AXT Inc (AXTI) 2006 Q4 法說會逐字稿

Good afternoon. My name is Lindsay, and I will be your conference operator today. At this time, I would like to welcome everyone to the AXT fourth quarter and fiscal 2006 earnings release conference call. All lines have been placed on mute to prevent any background noise. After the speakers' remarks, there will be a question-and-answer period. [OPERATOR INSTRUCTIONS] It is now my pleasure to turn the floor over to your host, Dr. Phil Yin, CEO. Sir, you may begin your conference.

Good afternoon, everyone.

Welcome to AXT's fourth quarter and fiscal 2006 conference call. I am Phil Yin, Chief Executive Officer and I would like to thank you for taking the time to be with us. With me is Wilson Cheung, our Chief Financial Officer.

Wilson will take you through a detailed financial overview of our fourth quarter and fiscal year results, and then I will give you my perspective of our markets and future opportunities. Following the conclusion of my comments, Wilson will provide forward-looking guidance and then we'll open up the call to your questions.

Wilson?

Thanks, Phil.

Before we begin, I would like to remind you that during the course of this conference call, including comments made in response to your questions, we will provide projections or make other forward-looking statements regarding, among other things, the future financial performance of the company and our ability to maintain profitability, control costs and improve efficiency, improvements in our manufacturing costs, improvements in our competitive position and our technology development, the impact of customer qualification of our products, new opportunities for our China joint ventures, improvements in our production processes, product quality and yields, cost and supply of raw materials, the impact of technology developments providing new markets for gallium arsenide and germanium substrates, as well as market conditions and trends. We wish to caution you that such statements deal with future events are based upon management's current expectations and are subject to risks an uncertainties that could cause actual events or results to differ materially.

In addition to the factors that may be discussed in this call, we refer you to the company's periodic reports filed with the Securities and Exchange Commission, and available online by link from our website for additional information on risk factors that could cause actual results to differ materially from our current expectations. This conference call will be available on our website at AXT.com through February 28th, 2008.

Also before we begin, I want to note that shortly following the close of market today, we issued a press release reporting financial results for the fourth quarter and fiscal year 2006. This press release can be accessed from the investor relations section of AXT's website at AXT.com.

Now, turning to our potential results. I'm extremely pleased to report that we had a great fourth quarter, exceeding our gross margin expectations and reaching profitability. Revenue for the fourth quarter of 2006 was $13.1 million, up 4.2% from $12.5 million in the third quarter of 2006.

Total gallium arsenide substrate revenue was $11.1 million for the fourth quarter 2006, up 4.7% from $10.6 million in the third quarter of 2006. Of that, six-diamond wafer sells were $4.9 million for the fourth quarter, consistent with six-inch diameter sales in the third quarter. Indium phosphide substrate revenue was $456,000 for the fourth quarter of 2006, compared with $340,000 in the third quarter. Germanium substrate revenue was $318,000, compared with $387,000 in the third quarter. Finally, sales of raw materials, primarily 99.99% pure gallium were $1.2 million in the fourth quarter of 2006, compared with $1.3 million in the third quarter.

In the fourth quarter of 2006, revenue from North America was 37%, Asia Pacific was 47% and Europe was 16% of total revenue. Two customers generated 23.5% of our revenue during the fourth quarter, while the top five customers generated 46.6% of our fourth quarter revenue. Gross margin was 38.2% of revenue for the fourth quarter of 2006. This included a benefit from the sale of approximately $730,000 in fully-reserved wafers which positively affected the quarterly gross margin by 5.6 percentage points. However, AXT also benefits from overall yield improvements in the fourth quarter as a result of growing longer ingots, better slicing and shorter cycle times. These improvements in yields contributed to a significant increase in gross margins.

By comparison, gross margins in the third quarter of 2006 was 27.7%. This included a benefit from the sales of approximately $802,000 in fully-reserved wafers, which positively affected the third quarter gross margin by 6.4 percentage points.

Selling, general and administrative expenses were $2.9 million for the fourth quarter, compared with $2.6 million for the third quarter of 2006. The third quarter SG&A was unusually low due to insurance carrier legal fee reimbursements. Research and development costs were $854,000 for the fourth quarter of 2006, compared with $392,000 for the third quarter. R&D costs included $424,000 of accrued severance for the retirement of our former Chief Technology Officer. Included in cost of revenue and operating expenses, was stock compensation expense of $173,000 for the fourth quarter of 2006. The amount of annual stock options we granted in October was less than the previous year so as to minimize the income statement impact on the FAS-123R.

