Global semicon industry update: 30 percent growth now on radar for 2010, says Future Horizons

Here are the excerpts from the Global Semiconductor Monthly Report, January 2010, provided by Malcolm Penn, chairman, founder and CEO of Future Horizons. There are a lot of charts associated with this report. The report also covers market trends. Those interested to know more may contact Future Horizons.

November’s IC sales continued the year-end rally, down just 2.4 percent on October, up 29.3 percent vs. November 2008. This confirms our earlier prediction that Q4-09 sales would be up around 6.4 percent on Q3-09, one of the strongest year end-closes on record - Q4 sequential growth is typically ‘zero plus minus 2 percent’.

This confirms that 2009 will come in close to our minus 10 percent forecast, most probably at minus 9.7 percent, setting 2010 up for a bumper double-digit growth year. Only a Lehman Brothers-type event can now derail the recovery, the future is bright, and not before time too. For far too long now doom and gloom has spoilt the chip market horizons. Industry faith has been stretched beyond the limit.

Ignoring the structurally (and typically) wild individual monthly fluctuations – which simply means no single month’s data is a good indicator of the underlying trends – November’s result places us comfortably within our minus 10 percent 2009 growth estimate.

Based on November’s WSTS data, it is now very difficult to see anything less than a 22 percent growth year for the semiconductor market in 2010 based on the current industry momentum (i.e. a fourth quarter growth of around 6.4 percent) and a very ‘average’ quarterly growth pattern for 2010. Indeed we are now starting to see the first industry guidance revisions that tend to indicate even this range might be low. If the current growth momentum holds firm, 2010 chip market growth could easily hit 30 percent.

Low double-digit growth is totally out of the question, growth in single digits an absolute impossibility, Figure E3. Either of these scenarios would need a very poor start to the year, which is simply not happening. Order books are strong, inventory levels are low, capacity is tight and demand is holding up. You could not wish for a better start to the year … what a difference from this time 12 months ago. Only a massive economic collapse can now spoil the party.

As we mentioned before, only a massive economic disruption like a Lehman Brothers bankruptcy can now derail the recovery and this is not being forecast by the economists. Quite the opposite, GDP data is trending more and more positively, with an upwards revision at the macro level more likely than not. This is not to say that the economic recovery is not fragile, it is far from out of the woods and many risks still remain.

Of the ‘not so good news’, to our minds the biggest single problem is the world’s financial systems remain unreformed and, worse still, unrepentant. This means the same issues that caused the global financial problem in the first place remain unchecked. In 1929, Wall Street’s shamed bankers jumped from their office windows. In 2009 they stood in line for their bonuses. From a chip market perspective, a sound economic base is important but the correlation is poor.

Whereas a collapsing GDP will trigger a chip market downturn, just as it did in the 2001 dot com bust and September 2008 Lehman Brothers collapse, the rates of recovery are independent of each other. For example, the economy recovered faster than the chip market after 2001 whereas the chip market is leading the recovery in 2009.

The extent of the market collapse can be gauged by looking at the peak to trough data, showing over one third of the chip market simply disappearing overnight. Except ASPs, despite this massive decrease in demand, ASPs help firm, in fact they rose a modest 1 percent. One year after the chip market collapsed, units and value have now recovered to 98 and 90 percent respectively of their Q3-08 (market peak) value, with ASPs coming in just 8 percent lower.

This is quite an extraordinary recovery, seeing as it took a full two quarters more for the world to exit recession. It happens though because there was no chip market bubble prior to the downturn.

With the memory market now in full flood of recovery – we can easily see an upside potential of a $60 billion market for 2010 – and memory prices increasing with barely a flinch from the market, 2010 is set to be a very good year for the industry. The only problem is that no one yet believes it.

Confidence has been shattered ever since the 2000 bust, with a glass half empty mindset dominating collective thinking. “Market growth is now single digit; ASPs will keep on falling; Where are the killer products to drag the chip world out of recession; We need to specialise, merge, narrow the R&D scope, cull the product line and above all dump all the fabs; outsource for capital and operating efficiency; etc”.

Well, to coin a phrase once used by Jerry Sanders III, “Nuts!” It was only 2004 when growth hit 28 percent just after an 18 percent growth in 2003. Better get planning now, it’s already too late.

