How much would it cost to replicate Complete Genomics' business?
Let's calculate: the company has an annual capacity of 5K genomes, which is being ramped up to 10K. Now, a HySeq instrument can sequence roughly a 20x genome per day; so to get to capacity of 10K per yr., I would need 32 HySeq instruments. At a list price of $750K, that would cost me ~$24M (list price).
I will also may want to invest in some computing power, although it is not clear that owning and maintaining a data center is more economical than using cloud service, such as Amazon. Software: open source...
So a decked-out 10K genomes/yr sequencing center would cost ~$30M, using Illumina technology at list price. $GNOM has 33mln. shares, which makes it less than $1/sh to replicate the business. Why is the company trading at such large mark-up?
Obviously, at present no one is nuts to build a 10K genomes/yr private sequencing center, only to compete with publicly funded (subsidized) sequencing centers, like BGI, Welcome, Baylor etc... The point is even if the $GNOM business were to make money, their capacity can easily be replicated at low cost.
Obviously they blew it; according to the company they had performance edge over ILMN last year, but instead of selling the instrument and to compete for lucrative market share, they went with this discredited fee-for-service business model, which gets them nothing.
As soon as insiders start selling (they haven't yet) I expect $GNOM to fall to the $1-3 range.
I appreciate very much biology specialists giving their opinions, with good justifications based on the costs.
My opinion was that they had a different product to offer as Illumina.
Since you wrote this post, insiders (and institutions) have been selling a lot recently.
I had bought them long ago, when they were new, and made a good profit. Then recently, I thought they were oversold, but did not do my "home-works" like I usually do.
My question is: can they be sued for giving wrong info?
Even if I do not trust analysts, how in the world could they set a target around $9.00?
Did they not ask the opinion of specialists in the field?
GNOM is on its way to $0, cannot compete with real companies like ILMN. Read length is 10 - 10 - 25 - 30, which is too short, even though they are linked. Try to do a circos plot with their data and you will see about 7,000+ chromosomal translocations in NORMAL diploid cells. Too all you geniuses out there: How many chromosomal translocations do diploid cells have. Please tell me the answer. Bio101.
You're right, the capacity could be reproduced for about that much. But, given ILMN recent leanings (about investing in hardware in the face of uncertain funding in the future), some companies may make the "rent instead of buy" decision, going, in the short term with GNOM. A more fundamental flaw in GNOM is their long term business model. Lots of companies (by which I mean pharmaceutical companies) will "test the water" with GNOM, see if there is value. If there is value (which is still to be demonstated), they'll buy (ILMN), not rent (GNOM). GNOM will still get the occasional one-off HIV grant, INOVA like project, but there won't generate the volume they need to be profitable. So, GNOM will never get the volume they need, given that this service/equipment is going to become essentially a commodity over the next 4 years. So, no short term prospects, no long term prospects.
ILMN is producing more cheaper hardwares and low cost services into 2012. I don't think GNOM can compare itself to ILMN in all means. GNOM has to fall for a few weeks before some people want to buy it.
KD, your analysis is pretty much spot on. What amazes me is the people who claim to be scientists (and no doubt they have "scientist" in their job title), but clearly don't know sequencing and probably have only read a paper or two on the subject. Probably a bunch of Affymetrix people, still thinking that arrays will have their day again. One guy claiming you need "200x" to sequence a genome, and another person saying Complete does 17x "because that's what's in their press releases" (clearly that guy hasn't read one issued by Complete in the past year).
The proof that your analysis is correct is that the short interest as a percentage of the float is about 25%.
Complete is a Complete Joke.
I have a B.Sc. and an M.Sc. in evolutionary biology and anthroplogy, and working on a third degree. What are your credentials? I have read the comments of others and I can tell you with certainty that both you and KD have no idea what you are talking about. Take your allowance and run along little fellas.
The problem with your argument is you don't understand how genomes are sequenced. I am a molecular biologst and have been working in the sequencing field for 15 years. When you read a 20X genome coverage, that is misleading. To sequence a single genome at high confidence of accuracy, you need something like 200X coverage. So the HySeq 20X coverage per day would take about 10 days to sequence a single genome. Thats assuming everything goes perfectly, which it never does. Illumina machines can be employed to do the work like you suggest, but you would need a facility of dozens of the instruments, dozens of staff, major laboratory infrastructure as well as computational and informatics infrastructure. My lab which pales in comparison to GNOM labs is in the mid tens of millions to set up and staff. Per year. The market cap of GNOM is way too low.
Let me ask the question differently: $GNOM records $31.4M "plant & equipment" on their latest balance sheet. With 33.6M shares outstanding, that's $0.95/sh.
My question is: why does the share price is higher than that? What is the proprietary advantage that cannot be simply replicated by a $30M with HiSeqs? If demand picks-up, why can't another fee-for-service provider build a $30M facility with ILMN technology?