芯源系統 (MPWR) 2006 Q3 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Monolithic Power Systems 2006 third quarter earnings conference call. My name is Latitia and I will be your coordinator for today.

[OPERATOR INSTRUCTIONS]

As a reminder, this conference is being recorded for replay purposes. At this time, I will now turn the presentation over to Mr. Rick Neely, Chief Financial Officer. Please proceed, sir.

Good morning, and welcome to the third quarter Monolithic Power Systems conference call. Michael Hsing, CEO and Founder of MPS, is with me on today's call. In the course of today's conference call, we will make forward-looking statements and projections that involve risks and uncertainties.

For example, our business outlook, including our business and financial outlook for the fourth quarter of 2006, projected fourth quarter net revenues and gross margins, our expectations for fourth quarter litigation and non-GAAP operating expenses, our target operating model range for gross margins and operating expenses, our activities and plans for production in our China test facility, our planned new product introductions and the various opportunities these present, our process development, design activities and relative competitive position, expected growth or declines in our product lines and geographic markets, anticipated outcomes of our pending litigations.

Forward-looking statements are not historical factors or guarantees of future performance or events and are based on current expectations, estimates, beliefs, assumptions, goals and objectives and involve known and unknown risks, uncertainties and other factors that may cause actual results to be materially different from the results expressed or implied by these statements.

Risks, uncertainties and other factors that could cause actual results to differ are identified in our SEC filings, including, but not limited to, our second-quarter 2006 Form 10-Q, filed on August 4, 2006, which is accessible through our website, and our third-quarter 2006 Form 10-Q, to be filed on or before November 9th, 2006.

Also, please note that during this call we will discuss net income and operating expense on both a GAAP and a non-GAAP basis. The non-GAAP financial measures exclude charges related to stock-based compensation and one-time litigation settlements and their related tax effects.

A table that outlines the differences between the non-GAAP financial measures to GAAP financial measures is included in our earnings release, which we have filed with the SEC and posted on our website at www.monolithicpower.com.

I'd also like to remind you that today's conference call is being webcast live over the Internet and will be available for replay on our website for one year. We would like to start this call by reviewing our third-quarter business highlights.

Following the business update, I will discuss our financial results. We will conclude by discussing our current expectations for the fourth fiscal quarter of 2006. We will then open up the call to your questions.

The business highlights start with third-quarter results for MPS that showed consistent execution as we reached revenues of $27.3 million. We had record gross margin of 65.6%, introduced our first high-current product to the marketplace on the BCD Plus technology and added $9.5 million to our cash and cash equivalents balances. The overall analog market environment for the third quarter was not robust, but MPS exceeded its revenue targets.

Our sales into the DC-to-DC segment met expectations at $18.8 million, as did our revenues CCFL products, which came in at $7.4 million. Our third-generation process technology, BCD Plus, is demonstrating excellent early results. We have demonstrated current-handling capabilities of up to 10 amps at 28 volts for new BCD Plus products. Michael Hsing, our CEO, has named this product family the MiniMonsters. We're about to release the first two products in this family, the MP 8666 and the MP 8668. These products provide small form factors and low-cost solutions for high-volume DC-to-DC applications that require high current and/or voltage handlings such as graphics cars, laser printers, large LCD TVs, PDP TVs, communication equipment and other similar markets. Other MiniMonsters in development target specific applications in the computing, networking and communication arenas.

MPS will soon be releasing our first four megahertz DC-to-DC step-down regulator that takes advantage of the low gate charge feature of our BCD Plus process. This product reduces our customer solution size by 70%, helping our customers achieve a very high operating frequency in the smallest footprint solution available.

This product highlights the integration capabilities of our BCD Plus technology as it integrates a 20-volt, 150-milliohm power switch, a four megahertz oscillator and a boost diode and loop compensation onto one chip. We believe a customer can use this product to build a complete solution with just seven external components, fewer components than using a non-MPS solution. This is a significant achievement in the general DC-to-DC market.

As we introduce more products with these technical advantages, we expect to open up new market applications and revenue opportunities for MPS. In the manufacturing arena, our Chengdu test operations hit full production in the third quarter, and, combined with a favorable product mix, enabled MPS to achieve a record gross margin. Tight inventory management and reduced capital spending drove the growth in cash flow, which increased cash and cash equivalents by $9.5 million in the quarter to $74 million as of September 30, 2006.

