Gorillas and patents:
The following is a thought provoking discussion from Bill Meade on the GG discussion thread regarding patents and litigation. Seems especially pertinent to smaller gorilla wannabes like GMST and CREE.
Subject: Gorillas and IP (not TCP/IP) was Re: PS: IDC patents and Markman hearing
From: "Bill Meade" <bmeade@webcom.com>
Date: Sat, 8 Jan 2000 11:39:06 -0700
>>Markman Hearing: (please, anyone that can explain this better I would
>>appreciate it) The court appoints a technical expert to determine the
>>validity of a patent infringement. The expert solely decides if the
>>patent is applicable in this particular situation, and if the patent
>>infringement is real. >
>
>Pretty close on the Markman Hearing, but it is the judge who rules, not the
>expert. Markman hearings are one of the ways the litigants have to present
>their experts' opinions to the judge, who then makes a ruling on the patent
>as a matter of law. Only after this legal ruling by the judge would any
>jury be involved in subsequent questions, like damages.
>
I'm a non-lawyer but I think we will see patent infringements become a
larger part of technology (and consequently the technology investment)
landscape. So I think this Markman thread is a great place to add a few
factoids (lawyers please keep me honest) and at the end, two questions
for gorilla investing:
Background info on Markman hearings:
- They've only been used for the past 5 or 6 years. I think it was the
Supreme Court that added the Markman process to a patent infringement
trial.
- The purpose of Markman hearing is to avoid having nontechnical juries
rule on technical issues like which claims of a TDMA patent are
infringed.
- There were also several other reforms to the patent infringement trial
process that have been recently introduced. A separate appeals court and
process for patents, for example.
- A markman rules on which claims in a patent are infringed, and what the
technical words mean. The rulings I've seen are in the 15 page (double
spaced) range.
- Based on a very favorable Markman ruling, the "winner" the ruling can
win a motion for summary judgement, greatly shortening the time to a
decision on patent infringement.
- There is a lot of "extra-technical" maneuvering that goes on via
Markman rulings. Basically, judges want parties to settle (much easier
for them than learning a new technology in every case). So, it is not
unheard of for parties to alternate "winning" Markman hearings as they
counter/sue and a judge jerks them around.
*Implication* is that it is substantially more risky than a lay person
would think, to read one party's patent, look at the other party's
allegedly infringing product, and then bet investment dollars on a simple
reading of the facts. Consequently, I would think that this kind of
strategy is not a gorillagame investment strategy.
Other related background info:
- A "special master" is, like Markman rulings, a recent role that has
been added to the US court system. Because the judicial system is
backlogged (because apparently the Republican congress won't ratify any
Clinton judge nominations), the special master role has been introduced
to offload some work from federal judges.
- Most special masters are former federal judges.
- The work that is offloaded from the judge is largely mediating between
parties.
- You don't just try to break a patent in court. All lawyers, when
things get to trial will tell you that your chances are 50/50 no matter
how good your case. So, in addition to legal maneuvering the parties to
a patent infringement will be scouring the world for prior art so that
they can trigger a reexamination of the other party's patent.
- When a patent is re-examined by the USPTO, it has about an 80% chance
of being reduced substantially in breadth.
- The median licensing fee that patent attorneys think in terms of is 5%
of the wholesale price of whatever the product/service is. This can go
up to 7% or 8%, but not often.
- A patent infringement trial is 2 trials in one. The first trial
determines infringement. The second trial takes place only if
infringement is found, and determines the magnitude of damages.
- The largest damages that have been realized from patent infringement in
the US are in the 1 billion range. So even if a gorilla had $50 BN in
sales for a blatantly infringing product, it is not likely that
infringement would cost them $5 BN.
- My observation (non lawyer) is that with expensive enough legal talent,
just about every patent will be demolished in the process of litigation.
There are some "silver bullets" that make it through and rake in big
damages, but they are less than 10% of the patents that are litigated.
- High powered legal talent when pitted against other high powered legal
talent will burn $1,000,000 per month for each side. Given that a trial
will take 18 to 24 months, this implies that litigating for less than
$50,000,000 infringements is foolish. Being foolish has nothing to do
with the decision of course, but there are underlying economics.
