Illumina’s Sorrow and DeSouza’s Woes

There’s an interesting article on Illumina in the FT. It’s well worth reading but I’ve listed what I think are the key quotes below:

• Within the next few weeks Brussels is expected to hit it [Illumina] with a fine of up to 10 per cent of revenues or about $450mn, for closing the Grail deal despite opposition from regulators.

• Dan Brennan, analyst at Cowen, an investment bank, said many investors are not "fans of the deal" [referring to the Grail acquisition] and it came up in almost every conversation he had with them.

• DeSouza said "The stakes are really high here in terms of human lives. This isn't just another deal: if Illumina prevails it can bring Grail's life-saving test to countries beyond just the US and the UK, which is what Grail is planning to do," he said.

• DeSouza said Illumina has a "huge amount of respect" for regulators but insists it will fight the EU's order blocking the merger in court, arguing Brussels has no jurisdiction as Grail has no revenues in the bloc. Any EU fine for "gun jumping" - closing the deal despite opposition from regulators - would likely be similar to the $300mn break fee in the merger contract it signed with Grail

• Illumina's former CEO Jay Flatley also questioned the economics of the $8bn takeover, which was negotiated in 2020 near the top of the market. "If they had waited a year then it would have been a $2bn acquisition," said Flatley, who was chief executive for 17 years until 2016 and then chair until 2021. "In retrospect, it would have been better to wait and realise that the market was sizzling hot at the time and therefore it was overpriced." It is a "huge disappointment" and investors want it spun back out, he said. "Some investors frankly don't care how much Illumina is going to get for it. In some ways it's kind of a sunk cost. If they can spit it back out, then the earnings numbers get back to where they should be," he said. "If nothing happens on this in the next year, I think the grumbling will probably get louder ... Francis is at the point of that spear."

Illumina is facing challenges on multiple fronts. It’s clear that their historic dominance of the DNA sequencing market is coming to an end. This is largely because fundamental IP is expiring allowing new players to enter the market.

Illumina acquired this IP with Solexa in 2006 for $600M. I’ve heard that in part they felt increased sequencing capacity would bolster their array business. Sequencing would be the “discovery platform”, identifying new SNPs. But the bulk of the market would still be arrays.

If that’s the case, Illumina got lucky, because at this point it’s clear that sequencing is a fundamental platform on which pretty much everything else in the genomics space is ,or will be, built.

Illumina executed well on the Solexa IP. The original Illumina Genome Analyzer was pretty much the Solexa instrument with a new logo on the box. But the Hiseq and NovaSeq are solid iterations on the approach that push throughput.

To date Illumina have had no realistic competitors

Over the same period (2006 until now) there has been no realistic competition in the market. Let’s briefly review:

• There was the ABI SOLiD, which produced color space “reads” and was only barely a “DNA sequencer”. Never gained significant market share. Eventually discontinued.

• There was Complete Genomics. Who (despite what others might say) had very poor data quality and we’re not competitive. Never shipped, acquired, merged into MGI.

• There was 454. Which was a solid instrument, but used an approach that didn’t scale. Discontinued.

• There’s Ion Torrent. Again, a solid instrument but with similar issues to 454 and a relatively high COGS. Small market share.

• And there’s Pacific Biosciences and Oxford Nanopore. Who have so far (though this may be changing) failed to compete on cost.

With fundamental IP expiring we’re now seeing some pretty realistic competitors enter the market: Ultima Genomics, Element, Singular, PacBio’s Osmo, MGI and others.

In many cases these approaches are built on expiring IP. Performance is similar to Illumina, and they provide realistic competition.

Without something new Illumina will lose market dominance as well has having to reduce margins.

What can Illumina do?

Well, they could have developed a fundamentally new approach. But big companies can’t do “disruptive innovation”. And unfortunately there’s no Solexa for them to acquire. And even if there was they’d be blocked from doing so.

They therefore seem to be trying two things to help them maintain margins:

• Bringing more of the backend (manufacturing) in house.

• Bringing more of the frontend (applications) in house.

On the manufacturing side they’ve announced that flowcell manufacturing via nano imprinting will be moving in house. The hope here is likely that they can reduce costs and thereby somewhat maintain margins. I suspect this will be only moderately successful.

On the frontend side I suspect they want to buy Grail so they can take more of the value add on the application side. This makes sense, they’re going to lose market share anyway, so who cares if they upset their customers. They see Grail as a potentially huge market which leverages their basic capability in sequencing and again allows them to maintain margins by taking more of the application revenue.

This seems risky, but at this point is there a better way forward? I honestly can’t see it. I suspect Illumina and other players will spend the next few years trapped in blood thirsty race to the bottom as each player attempts to gain and maintain market share.

The good news is this should work out well for customers, who finally should have a realistic alternative available to Illumina.