Finally, we reached a settlement agreement in our patent lawsuit with Micrel Corporation, resulting in a one-time charge of $3 million that enhanced our process technologies portfolio. Now let's look at the financials in more detail. On our profit and loss statement, as we mentioned, revenues for the third fiscal quarter were $27.3 million, a decrease of 8% from the third quarter of 2005.

This compares to net revenues of $26.6 million in he prior quarter, or an increase of 2% sequentially. Let me break down our third quarter revenue by product line. DC-to-DC product sales met our expectations, coming in at $18.8 million, up 2% from 18.5 million in the year-ago quarter. This was down slightly from our second quarter 2006 figure of 19.2 million.

Our core DC-to-DC products grew about 6% year-over-year, while the white LED driver segment of our DCD product area suffered from difficult market conditions and declined about 29% year-over-year. LCD backlight revenues came in slightly higher than our expectations at $7.4 million, a decline of 27% from the same quarter a year ago.

This result was consistent with our expectations of essentially flat revenues in this product area near term, as our second quarter 2006 backlight revenues were $7.0 million. Audio revenue made a nice recovery to prior run rates, coming in at $1 million, essentially flat to the 1.1 million recorded in the year-ago quarter and up from the $400,000 we sold in the second quarter of this year, as several new designs kicked in.

I would also like to mention that the higher-than-normal inventory levels in our distribution channel at the end of 2005 and the first quarter of 2006 have been resolved. Distributed inventory levels are in good shape. On the gross margin side, Q3 gross margin was 65.6%, compared to 64.3 in the third quarter of 2005 and 63% in the second quarter of 2006.

Our new test facility in Chengdu came up to full production in the quarter, so the small impact of gross margins in the first half of this year for that startup process are completed on schedule. This well-executed factory ramp, combined with a slightly favorable product mix, were factors that drove our gross margin to record levels.

Though we are proud of this achievement, we remind listeners that our long-term model is to operate between 58 and 63% gross margin and future results may not match our third-quarter results. Let's look at reported expenses. Our GAAP operating expenses came in at $18.5 million in the third quarter. This includes $2.8 million for patent litigation costs and $3 million for a one-time settlement for a total litigation expense of $5.8 million -- $12.8 million in R&D and SG&A expense, which includes $2.8 million for stock compensation expense under the new FAS-123R accounting rules. This result was consistent with our guidance at the beginning of the quarter. Compared with Q3 '05, GAAP operating expenses increased by $14.1 million.

This amount is comprised of actually a decrease in base litigation costs of $1.5 million, as we spent $2.8 million in Q3 '06, compared to about $4.3 million in Q3 '05. While the base litigation costs went down, we saw an increase in net settlement or legal judgment charges of $11.8 million from a year ago. We had recorded an $8.8 million legal judgment reversal in the third quarter of 2005, which, combined with the Micrel settlement charge of $2 million equals a total increase of 11.8 million year over year.

We also had an increase in R&D spending year-over-year of $1.9 million and an increase in SG&A spending of $2 million. Compared to the second quarter of 2006, our sequential GAAP operating expenses increased by $3.6 million.

Let's look at some more detail on the expense changes from a year ago. The major swing in litigation expense was due to the reversal of $8.8 million worth of the provision for a preliminary jury award in the O2Micro trade secret trial that was recorded in the third quarter of 2005. Base litigation costs, as I mentioned, actually dropped $1.5 million year over year due to reduced trial activity in our cases.

The increase in R&D spending reflects the company's continuing investment in growing our design teams and new product introductions. In addition, stock compensation expense increased significantly year over year due to the adoption of FAS-123R. Our SG&A expense growth year-over-year was split between headcount increases I the sales force and financial teams and in the increase in our stock compensation expense due to he adoption of FAS-123R in 2006.

Let's look at our expenses now on a non-GAAP basis. Excluding stock compensation and litigation costs, our non-GAAP operating expenses were $9.9 million in Q3 '06, compared to $7.9 million in Q3 '05 and up sequentially by $700,000 from the $9.2 million recorded in the second quarter of 2006. The $700,000 expense increase from the second quarter was primarily in R&D, which was up $485,000 and in sales and marketing, which grew by $255,000.

