Acadia. I still don’t think this drug will ever sell too much. There are a few D2 sparing 5-HT2a antagonists (nothing as pure as pimavanserin, given), but the company’s spending is a bit insane. The drug is probably great for PDP given the levodopa interference at the dopamine receptor, but what good is the drug for the other illnesses when you can use Risperdal, Abilify, Seroquel, etc. Re spending, I understand that one needs to invest in a brand, and it will stop at some point, but this is quite dramatic cash use. I am skeptical of the intellectual property of pima, so I expect a generic sooner than others.

Continuing our Pfizer (Merck next) R&D productivity analysis, what is fascinating is almost all of Pfizer’s drugs were discovered by other companies. I suppose it is not too surprising given the mega-mergers that have occured. Still, the idea that Pfizer has only created 11 drugs in the last 30 years despite spending >$100 billion on R&D fascinates me.

Warner Lambert/Parke-Davis: Lyrica, Ibrance, Lipitor, Neurontin, Accupril
Wyeth: Prevnar, Enbrel (Immunex/Amgen), Bosulif, BeneFIX, Refacto/Xyntha, Premarin, Pristiq, Effexor XR, Rapamune, BMP2, Zosyn, Tygacil
Pharmacia/Sugen/Upjohn/Monsanto/Searle: Sutent, Celebrex, Genotropin, Somavert, Xalatan, Detrol, Xanax, Xalkori, Bextra, Camptosar, Ellence, Detrol, Cleocin, Inspra, Zyvox, Depo-Provera, Aromasin, Medrol, Fragmin
Bristol-Myers: Eliquis
Eisai: Aricept
Medivation: Xtandi
Pliva: Zithromax
Agouron: Inlyta, Viracept
Merck Kgaa: Bavencio
UCB: Toviaz, Zyrtec
Anacor: Eucrisa

Unasyn – FDA approved in 1986
Cardura, Diflucan – FDA approved in 1990
Zoloft – FDA approved in 1991
Norvasc – FDA approved in 1992
Viagra – FDA approved in 1998
Vfend, Relpax – FDA approved in 2002
Revatio – FDA approved in 2005
Chantix – FDA approved in 2006
Xeljanz – FDA approved 2012

I assumed net margins on pharmaceutical products are 50% (well above what they really are but around what they are on a steady-state well-managed level). I assumed 50% of R&D is spent on discovering or running clinical trials for in-house projects (as opposed to clinical trials for acquired products). At a 3% discount rate, PFE made 32B for shareholders from 1991 to 2017 (16 years) investing in R&D. All of the profit was from 1992 to 2006, indicating very roductive R&D in the 1990s, creating monsters like Norvasc, Viagra and Zoloft. Net income of this “inHouseR&DOnlyCo” reached as much as $2 billion in some years. Net income turned negative afterwards, ranging from breakeven to negative 1 billion per annum since 2006.

If, instead of doing R&D, Pfizer invested the R&D cost (50% of reported cost) in a fund returning 5% after-tax, it would have generated 124B in value as opposed to the 32B NPV. I have never seen an analysis like this done (it has taken me the better part of 3 days).

I’ve been having a recurring dream about being able to go back in time and see the future and live the past simultaneously. I’m working at a hedge fund and I somehow know the future years in advance. I know that Facebook will get started in 2004. I know that Google is a great buy at or before the IPO. I know what drugs will work and what won’t (although I know that today). I knew the market will crash in 2000-2002 and recover in 2003. I guess this is what insider traders feels like?

Papers I’ve Read Recently
Idiopathic Pulmonary Fibrosis. Lederer & Martinez. NEJM 2018:378;19.
This is the kind of review article I like to see in the NEJM, especially as it covers emerging therapeutic options. Bravo!

cAMP signaling in brain is increased in unmediated depressed patients and increased by treatment with a selective serotonin reuptake inhibitor. Fujita et al. Molecular Psychiatry 2016.
Fascinating paper with important implications.

A paper in Molecular Pharmacology I cannot discuss.

Efficacy of Exome-Targeted Capture Sequencing to Detect Mutations in Known Cerebeller Ataxia Genes. Coutelier et al. JAMA Neurol 2017.
Nothing too exciting here, very standard in today’s diagnostic odyssey.

