Saturday, April 20, 2013
The Securitization of Patents
[I cross-posted this at Patently-O last week, but thought it might be of interest to a more general audience]
My forthcoming article in Duke Law Journal, The Securitization of Patents, argues that the best way to create patent markets might be to start treating portfolios as securities. A full draft is accessible at this SSRN page. The article makes four basic points:
- Aggregation and trading is not limited to non-practicing entities – everyone is doing it.
- Companies are trading aggregated patent portfolios as they do other patent instruments, either through sale or licensing.
- Aggregation is beneficial, even critical, for efficiency; this is directly contrary to the conventional wisdom.
- Based on the above, markets might be improved by applying securities treatment to patent portfolios.
[NB: I focus on portfolios, not individual patents. I also focus on sale and licensing of patents, not the initial patent grant. The paper explains why in more space than I have here.]
When I first wrote Patent Troll Myths, there was very little empirical data about NPEs. Since then, such research has exploded, with new data every week seemingly counting the number of NPEs and their cases. This data, though helpful, leaves a lot to be desired, I think. First, there is a rarely a real apples-to-apples comparison with the activity of product companies (and when there is, the comparison is not very granular). To that end, I’ve been developing a matched data set for my Patent Troll Myths data so we can test what real differences in quality and quantity, if any, exist. Second, the data largely ignores licensing practices, which can be quite similar. To be fair, licensing data is difficult to come by, but without it, normative determinations are difficult. Third, studies like mine, which look at the provenance of NPE patents, are rare.
These issues lead to my first point: aggregation is not just for trolls anymore, if it ever was. The public is becoming a bit more aware of this with new focus on privateering, the outsourcing of patent enforcement by product companies to licensing and assertion specialists. The idea that aggregation is just fine when a product company does it, but suddenly evil when those same patents transferred to a third party has never sat well with me. And regardless of moral considerations, the fact of the matter is that patent aggregation is everywhere.
My second point follows from the first: aggregated portfolios are being used as assets, and traded as such by all sorts of companies. This is nothing new; people have been writing about patents as a new asset classes for a while now. Transactions are getting bigger, however, and they are hitting the news. Perhaps no transaction better illustrates my point than the recent Kodak patent auction. First, Kodak offered its patents for sale as a financing strategy in bankruptcy. Second, the eventual buyer was a consortium including, among others, Microsoft (a product company); Intellectual Ventures (a licensing company, but also one that litigates, but also one that aggregates defensively; and RPX (a defensive aggregator). This one transaction is my argument in a nutshell: everyone is aggregating, and they are doing so in buy/sell type transactions for financial purposes.
My third point is that such aggregation is not always (or necessarily often) a bad thing. This is decidedly against the conventional wisdom. Companies with large portfolios surely have the ability to cause “royalty stacking,” but in practice this is less likely than if many separate parties enforced those same patents. Litigation looks much the same; regardless of the size of the portfolio, courts are just not going to hear a case asserting 1000 patents. Only a few (at most 5 or 10) patents will be at issue, and then the aggregator looks like anyone else. Similarly, in negotiations, the parties usually haggle over a few “lead” patents. This is little different than negotiation with the owner of few patents – with one big exception. When you come to terms with the aggregator, you can settle and license hundreds or maybe thousands of patents at once. Not so with single-patent owners. These folks line up one after another, asserting a few patents at a time. The biggest NPEs will often assert patents obtained by individual inventors; would product makers really rather that the inventors assert their own patents separately? Maybe, before a time when people figured out a viable mechanism for funding patent assertion, but now that individuals might seek funding for enforcing their own patents, a single aggregator must surely be a better option than many inventor plaintiffs.
There is one difference with aggregated portfolios, of course. When the parties are done haggling over the lead patents, the portfolio owner always has more to discuss while the small patent holder has none. But rather than being the greatest cost of the portfolio, a seemingly bottomless portfolio is its greatest benefit.