The increases in both categories reflect our major new product push in R&D and continued investment in building our sales channels. The third quarter from a bottom line aspect recorded a net loss. Our third quarter GAAP net loss was $1.9 million, for a loss of $0.06 per basic share, which includes a $1.8 million tax provision. This compares to net income of $9.4 million for the third quarter of 2005, or $0.31 per fully diluted share. The third quarter 2006 figure includes the impact of the recent implementation of FAS-123R, as MPS recorded a total stock-based compensation expense of $3 million for Q3 '06.

We had a tax provision of $1.8 million this quarter, despite a GAAP loss, due to our FAS-123R stock-based compensation expenses and its related tax effect. These expense are not deductible under tax rules until the options are actually exercised and an actual gain is made.

Our effective tax rate for GAAP purposes for the third quarter of 2006 was 81%, but adjusting for stock-based compensation and the timing of the deductibility, as well as the one-time legal settlement, our non-GAAP effective tax rate is approximately 36% year to date. Over time, we expect our tax rate will diminish, but for now that is the main cause of the high tax rate for the third quarter.

If you look at the net income on a bottom-line basis from a non-GAAP perspective, we actually recorded a net income of $3.8 million for the third quarter of 2006, or $0.12 per fully diluted share. This figure excludes total stock comp expense of $3 million, the one-time legal settlement expense of $3 million and their related tax effects of approximately $341,000.

In comparison for the third quarter of 2005, the non-GAAP net income was $4.7 million, or $0.15 per share, which excludes stock comp expense of $1.1 million, a legal provision gain of $8.7 million and a related tax effect of $2.9 million.

Now, let's look at some of the major changes to the balance sheet. Cash, cash equivalents, current restricted cash and investments were $74 million at the end of the quarter, up substantially from $64.5 million at the end of the second quarter of 2006 and up from $60.3 million a year ago. The increase in cash for the second quarter resulted from strong operating cash flow, reduced inventories and reduced capital equipment purchases for production as the Chengdu ramp has largely been completed.

Accounts receivable ended the quarter at $9.6 million, compared with $9.9 million at the end of Q2 '06 and $9.8 million at the end of the third quarter of 2005. Our day sales outstanding for receivables were 32 days at the end of the third quarter of 2006, a slight improvement compared with 34 days at the end of the second quarter and very close to the 30 days DSO for the end of the third quarter of 2005.

Our inventories ended the quarter at $7.4 million, or about 72 days of inventory. This compares with $9.1 million, or 84 days of inventory, at the end of the second quarter of 2006. As of the end of the third quarter of fiscal 2005, our inventories totaled $5.7 million, or about 49 days of inventory.

In the litigation area, there were some recent encouraging developments. On October 11th, the court approved O2Micro's dismissal of their lawsuit on the 129 patent against MPS's accused products. Finally, in the O2Micro case on the 722 patent, a trial date has been set in the Northern District Court of California for April 30th, 2007.

Let's turn now to our general business conditions. We see our business continuing to be fairly flat, with potential upside traction from our newly introduced products in the battery charger, high-voltage and high-current markets. We see positive developments in the LCD TV market and multiple new applications for our high-voltage product offerings, which go up to 40 volts, in the industrial and consumer segments.

Geographically, sales continue to grow in regions outside of Asia. Our European sales recorded another $1 million plus quarter and sales in Japan were 22% higher than the same quarter a year ago. We have seen our first $1 million-plus quarter from demand created in the US as we continued to make progress on major design wins and multiple first-tier US customers.

We expect to see the effect of these designs next year, as the penetration into these companies continues. In addition, we have started seeing good growth in our US distribution channel, establishing the acceptance of MPS's high-performance products. For the CCFL product line, we continue to project approximately flat sales for the fourth quarter of 2006 compared to the third quarter of 2006. We expect to introduce several leading-edge products in the fourth quarter for new applications and new markets. As we mentioned in the early part of the call, we will be introducing a high-speed, four megahertz 20-volt DC-to-DC step-down regulator for space-constrained applications.