Industry capacity
Q3-09 total MOS IC capacity was down 12.5 percent versus Q3-08, which in turn was up 6.2 percent on Q3-2007. Quarter on quarter growth was minus 0.7 percent, compared with minus 2.7 percent for Q2-09, minus 8.0 percent for Q1-09 and minus 1.6 percent for Q4-08. This dramatic slowdown in net new capacity is in direct response to the slowdown in Cap Ex that has been gaining momentum since the second half of 2007, i.e. well before the September 2008 market crash.

It should be remembered that there is a three quarter delay between a Cap Ex spend and saleable units out, plus at least a quarter equipment delivery lead time so Cap Ex in year ‘n’ drives capacity expansion in year ‘n+1’. As a result of the Cap Ex spend now growing much slower than the underlying demand, Q3-09 capacity utilisation rates hit 87.2 percent, up from 78.6 percent in Q2-09 and 87.5 percent in Q3-08. This bounce back was exactly in line with the prediction we made in our July and October Monthly Update Reports.

Just to recap, the sharp fall in Q4-08 and Q1-09 was the direct result of the uncertainty following the September 2008 financial crisis and the ensuing near-term inventory purge driven demand slump. It was not representative of the underlying trends.

With utilisation rates now back to their pre-Lehman level, we expect utilisation rates to improve again in Q4 with overall MOS IC capacity breaking the 90 percent barrier. Advance wafer fab production is already ‘Sold Out’ with Q3 utilisation rates reaching 93.5 percent. Leading edge capacity is now very tight indeed.

With Wafer Fab Cap Ex spend averaging around US$8 billion per quarter between Q3-06 and Q1-08, spending plunged dramatically in Q2-08 reaching around one quarter this average in 1H-09, recovering very slightly to $3.3 billion in Q3-09. Given the current front-end Cap Ex book-to-bill trends, this spending level will improve only slightly in Q4-09, exiting the year at just over $ 1 billion per month, half the previous 2006-07 level.

It must not be forgotten that this cutback was deliberately started well before the Q4-08 market meltdown, in a premeditated strategy to dramatically tighten supply and thereby increase wafer and IC average selling prices.

With November’s book-to-bill ratio now back to 1.0, the level of new front-end capital equipment orders has now been sizeably lower than sales for 39 consecutive months, from September 2006 through November 2009, aside from three brief incursions into positive territory circa Q4-06, Q4-08 and Q3-09.

2009’s Cap Ex spend now looks set to come in at around $ 16 billion, down 46 percent on 2008, which in turn was down 31 percent on 2007. That puts 2009’s Cap Ex spend at one third of its US$ 48 billion 2000 peak, basically at spares, upgrades and maintenance levels and no serious new capacity build.

No amount of productivity gains can offset this slowed investment, especially now the one-off 300mm conversion gain has been absorbed. Net new capacity addition is thus condemned to shrink even further during 2009, the capacity utilisation effect of which will be briefly suppressed in the first half of the year due to first half year seasonal demand and inventory adjustment process.

Unlike 2001, when the recession hit during a period of Cap Ex expansion, the industry was already 12-18 months into a capacity slowdown before disaster struck. For the first time in its history, the industry is entering the recovery in capacity famine mode. With the ‘Allocation’ word now back in the vocabulary, and the fact many previously IDM firms have gone decidedly fab-lite, it is not at all clear how the industry will respond or how tight things will become.

The interim period of ‘plentiful capacity in 2009’, has feed the illusion of plentiful capacity. This is now well and truly over, although few firms yet believe it; even fewer took the precaution of tying down their supply positions whilst the going was good. It is now too late; the era of cheap and plentiful wafers is over. Supply will get worse well before it gets better!

Market trends - growth prospects for 2010
Semiconductor applications were negatively affected by the recent economic downturn in 2008 and 2009 and the outlook was very gloomy at the beginning of 2009 getting progressively worse as Q1 figures were declared. The consumption of electronic equipment is, however, only loosely connected with the economy and many applications showed an amazing resilience in the turbulence of the economic storm triggered by the US housing market and the banking troubles.

The rest of 2009 showed a recovery, particularly in PCs and mobile phones units produced. Both of these applications categories make a significant contribution to the overall semiconductor consumption. Unfortunately electronic equipment was under price pressure as manufacturers tried to sell more in a depressed market that is already subject to stiff competition. This had a knock-on effect on semiconductor ASPs, but the end result was a semiconductor market that was down, but considerably betters than the industry believed at the beginning of 2009.