Income from operations for the fourth quarter of 2006 was $1.2 million, compared with a loss from operations of $971,000 for the third quarter. Net interest and other income for the fourth quarter of 2006 was $1.1 million, compared with net interest and other income of $744,000 for the third quarter. Other income in the fourth quarter included a gain of $1.3 million on the sale of $474,000 Finisar shares compared to a gain of $650,000 on the sale of 300,000 Finisar shares in the third quarter. At the end of December, we no longer held shares of Finisar Corporation.

During the fourth quarter, the company recognized a net income tax benefit of $1 million compared with a net income tax benefit of $862,000 in the third quarter. Net income in the fourth quarter of 2006 was $3.4 million, up $0.13 per diluted share based on $25.5 million weighted average shares outstanding. This includes approximately $0.05 from the gain on the sale of Finisar stock and $0.04 from our net income tax benefit, both of which will not repeat in the first quarter of 2007. In the third quarter of 2006, we reported a net income of $639,000, or $0.02 per diluted share in the third quarter.

Let's now look at our cash and the balance sheet. Cash and cash equivalents with maturities of less than three months, short-term investments and other investments in high-grade debt securities with maturities of less than two years, including restricted deposits, were $42.7 million at December 31st, 2006, compared with $23.1 million at September 30th, 2006. Our December follow-on stock offering added $24.1 million to the cash balance, which is net of underwriting expenses and prior to the exercise of the green shoe, which occurred in January.

Our cash used in operations for the fourth quarter was approximately $2.5 million. Accounts receivable net of reserves were $9.6 million at December 31st, 2006, compared with $8.8 million at September 30th, 2006. Day sales outstanding was at 68 days for the fourth quarter, compared with 65 days in the prior quarter. Net inventory increased to $20.3 million at December 31st, 2006, from $17.4 million at September 30th, 2006. This increase in net inventory is the result of growing demand from our tier one customers. Depreciation in the fourth quarter decreased to $260,000 as many pieces of our manufacturing equipment reached the end of their accounting usable lives. At December 31st, 2006, the company, including our consolidated joint ventures, had 1,022 total employees, of whom 819 worked in production.

And now turning to the results for fiscal year 2006. Total revenue for fiscal 2006 was $44.4 million. This was an increase of $17.9 million or 67.5% over $26.5 million for fiscal 2005. Gross margin for fiscal 2006 was 28.7% compared with 8.3% in the prior year. We are very pleased to report that we were profitable for the year, with GAAP net income for fiscal '06 at $944,000 or $0.03 per diluted share. This compared with a GAAP net loss of $12.2 million or a loss of $0.54 per diluted share for fiscal 2005.

This concludes our financial review. Let me now turn the call back now to Phil.

Thank you, Wilson.

2006 was a great year for AXT. The strong market conditions, positive customer reception for our product, and solid company execution of our plans, revenues in 2006 increased by 67.5%, and a net increase by $0.57 per share over the prior year.

During the fourth quarter, we were able to achieve tremendous yield improvements and cost reductions in nearly 12 areas of manufacturing. Specifically, in success in growing longer ingots and shorter cycle times in crystal growing, and less material losses during slicing. These yield improvements and cost reduction programs helped to drive our gross margins to 38.2% in the fourth quarter and allow us to achieve profitability sooner than we had expected. The last time the company posted sustainable profit was in the year 2000.

As we entered 2007, we are energized by our achievements if product quality, yield improvements and sales penetration, and we are optimistic about our prospects with increasing demand in each one of three market segments. Our most important achievement in 2006 was the improvement in sales penetration in the gallium arsenide market. Having resolved the quality issue, we were able to qualify a number of customers that have either reduced or eliminated shipment to AXT. In both the semi-insulating and semi-conducting markets, we exited 2006 with market share in the mid-teens, up from the low single digits in 2005 and we are qualifying our products at nearly every strategically important company in our market, opening up the door for further market share gains in the coming year. Among the qualifications that will have a positive impact on our results in 2007 will be the qualification of our substrates in a [bi-fed] process, used to reduce costs of cell phone applications. We are seeing an industry shift towards this technology, which combines an HPT and a PM transistor on one chip.

