Wednesday, November 7, 2012

Vringo Versus Google: Did the Jury Make a Huge Error?

Yesterday, the jury in the patent infringement case involving Vringo (VRNG) and Google (GOOG), returned a verdict. They found that Google's search products have indeed infringed on both the 420 patent and the 664 patent. The jury recommended a "running royalty" of 3.5% going forward. This was a great thing for investors, however, the jury's ruling on past damages makes very little sense to me and I'm fairly certain they may have made an EXTREMELY crucial mathematical error. Here's how the jury breaks out the past damages due to Vringo (from the latest press release on Vringo's web site):

  • Google: $15,800,000
  • AOL: $7,943,000
  • IAC: $6,650,000
  • Gannett: $4,322
  • Target: $98,833

When this damage assessment first came out, for me, red flags immediately went up. Why on "God's Green Earth" would AOL be required to pay $8 million while search behemoth Google is penalized only $15.8 million. I thought, the award to Google should be magnitudes higher given Google's dominance in the search market?!

So, I began trolling the message boards for answers. On Steve Kim's web site, I noticed "MikeHD25" (a long-time and reliable in-courtroom resource) posted the following:

First of all I am an american and have complete faith in our legal system but I believe the jury had a misunderstanding about the past damages. Read over their questions the last two days and compare the following math. see what you think

Before the jury starting deliberating the were informed that all damages are to be calculated from 9/15/11 for google as well as co defendants. The only piece of evidence they had to use for codefendents in consideration of damages is a slide that vrng provided with each co-defendant listed seperatly with the following damages. Keep in mind these damages represent 3.5% of 20% of total revenue from smart ads

  • google-$451,190,903
  • aol-$22,693,517
  • IAC-$18,917,570
  • target-$282,380
  • Gannet-$12348

Now keep in mind the judge after reading the verdict page by page very careful asked for it back after it was read aloud by the court clerk to double check something that he obviously didn't see the first time.

now look at the following calculations.

  • aol $22,693,517 x 35%= $7,942,730 damage amount $7.9mil
  • Iac $18,917,570 x 35%= $6,621,149 damage amount $6.6mil
  • target $282,380 x 35%= $98833 damage amount $98833
  • ganet $12348 x 35%=$4321 damage amount $3432

GOOG $451,190,903 x 3.5%= $15,791,681 damage amount $15.8 million.

The jury applied the running royalty rate to the past damage amount that VRNG was seeking. THEY DID BAD MATH and misunderstood the answers the court gave to questons. I don't blame them. They wanted to go home and felt they were doing what the judge instructed.

One question the jury asked was "if a running royalty rate is to be used are we to apply to each defendant" They were asking if they needed to apply the 3.5% to damages for each defendent.

The question today was "If we assign a lump sum amount to be awarded is this in addition to the running royalty amount" They were asking if they award a lump some do they need to apply the running royalty % to this lump sum

The judge double checked the verdict in my opinion because he thought it said 158,000,000 for google and when read by clerk he realized it was incorrect. If he noticed this right away it would explain why after the 5 min recess he recessed for the day. This is all speculation but makes way to much sense. If the judge is in charge of the exact damages and has the power to change them I would look for this to happen if he figures this out. Just my opinion. This post should get some donations. I don't think anyone has come up with this scenario.

Now, I suppose MikeHD25 could still be wrong. However, as soon as I read Mike's post, I jumped out of my chair as all the pieces of this amazing Vringo story came together. The jury has likely made a mistake, and it is a big one.

Not only does this post illustrate a HUGE mathematical error, but the jury also went above and beyond the minimum 20% award for past damages which Vringo was seeking. From reading case transcripts and exhibits through, I know that Vringo's star witness Dr. Becker had been pushing for a range of 20% to 40% in regard to how the Lang patents boosted Google's revenue. Well, using the math above, it looks like the jury awarded 35%. This is obviously at the top end of the range and I believe it shows the intent of the jury to penalize Google almost as much as possible.

Also today, after the verdict was read aloud, I wondered why Judge Jackson asked for a 5 minute recess and then never reconvened court. That event did not make any sense either. Given the report above, this event now makes perfect sense to me. Judge Jackson, very likely, began taking steps to fix this error.

So, for me there are two BURNING questions in the plot-twisting saga of "Vringo versus Google", both of which will likely be answered tomorrow:

  1. Is this truly a mathematical error by the jury?
  2. If so, how will the judge rectify this error?
Good luck everyone!

