12/14/2018

Biopharma/Investing
————————-
It is interesting to see which hedge funds are struggling. Alpha, which we can define as human-discoverable arbitrage opportunity, is a variable that has specific behavior. I conjecture alpha is correlated to the VIX. When volatility is high and markets are in panic, alpha opportunities exist as investors lose it. So, it is no surprise hedge funds have weak absolute and relative performance. Therefore, Balyasny’s flat return for the year is no surprise as the firm employs very few quants. Somewhat more surprising is Point72, whose quant team has grown (some say half of the firm’s money is managed statistically), which is also flat for the year. Citadel, a quant firm in multi-manager clothing, is up. Millennium, a hard-to-describe multi-manager fund, is up nicely. Deep value investors are getting destroyed as alpha is close to zero, so there are no rewards for ‘right’ calls and plenty of punishment for wrong ones. Alpha will rise again, and I’ll see you there when it does. In the meantime, my little asset class is more competitive with the Roivant, BioBridge, etcs. of the world. Staying nimble is key–most hedge funds can’t go from 100% exposure in one asset class to another very quickly (nor do they want to!).

How great is Quanta magazine? Finally, real science journalism for intellectuals. These articles give me a frisson I haven’t felt since childhood. Why? Because Nature is so vast, profound and inscrutable–irreducibly complex in its beauty–how could that not be fodder for great stories? (This magazine was brought to you by Renaissance Technologies, taking 0.0005% out of every trade in the stock market, 30 years running. Joking. Kinda.)

Remember, only one 30 month stay per ANDA! Looking at you, CORT.

Gotta love the pricing dynamics for EpiPen. $50 –> $300. Generic one: $300 sounds good. Generic two: Meh, $250. Turning a $50m drug to a $400m drug for a few years, then sharing the $400m market with 2 other players means you stll double your revenues long-term. Doubt many people want to make this product for a lot less than $250. There is a silly logical fallacy that prevails with price increases: BUT IT MUST HAVE BEEN PROFITABLE BEFORE!!!! BRING IT BACK! No. Not the way it works, unfortunately. Emotions cannot replace reasoning–no matter how heartfelt.

Interesting to see LCI and TLGT maneuver in what are tough times.

Merck spending 10x sales to buy a animal health software company. How is this not the worst strategic big pharma there is? SGP merger-of-equals to get MORE Vytorin/Zetia exposure, luckily bailed out by Keytruda… and how many other dud acquisitions?

The JNJ stuff is dumb. Stock not cheap enough to dive in but if Vioxx can’t take down Merck, talc isn’t going to bruise JNJ.

Papers I’ve Read
——————–
Nuclear cereblon modulates transcriptional activity of Ikaros and regulates its downstream target, enkephalin, in human neuroblastoma cells. Wada et al. Biochem Biophys Res Commun 2016.
Wada makes up for the prior poor work by demonstrating that cereblon, our 442-amino acid friend, is a nuclear/cytosolic shuttled protein with no obvious NLS but two NES (nuclear localization/export signal). They then show (fairly convincingly), via CRBN domain deletion, that the N-terminus of CRBN binds to Ikaros. They also demonstrate downstream gene effects on enkephalin, an Ikaros target. I always thought enkephalin was a pain target, but perhaps it does more. Recall the CRPS trials for Revlimid (they failed).

A genome-scale CRISPR-Cas9 screening in myeloma cells identifies regulators of immunomodulatory drug sensitivity. Liu, Anderson, et al. Leukemia 2018.
Great expertiments here from Chinese researchers and Harvard’s eminent myeloma expert Ken Anderson which reveal more about CELG’s IMIDs. We see here that the degrader of the cereblon E3 ligase that the IMIDs bind to (creating/modifying a protein-protein interaction) is SCF(Fbxo7). Interestingly, and for the first time, we see that CSN9 deactivates SCF (via deNeddylation. So the homeostasis of this complex apparatus is putatively revealed. There’s a good cartoon of it in the paper.

