Tuesday, May 31, 2011

Polysilicon Price Trends

A while back one of my customers mentioned that it would be interesting to compare trends in the solar business to those in the semiconductor business. Well, last week a sell-side report was issued where they paralleled the trends in the polysilicon market to the DRAM business. Prices for poly, if you haven't heard, are falling and this has implications - some good, some bad - for companies in the solar food chain. It doesn't look like these price drops are over as large capacity additions from GCL Solar and a Korean manufacturer are looming. If this is a fair comparison what does it mean? If you pull up a chart of a commodity DRAM play like Micron Technology from 2007 to mid-2009 you will see what excess capacity can do to a stock price.

Remember there 's a good and bad side to this story depending upon where you are in the solar food chain. Panel makers will benefit from the declines but the impact on poly companies and equipment suppliers could be problematic.

At the time of the report quoted poly spot prices were at $57.50/kg to $62.50/kg, cells were estimated at 0.90c/watt and panels at $1.33/watt.

I called a few friends in the industry to verify these trends and they said they can see poly hitting $35/kg, maybe $30/kg, but at the same time they are not convinced the Chinese can produce volumes sufficient to drive cost much lower. That said, one contact noted that GCL appears to be standing fast by their $20/kg target. Industry contacts have mentioned that that GCL is still buying higher grade to mix with their own to get quality up. Interestingly, GCL statements run contrary to that data point from my contacts. (As a side note, GCL and Canadian Solar announced a $77 million JV to build a solar wafer plant in Suzhou earlier today).

Call this, unscientific sampling…. So, your mileage may vary. Last week a friend heavily involved in the solar industry shared a story from one of their installer contacts. He had this to say:

"In recently provided post meeting feedback from YingLi, the company was pushing $1.50/watt pricing for June & September quarters, then upping their ASP to $1.60/watt in December. YingLi is reported as having explained that they expect demand to pick up and that they do not want to repeat pricing mistakes from 2009. One would assume the company is aware that any reset of guidance to investors will have implications in customer discussions, so it appears they are coordinating their communications - which is actually pretty sophisticated and bears factoring into the lenses thru which analysis is generated. "

Hmmm… That brings us to the issue in play: Is YingLi, and for that matter, all the other companies, right? Right now it appears as though the installers are willing to take the other side of that bet - which is worth noting as they are directly touching the market whereas YingLi and others are not. The games are just beginning. It's going to be very interesting to see who ends up on the right side.

Friday, May 27, 2011

Solar Module Inventories

As the module companies began to report the first looks at financial statements began to show what we all knew - the shutdown of Italy as it worked out its FIT issues backed up inventory, jamming warehouses, ships and manufacturing site parking lots with modules...

None of which, by the way, was evident in the analyst commentaries, including those who had been in China as recently as March & April.

Record Module Inventory Awaiting Installation As Prices Continue To Fall

"Global PV module inventory levels have reached over 10 GW in the second quarter of 2011, the highest amount recorded to date, according to the latest analysis from IMS Research."

One of my friends that works for a module maker began inquiring with sell-siders to find out what their channel checks had revealed. What he found was that several big name teams had not bothered to peform a check (stunning in light of the inverter companies pre-announcing) and had not heard this story line forming.

Anyway, most have gotten beyond that and are coming up with a real range of about 4GW to 6GW. The slow install period in Germany and the likelihood of a modest FIT revison come June 30th, should spur some demand in 2H11. Will it be enough? Not by themselves...and not if the Euro should move in a bad direction....

Friday, May 20, 2011

What Q1 Says About Semiconductor Capital Spending

With downgrades gripping the semiconductor equipment sector I thought I would share a bit of history about capital spending budgets. Again, I'm going to present some observations from the guru over at IC Insights, Bill McClean.

In this table we see how reliable Q1 announcements are in predicting the total amount and direction of capital spent for the year:

The amount of capital spending announced in Q1 seems to always set the direction for the year. The chart also shows that these Q1 numbers always under-estimate the final tally (up or down).

I believe herd psychology plays a big role in this. A few folks have described the cycles like this: "You tell me things are getting better (or worse). I tell someone else. They tell someone else. They tell someone else. Before you know it, things get better (or worse)." There's certainly more to it than just word-of-mouth. The march along Moore's Law is, obviously, the primary driver.