Disclosure: I am long VRNG. I wrote this article myself, and it expresses my own opinions. I have no business relationship with any company whose stock is mentioned in this article.

Friday, November 2, 2012

Vringo versus Google - Estimate of Jury Award

This article outlines my thoughts and speculation on the amount of past damages and future royalties which Vringo may be awarded in its recent case against Google. Investors and traders are sitting on pins and needles waiting for the jury to deliver its verdict on the patent infringement case. The verdict will likely come early next week, if not on Monday November 5th. As of the end of the business day on Friday November 2nd, the jury has been deliberating for one day.


First, from S&P, below are Google's quarterly revenue numbers:

In the spreadsheet below, the first step in computing damages is to remove Motorola's revenue from Google's Q2 and Q3 revenue in 2012 (since Google bought Motorola in August 2012).

Next, we calculate that 96% of Google's revenue come from AdWords as mentioned in this article (and numerous others that I have read).

Next, in this post from Google, it states 54% of revenue in the second quarter of 2012 was from outside the United States.

Once Google began infringing on the Lang patents, it was estimated that Google's revenue immediately increased 20%, so I have added that to the spreadsheet as well.

We're almost done... the royalty rate for which Vringo is asking is 3.5%

Finally, Judge Jackson has stated that Vringo can collect damages from September 2011 (when the suit was filed) to present day.

As you can see, when you plug in all the variables, the one time award should come to approximately $145 million for past infringement only.


Future royalties are a much easier nut to crack now that we have the above estimates. Let's assume Judge Jackson keeps the royalty rate stays at 3.5%. Also, in the interest of being conservative, also assume Google's going forward revenue stays completely flat. This way, we can take an average of the payout from the four full quarters (above).

Wow! If awarded by the judge, future royalties at a 3.5% rate, through the life of the patents, should come to a very tidy sum of $575 million. Please let me know if you see any items I have left out of this estimate.

Good luck and happy trading!

Tuesday, October 30, 2012

VRNG - "edvacourt" Full Post

Here's ED's full post regarding his update today (read from bottom to top). This doesn't include quite a bit of commentary he makes later in the thread:

Posts by "edvacourt"

Re: Re: Re: Re: Update, October 30 By edvacourt . 9 minutes ago . Permalink Go to topic

You got it ghart. I actually posted a first update around 2:30PM only to realize when I got back to my hotel that it had never actually posted. Haven't had this problem before today. I will say that Steve Kim is dead on with his article about the latest hail mary "dismiss everything' motion by Google. Judge Jackson has shown throughout that he wants the jury to decide this one and it looks like they may have it tomorrow at the EOD or sometime Thursday morning. 0 users liked this posts 0 users disliked this posts Reply Ignore User

Report Abuse Re: Re: Re: Re: Update, October 30 By edvacourt . 22 minutes ago . Permalink Go to topic

Last part: Some housekeeping notes: Google has no more witnesses. They have rested their case. I think since Judge Jackson would not let them talk about anything that would be considered expert testimony (which I think Google was perhaps trying to backdoor) Google just decided to pass on putting them up there. Dr. Carbonell will appear tomorrow as a rebuttal witness. The judge said that he anticipated getting to closing arguments and jury instructions tomorrow afternoon. He said it may not be possible and that he would not split up closing arguments (on different days). Each side will have an hour. The endgame is close people and I believe the question has moved away from whether Google infringed to how much is that infringement worth. 8 users liked this posts 0 users disliked this posts Reply Ignore User

Report Abuse Re: Re: Re: Update, October 30 By edvacourt . 23 minutes ago . Permalink Go to topic

Part 3:

7. One particularly damning area seemed to be when Sherwood puts up document # DX256, pg 15 the direct exam of Dr. Ungar. A slide comes up showing the top 50 webpages visited in August 2003 by people at home, work and at Universities. At #5 on the list is Google with 55M unique visitors and at #6 is Terra Lycos with 52M. Way down the list is Overture with 16M. This is important in that Ugone argues a comparison to the Google settlement with Overture, which was estimated at $291-365M based on 2.71M shares, is somehow irrelevant. The competitive equivalency between Google and Lycos (its parent) at the time of the hypothetical negotiation is quite relevant. This testimony a shaky Dr. Ugone though not quite as bad as Ungar. Sherwood did a good job linking Ugone to Ungar and further showing that both seem to be spoon fed much of their stories by Google. What amazes me is the lack of true due diligence related to this case that Google's expert witnesses appeared to have done or even admitted to doing. This serves only to further support the theory that they are following Google's talking points. Google's counsel Robert Wilson did a decent job trying to rehabilitate Ugone but he repeatedly led the witness in his questioning and was reprimanded several times by Judge Jackson. 2 users liked this posts 0 users disliked this posts Reply Ignore User

Report Abuse Re: Re: Update, October 30 By edvacourt . 31 minutes ago . Permalink Go to topic

Several big points by VRNG counsel Sherwood...

