Why is patent value so important?

For the patent system to serve its purpose toward the society, it is very important that only

“valuable” patents are granted exclusive rights. Granting of weak or invalid patents leads to welfare losses that are associated with a monopoly. If left unchallenged, the patent holder can accumulate private gains at the expense of consumers. The welfare losses are more than just the losses to consumers due to monopoly pricing. Third parties may be prevented from using a patented technology out of fear of infringement even though the invention was not novel in the first place. Thus, granted patents that are weak or invalid may potentially obstruct the diffusion of technology. Third parties have low incentives to challenge a patent unless they are directly affected by it (Farrell and Merges, 2004).

Lemley (2001) have, however, argued that some mistakes of the patent office may be

5 Evidence suggests all outcomes are equally likely as a result of the opposition procedure (van de Kuilen, 2013 p. 126)

overlooked as the most glaring mistakes are resolved by ex-post litigation. Lemley’s argument about rational ignorance of the patent office may seem prudent considering how few patents are ever litigated. But it is difficult to measure the social cost of weak patents that are barriers to diffusion.

The commercial and social value of a patent is thus important so that the patent office fulfils its purpose of encouraging only those innovations that are important for the society.

In the cloud of uncertainty surrounding commercial and social value of inventions, the field of patent value indicators looks for signs that indicate patent value. The next section reviews the various patent value indicators introduced by researchers.

3 Patent value and patent life cycle

All patents are not same in terms of their value. There is significant variance in their technical and commercial value. Empirical research (Pakes and Schankerman, 1986) has shown that half of the estimated patent value belongs to 5% of the entire patent population. More recently, based on survey data, Gambardella and Harhoff (2008, p. 79) found the mean value of a European patent to be about EUR 3 million. The median value was roughly EUR 300,000 and mode value only about EUR 6000.⁶ Owing to such skewed distribution of value, empirical research seeks to create measures that separate the wheat from chaff.

The literature (Trajtenberg, 1990) has differentiated between the private and social value of patents. The social value represents the net value created by the patented invention for social welfare. Private value is defined as the extra revenue generated by the patent over its lifetime for the patent holder.

For managers to make well-founded management decisions, it is important to have a fairly accurate estimate of the value of an invention. Applying and maintaining a patent is a cumulatively increasing cost (Figure 6). There are four major decision points faced by managers during the course of developing an invention. These are: (i) whether to file a patent application? (ii) Whether to continue with the patent filing following patent office actions? (iii) Whether to maintain a granted patent in force by paying renewal fees? (iv) How to exploit a granted patent? (Pitkethly, 1997)

6 Value estimated at 2003 euro level

Figure 6 Cumulative cost of a patent application (Source: Pitkethly, 1997)

The three classical methods of monetary valuation of patents are cost-based, market-based, and income-based methods. Cost based method relies on the estimate of actual costs incurred in creating the patented invention. The income approach attempts to quantify the discounted future cash flow forecast of the patent’s potential commercial use.

The market approach estimates value based on a similar transaction accomplished in the past. (OECD, 2005)

Economics and management scholars, however, are concerned with patent-based indicators that signal value (or other constructs of interest). This is because they are concerned with questions related to policy, organization, management and strategy. The treatment of patent information is therefore done at a more aggregated level. Patent-based indicators have been repeatedly proven in the literature to be correlated to patent value.

The patent value indicators have been validated by correlating them to (i) surveys asking inventors for perceived value of their innovations (Scherer et al., 2000); (ii) market or financial data (Hall et al., 2005); (iii) incidences of disputes (Harhoff et al, 2003); (iv) auction prices of patent transactions (Fischer and Leidinger, 2014); and more.

Patent value indicators churned out by researchers over time were put into four categories by Zeebroeck and van Pottelsberghe (2011). These were (i) Patent characteristics – these deal with the characteristics of the patent applications, for example, number of references in the patent document; (ii) Patent ownership – this category deals with characteristics of patent owners or applicants, for example, past portfolio size of applicant; (iii) Insider information – these indicators are theoretical and are created using qualitative interviews and surveys pertaining to the context of the invention; (iv) Filing strategies – they refer to filing strategies, patenting route selected, procedural indicators, etc.

Defining all the indicators in the literature is beyond the scope of this dissertation. The indicators used in works related to this dissertation are presented in the next section. They are separated into three categories, namely, patent characteristics, citation based indicators and procedural indicators.

A typical European patent application is divided into three parts. The first page contains general information about the patent like date of filing, family information, designated countries, name of applicant, name of inventors, technology classification of the invention, title, abstract, etc. The second part contains the full description, claims and figures. The third part consists of the search report appended by list of citations possibly added later. Patent characteristics and citation characteristics are obtained from the first and third parts of a typical patent document respectively. The procedural indicators are extracted from the European Patent Register.

3.1

Number of claims

The average number of patent claims can help in determining the breadth and scope of the invention (van Zeebroeck, et al., 2009). Experts opine that the legal proceedings related to patent infringements are invariably solved by an evaluation of the claims.

Typically, claims are divided into two types: independent or dependent. Independent claims are known to enhance the scope of an invention. Dependent claims usually provide more information and are built around independent claims. “…They operationalize what patent attorneys call fall-back options for legal disputes” (Reitzig, 2004). My own interviews have revealed that text of claims is more important than their numbers.

However, it is difficult to achieve a broad enough scope with just one claim. This makes publication claims an important factor in determining the quality of the patent portfolio.

The EPO allows up to 15 claims in a published patent. Any extra claim that a patentee wants to add has to be paid for (EPO, 2013). If an inventor is willing to pay for the extra claims, it can be inferred that the relevance of the patent in question is high or at least the applicant feels paying extra money will enhance the core scope of the invention. One of the major advantages of assessing the number of claims is the fact that this indicator is available early in time as soon as the patent is published. Empirical work has proven

“number of claims” to be significantly associated to patent value (Lanjouw and Schankerman, 1997; Bessen, 2008). A detailed discussion on strategic aspects of drafting patent claims is avoided here as “number of claims” is only used as a control variable for legal scope of a patent.

Number of technical classes

Each patent application is technically classified into specific technology areas related to the invention. It is a hierarchical way of assigning the category to which a patent belongs and is known as International Patent Classification (IPC) classes. Number of IPC classes attributed to a patent application has been used as a proxy for the scope, and hence the

market value of a patent (Lerner, 1994). It has also been argued in the literature that IPC classes have little or no association with the value of the patent rights (Lanjouw and Schankerman, 1997; Harhoff et al., 2003). Empirical literature has however used

“number of IPC classes” as suggestive of technological diversity of a patent (Guellec and van Pottelsberghe, 2000, 2010). Number of IPC classes has been used as a control for technical scope of inventions in some papers related to this dissertation.