Could this year be different? Maybe. The Japan disaster is wreaking havoc in many areas of the semiconductor supply chain. There's also some concern regarding final demand. Companies are openly talking about these items. Novellus Systems cited order pushouts during their April 27 earnings call. The company expects bookings to decline 10% to 25% in Q2. Lam Research guided their revenues to $745 million (+ or - $20 million), down from $813 million, on April 20.

In addition to the problems in Japan, you have the seasonal patterns that grip the sector every year. Q2 is typically one of the weakest of the year for the capital equipment companies. Bookings have traditionally peaked late in Q1, or early in Q2, and then bottom sometime in Q3. Semiconductor companies tend to get pretty aggressive with orders at the end of each year - busting the budget is one way to describe it.

Next week investors get to hear the story from Applied Materials - the gorilla of the industry. A weaker than expected view from Applied would certainly set the stage for a year that ends up outside of the historical pattern. It would also serve to validate those expressing bearish views.

Of course, quarterly perturbations are what they are…. Just noise…. Which reminds me, I need to make a post describing the bull case.

Don't touch that dial!

Tail Risk in the Semiconductor Business

For some bizarre reason the semiconductor business is drawn to areas that are prone to earthquakes.

Why do they do this? It's a question that has been asked time and time again but no one seems to have an answer.

About a month after the Japanese earthquake/tsunami Bill McClean of IC Insights, (one of my favorite and most accurate semiconductor industry analysts) released a research bulletin that summarized IC (integrated Circuit) manufacturing capacity located in seismically risky areas. What's staggering about this is that 90% of the world's pure play Foundry business, which includes companies like Taiwan Semiconductor and United Microelectronics, have their manufacturing capacity situated in risky areas.

From the report: "The bottom line was that for every month the IC fabs in Hsinchu (home of several major foundries) were shut down, the electronic system industry would lose at least $10 billion in sales!"

The Foundry business is not alone. A great percentage of the semiconductor industry is located in high-risk regions:

Here are a few more comments from the bulletin:

"...In 2010, almost two-thirds of worldwide IC industry capacity was located in seismically active areas."

"Ultimately, all that really can be said about the ability to predict devastating natural disasters is that everything is just “fine” until one day it isn’t. However, while these tragic events are impossible to predict, they are not impossible to plan for."

The shocks we have experienced from the leverage in the financial world have been described as "Tail Risk" . Events that you think couldn't possibly happen, happened. Obviously the financial markets are not alone. I realize I am stretching the definition a bit but it seems to me members of the electronics food chain have embedded their own tail risk.

Even if you disagree with me calling this tail risk, you have to admit that things like this speak volumes about human nature….. Will we ever learn?

Thursday, May 19, 2011

The Intel Treadmill

It’s been a busy week for Intel shareholders. As noted by Eric Savitz this morning, Goldman Sachs semiconductor analyst Jim Covello issued a sell recommendation - citing the usual concerns of overcapacity, tablet cannabilization and, of course, penetration in their primary market by ARM processors.

If you have been following Intel for any length of time you know that these are not new revelations.

The sell recommendation was also accompanied by a negative view of two semiconductor capital equipment companies, Applied Materials and KLA-Tencor. Lam Research, on the other hand, received an upgrade because of their exposure to the memory business.

Not to be left out, CNBC's Fast Money group (yeah, that's what it is) chimed in with their views:


Fast Money it is….

All of this chatter comes on the heels of yesterday's Investor Meeting. For anyone interested in Intel I would strongly suggest a review of the materials presented . You can find them here (a painless, free registration may be required ):


When scrolling through Stacy Smith's presentation you will see, starting at slide #26, their assessment of the market. Noteworthy, and management has mentioned on their last few earnings calls, is the growth they have been seeing in the emerging markets. They've even gone as far to mention persistent under-estimation of that growth by the analytical community. Be on the lookout for even more surprises on that front.

So, is Goldman Sachs right and is the industry heading for a downward? During this quarter's confessionals pushouts and cancellations were mentioned by a few of the capital equipment companies.

Some of this can be attributed to the disaster that took place in Japan. Overall though, these dislocations were noted as short term problems. The equipment companies I speak with are leaning decidedly toward the bullish side of the ledger after these short term issues are resolved.

Having followed this sector for many years I don't find it surprising to see and hear all this mid-quarter noise. We're in that window between earnings reports and a lot of the analytical banter is pre-positioning for the next leg up. Given that semi-related shares have been performing well a breather is to be expected. For opportunistic types, any weakness that comes from this "noise" should be welcomed with open arms.