Friday, September 27, 2013

PTO empirical data: US and UK

Special Session: Policy Perspectives on Trademark Data

Alan Marco, USPTO

TM case files are already publicly available.  Data from assignments.  One of roles of gov’t is to provide public goods, and open data is really important there.  Patent application information is coming, as are similar tailored data sets (time series of applications by technology).

Assignments: involve choices.  Main involve assignment of assignor’s interest (generic, change of ownership, e.g., inventor to employer); merger; security interest agreement; license; name change; correction; other.  Patent form: electronic choices are different than on the paper form.  You can write in whatever you want. Tried to create categories out of the answers, but full text is in the data.

Trading a patent/TM might be an indicator of value—those that have changed their name a lot or made corrections aren’t necessarily more valuable, but merger/security interest/license is some indication of commercialization/use. TM: 4 million property-level transactions; patent: 4 million post-grant.

Banks demanded rights to blue oval logo used by Ford as security for $23 billion loan.  Can Ford put up its goodwill as collateral? Is the mark valuable separated from goodwill?  Security interest recordation happens quickly; merger etc. may not.  No incentive to tell TM office until it’s time to renew the mark, since only the owner can take action at the PTO. So different actions have different recordation lag.  Ford recordation happened on the day it was executed. Real growth in security interest filings: more common than change of ownership.  Are they recording more or is this actual growth? We think this reflects actual trend to assign security interests in marks, in large part because banks as assignees have significant incentive to record.

In patents, similar overall trend though security interests don’t outpace change of ownership.  8-12% marks are involved in some type of recorded transaction each year, not counting release of security interest.  That’s a lot!  Patents, it’s 4-8% depending on how you define it—similar trend, lesser magnitude.  Big banks are big players: Bank of America 15%, JPM Chase 10%, Wells Fargo 10%.  New specialization in valuing brands for purposes of security interest.

Assignment and renewal are both value correlates. But part of this correlation may be artificial, based on incentives to assign.  If you want to renew a TM, only the owner can do that, so assignment is required.  So someone who doesn’t bother to renew might not bother to assign. 

So what?  Deals not primary drivers of IP value, but this is a signal.  More liquid and robust market for IP; valuation is getting better. Collateral use shows that trading/licensing isn’t the only way to get value.  Secondary markets: take some lessons from mortgage-backed securities (MBS). Financial firms have a great idea: If underwriters are writing insurance contracts for the banks, maybe that can be packaged into a security!  A lot of people he talks to in IP are talking about that with excitement.  Trading with perfect liquidity.  That worked out not so well in MBS.  An IP bubble?

Nicola Searle, UKIPO

TM lookalikes: a case study of what happens with TM research.  “Fast moving consumer goods” (FMCG) markets: food, cleaning, personal care products.  Sold by a third party, looks similar to brand owner’s product.  Head & Shoulders example.  80s: lookalike bottles in one shape; H&S changes shape and then so do lookalikes. 

Massive tension between brand owners and supermarkets.  Brand owners may not like going after major or even only customers.  UK: supermarkets have advance notice of packaging changes; Belgium: only 6 months notice allowed, by statute.  Many times the packages aren’t actually infringing.  Other goods/services can be involved—books with similar covers, singers who look like other singers.

Idea is that entrance of lookalike affects supplier’s market share.  Claim: affects innovation—lower returns to marketing/packaging innovation if it’s copied easily.  Not much reasearch in this area because it’s difficult to do.  DG Enterprise 2011 report found lookalikes didn’t affect market share or introduction of new products.  Did find that Spain was different, where lookalikes were relatively new.  Lookalikes may increase size of market instead.  Another argument: wasteful repackaging/innovating done just to avoid lookalikes.

Competition: use of advance notice to one’s own advantage.  Category colors: if a particular soap has a particular color, should other soaps be allowed to use that color?  Monopoly concerns, esp. when design is functional.  Consumer concerns: industry purports to represent consumers.  Could increase choice; could increase confusion.  What are the actual consequences?  Finally, signalling: used in a loose matter when we talk about TM.  Need a better definition; the idea that lookalikes confuse brand signals.

Research: brand owner interviews; consumer survey; sales analysis.  Industry-based surveys are open to a lot of manipulation, reflected in existing literature.  Looked for evidence of consumer mistakes, perception of common origin, perception of quality based on similarity to national brand.

Classic case in UK: Puffin v. Penguin—Penguin biscuits were copied and Puffin was the knockoff; lost the case because too close.  It was challenging to get participation from supermarkets.

Surveys demonstrating confusion don’t show economic impact.  Brand owners can be happy with consumer surveys—they tend to use them themselves in product development.  Eye tracking tech—track how much longer the eye lingers—but that doesn’t tell you about economic impact or why consumers end up choosing what they choose.

Sales analysis: people wouldn’t share confidential sales data; not enough to draw proper conclusions.

Overall: equivocal. A small, but statistically significant lookalike effect leading consumers to believe that similar looking products have similar characteristics and similar origin. Roughly equal numbers of consumers feel disadvantaged by a mistaken purchase as advantaged.  (If you don’t feel oppressed, are you really oppressed?)  You can’t say they cancel each other out.  Substantial majority had deliberately purchased a lookalike and found the experience advantageous.

Limited evidence that lookalikes spur brand innovation.  Not found that lookalikes directly cause brand owners to make changes to packaging.  Sales showed an association between an association between a reduction in sales of the brand leader and an increase in sales of the lookalike in a limited number of product categories. 

We end up with big question marks on data, actual impact of lookalikes. 

Q: does copying allow copycats to raise their prices?

A: no evidence.  Supermarket own-brand copies and independent copycats exist.

Q: what impetus?

A: her private impression: stakeholder pressure.  Had suggestions for new legal right/extension of TM.

Q: conclusions might question the validity of trade dress law, since they’re so inconclusive.

Another commenter’s perspective: private right of action allows it to be addressed case by case, instead of squeezing it all into the same category.  (RT: This risks big false positive errors though!)

Another comment: benefit of private right of action is unclear—if the problem is supermarkets, then there are collective action problems.

A: Ireland has a private right of action, not used much. Either it’s just used as a threat, or it’s not used at all. 

Asda’s big line is “chosen by you.”  The whole idea is that they have lookalikes, at lower prices, and thus provide consumer value.  Surveyed consumers about opinions and also asked them to look at products.  People with more experience with the product were less likely to be confused.

Comment: mistake that leads to satisfaction could keep consumers switched/lead to repeat purchase of the copy—take business away.

Q for Marco: does increased securitization distort incentives to patent/register TM?

A: it is a way to monetize. For the most part, in this sector, they need debt financing to produce—mostly manufacturers. And it’s a portfolio, including other intangibles not just marks. 

No comments:

Post a Comment