Chip Market Copes Well, Post Quake: Part 2

The electronics industry is trying to recover from the effects of the Japan earthquake, but this will not be as easy as it sounds. For a start, the supply chain was already maxed out before the disaster, so even if a firm wanted to place an order elsewhere there is little spare capacity to service these new requirements.

Secondly, there are issues of qualification. For consumer products, changing suppliers at a whim might be commonplace, but for other applications this inevitably involves a qualification and testing process, often requiring 1000 (i.e., three months) life tests.

There are also potential yield and reliability issues in changing wafer substrate suppliers and packaging materials, some of which manufacturers may take a risk on and some of which may be impractical. For example, change the airbag controller or other safety-critical MCU, and you almost certainly have to re-run the airbag qualification test once you get qualified samples of the new parts from the new sources, adding even further delay.

This is where the real supply chain risks, so readily cast aside over the past decade or so, are the most profound. We have squeezed inventory and just-in-time beyond all degrees off reasonableness.

Just as a 30:1 loan-to-equity ratio sowed the seeds of self-destruction for Lehman Brothers, whereby just a 3 to 4 percent fall in asset value triggered the company’s downfall, so the ridiculously low inventory levels in today's ever-tightening, just-in-time supply chain and equally over-leveraged manufacturing environment will trigger a supply shortage that will take several years to unravel.

The fact is, the IC production cycle, from order placement to final end-product delivery, is typically six months, and that is when everything works like a dream. That requires a global inventory level of at least three months overall; today’s industry, and especially the financial community, would fret if this started to come close to half that level.

Add to that the passion to outsource everything to an increasingly reduced, often single- or dual-sourced number of suppliers -- with no one responsible for overseeing the increasingly extended and complex supply chain links, just their own in-out interfaces -- means a supply chain balanced like an inverted triangle. On March 11, 2011, that triangle fell over.

What then is the likely impact on our forecasts? As mentioned earlier, the current industry robustness to these issues underlines our fundamental premises. Pre-disaster, the 2010 recovery and 2011 outlook were incredibly strong. We are thus keeping this still firmly in mind together with the fact we are not in the business of making headline-grabbing pronouncements.

The truth is that no one knows what the actual impact will yet be. Crises, by their nature, generate clouds of uncertainty, and it is far too early and too complex for these to disperse. Traditionally, businesses postpone capital spending and hiring plans until the clouds clear, but there is little evidence that this is the case so far.

The industry's biggest problems in 2011 were always going to be supply not demand related. The situation in Japan has simply exacerbated the underlying problems. As such, our instinct is that there will be little overall change to our 9 percent semiconductor market growth forecast for 2011, with any reduction in unit shipments offset by higher pricing. We are also mindful to keep with our longer-term outlook of 18 percent growth for 2012, slowing to minus 2 percent in 2013.