Wednesday, April 17, 2024

PacBio Plummets

PacBio announced preliminary earnings yesterday, and the nearly immediate result was a 50% plunge in their share price.  Along with the earnings, the company announced significant cost cutting.  The details of those cuts were not made available, but some clever tea leave parsers noted a significant omission from what the company said it would continue.  The ASeq Discord channel on PacBio absolutely blew up, with opinions ranging from PacBio is in a death spiral to PacBio must be for sale, with significant numbers of "Christian Henry won't be CEO by year's end".  

The worst news in the earnings is that sales are essentially flat.  Perhaps the most disturbing is that consumable sales are flat, even as Revio instruments keep being delivered.  A well-fed Revio should be consuming flowcells at about four times the rate of a similarly well-fed Sequel, so the flat consumables pull indicates that users are just running slightly larger experiments and not the grand campaigns that PacBio was banking on.  And that's on average: there are labs pushing Revio's hard and that means there are others that rarely use them at all.

The notable omission from the continued plans note is the ultra-high throughput long read sequencer, originally mentioned as a partnership with Invitae - a company with similar investors as PacBio that is now in bankruptcy.  Parking/axing this project makes sense; if a Revio with 4X capacity wasn't finding purchase, then a successor with perhaps another 4-8X doesn't make sense in this market.

On the other hand, the press release did call out the benchtop long read sequencer for continued development.  That instrument might be unveiled late this year (perhaps at ASHG?), though the company hasn't made any committments.  If it is sufficiently simpler and less expensive than a Revio, perhaps with one flowcell slot and less powerful compute and less extensive robotics, then hope would be a lower capital cost would enable more labs to get their own PacBio.  But the risk is that labs currently on the fence for a Revio will all go to the lesser unit, locking into the lower possible consumable use.

Also called out for continuation is the ultrahigh throughput short read sequencer, which will marry the Onso (ex-Omniome) chemistry with the Apton hardware.  I continue to be skeptical that PacBio will make significant inroads with the Onso platform; if I were in charge (complete fantasy for sure) I'd be exiting the short read side to completely focus on long reads.  As murky as the long read market is, the market for "not NovaSeq X" population-scale sequencers seems even dicier, with Ultima and Complete Genomics already in position.  PacBio's argument has been that the higher accuracy of Onso chemistry will win customers, but there hasn't been strong evidence that the market feels the same way.

Can PacBio ride out the storm?  Under the old cash burn they had about two years cash runway, assuming flat revenue.  They also have $1B in debt, though none of it is due for several years.  

But the real challenge is identifying markets for the long read tech and elbowing some space in short reads.  PacBio and its collaborators have done well identifying several potentially important applications, but none have quite caught fire - and Oxford Nanopore is always nipping at their heels.  For example, it was 2014 when I declared that PacBio had become the standard for microbial genome assembly - but people still publish lots of short read microbe genomes despite the cost per genome of running that on PacBio being a fraction of what it was.  HiFi data has clear advantages for metagenomics work.  I have an embryonic piece arguing that PacBio should be the standard for rare disease sequencing.  There's papers showing HiFi's utility for quality control of gene therapy vectors such as AAV.  PacBio 16S resolves species far better than short read 16S.  Iso-Seq reveals key details on splicing, 5' and 3' end variation.  

PacBio has tried to help many of these applications along.  Buying Circulomics and continuing its sample prep technology development has enabled higher quality HiFi libraries; linking with automation partners makes practical building more libraries in short order. The Kinnex payload multimerization kits enable more data per flowcell for applications such as 16S or IsoSeq. There's some loose ends - bringing their bacterial methylation analysis fully into the HiFi world would have some knock-on benefits for eukaryotic chromatin analysis methods such as Fiber-Seq. There's probably some more Kinnex kits worth launching.  But in general on the technical side, PacBio has done many things well - yet it hasn't translated to consistently growing sales.

On the non-technical side, PacBio cut a very bold marketing course with their name-act music concerts, something I in which expressed skepticism in the past. More recently, there was a very hard sell marketing campaign last month going aggressively for Oxford Nanopore users - a move that smacks of desperation now that we can see that PacBio was having trouble closing Revio sales (does anyone have an inside track on steak knife purchases by PacBio?) The acquisitions  were paid for with stock (correction from a commenter: Omniome was half cash, half stock), which preserved cash.  Perhaps the decisions that look the worst is taking on the debt rather than issuing more stock when the price was twenty fold higher than the aftermath of yesterday's plunge.

So where next?  One speculation is that PacBio is up for sale, whether management likes it or not. Illumina clearly isn't a player here! There's the persistant fantastical idea of Nvidia buying it by diluting its stock by about 1/5000th - but why would Nvidia bother?
} Rumors keep popping up that Roche - a one time PacBio partner who dumped them abruptly - will finally unveil a nanopore sequencer at the SFAF meeting in Santa Fe at the end of May - which perversely will be coincident with Oxford Nanopore's London Calling.  ThermoFisher is a routine candidate.  Taking PacBio private - as is happening with nanoString, might be an option as well.  Sadly for the company, the days of SoftBank throwing money wildly are probably never coming back - that free-spending helping stoke PacBio's fires but also gave us the embarassing spectacle of WeWork

And would an acquirer swallow PacBio whole or break it up?  The Circulomics business could easily fit within any of the major reagent houses - ThermoFisher, Agilent, QIAGEN, Danaher, etc. Thermo has a short read platform and it is in better financial shape than you might think (another inchoate piece); perhaps they'd only want the long read part.  Dismembering the company might be a way to get the best parts while leaving the investors holding the bag on the $1B debt - not a polite strategy but who expects Wall Street to be polite?