Another key product introduction from the MiniMonsters family planned for the fourth quarter is our first six amp and eight amp step-down regulator in a small form factor, three millimeter times four millimeter QFN package, which achieves greater than 90% efficiency for high-volume markets such as graphics cards, gaming consoles, large flat-panel TVs, laser printers, communication and point of load requirements.

Based on recent customer visits, initial feedback is very positive on these cost-effective, space-efficient solutions and we expect these products to begin contributing to revenue in the second half of 2007. Now let's turn to our outlook for the fourth quarter of 2006. Our revenue guidance is in the range of 25 to $27 million. Gross margin is expected to be in the upper end of our target range of 58 to 63%. We expect stock-based compensation expense in the range of 2.7 to $3.0 million. We expect non-GAAP research and development and selling, general and administrative expense in the range of 10.5 to $11.5 million. This estimate excludes the stock compensation estimates mentioned above. Finally, we expect litigation expense in the range of 3.5 to $4 million as we prepare for the '02 micro trial at the end of April 2007.

In conclusion, we are pleased to report that MPS is continuing to develop new products and technologies to address different market segments and expand our customer base. We have grown our sales channels and customer base, increasing our revenues geographically.

Our US channel continues to increase at an accelerated level due to our high-performance space-saving technology. In addition, our European customer base is increasing, with various tier-one customers increasing their long qualification cycles and projected to ramp in 2007.

We are managing our cash and working capital well while maintaining R&D spending to effectively support new product development and design win activity. Our litigation issues diminished significantly in the quarter with the settlement of the Micrel lawsuit, and, finally, we continued to innovate with our proprietary BCD Plus process, as we planned to introduce the MiniMonsters, with high current handling capability and great cost-performance ratios in very small form factor packages. We expect product families like the MiniMonsters to enable the types of unique solutions that have made MPS what it is today.

Now, we would like to open up the microphone and take your questions. So, operator, if you can open up the question lines, that would be great.

And thank you for that presentation.

[OPERATOR INSTRUCTIONS]

And your first question comes from the line of Simona Jankowski with Goldman Sachs. Please proceed.

Hi, thank you. The first question I wanted to ask your guys is on your DC-to-DC core business being up 6% year-over-year. Can you just give us a little bit more color on what is causing the growth rate to slow down so significantly and how you expect that to roll out in the next couple of quarters?

Simona, thanks for the question. As you noted, DC-to-DC is growing, not very fast, as it was before, so -- but we have done a good job of penetrating the current applications for our products. We found in the last six months we need to expand our product lines. We don't have enough different products to keep moving, and we also need to go into new applications. Thus, most of our emphasis has been on new applications and I'll turn it over to Michael to explain a little bit more about that.

Yes, let's talk about the existing products as a way -- in the beginning of this year, I said we have many products designed in the LCD TV and I'll start to see the results, and I'll see even a better year in the next year for the LCD TV market.

The older product that we just [about] released, and I was very excited about the high-current products, these are also for LCD TV applications and -- but we address a much larger market, such as the graphics cards, laser printers and other type of high-current applications, such as communications.

Okay, no, that's very helpful. Just to maybe get a little bit more detail on maybe the components of what caused the growth rate to slow down and which of those components might reverse, so normally we'd think of either inventory having built like in '05, and some of that I think was a factor in the early half of '06. Additionally, were there any competitive pricing issues, and also lastly, any issues of whether competitive design wins displacing some of the volume at your customers or maybe just not having as much of a pipeline.

I mean, can you maybe give a little more color that way so we can get a sense of which of these factors might reverse out within the next few quarters?

Simona, were you referring to DC-to-DC again?

Yes, just for the DC to DC.

As you said, in the DC-to-DC arena, while there was some inventory build in the first part of the year, we've controlled those. Relatively on the penetration side, one of the things we've found is in the areas in the -- in the product areas where we have a lot of sales coverage, in Asia and greater China, we have pretty good penetration ratios, but the problem is we just don't have enough different applications. We have very broad-based products, but also the corollary to that is they go into a lot of different applications. And we've got a good penetration in the applications that are done in the greater China region, for example, but there just isn't enough of them. That's only a segment of the world.

