Standard Essential Patents – SEPs – keep our digital world connected, but SEP regulation is currently a contentious and critical debate in Europe. To cut through the noise and understand what the European Commission’s proposals to regulate SEPs mean for the sector Euractiv’s Christoph Schwaiger spoke with Prof. Joachim Henkel, a widely published academic specialising in patents and standards.
CH: What are the issues surrounding Standard Essential Patent (SEP) licensing?
SEPs have a number of characteristics. In the larger standards, there are just many of them. Tens of thousands in the really big standards.
Secondly is their distributed ownership. For 5G, more than 300 parties have declared to own standard essential patents.
Thirdly, many of these standards are used in various fields. So they are enabling technologies. It’s not just smartphones that use 5G, but also cars, planes, and all sorts of Internet of Things (IoT) devices.
Standard development organisations bring these patent owners together. In the past, it was a simpler game with big telecom companies cross-licensing to each other. Now we have many more patent owners and many more licensees or implementers. Many of them with little understanding of the technology.
A fourth one is market power. If a standard becomes dominant, there’s no alternative to it, the technology essentially constitutes a monopoly. Every firm that wants to be active in the respective market needs it. And since the patents on it are not substitutes to each other but complements, each patent owner, in a sense, has a monopoly.
To prevent them from charging excessive royalty rates, which would be harmful for consumers, they have to make a FRAND commitment – the commitment to license their patents on the standard under fair, reasonable, and non-discriminatory conditions. But what FRAND actually means is very fuzzy. In the end, that has at least to some extent been determined through case law and other estimates.
Big firms know the technologies and can afford these legal battles, though it’s difficult enough for them. But for smaller firms, it becomes extremely difficult. For them, it is nearly impossible to assess the quality of the patents offered for licensing, what share of the overall standard a given portfolio constitutes, and what royalties are fair and reasonable.
Patents can be enforced on any level of the value chain (like at the chip level or product level). This makes it even more difficult to determine royalty rates because the reference base varies strongly. Given all these complications and the huge number of standard-essential patents, it is still not clear what FRAND is.
Patent holders have realised that the further downstream they enforce their patents, where the products are more valuable and revenues are higher, the higher the royalties they can demand. Even though FRAND should be independent of the licensing level. A further problem is that downstream firms for example in the IoT space are mostly ignorant about the technology, so it is impossible for them to assess if a licensing demand is FRAND or excessive.
That was a very long- answer, but it is a complicated question.
CH: So, there are clearly issues, but the European Commission is trying to solve them and it made some proposals. What did it get right?
One thing they’re doing right is to do something.
Issues related to a lack of transparency are becoming more pertinent the more we use standardised technologies outside the core of ICT.
I’ve been an expert witness in many cases. Every patent owner that I’ve come across says ‘My portfolio is of high quality and above average’. This is unlikely to be true for all the portfolios I’ve seen. So how does a potential licensee, in particular one that doesn’t understand the technology, evaluate this?
The reference to earlier licensing deals is common, but often not helpful. There is a lot of confidentiality, and patent owners typically present earlier licenses that are advantageous to them. Also, even if a look at earlier agreements might help to achieve non-discrimination, it remains unclear if these have been fair and reasonable.
The Commission proposal is complex but let me address two central aspects. The first is a process to arrive at a public aggregate royalty for a standard. This is important because implementers want to license the standard, not sets of patents, and need to know the cost of the standard.
Second, the Commission proposes a method to assess individual portfolios for their essentiality, as a guideline for how to split the aggregate royalty among the patent holders. This is not perfect since also patent validity and technical relevance should be considered. Also, each standard comes in different versions. But then, these issues arise also in each bilateral licensing negotiation.
In any case, compared to the situation we have today these measures are considerable improvements.
CH: What did the Commission get less right with these proposals?
There has been criticism of ambiguity and imprecision. Peter Picht from the University of Zurich for example has commented on this, as has the Max Planck Institute for Innovation and Competition.
