The Relief from Royalty Method and the Use of License Agreements

In a recent series of articles published in QuickRead, the weekly online news channel of NACVA, Robert F. Reilly discusses the “Application of the Relief from Royalty Method” in the valuation of intellectual property. Robert is one of the masterminds in intangible asset valuation, regular author and speaker, and Managing Director of Willamette Management Consultants.

In the article, Robert refers to “numerous commercial data sources that analysts may access to find arm’s-length intellectual property license agreement royalty rate data.” The main source to identify and download intellectual property license agreements is the SEC database where filers have to disclose “material agreements”. Among five other data sources, he also mentions MARKABLES. MARKABLES “.. is different from the other databases discussed above. Rather than drawing royalty rates from actual license transactions of intellectual property, Markables gathers its data from purchase price allocations published in SEC filings.”

Further on, Robert explains the application limitations associated with the use of commercial intellectual property license databases. Robert lists six particular limitations which analysts should be aware of:

1.??????“There may be numerous duplicate license agreements included in these databases.” True. Our own experience with one of the license agreement databases is that at least 20% of the license agreements are duplicates. We even found one license agreement listed five(!) times in the database. The reason for such duplicates is old database technology. Employees(s) in charge of identifying and gathering relevant license agreements do not realize at the time of filing that a particular agreement was already included in the database before, and filed it again. For any reason, the data vendor could not clean or eliminate the duplicates later.

2.??????“There may be multiple updates of the same license agreement in the database (i.e., another type of data duplication).” True. Sometimes, license agreements have a longer history of renewals and/or amendments. Such updates may – or may not - involve new terms, and adjustments of royalty rates. We found in one larger project that 75% of license agreements have predecessor or successor agreements, some even tripled. Some of the involve the same terms, and the analyst must decide if he counts this royalty rate twice. Most of them involve different, adjusted royalty rates; typically, royalty rates decline with the updating of the license agreement. For the analyst, the challenge is to decide which of the different royalty rates from this agreement is appropriate: both, either of the two, or a blended rated.

3.??????“.. some “license agreements” may actually be asset purchase agreements or other types of transactional agreements. Not every agreement is an intellectual property use license.” Very true. We estimate that approximately 65% of all license agreements sourced from the SEC database do not qualify as intellectual property use license, but rather as ancillary license agreements to support other agreements like asset purchase, share purchase, joint venture or supply agreements. The terms and royalty rates of such agreements are not between independent, willing parties and carry a lot of noise. For more details on this phenomenon, please read our white paper on the landmark case Amazon vs IRS here.

4.??????“Some of the license agreements may be between related parties (and, therefore, may not be arm’s-length agreements).” True. Our own experience with one of the license agreement databases is that 7.5% of the license agreements are controlled transactions between related parties. Some data vendors use a tag that qualifies such transactions as related party transactions, others don’t. Those who do may miss to identify all relevant cases. In any case, the license agreement itself does typically not allow to understand if the parties to the agreement are related or independent.

5.??????“Some of the license agreements may involve several different types of intellectual property (e.g., a trademark and a patent), making it difficult for the analyst to extract a specific royalty rate for a single intellectual property.” Very true. From our own experience with license agreement databases, we estimate that 72% of the license agreements involve several (more than one) different types of intellectual property, and only 28% represent pure play license agreements involving only one single intellectual property.

6.??????“The royalty consideration formula in the license agreement may be presented in a form that is not particularly useful to the analyst (e.g., a royalty dollar per 1,000 barrels of beer sold—rather than a royalty payment as a percent of licensee revenue).” True. From our own experience with license agreement databases, we estimate that 30% of the license agreements show royalty rates that are not calculated as percent of revenue and therefore not comparable. Typical alternatives for % of revenues are dollar per unit or weight sold or purchased, percent per profit, or percent per cost of supply.

There is one more limitation which Robert does not mention in his list.

7.??????The relief from royalty method requires the application of a profit split test. The profit split analysis splits (or allocates) some measure of owner / operator income and assign that allocated income to the intangible asset. Profit split is intended to check if – after paying royalties to the licensor – the licensee is left with enough return on his own assets. Typically, between 25% and 33% of the total business income is said to be allocable to the intangible asset via a royalty payment, the remainder amount is the reasonable portion that shall remain with the licensee.

The problem with license agreements is that none of them is tested or testable for profit split. It is simply unknown if the licensee generates enough profit from the agreement, or if the agreed royalty fee (and other terms of the agreement) was too high for him to operate profitably.

One indicator would be the effective lifetime of a license agreement, that is not its contractual term but the period during which licensed products under the license agreement were effectively sold. Interestingly, the majority of license agreements are not renewed into a second term, and a sizeable number do not even reach the end of their first contract term for premature termination.

In practice, the profit split test is only applied to the subject intangible asset valuation. However, it is not applicable to test the reasonableness or feasibility of effective royalty rates in license agreements. As a result, (some of) the royalty rates selected from license agreements for comparability analysis may have been financially unfeasible.

Given all these limitations, the selection of intellectual property license agreements available and useful for valuation purposes becomes rather small. For most sectors, it is almost impossible to identify a meaningful number of comparable agreements, if any. Your initial selection of comparable license agreements may melt down like snow in the sun, and finally even disappear. Numerous experts testifying at court can tell their stories how their royalty rate analyses were – rejected.

For all these – and more – reasons, we have created MARKABLES. As Robert stated: “MARKABLES is different from the other databases.” Please feel encouraged to find out where we are different, and if we are better.