CEO Christian Henry is clearly under great pressure to perform, but I doubt the Board of Directors is eager to jettison him.  Not only is that always expensive in terms of golden parachutes, but firing a CEO is not a way to raise confidence and share price, plus it would mean a potential period of uncertainty around direction (conversely, GE fired two CEOs in quick succession and seems to be climbing phoenix-like out of the ashes). But Henry probably will be on a short leash.

One way to look at things is the cost savings gives PacBio management two years to turn things around or sell itself - you don't wait until the money runs out to shop the company.  Each of the next earnings will be tightly scrutinized, and the rollout of the benchtop instrument will as well -- what sort of performance will it have (particularly on the compute side)? What capital cost? When will it hit the streets?  Analysts will also be watching for Jr cannibalizing Revio sales - or perhaps worse, anticipation of Jr delaying Revio sales.

It's also worth noting again that we understand the fraught nature of companies such as PacBio, BioNano Genomics and Singular Genomics because they are public companies; we can also peer into the stalling at Illumina.   Private companies such as Element Biosciences and Ultima Genomics may also be looking at dwindling cash but we can't get the details on that.  Biotech in general is struggling with financial issues largely around the current high interest rate environment - why invest in a risky venture when you can get solid, boring returns from a bank account?  Many companies will weather the storm - the value of genomics is clearly there - but how many current names will be only memories when the biotech sun appears again?  Check back in this space a few years from now to see.

[2024-04-17 11:34EST - as pointed out by an Anonymous commenter, Omniome was half cash, half stock]


Anonymous said...

Omniome acquisiton was 1/2 cash, not all stock, as per the press release:

Anonymous said...

Also - PACB market cap is now less than half of what they paid to acquire Omniome.

Anonymous said...

Seems like Henry went to PacBio with a clear goal: re-create a new illumina. Look at all of the illumina (friends) commercial leaders that followed. Also why they opened the SD office? IMO: This cookie cutter commercial approach led to this demise: they wasted time with the SBB platform because that’s what they knew how to commercialize. I bet PacBio would be in a different place today if they took the energy spent on short reads and reallocated to the long read business (also instead of massive party spend).

Anonymous said...

Um, ONT revenues and growth are ahead of Pacb based on recent numbers, so technically pacb is now snapping at ONTs heels, more so given their desperate recent marketing.

Anonymous said...

L O L. Waste of money.

Anonymous said...

Worst part is that they took on debt to finance the cash portion of the Omniome deal. The $900k in convertible notes they took on a few months before was burning holes in the pockets of the ex-Illumina exec team hoping to recreate their Illumina glory days.

Anonymous said...

Interesting comments on here about CH and others from Illumina simply recapitulating the patterns of thinking and behaviour they followed at ILMN. Ive seen this a lot over the years in biotech. In a way prior success is poison as it prevents analysis of a new situation on the merits, by new employees. Long term pacb people must be very upset.

Diverging onto Omniome also shows a lack of internal focus and, some would say, a lack of belief in the pacb long read tech.

But the core issue with pacb is the technology itself forces a large complicated and expensive machine, where it is hard to get a decent margin versus the nimbler boxes from others. So they only ever seem to sell 200-300 before having to make an incremental version, rinse, repeat.

Anonymous said...

Many pre-illumina takeover employees left within 1-2 years or got pushed out by this management

Anonymous said...

Yes, I am one of them, previous long-term employee. Christian and Mark had no interest in learning from us, and simply wanted to apply the formula (and managers) that they knew. We could tell them all day long that it was fundamentally a different business with different dynamic and economies, and they discounted our experience entirely. Nothing that has happened since has surprised me. It’s sad for the loyal PacBio customers who have invested in the technology and now have to deal with this business uncertainty again. I love the PacBio tech, and a sincerely hope it can stay in the market despite this poor leadership.

Anonymous said...

As a cost cutting measure, former CEO of Circulomics and current VP of Tech Dev, Kelvin Liu, announced his departure. Didn’t sound like this is voluntary. The Illumina crew continues to push out key long read talent.

Anonymous said...

the MBAs from illumina are clowns!

Anonymous said...

Rumored PacBio had a 1/3 RIF today. Hit R&D, Bioinformatics Dev, Application Dev, and sales staff.

Anonymous said...

195 employees were laid off today 4/28/2024, and it looks like PacBio is shutting down the entire San Diego operation (Omniome). Impacted employees received an invitation to attend an "all hands" meeting where it was announced that their employments will end by June 26 (WARN act) and that they should expect to receive a severance offer before end of June. Just last October they laid off about 75 employees with a 1-month salary severance. Six months later they terminate almost 200 employees more. Any guesses on the future of PacBio?

David Eccles said...

PacBio trying to pinch ONT customers is a weird punch. People who are already using ONT probably don't care so much about accuracy, which is pretty much the only advantage PacBio has over ONT [for a little while longer]. Apart from that, ONT has more resolution on base modifications, and is capable of longer reads, shorter run times, and lower minimum run costs. Other claimed advantages with PacBio seem to be mostly associated with sample preparation, which is not so much a technology issue as it is ONT dropping the ball [many years ago, if they ever had it] on their protocol development.

I think PacBio would be better off encouraging Illumina customers away from short-read sequencing: it's a much larger pool of potential customers, and long read benefits with similar accuracy are easier to sell. It does still leaves them with order-of-magnitudes-higher cost-per-read (vs cost-per-GB) as a hard sell [more reads are better, right?], but I consider fighting that battle to be a worthwhile endeavour.

Anonymous said...

When will ONT surpass Pacbio in accuracy? When they truly solve the homopolymer problem?