Home » Resources » Myth: Agreements from the 1990s Should Never Be Used as Comparables!
Dark Mode

Myth: Agreements from the 1990s Should Never Be Used as Comparables!

website img – 11-20-23_myth_1990s

Myth: Busted!

I myself have utilized the criterion in the title of this post as a reason to reject license agreements—on the sheer basis that they are too old to be considered good comparables to intercompany license agreements being put in place today. Arguably, the economic conditions were significantly different throughout the decades, so how can we possibly compare decades-old agreements, some of which have expired, to those of today? But if we look at RoyaltyStat’s statistics tool, we see some very interesting data.




The first thing that strikes me when looking at the data in the tables above is that the median of licensing agreements in the entire RoyaltyStat universe has increased by 1% over the last 36 years, from 5% to 6%. And during those years, there hasn’t been any fluctuation in the median to boot. If we look at the same data in chart form, we can more easily see the trends in the quartiles (see below).


Looking at this, the trend shows that the quartiles have tightened up through the years, looking at the broad picture. That said, should we automatically exclude agreements from our set, just because they’re older?

Given what the data show, I would argue that we should keep these older agreements and employ an interquartile range to remove any outliers. Why? Because there are more important comparability factors that we could be using, such as the industry and the type of intangible being licensed (a topic for another post, but also based on RoyaltyStat’s Analytics and Statistics Tools). Perhaps if you’re drowning in comparables, and have exhausted all other rejection criteria, you might choose to limit your search to agreements within the same inception window as your intercompany agreement, but I would argue that I’ve never seen a royalty CUP search where I was overwhelmed by too many perfect comparables, given the uniqueness of the property being transferred. In other words, there are better ways to filter your results, it would seem.

Myth: Busted