I've been asked by some of my clients to provide my personal point of view on an recently published analysis research from Societe Generale. They transferred the analyst coverage of AB Science to an analyst in London call Jason Smith per the publication of the bank.
Societe General has a buy rating on the stock but they reduced the PT from 24 to 21,8euros despite the data of Pancreatic cancer. That probably is at the origin of the drop of roughly 10% of the stock in the last two days. I thought Christmas gift was on December 25th, thanks to Societe Generale and Jason Smith it comes 2 weeks earlier for Long term Investor. Thank you.
More seriously, I was surprise by the low quality of the research. That is of course my personal point of view. I will try to explain a little bit below, what all that is insanely stupid and low quality in term of analyst coverage.
I will try to summarize some of the weakness according to me.
=> SG recognizes the well received data in pancreatic cancer and point to the two subgroup success . They raised the "significant binary event risk surrounding the regulatory approval of the drug in this disease" : 1) Failed overall population 2) Subgroup not specify formally in the protocol 3) "Not aware of cancer drug which failed phase 3 primary endpoint but were approved on the basis of post hoc subgroup data."
First of all : No mention of the incredible Hazard Ratio and the P-value that solidify the value of the two subgroup. A bit surprise by that... to say the least.
Also, they kept mentioning "approval" "approval" ... They should specify "conditional approval" and not full approval which changes totally the situation.
AB Science is most likely pursuing conditional approval on both or each subgroup with - as a condition- to do the confirmatory study on those subgroup as overall population. Based on both the Hazard Ratio and more important the p-value, the success of the study is an almost sure bet and the gravity and lack of treatment of that disease increase the odds of potential conditional approval.
Also, the analyst raised that they aren't aware of any drug being approved (even if we are talking about CONDITIONAL approval) with overall population missed and post-hoc subgroup data. I will say one word : Deferiprone (or Ferriprox) that has been approved on Retrospective, non randomized study. That is pretty amazing isn't it ? Little more research would have been welcome by SocGen on that.
Also they didn't mention the GIST chance of approval. Most likely they would have raise concern that they don't recall any product approved after Phase 2 study ? Well here we can bring Brentuximab (Adcetris) or Vincristine (Marqibo) or Carfilzomib (Kyprolis) that all got approved (conditional) despite only a Single Arm phase 2 study.
So on that little statement from SocGen, I'm surprise by the lack of research on historical occurrence but also Im disappointed that they cannot even imagine the possibility that AB Science could simply do a confirmatory study and get overall population (being only the subgroup) successful ? And based on the Hazard Ratio and the P-value the success of the study would be more then sure and probably it would be stopped before the end and makes approval most likely in 2-3years from now top.
So bottom line :
- Lack of research on other cases
- No projection into models where a confirmatory study could be done
- No mention or analysis of the Hazard ratio and P-value that are amazing (not even a mention on the fact that we also AREN'T AWARE of any drug with such hazard ratio and p-value in pancreatic cancer?)
Now there is one funny statement on that research :
"Increased visibility in pancreatic cancer is offset by the slower ramp up of sales in the veterinary indication".
That one is pretty fun to read. I think there is no need to even comment that.. People that are little bit clever will understand how insane that statement is... (Hope so). Just a hint : Pancreatic cancer market is a multi billion dollars market, Veterinary is a multi million dollars... Second hint: HR<0,3 and p-Value < 0,0001..... Enough?
Now let's see the conclusion of the study :
" Approval Stock >25euros, Not approve Stock <15euros"
So Approval means market of multi billion within reach of AB Science but the stock would be value > 800Milions euros... only ?
So not approve .... means that it is one and fail attempt? Especially when we know they didn't go for the FULL APPROVAL, but CONDITIONAL APPROVAL which imply that they will do a confirmatory study with the subgroup only most likely ? How come the analyst don't imply that occurrence to ? (That again would point out to the HR and the p-value which provide the chance of success of a confirmatory study).
It is a big surprise to see such a poor analysis from SocGen that definitely don't get it with AB Science and really demonstrate the ability to provide a true long term analysis with real deep research.
I think it would have been fear and transparent that the new analyst that covers AB science also disclose HOW MUCH time he spent on analyzing the company and discussing with management of the company as there is no mention about the data itself or the potential strategy and options ahead. It looks like a quickly written analysis without any deep research behind. Of course that is my personal point of view but mine is backed by the data itself, the p-value of those data and historical drugs being approved in condition similar to them and also on Long term vision.
It is even more of a surprise to see the market selling on such poor analysis, makes me wonder if one day the market will really get it with AB science or will continue to see the train pass and miss every upside move...
Hopefully, true investor will see behind that a nice gift for more buying at lower price. The long term prospect of that biotech is more solid then ever and the data we saw so far just demonstrate it... It is just sad that the market still don't get it. In the case of SocGen, let's say the new analyst being in London didn't have enough time to really study the case and just did a brutal DCF model with highly conservative hypothesis and you need to have insanely huge courage to reduce PT on data like that, you can at least give them that... :)
I won't elaborate much on that research as it is still at the end opinion against opinion, and we will see in the long run which opinion was true :)
Disclosure : Long AB Science
Disclosure: I am long OTCPK:ABSCF.