So one of the things we're seeing is we're getting good growth in other parts of the world, but our sales force is still young and junior and small in those areas, so we're trying to pick up the growth there. I think the second thing that we reverse out, as you mention, the inventory things are clearing themselves out, but the place where we have a lot of price issues, and it's a small part of our DC-to-DC revenue, it's our white LEDs.

As we said before, it's 10 to 15% of our DC-to-DC number, and we've had significant pricing pressure there and there's other competitors of ours who have tremendous impact to their results because of the pricing in that arena. For us, it has a small impact, but it does show up in our DC-to-DC growth number.

Michael?

Yes, in general, we do see price pressures, and in the past we can maintain our gross margins, and at the same time we're facing the price pressures. To maintain our business models, what we have to do is for the existing product we keep finding the new product market and also we grow geographically, and at the same time we introduce new products.

Sure, and just last question. You continue to do a very, very nice job on the gross margin side. Can you just explain a little bit, how much elasticity is there on that line item in terms of what you be able to get into a broader range of design wins if you were to take a slightly lower margin that's more in the middle of your range, or is there any elasticity in the follow-on volumes that you'll sell more if you had received the design wins versus some of your competitors that you might be second sourced with.

Yes, the answer is that we -- our business model still have 58 to 63%, so for our gross margins. And we are looking at them since now we are above our model, we are looking at the opportunity that we can expand our revenues. So as can do [operation picketing], which more provide our cutting down the number of shipping days and also reduces some of the cost, but we now have more opportunities for the other products.

Yes, I think to summarize, also, the best way for us to use the gross margin leverage is on new designs. Once you have a design, we typically get a pretty good portion of the revenues, so lowering prices doesn't do much in existing designs. It's better to use it in new design wins.

Got it, that's very helpful. Thank you very much.

And your next question comes from the line of Tore Svanberg of Piper Jaffray. Please proceed.

Yes, thank you, and good afternoon. A couple of questions. First of al, could you talk a little bit about how much of your testing is now done by Chengdu versus what's still external?

All of our final testing is done by Chengdu, and in fact we were using some outside contractors and we moved that all to internal this quarter. We're still in the process of transferring the sort test, which is the one you do right after fab. We're wrapping that up in Chengdu this quarter and ramping it down in Los Gatos. So by the end of this year we won't have anything in Los Gatos except R&D and Chengdu will have all of our back-end testing, including sort.

Great, and you mentioned the MiniMonsters and also other BCD Plus products ramping second half of '07. Could you maybe give us a little more color what type of percentage these products could contribute in, let's say, as we exit the year?

Well, we can't -- since we targeted some of the applications which are of the very large market side, such as the graphics card, the laser printers and the flat-panel TVs, and it is since the market is very big, it is a cost for us to predict that. And if we have a -- you know what -- if we have one bigger design wins, then it has the potential to [hit] very big, but at this time we're not speculating anything.

Okay, and if you would go to guidance, you're expecting revenues to potentially decline in the fourth quarter, yet you're -- based on my math, your op expenses are going to be up sequentially. Could you elaborate a little bit on why we should see operating expenses increasing as much?

Yes, I'll cover the -- the operating expenses are going to grow sequentially a little bit. In the areas where we're growing them, we're still hiring new R&D people and we're still growing out our sales and marketing channels. One of the issues is the areas of the world, Japan, Europe, US, where we have a lot of potential but we only have handfuls of sales guys, we have to increase our sales channels expenditures, so we're doing that to invest so we can have the coverage on our new products.

And the R&D, we have so many products stacked up, we have to bring contractors in to do our layouts right now. I guess that's good news, but that costs a little bit more money. So we're going to spend the money to get the products out, and that's really -- the time to market is what we're focused on and we think that the small increases in R&D are justified.

Great, just two more questions. First of all, on legal, is the only outstanding issue at this point the trial with O2Micro in April?

Yes, that's the -- there are some smaller items, but that's basically it. There is a lawsuit open with Linear Technology, but it's very early stages. There isn't really anything going on there.

Great, and finally, your tax rate has been a little bit all over the place this year. What sort of rate should we assume in Q4 next year?