We're pleased to report that we are qualifying our substrates of tier one device manufacturers, either directly or through [inaudible] foundry companies. Because of these qualifications, our 16 sales in the fourth quarter of 2006 were comparable to the third quarter and we expect to see similar results in our six-inch substrate sales in the first quarter of 2007. However, we expect to complete his qualifications in the first half of the year and the to resume growth in our six-inch substrate thereafter.

We are very well-positioned to benefit from this transition in the second half because of our unique ability to add capacity cost-effectively and because of our low-cost manufacturing capabilities. Our ability to scale capacity cost-effectively aided in the sales penetration in 2006. Gallium arsenide [inaudible] utilization during 2006 was about 90%, with demand continuing to grow. The strongest demand continues to come from the larger diameter semi-insulating substrates, such as six-inch substrates that are expected to grow at a market average annual growth rate of 24%, through 2010. In fact, six-inch substrates now make up greater than 50% of the total merchant market and it's projected to grow 70% at several market leading companies such as RFNP, IQE, Sony, and [inaudible] completed or in the process of converting to six-inch diameter wafers.

Prompted by strong demand and increasing constraint, we undertook two expansions of our six-inch, semi-insulating substrate capacity in 2006. During the fourth quarter, we completed an increase in our capacity for six-inch substrates by 50%, and we are currently in the process of an additional 40% increase to our six-inch capacity to be available in the first half of this year.

Our ability to undertake these expansions with minimum CapEx requirements is very unique to AXT. Unlike our competitors, AXT builds its own crystal-growing furnaces, using a highly-proprietary design. These furnaces cost far less than if they would if purchased commercially. Further, we own the largest compound semiconductor manufacturing facility in the world, with plenty of room for expansion with much of the infrastructure in place. And we manufacture in China, where labor costs are substantially lower than our largest competitors. These advantages allow us to expand our manufacturing in a matter of months and for a significantly lower cost that our competitors. Therefore, we have been able to meet the increasing demand for our products and benefit from a very constrained market.

With the gallium arsenide market estimated to grow from $2.5 billion in 2005, to approximately $4.5 billion by 2010, the strongest demand continues to come from the gallium arsenide wafers used semi-insulating applications, such as power amplifiers and switches for the mobile handset market. The mobile handset market is a key market for us and according to Strategy Analytics, it's estimated to grow from 986 million units in 2006 to an estimated 1.29 billion by 2011, with the substrate portion of it growing from 113 million in 2005, to an estimated 244 million by 2010. Further, the content of gallium arsenide in each handset is also increasing, because new products, like the iPhone. offer more functionality, such as Internet access, music downloads and enhanced digital camera capabilities.

On the semiconducting side, where our substrates are used in applications such as L.EDs, the merchant market is estimated to grow at 13% of $99 million in 2005 to an estimated $179 million by 2010, with a move to large diameter substrates now progressing quite rapidly. The end market for LEDs was $5.2 billion in 2006 and is projected to grow to $8.8 billion by 2011, with approximately 51% gallium arsenide base for red and 39% nitride-based for the green and the blue. Ultimately the driver in the LED market is the move to replace traditional incandescent and fluorescent light bulbs with those made with L.EDs. LEDs are already gaining in popularity in traffic signals and automotive and electronic billboards. We expect the application for their use will continue to grow. In fact many high-end homes are now being fitted with LED lighting.

Now, turning to germanium, this continues to be an interesting part of our business. While our revenue in this area is small comparatively, it experienced significant growth in 2006. We are pleased to report that in the fourth quarter, we received very encouraging results on the qualification of our concentrated portable tag, multi-junction solar cells for directional applications at a major North American supplier. We are also anticipating qualifications for space and commercial applications for the same supplier. We have begun working with this customer to further enhance our technology and process, and we expect that this partnership will lead volume production in 2007.

