Mark Harris, writing in IEEE Spectrum, has an extremely well researched and insightful essay on The Lowballing of Kodak’s Patent Portfolio.  Harris compares the difference between the selling price of 1,730 of Kodak’s patents and the much higher earlier valuation – the amount their consultants had told them to expect.

As a Certified Patent Valuation Analyst, I know there are many different ways to value a patent portfolio, but they all boil down to a discounted cash flow analysis of how much licensing revenue you believe can be generated from it.  How much should one pay for a portfolio which should generate $100 million per year in royalties over the next five year?  Superficially, it’s a simple question.

Scratch the surface, though, and you discover a lot of hidden variables.  For example, pre-existing cross licensing agreements may place certain prospective licensees out of bounds for a given licensor – but not for others. Thus, the plausible expectation of the $100 million that you could generate is far too high for your prospective buyer, and he will value the portfolio at a lower amount than you have.  Furthermore, assumptions about the annual cost of obtaining that $100 million (mainly litigation), or a reasonable time period for negotiations to take place, often lead to significant differences in a valuation.

This is why valuations of those 1,730 Kodak patents ranged from $1.8 to $4.5 billion. But the various consultants who came up with these different numbers overlooked one of the fundamental truths of economics: The worth of a thing is what it will bring.

Nicholas Cage was painfully reminded of this fact when he sold his Rolex Daytona for $500 in Leaving Las Vegas.  When a buyer knows you’re desperate for cash, expect him to make a very low offer.  That’s precisely what happened to Kodak.  The bidders knew Kodak needed the sale to allow them to refinance their debt while in Chapter 11, and thus secured not only the original 1,730 patents, but a license to approximately 20,000 additional patents Kodak was hoping to retain.  In the end, the patents sold for only $94 million, or 2% to 5% of the amount Kodak was hoping for.

I have one quibble with Harris’ analysis: A filing to the bankruptcy court claims that $433 million of the $527 million payment from the IV/RPX consortium was in fact for a license to the approximately 20,000 patents Kodak wasn’t selling, meaning the 1,730 patents sold for $94 million.  While legally indisputable, I believe this simply reflects another valuation exercise, and the $94 million is artificially low.  But one certainly cannot claim that the whole $527 million was for the purchase of the patents, which is what numerous commentators said or implied at the time, so kudos to Mr. Harris for doing the digging to bring this fact to light.

The same thing can happen in reverse, of course.  Rockstar and Google may have been irrationally exuberant when they were bidding on the Nortel patent portfolio, driving the selling price past the $1 billion valuation to a staggering $4.5 billion.  It’s not hard to find people who think Rockstar overpaid, but maybe they’re just a better judge of value than the critics think.

The moral of the story:  No matter how accurate your valuation model, when the time comes to actually sell your portfolio, many additional factors can come into play which can significantly affect the final selling price.   Caveat venditor – let the seller beware!