The first industry to squeal was automotive, Nissan closed six car factories because of chip shortages. The second squealer was the telecoms equipment industry with Ericsson saying it had lost $400 million in revenues because of chip shortages. The latest squealer is the smartphone industry.
According to Associated Press, Sprint Nextel hasn't been able to supply HTC's 4G phones because of shortages of ICs; Verizon can't get enough Motorola Droid phones for the same reason; and there's a three week wait at retail for the Apple 4.
It seems none of the present generation of semiconductor CEOs has the balls to do what their predecessors did - invest in the downturn.
Instead, in the downturn of 2008 and 2009, CEOs chopped capex 31% and 41% successively. Result: Massive shortages in capacity in 2010.
A prime example was NXP which said in its Q2 results that Q3 revenues would be flat with Q2's.
"They (NXP) now face a flat sales growth quarter in the traditionally busiest quarter of the year! How to shoot yourself in the foot in one simple lesson", comments Malcolm Penn, CEO of Europe's top semiconductor analyst company, Future Horizons.
In that same Q2 earnings report NXP wrote: "Our wafer factory in
Now you get major customers of the semiconductor industry whingeing:
"We continue to see challenges in procurement of components this quarter," says Cisco CEO John Chambers, "supplier lead times now appear to have stabilised, but are still longer than we would like."
"It's an industry-wide, global problem that won't be resolved over the next three months," says Alcatel-Lucent CEO Ben Verwaayen.
TOMORROW MORNING: Top Ten Semiconductor Equipment Suppliers
I'm not sure NXP is a good example - if they had not cut investment and operating costs they would have gone bankrupt months ago. Same with a few other semi companies probably.
Besides, these OEMs can complain, but they are also the ones who cut inventory drastically and are now struggling to restock. The semi guys are left to take the brunt of the bullwhip effect. And the best solution for that anyone can come up with to date is to go fabless. Hence IDMs closing fabs. I am sure NXP's statement that they will add capacity means they will be going to the foundries rather than building more of their own produciton lines.
Dave's comments don't really add up.
NXP still has about $1 billion of cash, keeping those fabs open would of course have eaten into that cash but they would have saved on restructuring costs. And now if they weren't constrained by capacity they would have had significantly higher sales and been able add to their cash.
As to going to foundries, this is not an option for a lot of their remaining products which are dependent on NXP specific diffusion processes (it was mainly their Mobile and Home businesses that used industry standard CMOS processes).
I'm sure you're right, Former NXP Employee. If you constrain your ability to grow output in a semi upturn, you miss your chance for windfall profits, gaining market share, building up cash balances and growing the business. Now is a once-in-a-decade chance to grow semi businesses rapidly, and very profitably, and where are ST and NXP? They're shut out of that opportunity because of lack of capacity. ST's shares have sunk to €5.55 - their lowest-ever. Why are ST and NXP lacking capacity? Answer: because of poor judgment by their CEOs. They cut too far in the downturn, and didn't re-invest in capacity early enough in the upturn. Result: they miss a huge opportunity to benefit their companies, their shareholders, their employees and their customers. And what do companies pay their CEOs for? Answer: good judgment.
In fairness to these folks, were we not all singing to the tune of "It'll all end in tears again!" a few weeks back when discussing the risk of investing in massive capex expansions in light of the inevitable, cyclical bust?
As Dave said, we get the sharp end of the bullwhip. If these CEOs invested in massive capex expansion and it all goes pear-shaped 6 months on (note how the "markets" are slumping again and getting jitters about consumer confidence and spending), then you can imagine the flack they'd get. Damned if they do...
The problem, as someone pointed out on a previous thread, is that the semi industry has never been good at inventory control (even Christmas can let us down!). Partly, I guess, because we are in a game where next big things can suddenly come out of nowhere and can't be planned for. Partly also, I guess, because we get the ("bullwhip" of) increasingly amplified effects of OEM's inventory shenanagins backwards down the chain from the market, and can't plan for those.
Some lucky punters might the expansion/demand timing right sometimes, but the house always wins in the long run!
You're absolutely right, The Baron, judging the timing of the upswings and downswings is horrendously difficult, probably impossible. But CEOs are paid handsomely for their judgment. If they judge things wrongly - and it is patently obvious who has judged wrongly in the current upswing - then the guilty CEOs should expect some flak - from customers for shortages in supply, from shareholders for not capitalising on growth/profit/cash-building opportunities, and from employees for missing the chance to grow the business and create better career prospects for them.
P.S. The Baron, You're also right in saying that it's too late in the cycle to build capacity now - when it comes on-stream it'll be ready for the next downturn. The CEOs who missed the up-cycle investment window can do little about it now.
I'm not even sure CEOs need to be paid handsomely for these sort of judgements. Due to the well proven cycles of our industry over 40 or more years it's hardly rocket-science any more :
"We are in a boom - it WILL go bust."
"We are in a bust - things will recover and go boom again."
Hands up who thinks it's a good idea to invest in a boom ?
So all we are left with is deciding how much to invest during a bust. Of course you may not want to bet the company on this but many companies have quite a lot of money sitting in the bank which could be used carefully. Also this is the time there are bargains to be had on capex expenditure.
That said those who have bet the company have been the big winners over the years. TI (several times), Cypress, Xilinx, Samsung and TSMC are just a few examples of those who believed that what goes down will bounce back up again within a predictable time and match investment to this cycle.
I agree Mike, it should be obvious. So why did so many CEOs fail to invest in the downturn? IMHO, in the case of ST, it was because they were so wedded to their strategy of reducing capex as a % or revenues in their pursuit of fab-lite, that they failed to factor in the need to have sudden increases in capex as a % of revenues if they were to take full advantage of market upturns. If those upturns aren't exploited then, as you are only too well aware, the company starts growing less than the market, margins fall and customers go elsewhere.
"But CEOs are paid handsomely for their judgment."
I say "Bluff"! CEOs get paid handsomely regardless of what they do; in fact, if you want a seven digit payout/off with megapension perks then ****ing-up royally does it everytime.
I certainly wasn't trying to defend this pair for not anticipating some sort of upswing. I get as steamed as any other watching CxO uber-folk rambling interminably about how uniquely wonderful they are and why they are worth so much money relative to us massed peons - only to watch their shoulders invariably slope to vertical as soon as it all goes a bit skew-whiff and an bit of handsomely paid-for responsibility and accountability is required. "The bucks stop here but the buck stops there".
Sadly, businesses of the world seem liberally peppered with this sort; from FTSE-100 titans (as city types like to call each other when spouting their "egotripe") to senior management in SMEs.
Anyway, I'll weesht now as I'm diverting the "when best to capex?" discussion via one of my many pet peeves...
I'm all for making capex and expenditures a percentaqe of revenues but it has to be done across one business cycle and unfortunately for semi companies that cycle is almost a decade long.
A similar thing in reverse is happening now in the games software industry where there is a roughly six year cycle and overinvestments were made during the good years and now they cannot survive the downturn with several going bust.
Wise words, The Baron, unfortunately the sort of people who get appointed CEOs - as opposed to those CEOs who found and grow their own businesses - tend to be the sort of people who put their own self-interest above the interests of the company.