to the others.


  • Customer A ordered products the production of which requires 2,000 type 1 semicon-ductors.
  • Customer B ordered products the production of which requires 8,000 type 1 semicon-ductors
  • Customer C ordered products requiring 10,000 type 1 semiconductors.

The total cumulative purchase volume of type 1 semiconductors thus is 20,000. Customer A's quota is 10%, customer B's quota is 40% and customer C's quota is 50%.

If, in a given period of time, a total of only 10,000 (instead of 20,000) semiconductors is available to supplier X then customer A gets 1,000, customer B gets 4,000 and customer C gets 5,000 semiconductors, each in line with its quota of 10%, 40% and 50%, respectively.

In this scenario, typically it will be difficult to argue that either A, B or C are being discrimi-nated against. The pain is shared evenly amongst all of them.

No pain no gain

However, what if adjusting A's quota by just a few more units could avoid causing maximum damage to A, e.g. in the form of a line standstill? Can A then request to get a better deal? Can A even bring a claim for the supply of those additional units, or for damages for failure to supply those additional units, based on abuse of dominance (“refusal to deal”)? Or can the supplier X retreat to the argument that if A were favoured (compared to the share/quota that it would be entitled to under a strict pro rata approach) this would in turn discriminate B and C? Probably not.

Allocating short supplies in times of crisis is not (only) about quotas. It's about case-by-case justice. If, under the circumstances of the individual case at hand, adjusting A's quota can avoid serious distortions of competition on downstream market levels where A, B and C compete, treating A different from B and C is not discriminatory; it is rather a means of avoiding the abuse. At least this is true where only minimal adjustments (of, in the example above, say 10, 20 or 100 units) to an otherwise rigid system (i.e. a strict pro rata approach) can avoid maximum damage.

Competition law unchained

So what if, in the examples above, conversely supplier X wanted (but did not necessarily have to, see above) deviate from a strict pro rata approach, you may ask? Doesn't this pave the way for arbitrary results? Not necessarily. Competition law does not force even a domi-nant supplier of goods in short supply to treat the unequal equally. Alternative allocation methods can be justified so long as they are backed by objective commercial reasons.

For example, during the oil crisis back in the 1970's (see above), competition authorities and courts both on the EU and national level have held that “having regard to the objective of the GWB [German Act Against Restraints of Competition] directed towards the freedom of competition” preferring existing or regular customers over stray customers is legitimate and not discriminatory, even in times of crisis (see e.g. Higher Regional Court Berlin, loc. cit. – AGIP II; CJEU, decision of 06.03.1974, court reports 1974, 223 – Commercial Solvents; de-cision of 29.06.1978, NJW 1978, 2444 – B.P.; Bundeskartellamt, activity report 1979/1980, official record 9/565). Notably, this was true even despite the fact that oil is a commodity that can be allotted comparatively easily. Not surprisingly, in another case the Higher Re-gional Court of Stuttgart (decision of 30.04.1980, WuW/E OLG 2700, 2702 – Modelleisen-bahnen) therefore rejected a damages claim for (alleged) refusal to deal, arguing that this case concerned highly customized products the production and sales of which was closely interlinked.

Chips (or modules containing a complete chipset) are also customized to meet the specific demands of e.g. automotive or electronic devices. Transposing the above case law to the cur-rent chip crisis therefore means that also a chip (or module) supplier can likely invoke legit-imate commercial reasons for treating one customer differently from another (that is to say deviating from a strict pro rata approach). While applicable competition law rules will re-quire a careful balancing of interests in each individual case, generally such reasons may in-clude, but are not limited to:

  • The turnover generated from / profits made with one customer as op-posed to another, its payment history etc.

    If regular customers can be favoured over stray customers (see above) without dis-criminating one against the other, more generally this means that some customers are more important than others, commercially-speaking – and that competition law should not prevent you from factoring this in when calculating quotas. One way to do this could be to use a sliding (as opposed to a strict pro rata) scale that rewards e.g. the duration of the customer relationship, its profitability or its payment history. As a positive side-effect, doing so could also help avoid that some customers may feel inclined to inflate their (fixed or rolling) forecasts just in order to secure a “larger piece of the cake”, as is the case under a strict pro rata approach.
  • The customers' economic state ex ante (prior to the current crisis)

    In the same vein, some customers may have been in distress, and may not have been fully utilized, even before and irrespective of the current chip shortage while others (primarily small volume/high margin suppliers) may come away empty-handed if measured exclusively pro rata against larger players. It is difficult to imagine that a competition authority or court would find it discriminatory if some (limited) quanti-ties were deducted from the former and allocated to the latter instead (or any other customer of over-proportionate economic importance to the supplier, see above). The same is true for reserving upfront a limited number (e.g. 1-5%) of the total avail-able supplies for small scale customers and allocating only the remaining quantities pro rata. 
  • The fact that some customers may be ‘system-relevant'

    Last but not least, as seen in many crises before (take e.g. the financial crisis of 2008/2009 or the current pandemic) some companies may (claim to) be system-relevant, which would allegedly entitle them to a larger quota than would be allocat-ed to them under a strict pro rata approach. ‘System relevance' is a slippery slope to take in this context, given in particular that it is not a legacy competition law argu-ment that authorities and courts are used to dealing with. Also, the notion of ‘system relevance' is in itself blurry – a truck maker, for example, will most likely not qualify as ‘system relevant' just because trucks are, amongst others, used to deliver food to the grocery stores; if that was true, the same could be argued for passenger cars be-cause passenger cars as well are used not only for transporting passengers but also e.g. human organs. On the other hand, competition law has always been, and re-mains, flexible enough to address new theories of harm, as well as new arguments to rebut those theories. Take for example the ongoing sustainability debate. Also, when distinguishing products that are indeed ‘system relevant' from those that are not, reference can be made e.g. to the applicable rules governing foreign direct invest-ments (‘FDI'). All in all, it therefore appears that, while caution should be taken in making use of this argument, there may be cases where indeed (backed by an official certification, as the case may be) favouring a ‘system-relevant' customer over one that is not can be justified and is not discriminatory.

Wrap up: risks and opportunities

As with any crisis, the current chip crisis bears risks and opportunities. The risks are (most-ly) covered by taking a pro rata approach to allocating short supplies. However, the oppor-tunities are not. Considering further that even a pro rata approach is not risk-free, there are good reasons to develop a fresh view on how to deal with the ongoing crisis. Suppliers may want to think about refining their current allocation method so as to better meet their busi-ness objectives, while customers may want to consider bringing a competition law argument aiming at securing a better deal (better compared to a strict pro rata approach). As this arti-cle has shown, competition law leaves some leeway in both directions – and those making use of it will likely emerge from the current crisis even stronger than before.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.