In addition, we also have been asked to supply germanium substrate samples to several European portable tag suppliers. We believe that the great efficiency of CPV triple-junction solar cells together with the shortage of polysilicon, are driving germanium substrates to become the material of choice for the portable tag applications. Based on strategy and analytic study, the total germanium substrate demand from portable tag markets will grow from 32 million in 2005, to approximately 126 million in 2010. Of this, the terrestrial portion is expected to grow from 10 million in 2006 to an estimated 68 million by 2010, assuming only a conservative 1% gallium arsenide germanium triple junction CPV penetration in 2010. In addition to terrestrial applications, space-based germanium solar cells is another driver, with a substrate growing from 49 million in 2005, to an estimated 58 million in 2010, based on satellite installations from 178 to 200, respectively. Thus, the total market for solar cells using germanium substrates will be approximately 126 million by 2010.

With our own raw material source for germanium, as well as our proprietary crystal growth process, we are poised to become a significant substrate provider to the portable tag industry. Despite its light revenue growth in Q4 for indium phosphide, we do not see any major increases going forward, as the industry has not returned to the good old days of fiberoptics. Thus, when you look at the total available market for substrates for gallium arsenide and germanium, it will grow from 290 million in 2005, to an estimated 580 million in 2010.

Another important area of focus for us in 2006 was our raw materials business. In January of 2006, we formally introduced our product offering in order to further differentiate our value and competitive advantage in the market.

Today, we have five joint ventures for which we offer raw materials such as gallium, four-ninths, six-ninths, and seven-ninths pure, arsenic, four-ninths, six-ninths pure, germanium, and pyrolytic boron nitride components, crucibles, and other related MBE parts. In addition to providing a strong competitive advantage, they also provide us a key strategic benefit because they allow us to have more control over our own raw materials costs and supplies, especially as demand has increased dramatically over the past year. These joint ventures are providing critical materials at predetermined costs and ample supply. To address these increasing demands, every one of our joint ventures has completed expansion capacity projects. The capacity expansions range from a 25% increase to more than 100% increase. With these expansions now in place in 2007, we expect to expand our raw material sales efforts and explore other new investment opportunities as this part of our business becomes an increasingly important strategic differentiator.

In closing, when I joined AXT in March of 2005, I pledged that AXT would once again be recognized as one of the leading suppliers of compound semi-conductor substrates. In the last 19 months, AXT has made tremendous progress towards this goal and now with six quarters of sequential growth in our revenues, market share gains in every area of our business, positive gross margin progress and a return to profitability. As we enter 2007, our opportunities appear brighter than ever. Our sales momentum is strong and our markets continue to show solid growth potential. We are poised for a great year in 2007, and we look forward to reporting to you on our progress.

Finally, before I turn this call back to Wilson, I would like to thank the employees of AXT Worldwide. Our success with not have been possible without their efforts, their motivation and their dedication to excellence.

I will now turn the call over to Wilson to discuss our forward-looking guidance. Wilson?

Thanks, Phil.

We estimate our revenue for the first quarter will be between $13.1 million and $13.6 million. Also, we estimate that our net income per diluted share will be between zero and $0.03, which takes into account stock option expense of approximately $200,000 and our increased diluted weighted average share count of approximately 31.5 million shares. This concludes our prepared comments. We are now happy to answer questions.

Thank you. [OPERATOR INSTRUCTIONS] We will pause for just a moment to compile the Q&A roster. Your first question is coming from Dave Kang from Roth Capital. You may go ahead.

Good afternoon, gentlemen. The first question is regarding your gross margin. How much inventory reserve do you have and what are you assuming as far as the first quarter inventory reserves being used?

Hi, Dave.

In terms of guidance.

It's Wilson here. We still have about $9 million worth of fully-reserved wafers sitting in our inventory, and as far as projections are concerned, in Q1, we expect somewhere between a quarter million to no more than half a million dollars for sales for the first quarter.

Got it. Now, all of the $9 million, I'm assuming not all of it are usable, correct?

That is correct. I mean, as we progress later into the year or even further into 2008, or 2009, we should expect the sale of fully-reserved wafers to gradually decline.

Got it. Now that you are quite profitable, how much NOL do you have?

From the federal perspective, we still have about $120 million.

Oh, okay. So it's going to be a while then.

Right. Well, you should still expect to see some income tax provisions in our P&L, primarily for our China operations.

Got it. And then regarding your proceeds from December, have you guys decided -- I guess you guys were looking at two options. One is to make an acquisition or increase stakes in JVs. Have you determined which way to go on this topic?