Review of Neurological Implications of von Hippel-Lindau Disease. Dornbos et al. JAMA Neurol 2018.
Just catching up on a rare disease.

adnexal. the area consisting of the ovaries, fallopian tubes, their support structures, etc.
endolymphatic sac tumor (ELST). common tumor of VHL disease, fairly benign.
hemangioblastoma. common tumor of VHL disease, fairly benign.
saltatory. growing in an unpredictable pattern.


Biopharma & Investing
In demonstrating PFE and MRK to an inmate here, I was reminded of the poor R&D productivity of big pharma. I’ll do a little more of an in-depth analysis but if you take the aggregate R&D spend of these two “industry leaders” and evaluate what projects ended up being successful, what are we left with? Only compounds definitely from within count. Buying Schering who bought Organon and lucking out on Keytruda doesn’t count. I’ll try and trace back the origins of each compound.

Organic Merck: Januvia, Cervarix, Isentressp
Organic Pfizer: Ibrance?, Lyrica, Xeljanz, Inlyta, Chantix, Celebrex, Neurontin, Lipitor
Inorganic Merck: Keytruda (Organon/Schering), Remicade (Centocor), Simponi (Centocor), Vytorin (Schering), Zetia (Schering)
Inorganic Pfizer: Prevnar (Wyeth), Enbrel (Wyeth), Sutent (Sugen/Pharmacia)

This analysis will continue. Anything I missed?

–Very glad to see the WSJ take Greenlight down a notch. There’s a joke in there about me. Well, I guess the jokes on Einhorn. He passed on my company and I project that is up about 4x. Meanwhile his Greenlight fund is… uh… not doing so well, apparently. I’m far better off than he was at my age, and at the rate he is losing money and I’m making it, I’ll be wealthier than him soon even without the 15 years he has on me. Keep shorting NFLX and AMZN bro, you gotta be right someday.

Book Review
The Dhandho Investor – Monish Pabrai

This is one of the worst books I’ve ever read. Most investing books suck, and Pabrai breaks new ground in the arcane field of suction engineering. Pabrai is what I would call a weak-form Buffett clone, and appropriately comes off as embarassingly clueless. The only property of this work worse than its content, which lacks one original thought, is Pabrai’s putrid writing style. I lost count at how many times he printed his meaningless platitude “heads I win, tails I don’t lose much”. Pabrai believes he’s doing the world a favor by writing, but I suggest he finds a new hobby. As if the secrets of wealth are contained in 180 pages of regurgitated “value investing basics”, Pabrai’s arrogance is intolerable. He doesn’t mention that, like any investment strategy, value investing can become wildly overcrowded and ineffective. Why are all the value hedge funds doing so poorly? I guess they haven’t picked up this classic.

Contradicting himself repeatedly, Pabrai screws up basic calculations, doesn’t explain (frequently being wrong) key inputs to important formulas. He stresses simplicity, but misses the point that investing, in fact, is anything but simple. Take Buffett. I consider myself a Buffettologist and have learned a lot about this legendary investor. From what I can tell, Mr. Buffett and Berkshire Hathaway have been a leveraged, long-only beta bet on American equities. It has worked well. The false conclusion drawn by Pabrai and his sorry acolytes is that we should be like Buffett. America has changed. We are a very leveraged country and much larger, among other differences. I have no idea if the great bull run of the 60s-present day will continue. Neither does Pabrai. The US may be the next Japan or Russia (two leading countries in the 70s), or worse. The question of ‘why is it so easy?’ apparently never enters Pabrai’s mind. The answer is because it isn’t. Nowhere does Pabrai suggest holding cash in a portfolio or entering into more complex transactions.

Even in relative bright spots, like a fairly pathetic tour of his successful investments, Pabrai’s level of ignorance is astonishing. His funeral home example excludes the company’s gigantic debt load from his calculations, happily allowing him to pronounce his purchase of the company’s stock at 3x cash flow. Yes, I too can ask Bloomberg to print me out a list of enormously overvalued companies ‘trading at low P/Es’. I was perhaps this ignorant when I was 17 years old, in my first year on Wall Street. Some examples, like his Level Three convertible bond “analysis” are superficially fun, but show the Buffett sycophant syndrome. He buys these bonds because Buffett buys the bonds. Okie, dokie. Pabrai’s application of the Kelly criterion is also laughable, often humorously suggesting to put 90% of one’s portfolio in an equity.

