Lessons from my worst investment

How I played one of my most researched positions all wrong, despite correctly predicting the outcome ahead of time

Today I wanted to share the story of my worst investment in the hopes you can take away some lessons for your own approach and process.

First, some background

From 2013 to 2018 I could do little wrong investing in (mostly) small cap life science companies, through a mix of luck, perseverance and bull market tailwinds. I grew my portfolio to a respectable level for someone who had started at 0, grown up without a father, worked a day job since was 15, suffered financial insecurity and had no formal investing background. It was empowering. I should have started this far sooner. But ultimately I became overdue to make a bad call, which happened.

Our world champions success but what about failure? Why do we not talk about our missteps more often - if we don't we risk repeating them. They are also far more interesting from a behavioral economics perspective as much of our actions are at times completely irrational from otherwise rational people (I'm rational a majority of the time, I like to think, but not all as this story will show).

Onto the story

In 2018 I drafted (my first and only) investing report on Achaogen, a SF based company working on antibiotics for complicated infections effective against superbugs. It's rushed and 40 pages of a lot of confirmation bias. If you really want to check it out, it’s here. At the time I thought it was sufficient work and enough of a story to at least sell myself on holding a position through a risk-on period with a company I felt was misunderstood. I even sent it to the CEO, Blake Wise for thoughts who basically told me I understood the company, market and future.

I'm not formally trained to write a report I just wanted to try a scrappy one for fun. I like writing and collating ideas. I do this stuff in my head but seeing it I knew would force me to improve and reflect. Well here we are. Ultimately the company ended up going bankrupt, getting de-listed, ruining many portfolios and causing 100s to lose their jobs, despite spending 1B$ to develop a drug that (still could!) help save many lives. That part can be true, a company can have a potentially highly useful and important asset, but that alone is not nearly enough to justify an investment. 

How I predicted $AKAO would pan out

How I actually played $AKAO

I was precisely correct about what would happen from my analysis report and failed to act on the losing scenario as it presented itself. My downside prediction range of 7-9 was exactly where the stock went, but instead of deciding that I was wrong I continued to hold my position eventually exiting for a close to 70% loss (my worst at the time and still to this day). Incidentally it wasn’t enough to be prepared and have a prediction, I had neglected the step of committing to what I would do in the losing scenario, so I clung to hope. Let’s give some more background before we go through the whole story as I think it will become clear why I failed to act.

Many post-mortems on the company were written by analysts and media including the below if you’re curious. We all thought the company was a special situation and an eventual success (emphasis mine highlighted below). We were wrong, not just myself, but tenured industry pros who have been commenting on the sector a long time.

Here's a short list of ways I was very wrong on this report and my thinking on the now defunct $AKAO:

  • I was clearly dead wrong on market need - overstated by CARBx/industry/media (MSM has been hamming up this problem for a decade with apocalyptic implications). I love the idea of being part of something that helps humanity so naturally I was interested in following the story of next gen ABX just like I already was interested in following next gen cancer/rare disease treatments. I should not have been for a myriad of reasons.

  • Wrong on thinking the FDA would decide to greenlight small N trial for key drug-resistant BSI infections (I read one too many stories on people dying from them and felt like they were going to use LPAD, guidance docs the FDA themselves had drafted to provide regulatory support to enable an expedient approval to save lives). What is obvious is the FDA is a bureaucratic, risk-averse organization first and foremost, even under arguably extraordinary circumstances (which I believed were present here). Pressured actions due to rare disease patient advocacy groups (what another company, Sarepta Therapeutics, greatly benefited from) were never going to happen from people who were at their death bed by the time it was known they had a hard to treat infection. There is no advocate group for this because it is an acute and not chronic issue. Their only hope was organizations like CARBx trying to advance ABX care and scientists/MDs speaking on behalf of patients. This is important, but likely not as persuasive to act with sense of urgency as parents and advocacy groups pleading their case for compassionate approvals of new therapies with FDA leadership.

  • Was wrong that Gates Foundation had any special insight on this company - they are not necessarily the top brass at life sciences and appear to frequently throw money at a lot of ideas, hoping some work. Never follow billionaires into investments. They can burn $ you can't.

  • Was exactly right on the downside scenario if just CUTI approval I hit that price target exactly -- and no BSI - hey got one right:

  • Right on the superbug trend - it is definitely real I have talked to many people in space MDs and research scientists alike - IS IT economically viable is a diff question ...so we have to of course consider this greatly in pre revenue. I ignored this plainly obvious pitfall.

  • Wrong on the fact that poly-drug combos DO freq work on these bugs so many situations may not be so dire (I learned this only after talking to more specialists in space, the one I talked to initially was overconfident, I should have spent time with 2-3 SMEs min).

  • Wrong on economic potential for them (last line is never used, and the bugs aren't evolving that quickly).

  • Failed to properly estimate cash burn of a newly commercial stage biopharma company within the context of their first drug launch: in hindsight this should have been sufficient to stay away. They were not capitalized for this close to sufficiently.

  • Wrong on picking an arguably dated platform - I am not a scientist and only discovered this after when I spoke to some lab nerds - they said you may not want to be in that name, their tech is a gen old. I should have sold that day. I was too committed now, and thought that perhaps the time factor of working through the trials was being discounted, since in my head the need for new ABX was highly time sensitive (again, buying too much into media narrative).