And that is my fourth point: when parties are trading portfolios, the haggling should be over price instead of quality and infringement. In a large enough portfolio holding patents directly related to a particular product, there will surely be some number of patents that are both valid and infringed. The question is how many, and how much it will cost to find them. A central thesis of my article is that treating portfolios as securities will help lower transactions costs in a variety of ways by limiting the litigation costs of finding those infringing patents and instead better pricing patents in the market. For you legal sticklers, I didn’t just make this up: the paper looks at portfolios under the Supreme Court’s famous Howey test and concludes that such treatment is at least plausible under the law.
How might securities laws benefit markets? Not in the traditional “public offering” way. I suspect that most transactions would be excluded from the registration requirements. However, such transactions might be regulated as dark pools, and require clearinghouse treatment that makes such transactions public. Further, stock fraud laws might require the disclosure of information that might affect portfolio value. For example, patent holders who know of anticipatory prior art might be required to disclose it rather than keep it secret. Perhaps most important, accepting that portfolios are simply financial transactions might drive efforts to develop objective portfolio pricing. The goal of such pricing schemes is to determine a portfolio’s price even though the parties cannot agree on the price of any of the particular patent in the portfolio. I examine several pricing strategies that might work (and several destined to fail) in the paper.
There is obviously much more in this paper than I can write here. I detail my arguments in the full paper.
TrackBack URL for this entry:
Listed below are links to weblogs that reference The Securitization of Patents:
I'm somewhat confused by the title. Regarding the securitization of patents -- do you mean to say that patents should be securitized? That is, should patents be contributed to an entity whose interests (securities) are sold to investors? This would be the ordinary meaning of securitization. However, it seems like you are talking about something else entirely -- treating patent portfolios as securities. If you are in fact talking about treating patent portfolios as securities, I would suggest coming up with a term different from securitization.
Posted by: securitization? | Apr 20, 2013 10:13:32 PM
Yes, that's what I'm saying about patent portfolios. I'm sticking with the title, because I think it's already happened. As I discuss in the paper, putting patents into an entity whose stock is sold is the easy case (for obvious reasons). I realize that's what people call the whole "put it into the entity" thing, but that's a key point I'm trying to make.
Posted by: Michael Risch | Apr 20, 2013 10:17:31 PM
Interesting idea but you should have co-authored the piece with a securities prof. Your securities reg stuff is greatly oversimplified and misses numerous issues. For example, you have no discussion of blue sky law, whether a trading market will have to register as an exchange, whether players in the market will have to register as broker-dealers, whether downstream purchasers of patents are likely to fall under the definition of underwriter, etc. These issues undercut your argument that application of securities regulation will reduce transaction costs because compliance/structuring around is expensive.
Also, note that section 4 of the securities act was amended to add subsections (a) and (b) so references to 4(1) and 4(2) should now be 4(a)(1) and 4(a)(2).
Posted by: securities gal | Apr 21, 2013 12:16:42 PM
and who says students can't pick a good article...the title is sooo catchy!
Posted by: John Harris | Apr 21, 2013 2:55:45 PM
John Harris - would that it were so! I've had some sweet, sweet titles over the years that got little law review love.
securities gal - thanks for your comment. This is why I wanted to post here as opposed to a patent only venue. I did collaborate with securities folks on this - though admittedly primarily on the Howey question.
Your comment makes clear that I need to be more precise about what I am and what I am not claiming. I had thought about many of these issues; I even had a blurb about underwriting and Rule 144, but I opted against it for length. I suppose the heart of it is that I do make some simplifying assumptions on both the patent and the securities side. The primary assumption is that these transactions will all be private, and thus most of the requirements will be exempt. Where folks need to raise public money for portfolios, I would expect them to securitize in the standard form, which others have already written about.
I should make my assumptions and limits more clear, and perhaps then focus on the areas that won't be exempt and dig a little deeper. The article is already longer than I wanted, so I have had to cut some corners - maybe too many for you. If I really gave all of these the full treatment, this would be a hundred page article. Maybe that will be the next one, after the idea takes off (I dream hopefully).
Your comments raises a couple of general issues, though, that I think are interesting:
1. Does the NPE who sends 10,000 demand letters to end users for a settlement count as a private placement or public offer? Perhaps they don't count anyway, because they do not contemplate a portfolio transaction.