1. Sherwood establishes that Dr. Ugone has never negotiated a patent license 2. When Ugone has testified as an expert witness for a Plaintiff he has always recommended a running royalty rate but fails to do so in this case. 3. Ugone testifies that he is not technically astute and took most of his conclusions on the patent technology directly from Dr. Ungar. 4. Ugone agrees that a running royalty=rate X base but instead has arrived at his argument using representative transactions where he claims the parties somehow understood the value of the Lang patents. Actual testimony discussed below shows this to ignore reality. Sherwood points out that the GA Pacific method has many more areas that must be vetted before making a damages assumption including the value of the technology to Google. 5. Sherwood establishes that Ugone has done no analysis of the use of the patents in suit by the defendants. This morning Ugone said the damages by Ungar were to high but Sherwood is able to essentially have Ugone agree that he offers no royalty rate calculation or calculation of base. 6. Ugone made a statement Friday that Google had to supplement the technology of the 664 and 420 patents in order to make them useful though he is shown to have no evidence to support that. He has also testified that he is not an expert on the technology and was getting that information from others, chiefly Ungar. 9 users liked this posts 0 users disliked this posts Reply Ignore User

Report Abuse Update, October 30 By edvacourt . 56 minutes ago . Permalink

One last try here...if I didn't know better I'd say yahoo was in on some anti-NPE pact with Google!

Part 1: Dr. Ugone testified today that Dr. Becker's report did three things which should disqualify it. It overstated the claimed royalty rate. Second, it overstated the royalty base and third, the damages failed the reasonableness test. To substantiate his points he cites Lycos being sold for $95M in 2004, the 2008 purchase by Google of the Carl Meyer Patents for $3.55M and the fact that Lang bought his own patents back in 2011 for $3.2M. In short Ugone bases his argument on a series of inferences that look quite shaky under cross exam. Vringo counsel (Sherwood) does a good job of taking the jury through Ugone's history as an expert witness, over 200 cases in all and numerous times testifying at the behest of Google. In this case alone Sherwood estimates Ugone's fees to be roughly $310,000 which Ugone does not dispute. Sherwood also states that of Ugone's 5-6 published works, three of them are about how to provide expert testimony. 95% of Ugone's work we hear is expert testimony.

Thursday, February 23, 2012

ImmunoCellular Therapeutics

The current biotech industry is riveting. Lately, I'm reading about all kinds of cool breakthrough cancer technologies taking the industry by storm (do yourself a favor and Google $PCYC and $GHDX. Full disclosure: I own both of these). While I am by no means an expert in this area, I continue to study the industry and I find the science amazing. Recently, I read a Seeking Alpha article on a company called ImmunoCellular Therapeutics (IMUC.OB). They specialize in immunotherapy, which is basically stimulating or training a patient's own immune system to fight whatever horrible disease he or she is facing. The company's lead drug candidate is ICT-107. Recently, Phase I results were simply amazing in it's ability to extend the life of patients facing an extremely deadly brain cancer called Glioblastoma. According to Wikipedia, with the current standard of care (SOC), "the median survival time from the time of diagnosis without any treatment is 3 months, but with treatment survival of 1–2 years is common".

What is so interesting about ICT-107 is that, in Phase I trials, 100% of patients treated survived past the first year and 40% were disease free after three years. Without the drug, 6% of patients lived past 3 years.

This story, about a man suffering with Glioblastoma, brings these Phase I results to a personal and tangible level. The article quotes Dr. Joseph C. Landolfi, Director of Neuroncology at JFK Brain Tumor Center:

“The brain tumor trial is testing a vaccine that is developed from the patient’s own blood,” Landolfi said. “We attach proteins to the patient’s immune cells and then inject it back into the patient, which then directs the body to attack the proteins on the tumor. It essentially mounts his body’s immune system to attack the tumor.”