Dave, this is Phil.

Hi, Phil.

Not really, we are exploring all of these opportunities basically. It's an ongoing thing.

Okay. One thing, if you can talk specifically regarding customers. For example, I think you mentioned that at our conference, you expect to ship to Skywork fairly soon. Did you mean March or June quarter?

I think I mentioned this before during the conference, that the materials are in burn-in, live testing. They tell us they expect that to be completed sometime in the first half, second quarter, basically. So we're looking probably first half probably volume productions. Once that's obviously -- if it's successful.

And then will that be enough to start to show more meaningful sequential growth going forward, because obviously in the December and the March quarter, things are almost flattish?

Yes, well, I mean, it's a new customer to us, right? So it's added revenue. So obviously, yes.

Okay. But then, don't your parts get in to Skyworks through some epi players.

Yes, I mean, this is a direct qualification. You are absolutely right. We have epi-foundry people that are also qualifying our substrates at Skyworks and other customers also.

Got it. And then at our conference, you mentioned that you guys are being more selective. Was that another country manufacturer as far as revenues coming into -- coming in at the low end? Obviously you are focusing more on the bottom line now. How much more could you have done more if you weren't so choosy focused on the bottom line? I guess that's my question.

I think from a revenue standpoint, yes, we could have done more because we did refuse some orders because of low pricing. We didn't want them to affect our margins. You are right. Revenues could have gone up, but, again, it's a different story now. We are in the position of strength, we can -- yes, we just want to keep -- maintain our margins, that's the bottom line.

Sure. And lastly, as far as gross margin, the great numbers, very strong improvements sequentially, if you strip out that inventory reserve, it's still well over 32%. So I guess the question is: Where do you go from here? I mean, what would be your long-term gross margin target?

Well, definitely one thing for sure is that we still have a lot of cost cutting programs going on in place, so like I mentioned in the Roth conference there, that there's still room for improvement in terms of manufacturing efficiencies, but a lot of the leverages that I mentioned earlier about the gross margins are still going to be valid. For example, the better product mix that I mentioned, the depreciation expense savings because a lot has reached its accounting useful lives -- these are all permanent savings that will impact 2007 gross margins. And I would presume that going forward in the next quarter you should expect even with the lower sales of fully-reserved wafers, the gross margins will be sustainable in the low 30% range.

Wilson, just quickly on stock comp, can you break out between the cost versus OpEx? I know it wasn't that much but --

Yes. Basically I was at $173,000. About $16,000 is in the cost of revenues.

Okay.

About $31,000 is in R&D and the rest, $126,000 is in SG&A.

Got it. Thank you.

Hey, Dave?

Yes?

Just to add a little bit on the the qualification, things like that, as you know, the industry is switching to bifed technology. We're in the middle of qualifying a lot of our six-inch substrates, either directly or indirectly to epi foundries in that process. So, that's obviously going to be a positive because it's not going to be -- the bifed switch is not going to be a negative effect on volume. In fact, I think it's going to drive volumes even up, because that's going to reduce the cost of the handsets. They will be able to sell more phones to the emerging markets in the developing countries. So, we see that as a positive thing and it should drive gallium arsenide substrates even further.

Got it. Now, Wilson, just quickly. On cash flow, what was that? I didn't get that number.

The cash used in operations was $2.5 million.

Got it. Thank you.

Thanks, Dave.

Thank you. [OPERATOR INSTRUCTIONS] Your next question is coming from Pierre Maccagno. You may go ahead.

Congratulations on the quarter Phil and Wilson.

Thank you.

Could you qualify a little bit better the opportunity at Skyworks. If I understand, it's mostly for the switch opportunity, correct? And how big of a revenue run rate would you have with them once you qualify.

Pierre, those are hard things to tell, right. I mean, once you qualify you know they will not put you on at thousands and thousands of wafers per month, right? They're going to go slow. Probably start out at 500 or 1,000 wafers per month and then ramp up to see consistency of the yields, quality of the materials, everything -- on-time shipments. You have to remember that this was a customer that went to zero, right? They let us back in the door. So it's all going to depend on the consistency of the product and from the pricing standpoint, obviously, we're competitive, or else they wouldn't have even qualified us. So we do see, or we anticipate the revenue to grow, with this customer. To how far, or how much, that remains to be seen.