Yes, I'm shooting again for the mid 20s. I mean, we should definitely be -- the statutory rate, the non-GAAP statutory rate we're at now, which is 35, 36%, that should get down to the mid 20s. Again, because of the unpredictability of stock comp expense and how much is actually realized, I can have no way to predict that.

I don't know what our book rate will be, but our base rate on a non-GAAP basis without stock comp will be -- again, as you know, stock comp is a timing issue. Once it gets exercised, eventually you get the tax back, so the base rate will probably be in the mid 20s.

That's helpful. Thank you very much.

And your next question comes from the line of Eric Gomberg with Thomas Weisel Partners. Please proceed.

Thank you. Could you maybe comment a little bit about what you're seeing in terms of demand? You in the prepared remarks said that you believe that your inventory at distributors is now healthy. So it would imply that you're seeing perhaps something a little bit weaker than seasonal for Q4. What are customers out there telling you?

I mean, in terms of inventories, we look at our inventories relative to -- we project our Q4 demand versus inventories and they're at a much healthier level, meaning much lower, than they were earlier in the year. So we should be able to see, if there is any increases in seasonal demand, we would be able to see it flow through.

However, we don't see any particular drivers for that, and I'll ask Michael to comment.

Yes, okay, under inventory, we just adjust our inventory according to what the Q4 is. We don't see a particularly weak or strong Q4.

Okay, because you said CCFL should be flat sequentially, which implies that DC-to-DC is dropping a little, sequentially, and I'm just I guess a little surprised at that.

Well, s we'd mentioned, one of the segments, we had said that white LED was between 10 and 15% of DC, and in Q3 it was on the low end of that range, maybe 10%. It may -- as we continue to see the price pressure in that area, it may decline a little bit more in Q4. So overall DC is most of all flat, is really how it'd look in almost all of the markets.

Yes, the DC-to-DC is mostly flat from Q3 to Q4, with a few --

Few hundred K up and down. Whether CCFL is 7 million instead of 7.4, those fluctuations of a few hundred K.

Still we're looking at an [arrow], yes.

Okay, that's fair. Just on gross margin, obviously you did a fantastic job this quarter. Why would it decline by a few hundred basis points in Q4, particularly if white LED perhaps declines as well? I understand your long-term target is 58-63 and you continuously guide there, but why else would gross margin be coming down in the fourth quarter?

Well, we had a record gross margin this quarter, so we haven't been shooting for record gross margins. It's just great operational execution, but we guide the upper end of the range. We don't plan for 65%. We'd actually like to be more aggressive on the revenue side and we've had sales meetings to that effect. On the other had, there's other parts of the general DC-to-DC market always has price pressure. There's more second source people out there that used to be a couple of years ago. So there are general price pressures that you always have to keep in mind.

A lot of the improvements we're getting from Chengdu you're starting to see, but we're not going to repeat those improvements from Chengdu every quarter. That's going to load in and we might get a little more improvement there, but we'll also be offset by continuing market pressures from second sources on some of the more basic DC parts.

That's right. The company priority is to grow the revenues now. And when we set up the Chengdu operation, we have a large production capability now with very good cost. So we're looking for new opportunities and that we can grow the revenue, now.

Okay, if I could just ask one more, the BCD Plus and MiniMonsters, et cetera, you were reluctant to project what portion of revenue could become in the second half of next year. Just wondering, from a supply perspective, are you at all constrained to how fast it could ramp should it be accepted by customers?

And then, on top of that, just remind me, it's a smaller die size, so potentially it's comparatively cost effective and potentially very interesting from the cost side. Is that accurate?

Yes we offer, substantially, cost reduction to the existing solutions and/or reduce the price of the application size, as Rick mentioned a while ago, by 70%. So we give our customer values. Now, in terms of our production, the limitation, we don't have a limitation. The Chengdu testing facility and as well our fab side, they are all fully supported for our next year growth.

That's correct. We sized Chengdu for $200 million worth of business. We don't have all the capital equipment in there, but we have a building and so forth. We can go pretty rapidly up to that level without any problems.

But in terms of BCD Plus, there's nothing particularly different in the process that 12 months from now you might say, we have wins, but we can't ramp them, because the technology is too different?

No, we have as -- we announced our BCD Plus technology in March, in Q1 this year, and we reduced a couple of products, and they are in production now and they are running very smoothly and we don't foresee that could be a problem.