I hate to criticize other investors or beat my chest and imply “I’m richer than you” or “I’m smarter than you”. But I must implore you to ignore this book, or read it with an eye to correcting it. I’m not sure who sent it to me, but you hurt me. Why are you hurting me? One should read about great investors like Buffett, but try to abstract the governing dynamics of their actions. “Buffett started an insurance company. So should I!” “Buffett bought a lot of American Express during their crisis, I should do something like that!”. These are all unoriginal and overly simplistic thoughts that will teach you nothing and lead you astray. Investing actions are a manifestation of your investing theory. If you discover and understand, and then extend theory, you can create actions that aren’t simple acts of mimicry.

There are exceedingly important questions the advanced investor has to ask: are there times when I can discount future cash flow at very low (close to zero) rates? Is investing a “zero-sum game”? What is the relationship between arbitrage and crime/ethics? What can we infer from liquidity? Pabrai reminds me of the old SNL skit “Deep Thoughts”, the modern day equivalent of the philosoraptor. Except, even those constructs grope for a further truth and a refinement of technique. Not Pabrai, he’s got Dhandho.

Papers I read today and yesterday

The protocadherin 17 gene affects cognition, personality, amygdala structure and function, synapse development and risk of major mood disorders. Chang et al. Molecular Psychiatry 2018,23:400-412.
Interesting paper on a new target. First thing I ask myself when I read something like this is, how many AA is PCDH17 and is there a crystal structure?

The Lymph Node and the Metastasis. Tjan-Heijnen & Viale. NEJM 378;21.
Another “Clinical Implications of Basic Research”. Oh boy. Here we debate the irrelevant question of how tumors seed metastases. I don’t think it matters.

Correction of a splicing defect in a mouse model of congenital muscular dystrophy type 1A using a homology-directed-repair-independent mechanism. Kemaladewi et al. Nature Medicine 2017.
I reread this paper. It actually is a bit of a breakthrough, delivering CRISPR through AAV9 and selecting just the right PAM to excise/correct/create just the right donor splice-spite mutation in post-mitotic tissue. The animal data is impressive and augurs well for humans. I’m a little wary of the immunogenicity of the humorous delivery of gene editing through a viral vector. I also have questions about what drives expression from the episomes, but you can’t argue with the data!

Transglutaminase 2 overexpression induces depressive-like behavior and impaired TrkB signaling in mice. Pandya et al. Molecular Psychiatry 2016
Not terribly impressed here. Maybe I’ll re-read it.

synapse. The gap between neurons where chemical or electrical messages are sent. Understand the difference between presynapse and postsynapse.
amygdala. Bilateral brain structure in the temporal lobe responsible for a variety of functions including memory and emotion (not just rage as we’re taught in school). Close in proximity to the hippocampus.
hippocampus. Bilateral brain structure in the temporal lobe responsible for memory consolidation among ther functions.
temporal lobe. Major area of the brain responsible for interpreting sensory input and many other things.
dendrite. branched extensions of neurons that receive impulses/signals.
dendritic spine density. the arborization of dendrites is thought to be important for various illnesses. the more dense the better.
GWAS. genome-wide association studies. The study of entire genome (or exome) and its correlation to some phenotype. Beware multiplicity statistical errors, which are frequent in GWAS.
NHEJ. Non-homologous end joining. Contrasted with HDR, NHEJ allows for double strand break correction without a template. More error prone than HDR.
Protospacer-adjacent motif. A target region for Cas9 nuclease activity.
Single guide RNA. Guides CRISPR to nuclease site.
spliceosome. Cellular apparatus that removes introns from RNA.
donor splice site. 5′ site of intron to be spliced out.
post-mitotic cell. Cell that will no longer undergo mitosis. E.g. neurons.
hemagglutinin-tag. Similar to other tags like His6, FLAG, SUMO etc., tag proteins with short amino acid sequence to assist purification and identification. Used FLAG for my first proteins until I realized it really is a research tool. I believe there is one FDA approved drug with the tag intact, however (LOL!).
syngeneic model. An animal model using a tumor allograft from the same species/background to ensure immune competence.