But we are not anticipating any slowdown in terms of manufacturing and we are still keeping the 40% expansion.

Oh, yes.

And how is the -- is there any qualification process going on RSMD or are they still waiting to see?

We are still dialoguing with them. We had a conference call with them. I know they had an internal meeting in regards to the AXT. I think they are listening as we speak right now because they say they listen to all of our earnings calls. We anticipate some action, I think before the end of the year is out, definitely.

Okay. And pricing in general, how are we seeing that dropping or -- ?

You know, I think six-inch is stable. Four-inch is stable. There's always pressure to reduce cost, as you know, in every industry or any type of product and with the material components being such a big portion of the process see all -- all the suppliers are going to get pressured to reduce pricing. I think the good news is that six-inch is constrained. We're the only ones that are increasing capacity. We saw that four-inch was beginning to constrain. We saw a little bit of that in the fourth quarter of last year.

So I guess it's hard to really give you a definitive answer, Pierre, but from my standpoint, I hope it stays flat and doesn't go down any further. But if it does, obviously, we are very competitive because of where we manufacture in China with the low cost, so I think we're going to be very competitive and still maintain our margins.

Do you consider yourself the price leader now?

We're not the price leader. We were, but not anymore. In fact, we don't want to be associated with being the price leader.

Okay. Thanks very much.

Thanks, Pierre.

Thank you. [OPERATOR INSTRUCTIONS] Your next question is a follow-up question from Dave Kang. You may go ahead.

Yes, regarding OpEx, or more specifically on SG&A, I guess you had a one-time gain. So what will be the real SG&A going forward? And can you just talk about the leverage there?

Sure. I -- as I mentioned in my previous presentations, I think that the infrastructure that we have right now is -- we are able to support a revenue framework of between $15 to $20 million. So excluding some of those one-time events, I think if you look at the next few quarters we should be expecting somewhere between $3 to $3.5 million in SG&A going forward. Even as we get into -- when we reach about $18, $20 million, we are only expecting SG&A to be maybe between $4 to $4.5, but you won't see a tremendous amount of increase, because pretty much we have the fixed costs covered.

And can you just add a little bit more color in terms of your comments about germanium. I guess it's still lumpy, but it sounds like you could see some traction with the American customer going solar -- I mean, both a terrestrial and space, as well as some potential European customers.

Yes, well, with the way the energy is these days the -- obviously, especially with the triple junction right now -- I attended this alternative energy conference in the beginning of the year, and basically triple junction solar sales is going to grow way, way double digits. In fact, the demand for germanium from a terrestrial standpoint, from a square-inch standpoint is going to grow estimated over 200% of compounded annual growth rate.

The wafer market itself, from a dollars standpoint, is supposed to grow at about a 21% compound annual growth rate, all driven by the efficiency of triple junction germanium. And as you know with silicon, you are in the low mid-20s and with the triple junction, CPB concentrator portable tag solar cells, they are hitting 37%, 38%. In fact, I believe Spectra Lab announced 40.7% efficiency about a month, a month and a half ago.

And then the other -- I think the other metric to use is the local cost of energy, they call it. The local cost of energy in the U.S. is around $0.1358 per kilowatt hour. And the Department of Energy's goal for all of these projects that they are funding is around $0.08 to $0.10. The expected efficiency that they are looking at is really 50%, but they are also saying that when you hit 45%, that's when you become very, very competitive on the order of you will be able to hit energy levels around $0.13 to $0.15 per kilowatt hour. So, we're not that far off for the market to start exploding.

Okay. All right. I guess we'll just stay tuned for that.

Yes.

Thank you.

Thank you.

Thank you. Your next question is coming from John Evans from Wells Capital. You may go ahead.

My first question is, can you talk about R&D expense and kind of the number in the fourth quarter and how we should think about that going forward?

Sure. As I mentioned, John, the $854,000 R&D number did include a severance for the former CTO. So if you strip that out, we are in the mid-$400,000 range. I think going forward, for 2007, you should expect that to be relatively stable in the mid-$400,000 range, until we get to maybe $18 to $20 million revenue range. That's when you will start to see that starting to ramp up a little bit, maybe in the low to mid-$500,000.