Patenting Trends:
- Patenting is up by about 20% per year in the US.
- Patenting in the IT industry is going up faster than patenting in
general.
- Patents are becoming so expensive that high-tech companies are being
forced to change from a cost-center to a profit center business model.
IBM for example, went from $90 million in IP licensing in 1990 to $1 BN
in licensing in 1999 (not including IBM's new business model of granting
cross-license patent rights with large component purchase deals with
companies like Dell).
- As IBM is ranked #1 in patenting and in licensing, IBM is the "gold
standard" in IP today.
Gorillas and IP (not TCP/IP).
The Intellectual Property (IP) strategies of gorillas are shifted in the
direction of paying for (rather than developing) IP. Cisco (placed in
the 130+ range in US patenting) buys IP with companies. Qualcomm is
ranked around 110th (I don't know their IP strategy but it seems from
recent posts that they shed the atom-moving part of their business in
favor of bit-creating which implies they are a gorilla outlier by trying
to develop their own IP). Microsoft (placed near 50th in US patenting)
buys IP with companies ($9.8 billion last year) as well as paying when
sued. Intel is ranked near 20th and has a more traditional
develop-in-house and cross-license IP strategy.
IP strategy is traditionally a question of cost minimization. However,
as the GG list is detecting already, in the future it is possible that
litigation might influence who becomes a gorilla/king, monkey/serf. For
example, the nervous comment on Dell's patents in the business press and
Dell's response to questions "that litigating is under study". The issue
at stake here is whether Dell will remain a king, or via IP, it will
leverage itself into being a true Gorilla.
*Note* If companies crack the code on how to integrate IP into defacto
and/or proprietary standards, the duration of a patent has recently
become 20 years from date of filing so there is a lot of room for
supporting gorillaness at the back end. The problem with making a
gorilla with IP takes place at the "front end" of patenting. If a
gorilla wannabe does not have a patent yet, it can't use IP leverage in
helping the industry stampede.
So, patenting is *potentially* a new source of imposing something that
looks and feels like proprietary advantage. This raises two big
questions for me:
#1 Will gorilla wannabes crack the code on using IP to substitute for
proprietariness of standards?
- The time requirements of IP making a difference in a stampede to a
standard implies some legal firm can overcome the equation:
tech time = log(log(legal time))
so it would take some very creative legal strategies (declaratory
judgements on patents in-process?) to bring the legal system response up
to the bandwith required to affect infrastructure swap-outs. If some
firm could do this, it would try to patent it as a new business method,
thus becoming the only legal firm that could affect gorilla/king
outcomes. Having "the mother of all tech law firms" develop does not
*feel* possible. But it seems like an implication.
Theoretically, we should be on the watch for a tectonic shift from
proprietary standards (which all tech markets are avoiding) to IP-based
standards. If a futures or auction market for IP rights were to form,
there would be enough bandwidth in the system to affect technical
standards. This is crazy, but no crazier than the way the patent office
works today. And note, the government already auctions off other
monopolies.
- Will the legal system be more likely to affect gorilla/king status of
industries that have looked like they are "bowling alley forever" but
which suddenly tornado?
#2 Can IP patent pooling be used to convert defacto standards into
tacitly-proprietary standards?
- The original sin of US trusts which triggered "trust busting" was
patent pooling. Acquisitions were used to build huge patent portfolios
that locked competition out. *Mental note* Patents are government
granted monopolies. Not too surprising that monopolies could be combined
by entrepreneurs into oligopolies. If granted patents (yesterday's
technology) can be brought to bear on which company gets to be a gorilla,
IP management could become a proxy variable for future gorillas.
- This implies patent portfolio managers who are today totally saturated
in trying to figure out how to sell what is in their portfolios
(yesterday's technology), might be able to capture more wealth by
steering IP development (tomorrow's technology). Who is Qualcomm's IP
manager and does s/he "get" gorilla game?
These are my "big 2" questions on IP and Gorillas. There must be more
and better. Anyone else got questions about the relationship of IP to
gorilla investing?
Bill Meade |
