Paris, 9 June 2021 – Financial valuation of intangible assets is a complex process. In the case of patents, before applying one or more method(s) for estimating a value, a qualitative evaluation step is necessary in order to capture patent imperfections (such as claim scope or the number of protected countries). Commonly called “scoring,” this method consists in analyzing a multitude of criteria in order to obtain an index reflecting a patent’s potential. This article focuses on the main methods used to perform this type of assessment to date.
There are two standards for the monetary valuation of patents: the German standard DIN 77100 and the Austrian standard ÖNORM A6801, both of which were published in 2011. These standards notably define legal, technical, and economic influencing factors to be taken into account when qualitatively valuing a patent. However, the complete list of criteria to be considered remains at the discretion of an expert evaluator, who, once the context of the valuation has been specified, is free to choose the method that is to be applied.
This lack of a specific normative framework for the qualitative aspects of patents has led to the creation of different scoring methods that can be classified in the following two categories:
The main “manual” scoring methods were defined in the early 2000s. The most well-known is the IPscore® tool, originally developed by the Danish Patent and Trademark Office (DKPTO) and later acquired by the European Patent Office (EPO). This database, accessible free of charge on the EPO website, provides a qualitative assessment of a patented technology based on a total of 40 questions grouped into five categories (legal status, technology, market conditions, finance, and strategy) which are evaluated on a scale of 1 to 5. The answers to certain categories of questions make it possible to calculate, on the one hand, the net present value specific to each evaluated patent and, on the other hand, the percentages of opportunity and risk that are associated with each evaluated patent. It is therefore not, strictly speaking, a database that can be used to assign a quality score to a patent, even if a certain number of relevant criteria are taken into account.
Moreover, given the high number of elements to be evaluated (sometimes requiring very specific strategic and financial information from the company exploiting the patent), IPscore® is time-consuming and difficult to use by neutral expert consultants.
As a general rule, “manual” quality assessment methods used by patent valuation experts result from their own combination and weighting of criteria, including, most often, patent validity, claim scope, the ability to demonstrate infringement, or geographical coverage of the evaluated patent. When examining the details of the application of these methods and the calculation of each criterion, an almost total absence of metrics (i.e., indicators derived from statistics that can be determined automatically, for example, by using a patent database) is observed. The subjectivity of this type of method is thus increased, as quality scores may vary according to the degree of appreciation of the expert evaluator.
Alongside these methods, i.e., still at the beginning of the 21st century, economists began to publish work demonstrating the impact of certain information available in patent databases, which were expanding rapidly at the time. A certain number of parallels were drawn between criteria, such as the number of patents citing an earlier patent (in their description or search report) and the earlier patent’s market value. From patent citations, and more specifically from the dispersion of patent classifications (such as IPC1 or CPC2) among citing (post-patent) and cited (pre-patent) patents, the following indices were defined: patent generality (Hall, Jaffe, and Trajtenberg, 2001), patent originality (Hall, Jaffe, and Trajtenberg, 2001) and patent radicalness (Shane, 2001).
In the United States, this period also corresponds to the development of intellectual property investment banks, which were the first to develop automated methods for assessing the quality of patent portfolios. The main example is the IPQ® (Intellectual Property Quotient) tool developed by Ocean Tomo. This method, designed for U.S. patents, notably consists in linking the value of a patent to its renewal period, i.e., measuring the chances that a patent will be maintained in force over time. The details of the metrics used for this evaluation have never been disclosed by Ocean Tomo, which has led to a certain reluctance on the part of expert evaluators. Other American companies, such as PatentCafe (or Pantros IP, now acquired by IP.com), have disclosed all of the indicators used in their automated patent quality assessment method. In this case, it is the very definition of the metrics, their large number, and their use(s) that have raised questions among patent attorneys.
Much more recently, the main private sector paying databases, such as Questel Orbit, PatSnap, PatentSight (LexisNexis), and Derwent Innovation (Clarivate), have developed automated patent quality assessment tools. Their principle consists in calculating a “technological” score (depending, in particular, on the technical field of the invention and citations) as well as a “market” score (depending, for example, on the geographical coverage of a patent family measured according to the GDP of the protected States).
Some methods also use the remaining life of a patent to provide a final score in the form of an index, to be compared with a score of “1” corresponding to an “average” patent. Only PatSnap offers a method that goes so far as to display a dollar value when evaluating a patent family (the amount being calculated in particular via comparisons with past transactions relating to patents of the same technical field).
These recent developments in patent databases have the merit of putting the question of patent quality assessment back on the table, as platforms provide a rapid method for reviewing large portfolios for optimization and/or transaction purposes. The main criticism that can made with regard to these methods remains the “black box” aspect, due to the lack of justification for the calculation of indicators and their combination in order to obtain the final quality score. In addition, the lack of evaluation of essential criteria such as validity and claim scope is a source of reticence on the part of expert evaluators, who often prefer to use their own method.
To summarize, existing patent quality assessment methods fall into one of the two categories defined in this article. These two visions of how to rate a patent conflict in several ways. On the one hand, we have qualified experts who rarely take advantage of the strategic information contained in patent databases, and, on the other hand, we have companies that try to provide automated solutions, even if this means leaving out certain criteria that are nevertheless essential to patent valuation. The choice between these alternative solutions often depends on the amount of time and resources available.
Taking these different parameters into consideration, Regimbeau has developed a new method for assessing patent quality that could be qualified as hybrid, as it involves a combination of criteria measured alternatively in a computerized manner and using human judgment. This method will be the subject of a further article to be published in relation to this theme.