  • Wrong that every bio with approved assets or IP can have SOME value, they just bleed out forever (AKAO went to 0 and while I sold for a loss and recouped some funds I held far too long as I thought they would find a way forward). I had sunk cost fallacy having done a whole report and holding through a lot of market back and forth.

  • Wrong to do the contrarian thing and buy out of favor market like ABX (it is so profitable in tech!). Contrarian without full understanding is just a recipe to lose big.

  • Wrong to listen to a person at JPM healthcare conference who said AKAO was a winner - I really trusted them -- I trusted all the people I met in the space perhaps too much -- this is on me. Unlike tech everyone in life science sector is frequently wrong, it’s just so much harder.

    This is just a sample of media stories covering the unmet need of ABX-resistant superbugs. There was plenty of concerning infection data too. This can all be accurate, and a company and sector still potentially a zero. 
  • I expected the FDA to do the right thing in the near term and grant superbug infections for BSI as rare disease and consider them through that light. CARBX the organization that lobbies them kept a strong narrative here and becoming close to these people made me too optimistic. I thought this could reprice a lot of these companies. The FDA even drafted guidance docs to help speed these therapeutics through (which were never used). Again too many “what ifs” from regulatory here. Would the FDA really punt on a therapeutic that would save lives in the near-term because a basically impossible to design trial wasn’t perfect? The answer, is yes. The concerns here are still real and many stories go into detail why the FDA set a worrisome precedent (in addition to chilling investment in an entire sector).

  • I felt I had developed an eye for finding life science names and felt good about the process. And I actually was doing well - had called a few 10xers and saw CAR-T and gene therapy potential and names early. What is wild is I felt like it was too easy (cell therapy/car-t/crispr/etc). I thought “no this is too easy I need to go looking for things no one else is following” which lead me to ABX and AKAO (no one was talking about this space). Which does work but ONLY if you are deep deep in a sector. It doesn’t matter if you’re not first or if the idea is original, as long as it’s good and a company continues to execute. Don’t overcomplicate what is working.

  • I stuck with them despite ADCOM vote no. I actually could have cut my losses here without much pain at all, a reason I was OK with having a position in this name, there was not much risk to roll the dice, however my lack of ability to already know my course of action caused me to continue to hold unnecessarily even when it should have been a clear “get out” situation.

  • Media continued to give hope to this story, but the writing was on the wall and all of us should have realized how this would end much sooner.

My friend Morgan wrote a post on degrees of confidence. This is a perfect example of me thinking my tech industry analysis skills were transferable. I had done so well for so long I figured I could win here too with similar pattern recognition and thought processes. It’s true to some extent (p-values are p-values) but obviously there is so much more to analyze here.

Given I also spent time writing a report on a stock, because it's something I always wanted to do, I shared it with others who were interested (a lot of people kept asking me for notes on the co since I tweeted about them). Including personal friends. So now I suddenly feel my fate is also tied to this company. Which meant I didn't just take the loss post FDA downvote, I kept following their news/press releases / FDA follow ups/initial sales efforts in the hopes they would turn this ship. I became emotionally attached as I am high empathy, and felt if others who bought into my idea were to lose I should continue to hold with them. The wrong attachment leads to destruction. Being able to be wrong in public is important if you want to share ideas.

How I’ve moved forward

In the years since then I have more than recovered from this (and other) losses and my biotech portfolio has reached new highs. With a better assessment of risk management and sticking to plans things are more consistent and lower stress. In fact the last 2 years have been some of my best yet with far less swings and more predictability:

More than that, due to this experience and wanting to badly learn more from an insider perspective, I was lucky enough to get the opportunity to work at a genomics firm for a year and see the industry not as an investor but as an operator. This has greatly helped me refine how I approach investments in the sector.

Of note, this crisis still absolutely looms. There have been ongoing stories of concern such as below in the NYT. I do fear one day we may face a situation perhaps not as dire as a pandemic, but equally painful for those involved if we do nothing. I don’t believe I was wrong to identify the opportunity, moreso about everything surrounding a complex space and blind optimism a drug with unmet need would be carried forward (vs chilling the entire sector, which is what the negative FDA vote did).

One reason why this is important to me, and I continue to study life sciences as a tech industry leader

If we get enough tech people to also understand biology we can accelerate so much incl life extension, computational biology and all the things at the intersection of tech & science. That is the future of everything health and tech, and truly an exciting space. But we can’t get here unless more people in tech learn about new areas. Without some of us putting in the blood, sweat and tears we run the risk of inadvertently funding the next Theranos in the future, which hurts trust in progress and takes money from entrepreneurs and scientists who understand this is an industry where you don’t cut corners.

So the record for those keeping score, my fav correct biotech company calls (off top of my head):

$JUNO, $KITE, $NTLA $SRPT, $RLYP, $TRIL, $MYOV, $MRNA, $NVTA, $KALA, $CUTR, $PACB, $RVNC, $VEEV $OCUL $CPXX

My bad calls:

$AKAO

Remember, if not careful it only takes one bad idea to break your portfolio or reputation (this part is silly as many appear to not understand losers are part of the game, and it’s how you play them that defines you). Anyway guess which company above people still hound me on? Not Kite, which was acquired for 11.9B and I’d been sharing notes on them before it ran several hundred percent. Of course not. It’s the one I was wrong on and believed in management for too long.

Anyway, this is not a post against owning individual stocks and taking on risk. It’s to spend more time and get comfortable before you do such things. Only make individual company bets with proper mindfulness and preparation …and willingness to follow your plan.