Patent trolls have cost innovators half a trillion dollars
arstechnica.com
arstechnica.com
Patent sector profits rise to half a trillion.
From what I would tell, IV operates a lot like an insurance company.
They hoard 10s of thousands of patents which they enforce through 1 of 1300 subsidiary companies and larger companies like Microsoft, Samsung, HTC, etc. can pay a monthly or yearly membership fee to IV in order to "lessen the chance of being sued by IV or one of the child companies".
Of course it is cheaper for all of them to pay their membership fees and IV enforces this behavior by violently attacking anyone that decides they don't want to pay up.
There is an implied threat[2] to not playing along.
(this isn't me postulating, the "When Patents Attack" piece covers this).
In turn, investors step in and invest billions into IV to get a cut of the growing patent market, like I might invest in a fund.
I understand financial and legal systems are complex, but what IV has created seems to me to be this dystopian-esque octopus of circular self-leveraging that is designed to be impossible to untangle.
[1] http://www.thisamericanlife.org/radio-archives/episode/441/w...
If you dig into the dataset you can see that patent trolling drastically increased after 2005, given the stock market trends over that period it's not surprising that companies would lose a lot of value after a litigation event in a market defined by lack of confidence.
Also, as the case progresses the market will price back in the EV of the litigation, how do they account for that? A lawsuit is big news, a technical maneuver in court that absolves a lot of liability may not because of the complexities involved.
There is quite another type, alive and well in our political and legal system, that have no problem stealing or ruining thousands for a few dollars gain for themselves. I well and truly wish these folks hadn't been born.
I feel as soon as humanity grows out of the stage where this kind of behavior is tolerated, we'll be well on our way to the utopia that most of us hope for. Once you nip this impulse in the bud, it won't matter if the systems are socialist, capitalist or whatever, they'll probably work just dandy. This is why we can't have nice things.
An insurance company that insures against its own actions is usually called other names.
Don't blame the Supreme Court. They were paying attention 39 years ago with Gottschalk v. Benson.
http://supreme.justia.com/us/409/63/case.html
They reluctantly outlawed patents on software for general-purpose digital computers. I say reluctantly, because they really thought Congress should be addressing the question.
"The technological problems tendered in the many briefs before us [Footnote 7] indicate to us that considered action by the Congress is needed."
Confusingly, they were overruled by a lower court (Federal Circuit: In re. Alappat, State Street Bank & Trust v. Signature Financial Group). More on this at http://ourdoings.com/ourdoings-startup/2011-07-28
One interesting note in all this is that even in 1972 they knew that it was impossible to expect the patent office to rigorously review software patents. Gottschalk v. Benson quotes the President's Commission on the Patent System as follows:
"The Patent Office now cannot examine applications for programs because of a lack of a classification technique and the requisite search files. Even if these were available, reliable searches would not be feasible or economic because of the tremendous volume of prior art being generated. Without this search, the patenting of programs would be tantamount to mere registration, and the presumption of validity would be all but nonexistent."
Instead of hope just contact your representative: http://www.usa.gov/Contact/Elected.shtml
We have enough people on HN to pull it off.
> The theory is simple: a company's stock price represents the stock market's best estimation of the company's value.
This is contrary to all of the very little bit that I understand of the stock market. Investors attempt to find companies which are undervalued, and invest in them; by definition, that would mean that the market is not accurately estimating the company's value. A company's "real value", however you care to measure that, might be some kind of attractor in the market graph, but that's about the best I'd give it. Not that I know anything about this stuff.
> If the company's stock drops by, say, two percent in the days after a lawsuit is filed, then the market thinks the lawsuit will cost the company two percent of its market capitalization.
I am more certain of four problems here. One, that companies can take a stock hit for a variety of reasons, all, some, or none of which may be related to a patent issue. Two, although I can't be arsed to find an example at the moment, I'd wager a fair amount that there have been no shortage of data points where a company's stock market value rises, or stays level, in correlation with a patent issue. Three, "the market" doesn't "think". The market is simply an abstract construct composed of a huge number of individuals following individual goals and algorithms. Anthropomorphisation of complex systems often leads to faulty conclusions -- like using those systems to evaluate the value of individual elements. And fourth, since we can't accurately tie the company's value to its stock price in the first place, then concluding that a two-percent loss of stock price represents a two-percent loss of the company's fortunes is erroneous.
> But with a large sample of companies, these random factors should mostly cancel each other out...
And I am very certain that this is an error. Why "should" these random factors cancel out? Have there been any studies finding that they do, or are there any lemmas in statistics that supports this? Aren't the accumulation of errors in statistics usually represented as minima and maxima values, i.e., "we think the value could be here, but it could be as low as there or as high as there..."?
I like the conclusions but I'm uncertain of the claims.
>This is contrary to all of the very little bit that I understand of the stock market. Investors attempt to find companies which are undervalued, and invest in them; by definition, that would mean that the market is not accurately estimating the company's value. A company's "real value", however you care to measure that, might be some kind of attractor in the market graph, but that's about the best I'd give it. Not that I know anything about this stuff.
It's a tautology that the value of something is defined by what people are willing to pay for it. When the markets are open, shares of a large company are changing hands hundreds to thousands of times per minute. It sounds like you are suggesting that a company has some kind of "real value" that's divorced from its current selling price -- if so, how would one determine what that is?
The efficient market hypothesis (which most pros believe is accurate, at least in its weakest form) tells us that it's not possible to pick stocks that are "undervalued." The EMH dictates that all information relevant to the value of a company is incorporated in the price, so it's not possible for any one investor to rationally discern that a particular stock is undervalued.
There's a lot of back-and-forth on the efficient market hypothesis, but (other than market-wide run-ups based on bubbles / inflation / loose credit) by and large it holds with small opportunities for arbitrage. More likely you'll see manipulation (cornering markets, playing "penny stocks").
Regarding stock performance following an event, this can be statistically controlled in a well-designed study. I'm not saying that the referenced paper describes one, or that it doesn't, just that this is feasible.
Control can come both from a well-selected sample size, and including other control variables in your statistical model. "The law of large numbers" as applied to random samples will ensure the normalizing tendencies. A cogent argument could be made that stock markets and industry segments aren't particularly amenable to random sampling, though as I said above, other statistical controls also exist.
Maybe the Government should change the law and require all patent disputes settled via mediation before anything can even go to court.
Particularly small competitors, who cannot afford to go into a legal battle with a large multinational. Android is kind of a special case in that Google has a warchest of patents and a lot of money. Most companies would be forced to simply roll over because of the expense of five years of legal proceedings.
Similarly, five hundred dollars in losses to the power company translates into fifty bucks at the recycling center for stolen wire. I see no moral difference.
There are a million definitions for 'patent troll,' and all of them stink, capturing some good activity along with the bad.
The problem is the number of junk patents out there (obvious, prior art, etc.).
Robert Kearns pitched his wiper system to Ford and Chevy, but they preferred to build it without paying him. "NPE = troll" means he was the villain. (He should have, what, kept his invention secret until he built a rival car company?)
There are several other definitions of 'troll.' Lawyers who own patents. Groups who buy patents from others. All of these catch some sheep with the goats. (I'm leaving these as an exercise for the reader.)
It's far more direct to attack patents rather than people. Attack the patents which are obvious or prior art.
As a bonus to this approach, it smacks less of a sort of blind bigotry against the poor, lawyers, or investors. (Even if you do hate one of those groups, attacking the problematic patents directly affords a way of being discrete about it.)