The trucks that carried the last products from the onetime tech highflier Aetrium rolled out of North St. Paul the other day, ending a long era for this industry in the Twin Cities.
Boston Semi Equipment of Billerica, Mass., acquired Aetrium’s semiconductor test handler product line in early 2014 for no cash, just a royalty on future sales. It was the last product line built by the old Aetrium, a public company by then under the effective control of a hedge fund manager from Connecticut.
“We offered everyone in Minnesota a job here,” said Kevin Brennan, vice president of Boston Semi. “We really wanted to do what’s right.”
In the end, five Minnesota employees will move to Massachusetts. That marks the end of a very long slide in employment, all the way from 530 in 1998.
The end of Aetrium serves as a reminder of how destructive market capitalism can be. There are obviously still semiconductors being manufactured in huge volumes, but the industry followed a classic pattern of dramatic growth followed by slow growth and then painful consolidation.
A little bad timing or bad luck in a consolidating industry can leave smaller players behind to slowly wither. And that’s how 525 jobs can evaporate.
I know that 1998 payroll number without looking it up, because it was my job to know such things then, as Aetrium’s vice president of corporate planning. Seeing the company wither to nothing is as disappointing as hearing the hometown high school just closed.
A school is probably the right metaphor, too, as Aetrium provided the equivalent of a great on-the-job MBA. Aetrium CEO Joe Levesque and CFO Darnell Boehm were as hardheaded as any in business, but they were also generous teachers.
It made a lasting impression on me that in every downturn in that famously cyclical industry, Levesque made sure to cut his own salary first. That unforgettable lesson in principled leadership is one reason the Aetrium experience still seems worth at least 1 ½ Harvard MBAs.
It was an exciting time to be there, too, because selling equipment used to build and test semiconductors was among the hottest industries of the 1990s.
There was no reason to think it would cool off. The computer and electronics industries were hot, too, of course, but semiconductors tested on Aetrium-built equipment were getting used in all sorts of new applications, from a Timex watch to the gas pump in front of a Holiday Stationstore.
My job included talking to shareholders, so Levesque invited me to the quarterly investor conference call in July 1996, a few days before I formally reported for work. I sat in the boardroom while he and others fielded questions from shareholders about a quarter with about $17.3 million in sales and 37 cents of earnings per share.
Certainly no one on the phone that morning could have guessed it was the high-water mark. Earnings per share never again got to 37 cents.
As it turned out, 1996 was the turning point. The semiconductor business had its booms and busts, but had for 20 years been growing on average about 20 percent per year. Since 1996, the average revenue growth has been closer to 4 percent.
When that happens in any industry, the smart thing to do is to find a merger partner and fast.
Dealing with fewer equipment makers is certainly what the customers, the chipmakers, came to want. With the era of fast growth over, they needed to get more efficient and wanted to lean on just a handful of trusted suppliers.
A round of consolidation is what the equipment makers like Aetrium needed, too. To be a survivor meant getting big enough to be able to afford the kind of technical sophistication, global customer support and cost efficiency that the biggest chipmakers would demand.
It would’ve taken a rare insight to grasp all of that in the fall of 1996, so instead all sorts of growth initiatives got funded across the industry that in retrospect probably look foolish.
I was gone by then, but my former colleagues did seek to get out of the strategic box they found themselves in. Aetrium didn’t have the balance sheet to buy anything big, so the choice was to sell. Aetrium hired a top-shelf investment bank to find a buyer in 2004. Then the overall industry softened again, and no deal occurred.
Stuck going it alone, a good year by the end of last decade was $17 million in sales, what a good quarter would’ve been in 1996. Of its closest peers, only Aetrium had managed to remain standing.
The first one to go, in more or less a distressed sale, was bought by Micro Component Technology, then of Roseville. Micro Component was itself then finished off by the Great Recession. A successor company with a small office in Minneapolis took over some of the assets and now builds its remaining products in Malaysia.
An activist shareholder from Connecticut led a group that agitated for a new Aetrium board and finally got one in 2013. In looking through the public filings, it’s clear that all the dissidents wanted was a corporate shell with a lot of ability to shelter income from taxes, by using what’s called net operating loss carryforwards.
The company got renamed ATRM Holdings, shed the test handler product assets and acquired a modular housing business last year.
Meanwhile, Brennan from Boston Semi said his company intends to “breathe some life” back into the Aetrium product line, though he added that the Aetrium brand name may soon get tossed overboard. That would close the book on Aetrium.
It’s easy to look back on all this and wish for a different outcome. It would have been far more satisfying to see the company simply acquired, much like Chaska-based FSI International managed to do in 2012.
FSI built equipment for a different part of the semiconductor manufacturing process, but like Aetrium it was far too small for a consolidating global industry. At $6.20 per share, its sale to Tokyo Electron Limited came at less than a fifth of the peak valuation of FSI stock back in the go-go year of 1995.
That price may be disappointing to those who remember the boom years. On the other hand, FSI was spared the indignity of being kept alive by a hedge fund manager for its tax benefits.