Transitional Cree seeks next big thing

When Cree warned in mid-July that its fourth-quarter profit would not meet initial expectations, investors were spooked and its stock price closed down 25% on the previous day s valuation at just under $18.

This is not what we ve been accustomed to hearing from the Durham, NC, company – since 2002 almost every quarter has brought record-highs in revenue and solid profits as its fortunes followed in the massive upswing of the cell-phone handset business.

That upswing is still in full force, with almost a billion phones expected to sell this year – all of which will feature GaN-based LEDs in the keypad and display backlight modules. The problem for Cree is that, unlike in the RF space, where GaAs chipmakers have benefited from the need for more complex, higher-value components, LED backlighting is now largely commoditized.

Greater competition has been accompanied by protective measures from rival chip manufacturers and the resulting drop in average selling prices is now outweighing any increase in unit sales – and hitting Cree s margins. In the company s most recent investor conference call, CEO Chuck Swoboda said that he expected this market to remain relatively flat.

But there are other factors that are squeezing Cree – some apparently fleeting, but others are longer-term trends. The first is production. A quiet period of relatively low fab utilization in early summer was followed by a sharp spike that the company failed to predict and had difficulty coping with. Demand has since tailed off again and these sudden variations in manufacturing volumes make efficient management of a chip fab harder.

The second squeeze to Cree s bottom line is intentional and has come from its rising research and development spending. Even after some considerable stock compensation expenses are taken into account, this has grown by 25% in only 12 months, coming in at nearly $51 million or 12% of total fiscal 2006 revenue.

Now in the middle of a strategic transition, this is all part of Cree s push to reinvigorate its business and reproduce the record-busting performance that has been wooing investors over the last few years. At the moment, that means absorbing the costs of ramping up the production of Schottky diodes for applications in power switches, as well as the XLamp packaged LED components.

The power device market is going to be an important one for Cree. Having now officially opened its new production facility in Research Triangle Park, NC, where it will manufacture these chips, the foundations are in place to exploit the global drive to reduce energy waste by replacing relatively inefficient components with those based on GaN and SiC.

In the latest quarter, sales of high-power electronic devices were just under $5 million, representing a 29% sequential rise and up from just $1.8 million in the equivalent quarter of 2005. Swoboda says that it will take time to build both the Schottky diode product line-up and the associated power device "brand". He estimates that, in terms of their commercial maturity, the power products are about one year behind the XLamp.

Also fitting into this future investment category are Cree s "Colorwave" lighting modules for large-scale liquid-crystal displays. Although some rival products have been commercialized, this is, for now, a long way from being the volume market for HB-LEDs that has been envisaged. Cree s own experience probably explains why. Although the backlight development team has met every technical target that it has been set, there is still a problem: cost.

Cree s LCD-making customer needs the LED backlight at a lower cost than the US company is able to provide it for at the moment and this could hamper its commercialization. The expected date for initial deployment of Colorwave modules in production LCD TVs has now slipped back into 2007.

In the short term, Cree s shareholders may have to put up with lower margins, squeezed profits and flat sales figures (guidance for the current quarter is $106–110 million compared with $106.7 m in the period that ended on June 25). The company even admits that forecasting the next two quarters is tricky. But if Swoboda and the gang can bring the right products to market at the right prices, then investors playing the long game will reap the benefits. "We are in the middle of an exciting transition," said Swoboda. "We are trying to build a much larger business and to deliver real energy savings."

Quite how large Cree becomes will depend on many factors aside from its own execution. Sticking his neck out a little, Swoboda says that the goal is to increase revenue by 60–70% over three years. That would mean an annual revenue of $700 million in fiscal 2009 and would be great news for the wider industry.

Not that all of that increase in revenue is likely to be a result of organic growth. Cree now has $376 million burning a hole in its deep pockets and its acquisition of zero-micropipe SiC substrate developer Intrinsic this summer seems likely to be followed by similar deals that will broaden its net.