The performance of the electronic equipment and semiconductor market in the face of economic trouble has led us to the conclusion that growth over the next five years will be relatively strong overall, but the automotive and consumer market will take longer to recover. The seed corn of this growth is the increasing affluence in some of the developing countries, which will allow rising consumption.

The GDP growth of China and India, as two examples, was positive in 2009 compared to the more beleaguered developed countries of Europe, Japan and the United States. The consumption of electronic products in the East is growing fast relative to North America and Europe. Another factor that should not be underestimated is world population growth, which has more than doubled in the last 50 years. China and India combined make up approximately 40 percent of the total world population and this is set to increase going forward.

The total semiconductor market in 2009 is $205.2 billion and the market forecast for 2014 is $385.2 billion.

Key growth drivers
The key growth drivers for some of the major segments are as follows:

PCs
* Lower cost PCs are driving the market and we expect this trend to continue.
* The netbook phenomenon has spurred growth but this is not all gain for the PC and semis industry as it sets consumer expectations for low priced items and takes away from sales of (other) lower end laptops.
* 64 bit cores will trickle down to home PCs. These cores will give an improved PC experience for games and high definition movies especially for video editing.
* The PC will become more poplar for media playback in the home as it offers more flexibility than standard consumer items including DVD and Blu-ray players The availability of DDR3 will help on upgrades as it can improve performance for some applications.
* The trend to more fashionable, thinner, lighter and greener portable PCs will continue, helping new and replacement sales.
* Sales of PC will comprise high growth from developing regions and replacement sales in more saturated regions.

Automotive
* Semiconductor content is rising in the average vehicle in the longer term – pervasive use of electronics for all systems in the vehicle.
* Safety systems including new developments where accident avoidance systems are being developed for automatic braking to prevent impending collision.
* The car will replicate some of the entertainment and connectivity of the home. Ideally, the consumer would like to have in-car access to the full range of entertainment available in the home.
* Lower cost and more fuel efficient vehicles proved popular in the downturn but fuel efficiency is here to stay, which will require more electronics for control.
* Easing of credit restrictions will help growth in developed countries from 2010 onwards.
* Economic development in Brazil, Russia, India and China (BRIC countries) will spur automotive growth as more consumers in these regions become more affluent.
* New Asian automotive manufacturers (ex Japan) will become increasingly important producers in world as well as domestic markets.

Consumer (integrated flat panel TVs, STBs and games consoles)
* Consumers still have appetite for innovative products and are prepared to buy at ‘value’ pricing even during an economic downturn.
* Unit growth will continue but revenue will be much slower in rising as cut-throat pricing continues which will also be reflected in the associated semiconductor market.
* New innovations are needed to spur growth and this will be increasingly difficult as there is still a lot of concentration on cost reduction and manufacturing efficiency rather than innovative developments.
* Next generation consoles are likely to be released in 2011 and this will generate consumer interest. Interest in 3D displays, virtual reality and sensory feedback may feature more strongly and help growth.
* High definition video will encourage many consumers to upgrade for improved resolution and richer colours – ‘an altogether better experience’.
* More converged devices will appear - possibly larger thin screen navigation systems able to accept and receive phone calls, send cellular text, and play videos.

Mobile phones
* Mobile phones are becoming increasingly popular in developing countries and this will feed growth in the next five years, although world saturation will still occur in the more distant future.
* The market is becoming more commoditised and more phones are using applications standard products rather than custom chips. This development is helping reduce the end price for mobile phones and will spur unit and revenue growth.
* New mobile phone architectures that can handle multiple radio protocols efficiently and at low cost will help reduce the cost and increase penetration of mobile phones.
* Multiple phone ownership may become more prevalent with phones for emergencies, smartphones for business and fashion phones for leisure.

The mobile phone is the prime candidate for a converged device and is cramming in new functions year-by-year. Health monitoring, for example, could include heart monitoring, insulin analysis and pollen count indicators. These would feature in some new specialist ranges of mobile phones.

There is considerable potential for renewed growth over the next five years with a increasing number of ‘new consumers’ entering the market for the first time in developing economies. Electronics is becoming increasingly pervasive and most equipment is no longer in the luxury goods category and this will help increase penetration in the world market.