Another criticism that was raised concerns the essentiality test, which is the basis for assessing the shares that the various portfolios constitute in the standard. That assessment is quite burdensome given the large number of standard-relevant patents. But the expected cost is small compared to overall royalties from standards, and patent pools perform similar analyses. Unless there are more efficient ways to increase transparency, I think we need these tests.
One suggestion that Taraneh Maghamé and I have made in the IAM Magazine, and Jorge Contreras before us, is that the patent owners could among each other determine the IP shares that each owns in the standard. After all, they know the technology much better than implementers or courts. But if patent owners are unwilling to get there, to provide that transparency on their own account, there will have to be alternatives.
CH: On the aggregate royalty, is it even feasible to implement the total aggregate royalty proposal across diverse standards and industries?
Well, let me ask you the opposite question. What do we have now?
The same questions matter in each licensing negotiation. It’s the same story. If a patent owner wants to license patents to whatever implementer on whatever level, the same things matter. There are different flavours, different generations of a standard.
There are different application fields. And in bilateral negotiations, all these aspects have to be considered as well. It strikes me that it’s simpler to do this once and for all for the entire standard than to do it in each bilateral negotiation.
CH: During the conference by Charles River Associates, a panellist described the situation now as not ideal. What would be the effect for the regular consumer if these issues are not fixed?
Ana-Mariya Madzhurova from Fairphone made an excellent point, and I have heard the same story from another small smartphone maker. They’re in a very difficult situation because they can’t afford to go to court, despite the fact that they are strongly discriminated against in terms of royalties. As a results, their products become more expensive for consumers.
More transparency will help SMEs, but also larger firms, in particular in the IoT space. Downstream innovation may be hampered if firms can’t calculate their cost ex ante, and the risk of unexpected royalty demands, or even legal action creates harmful uncertainty.
Another thing is the lack of technical understanding downstream. I have spoken with small IoT firms that lack the knowledge to assess licensing proposals. And this is aggravated by the fact that licensees typically ask for an NDA that forbids the implementer to speak even with its supplier.
Sometimes small firms consider going back to older or less suitable standards where licensing is presumably simpler. This would harm consumers because good technologies would not be used as broadly as possible.
Owners of standard-essential patents are these days focusing on licensing downstream. My research suggests that the main reason is money: On the device level or even user level, also high royalties appear small in comparison to product value.
So, this effect tends to push up royalties. It is true that the proposed regulation addresses neither the amount of royalties nor the licensing level in the value chain explicitly (maybe it should). But the transparency it brings might counter the royalty-increasing effect of downstream licensing. For consumers, if inputs get cheaper products usually become cheaper.
As a final point, the transparency that the regulation seeks to create should discourage the business of patent assertion entities. These firms don’t produce or invent anything, they just enforce patents. Often, these are small portfolios transferred from larger patent holders, who seek to increase their income in this way. More transparency should help to make this practice unattractive.
CH: Time is pressing, so we’ll do two final questions with one-minute answers. First: How can we make SEP licensing more transparent?
My wish would be for patent owners to agree among themselves on who owns what share of the IP on a standard. This would be a simple and consistent way of apportioning the aggregate royalty.
CH: Lastly, did you spot any things you wanted to comment on or hear misconceptions that you wanted to clarify from the panel discussions that you weren’t part of?
Many important legal questions were discussed. For example, is the proposed regulation limiting fundamental rights or not? I think it is crucial that the legal ambiguities in the proposal are ironed out.
A comment by Rebecca Porath from Intel concerned patents that happen to become part of a standard without the owner’s agreement and even knowledge and without a FRAND commitment. Should such patents be subject to the proposed regulation? I think the answer is not clear.
Patent owners must have the possibility to withhold their inventions from a standard.
On the other hand, excluding such patents from the regulation might invite firms to act like Rambus did some twenty years ago, sneaking patents into a standard without a FRAND commitment and then seeking excessive royalties. However, I think these issues can be solved because technically important patents, the ones that firms really care about, aren’t likely to be built inadvertently into a standard.
In sum, the proposal needs to be polished. I hope that the next steps in the process will be used to fix open issues, without watering down the intention of the proposal.
The interview has been edited for clarity and brevity.