Secondly, you talk about your cost advantage? And I guess, could you help us understand maybe better in the six-inch? You are the only one adding capacity. And is there any way that you can give us any kind of insight into how much of a cost advantage you have over competitors and maybe where you think wafer prices would have to move before other people would be able to add capacity profitably?

Okay. This is Phil.

Yes, there are probably four main factors that we're able to add capacity very, very cost effectively and very, very timely. The first one is that -- and I mentioned this in my statement, is AXT is the only compound semi-conductor manufacturer that designs and builds its own crystal growing furnaces, VGF furnaces. Competitors have to purchase furnaces on the outside market.

For instance when we did our 50% expansion and the 40% expansion that we are doing now, we literally added -- and I want to add the word plural, hundreds of furnaces. This particular crystal growing technology, is not like silicon, where you're growing ingots or we call them [Buhls], that are, let's say, five feet long and you're getting thousands of wafers per ingot. In this particular situation, because it's a compound and it's not an element, the crystal growth time is very, very long. The ingots are very short. So the wafers per unit time for each ingot unit are very, very small. That's one.

Two, again, a commercially-available furnace, let's say a six-inch, costs on the order of about EUR500,000. So, that's about $600,000 per furnace. As I mentioned, we build our own. We are literally orders of magnitude less, right?

Now, if you are going to add capacity because of the long cycle time, because of the short ingots, you don't add one furnace, because that's only going to give you 40, 50, 100 wafers per ingot. So you literally have to add multiples and multiples of furnaces, such as 25, 50 or 100 furnaces, to make any kind of difference in your capacity increases. So, I mean, do the math, right. At $600,000 a pop, it's quite a chunk of money. Now -- and most importantly, I think it's the return on capital.

Right.

All right? Return on capital because you have to go to the board to get approval to do this. So if you do a quick back-of-the-envelope calculation -- let's say if you take the length of time to grow an ingot, the number of slices you get per ingot, you take the average selling price of that particular wafer, you take what it costs you to build one furnace, and what it costs your competitor to buy a furnace. The return on capital for AXT would be approximately two, maybe two and a half months. The return on capital for our competitors if they have to buy the furnace on the outside would be approximately two years.

Right.

So that's one of the reasons. Now there are three other reasons.

The second reason is we manufacture in China. So our labor there is less than 10%. When we were in California, it was around 30%. Our competitors are in Japan or Germany. So, there's -- that's a no-brainer. Okay? Third, again, as I mentioned, we are the largest compound semi-conductor facility in the world. We have 190,000 square feet. 90,000 of it has all the infrastructure in there. That's where we are able to add the 50% and the 40% expansion in very, very little CapEx and timing.

And fourth, but not least, we have our raw materials supply. We are completely vertically integrated. All of our five joint ventures supply all of our raw materials and I don't know if you are aware, but there's a big constraint right now in gallium and a big constraint in arsenic. In fact, gallium, four-ninths gallium has increased from the low-$300s to the mid-$400 per kilos and arsenic is expected to go up 30% in pricing. So those four reasons are one of the reasons we can add capacity and our competitors cannot add capacity that we know of.

So --

Did I answer your question?

Very -- very well, thank you. Can you just talk then -- I mean, so if we think about six-inch pricing, I know your customers may be on the call, but with all of those backdrops, why wouldn't pricing start to move up?

Why didn't the pricing move up?

Yes.

It's why didn't silicon prices move up, right? I mean, competitors, your customers, they are making devices. Materials is a big, big part, a big percentage of their costs. The first guys they start to beat down on is material suppliers. I mean, right now, I guess my comment is I think the prices are going to stay flat the way they are.

Okay. Thank you for the insight.

Thank you.

Thank you. [OPERATOR INSTRUCTIONS] You have a follow-up question from Pierre, Maccagno. You may go ahead.

Yes. Can you tell us the CapEx for the quarter and also what CapEx you expect for '07?

The CapEx for Q4 is a little over $2 million, and we expect the entire year to be around $5 to $5.5 million.

Okay. Thanks.

There appears to be no further questions.

Okay. Well, thank you, everybody for participating in our conference call. And we look forward to seeing you again in the future. Thanks for very much. Bye-bye now.

Thank you. This concludes today's conference call. You may now disconnect. Have a great afternoon.