A disclaimer: by no means am I a regular buyer of over-the-counter companies. And I despise hypers and promoters. However, I will tell you that today I bought 1000 shares of IMUC, based mainly on Phase I results for ICT-107. This stock is very risky, so this is money I am willing to throw away for the potential gains it could bring. As of right now, the company has enough cash to fund operations through 2013. If Phase II results are as promising as the Phase I, obviously this will change. An excerpt from illustrates what is happening now with Phase II trials:

ImmunoCellular provides update on Phase II ICT-107 Clinical Trial

ImmunoCellular Therapeutics has provided an updaet on clinical trial site activity and patient enrollment for its Phase II clinical trial of ICT-107, the Company's dendritic cell based cancer vaccine candidate targeting multiple tumor antigens for the treatment of glioblastoma multiforme. The Company has initiated the trial in 23 centers and has received Institutional Review Board approval from a total of 24 trial sites. The trial is expected to enroll approximately 160-200 patients to treat 102 patients with HLA-A1/A2 immunological subtypes. There are 115 patients enrolled in the study to date, ahead of the Company’s schedule. Enrollment for the trial is expected to be completed by the second quarter of 2012 and an interim analysis is expected when 50% of events have been observed. The Phase II trial of ICT-107 is a double-blind, placebo-controlled, 2:1 randomized study designed to evaluate the safety and efficacy of ICT-107 in patients with newly diagnosed GBM.

Technically, volume on the stock has exploded since early January. Also note that the stock seems to be holding above its 200 day moving average. Again, with stocks of this nature I am always very suspicious of manipulation and hype. So, it's definitely buyer beware. However, I feel that the trial results show that ImmunoCellular Therapeutics offers an exceptional risk to reward ratio *if* properly balanced in your overall portfolio.

UPDATE: Check out this interview with Dr. John S. Yu, the Chief Science Officer at ImmunoCellular.

Notable Insider Buys include the following:

Socius Capital Group - 2/13/2012
Terren S. Peizer

Ayer Capital Management - 1/20/2012
Jay Venkatesan

Information on the Phase II Clinical Trial for ICT-107 is available from here.

Tuesday, February 7, 2012

Proof That Valuation Metrics Matter

A Valuation Scoring System for S&P 5 Star Stocks

How much do valuation metrics matter in stock picking? It is a very important question considering how far the S&P 500 has come since the October 2011 lows. When scouting trades, my bread and butter has always been identifying high probability chart patterns and reviewing fundamentals to make sure a value component exists. However, at these lofty S&P levels, juicy chart patterns seem to be few and far between. Once a market begins trending in a stable low range environment (usually after a period of high volatility), a trader can easily get left behind simply because traditional setup patterns cease to exist.

So, to address the "lack of traditional chart patterns" in the market, I created a simple system that pushes fundamental analysis to the forefront of my stock selection criteria. I'm writing this blog entry because, along the way, I discovered something pretty cool which hopefully you can use. Let's get to it!

Via my Ameritrade stock screener, I created a list of 75 stocks rated "5 Stars" by Standard and Poors. This is S&P's highest rating and will serve as "safety filter" on my universe of stocks. Next, I loaded the stocks into Finviz here. My objective was to find the cheapest 5 star S&P stocks using various valuation metrics. I selected my metrics within Finviz, ran the screen, and downloaded the following spreadsheet (screenshot):

Using a simple ranking and scoring system, I gave each stock a rank/score for each valuation metric. For example, $HFC had the lowest Price/Earnings Growth (PEG) metric so I gave that stock a "1". The stock with the highest PEG was $CBL, so I gave it a "75". In other words, I computed a score for each stock across these 7 different valuation metrics:

  • Price/Earnings
  • Forward P/E
  • PEG (price to earnings growth)
  • Price/Sales
  • Price/Book
  • Price/Cash
  • Price/Free Cash Flow
Then, I added up the total scores for each stock to get a "Composite Valuation Score" (see screenshot):

Finally, I sorted the stocks by my Composite Valuation Score. The lower the score, the cheaper the stock. I decided to add Year to Date performance to my spreadsheet and do a correlation between composite valuation and performance. And guess what I found? Yep, valuation does matter! The stocks that had a lower composite valuation score have *significantly* higher year to date performance. In other words, it pays to search for value.

Here are the top 20 cheapest 5 star S&P stocks:

Please feel free to ping me on StockTwits at PTSD_Trader if you would like a copy of the spreadsheet. I can't seem to figure out how to upload it via :)