2. I need to be more clear about "market" - by patent market, I mean the private "market" - not the kind of public trading market we think of with most securities. And that is a fundamental difference. If the patents are "traded" publicly, I'm inclined to say that such a market might have to be an exchange, with the costs that that entails. But that's not really what I'm talking about here and I should be clear about that.
Finally, mea culpa on the lazy citation. Old habits die hard. I think I've yet to cite to 112(f) like I should instead of 112(6) in patent law.
Posted by: Michael Risch | Apr 22, 2013 7:45:43 PM
You're free to call the paper whatever you like, and I think it's a really cool topic, but if you going to skip analyzing fundamental securities law issues, you should at least get the title right. If you demonstrate a misunderstanding of the term "securitization" and then also fail to address the securities laws relevant to your thesis, that's a bit of a double whammy.
Like perhaps many others, I'm hesitant to apply securities law concepts in an area where the legislature has provided broad legislation. To get me over my hump, I'd actually like to see the key provisions of security law analyzed, especially in light of the inappropriate use of the term "securitization" (which already implies a lack of knowledge of the subject matter).
I hope these pointed comments help push your project in the right direction. It's nice to see an article in a top law journal that is actually concerned with the law.
Posted by: securitization? | Apr 23, 2013 1:35:19 AM
The general issues you point out further demonstrate that you don't know what the h your talking about when it comes to securities reg.
As for your general issue 1, the real issue would be whether the demand letters fall under the '33 Act, sec 2(a)(3) definition of offer. If yes, there is no doubt that making 10,000 offers would be a public offering under Ralston Purina (the seminal case on the meaning of public offering). Incidentally, private placement is not a term used in the '33 Act or its rules. The real issue is whether such an offer would be exempt or non-exempt from the registration requirement of section 5. Additionally, I have no idea what you mean by, or the relevance of, the possibility that "they do not contemplate a portfolio transaction."
As for your general issue 2, the question is not whether the market is public or private but whether the market falls under the '34 Act sec. 3(a)(1) definition of exchange, a definition that makes no mention of public versus private markets.
Posted by: securities gal | Apr 25, 2013 12:58:30 AM
Sorry I didn't respond earlier - I normally get an email when there is a comment on one of my posts, but I didn't get that email. I only looked back to make sure I was capturing the comments as I move to the next phase of the draft.
To respond, though, a couple points:
First, securitization -
1. Yes, I know what securitization is. There are articles on this for patents, and and I believe I cite them in this paper. What I mean by the term here is "the transitions of patent portfolios into securities in and of themselves." Perhaps I'll put that early in the introduction so people don't tune out early on.
2. Your point that "Like perhaps many others, I'm hesitant to apply securities law concepts in an area where the legislature has provided broad legislation" is well taken. It's one I heard when I workshopped this. I have a section about why patent law doesn't get the job done, but I hear you that I need more detail about how the securities law will work to convince you. I'm working on that.
Second, securities gal -
3. "they do not contemplate a portfolio transaction" means that the 10000 demand letters are usually about a single patent, which would not be a security. But maybe they might.
4. Yes, I understand how public offers work, which is why I raised the point in the first place. I realize that you are skeptical, but I know how the exemptions work - I was using "private placement" as shorthand to refer to all sorts of exemptions 4(a)(2), 4(a)(6), Reg. D (some of which I mention in the article). At the comment, I hadn't thought through which of these might apply - I'm not convinced that a settlement demand is "public" in the way an "offer to license" would be, so it might depend on the language. I agree with you that the 10000 demand letters probably can't really remain a "non-public" offering, but that's actually a good thing for patent policy.
5. Yes, I know what an exchange is. My point by saying that these are private transactions is not that it was a private market, but that there is NO market. It's all person to person/company to company. So, the thought of an exchange never crossed my mind. I am hearing that I should say that explicitly, so folks don't think I don't know what I'm talking about. It turns out there is a marketplace for individual patent licenses now (IPXI), but because it is for fixed unit licenses, it falls under commodities. If that expanded to a full on trading marketplace, then it might well be an exchange. I'll mention that as well.