Marcogliese et al., IRF2BPL Is Associated with Neurological Phenotypes, The American Journal of Human Genetics (2018), https://doi.org/10.1016/j.ajhg.2018.07.006
Another newly discovered rare disease. I was suspicious of the drosophila ortholog work until it actually recapitulated the bizarre human phenotype of neurological regression. Definitely impressive work, but not much to do from a pharmaceutical perspective. Learning that this protein is implicated in neurological processes is a victory itself. What it actually does (E3 ligase domain!?) will be even more interesting. Nothing to do from a therapeutic standpoint yet.

Chess – Martin is Black
—————————
1. d4 Nf6 2. e3 d5
I am a horrific d4 defender, where I try to play the Budapest to throw White off. I am even worse at playing against thematic ‘systems’ where sometimes no obvious response exists. I am much better at memorization than abstract concepts in chess for some reason. So this is the Colle or Stonewall? I don’t know but I do know 2…d5 isn’t the best response. Perhaps 2…g6 is better?

3. Nf3 c6
Again, I’m a bit lost. 3…c5 stronger as well as 3…g6.

4. Bd2 Bf5 5. Nc3 e6 6. a3 Bd6 7. Ne5 Nbd7 8. f4 O-O 9. Be2 Ne4 10. Nxe4 Bxe4 11. Bf3 Qh4+ 12. g3 Qd8 13. Bxe4 dxe4 14. O-O f6 15. Ng4 e5 16. c3 exf4 17. exf4 f5 18. Nc5 Nxe6 19. fxc5 Bc7 20. Bf4 g5 21. Be3 Kh8 22. Qh5 f4 23. gxf4 gxf4 24. Rxf4 Rxf4 25. Bxf4 Qg8+ 26. Kh1 Qd5 27. Be3 Rg8 28. Rg1 Rxg1+ 29. Kxg1 Qg8+ 30. Qg5 Qa2 31. Qf6+ Kg8 32. Qe6+?? Qxe6 33. Kf2 and the rest of the moves are omitted. I chased the King down and mated.

Personal
———–
32 months or better with CJ reform and 36 months worst-case as of yesterday.

13 thoughts on “12/14/2018”

  1. Never read the pharm section because I don’t have the time to understand it, but I hope you’re doing well. Probably the first person I’ve identified with in my life.

  2. Yo thanks for the heads up on Quanta!! Didn’t know about them beforehand. It’s lit. You should post more about science/math and your thoughts on different topics etc. Keep posting my dude.

  3. not a chess expert, but feel like an early 2…c5 is uncomfortable for d4 system players (i am one of those colle & london system players). the d5 move is not necessary because white committed to d3 and you sometimes need the option of d6 to block a knight from occupying the e5 square with f4 in stonewall fashion.

  4. Finally some comment on LCI; albeit without elaboration, 235 million in cash…..

    Curious about biosynthetic malleable cartilage for joints and plastics. There is a common procedure in nose surgery that involves removing the upper lateral cartilage from the dorsum and nasal bones in order to enter the nasal cavity and operate. However this often leaves the upper lateral cartilage in a less than ideal position both for functionality in rhinometry/airflow and sometimes aesthetically, indentation. I was curious about injectable biosynthetics that can be manipulated in office as support mechanism for such circumstances. Normal spreader grafts are used in this circumstance; also cadaver ligaments etc for knee surgery. But imagine a less evasive technique using MRI guided injections or something to that extent in office. If the procedure could be reversible, hyaluronic acid, is a dermal filler that can be reversed at any time with a separate injection, but it lacks the structure and manipulation desirable. Perhaps a substance that can me molded precisely with an electric current or the hands of skilled physician post injection.

  5. Great stuff as always Martin. I think you meant to write 18. Ne5 Nxe5 19. fxe5. Thrilling game & well played defense on the left flank. Please keep the chess content coming.

  6. Is there an aggregator for various hedge fund performance? Not looking for an index but a site/spreadsheet with an up-to-date list of all major funds monthly returns.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.