6. Now, what I did mention is a clearinghouse under Dodd-Frank, and I admit that the paper is very thin on that. I plan to bolster. I did mention IPXI here as a way to publish pricing - maybe this is where the exchange idea came from - I'll clarify that. I was only using it to show the benefits of reporting of transactions.
Posted by: Michael Risch | May 4, 2013 10:27:14 AM
This paper seems to be provoking a lot of boundary maintenance responses, for some reason.
Posted by: Bruce Boyden | May 4, 2013 11:34:51 AM
Bruce, here's a possible explanation: it placed very well, and exceptionally well if it were a real securities piece. Yet for a securities article, this one is done quite poorly. (For a patent article, I don't know.) So it touched a nerve: those of us writing in sec reg work hard to get our ideas out only to have 2Ls pick an article like this one. Now that the article placed well, we're either going to love it or hate it, there's no middle way. Same thing happens in every faculty workshop for a paper that overplaced.
Michael deserves a great placement - after all, he's had a number of good articles underplace - and he has the thick skin to put up with the push back. But he'll get it wherever he presents the paper. I'll save this post as a reminder that in order to make claims that transcend one's field, the author (I?) should know both fields very well.
Posted by: Another securities gal | May 4, 2013 12:38:39 PM
Well said, and well taken another securities gal. And it is a lesson well learned by me as well, though I do push back more on the notion that I didn't think about stuff than that I just didn't know it. There is clearly a precision and breadth that the securities folks wanted to see even if I discarded something; I just didn't grasp that aspect, and I agree that coauthoring with a securities person might have cured that (and brought even more securities ideas). Plus, the conversation has given me some good ideas that I didn't think of.
I'm presenting this at LSA on June 1, for anyone who wants to come and tear it apart in person...
Posted by: Michael Risch | May 4, 2013 12:47:43 PM
Well said and well taken, Another securities gal. My primary hackles are raised at the notion that I just didn't know the basics, which I do. I thought about many of the issues I've read here (though not all, admittedly), and chose to avoid them with glib shorthand; that did not go over very well and is definitely a lesson learned by me. I did view this as primarily a patent article and an essay at that, and coauthoring with a securities professor would have changed that. But, there's a reason I posted here, so I'm happy to get the comments and improve the paper - maybe I'll change a couple minds in the process.
I'm presenting this at LSA on June 1 if anyone wants to come and shred it there in person...
Posted by: Michael Risch | May 4, 2013 12:56:14 PM
I've though more about securities gal's comments on and exchange and my response, and I take it back. While I had person to person transactions in mind when I wrote the paper, there is a primary market/exchange type transaction worth noting: the patent auction. This raises a raft of issues, many of which securities gal noted. For example, is the auction really a public offer just because people know about it but where only a select few can participate? I'm sure there are parallels in securities law to look at, and maybe it wouldn't matter in the end if there's an exemption. Also, is the auction house an underwriter (unlikely), an exchange, or a dealer or broker?
On the one hand, it wouldn't break my heart to see a prospectus in these transactions. On the other hand, registration for a single portfolio purchase seems wrong (which is why I've argued that exemptions will apply until now).
In any event, I'll flesh this out in the next draft - thanks for the comments.
Posted by: Michael Risch | May 5, 2013 12:20:51 PM
I've uploaded a new version of the paper, complete with a new title, which apparently everyone (including patent folks) hated except me. I still think it was clever, but I accept that I don't want to spend precious space explaining why its a play on words when I could be getting to my point.
This draft is nearly 15 pages longer. It comes to most of the same conclusions about exemptions, but does add a lot of meat on the bones of the securities implications. I still made some tradeoffs, though - as my securities colleague noted when we discussed the draft a couple days ago, the securities part of the paper alone could be 50+ pages, and I agree. Maybe when the idea catches on, someone will do that paper :)
In any event, here it is. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2227103
Posted by: Michael Risch | Jun 7, 2013 10:12:24 AM