And your next question comes from the line of Ross Seymore with Deutsche Bank. Please proceed.

Thanks, guys. Just a question on the flat to down eight guidance, and I know you talked generally about flat, so it sounds like you're putting in a little bit of buffer just in case. But, as you look at it on an end market perspective versus a Monolithic Power Systems-specific perspective, does it seem to be more just the widespread weakness most analog companies are seeing, or is there something a little more specific to you guys?

We've seen it in the last couple of years, we all see the flat Q4. So we don't see anything different in this year versus the last couple of years.

Okay, if I look back, though, you guys were up, like, 10% last year sequentially.

Yes, that was mentioned in prior calls. When we looked backwards, once we got into Q1, the $3 million increase quarter over quarter, Q3 to Q4 last year, was inventory build, it was all [60-vale] product that was built in anticipation of new notebooks taking off or potential shortages, and in fact they ended up causing our Q1 of this year to be quite low.

So the comparable number last year was probably flat, if we controlled the inventory build. Now, it's down slightly year over year from the issues we talked about before, that our CCFL revenue is the main component of that being down year over year.

Yes, we can go back to the -- I don't recall particularly the product mix. I remember it mostly, as Rick said, as CCFL, but some of the little DC-to-DCs in Q4 last year. So that was kind of an oddball. But if you look at the other years, we're pretty much flat on the Q4 and this year basically the same as the other years. We don't see it particularly strong or weak.

So I guess that follows on nicely to the next question -- in the first quarter, and I know you're not guiding for the first quarter, but generally speaking you've been anywhere between flat and down 25%, so a pretty wide range. Do you see any problems with the logic of assuming since you're not really going up in the fourth quarter, you might come down somewhat seasonally, but thinking you'd be down way in the 15 to 25% seems a little bit dramatic and probably too extreme as we look forward into the first quarter of next year.

Well, s you said, we don't give forward guidance into the 2007, but if you take out the highs and the lows, a couple of years ago we had a strong Q1 based on some unusual design wins at one customer. Last year, we had an unusually low, because of an inventory build. So if you average it out, 10 to 15% would be the expectation I would use. That's just from a seasonality point of view, 25 is too much and zero is too much. So we would expect 10 to 15 on a generic basis. Though, again, that's not a forecast. That's just what --

What the historical shows.

Right, okay. And then the last question is something you've talked about a little bit and some of the limitations to growth being mainly either that you're not touching enough customers or you need a wider product range or greater regional reach. All of those sound like more OpEx to me, and I think you've gone down that road a little bit, but what sort of way should we get comfortable in the rate at which OpEx is likely to increase going forward relative to sales?

Well, as we had mentioned, we're trying on the R&D side, because there's a difficulty of just hiring and finding the people, so there's always that one that slows you down o that side. But on the operating side, we've said that we a sequential of about $700,000 this quarter. We're looking forward next quarter to maybe 400,000 or $500,000. And we're trying to focus that actually in our R&D sales and marketing. We think the sales and marketing should continue to grow from the beginning of the year, but we're keeping on -- the place where we're trying to hold back is G&A.

We're been able to reduce -- we only increased G&A 70,000-some this quarter over quarter. So we're keeping it sequentially G&A flat, but we are going to grow -- again, a few hundred K a quarter is what we've been doing in both of those areas, and I would say we would continue to do that in sales and marketing. Probably wouldn't do as much of that in R&D.

Ross, MPS doesn't have a habit of building the largest sales organization in any region before the revenue kicks in. So we will do it very proportional to our revenues. And, as now, all the revenue sites are from different geographic areas. Europe, US and Japan, they are all way understaffed now. So the headcount is still very small, one or two persons in each region.

Okay, that's fair. And you guys have done a great job of that in the past, so just wanted to make sure it'd stay kind of like that. I lied. There's actually one more follow-up. On the litigation side of things, spending now to prepare for the April 30th date, I think, is what you cited for the next O2 case. Given that that's two quarters out from now, is it you have to spend now so that you don't spend in the first quarter? I guess now that we're down to only one case, I would have expected the litigation expense to drop a little bit more. How do we think about that as far as it plays out over the next couple of quarters now that you're just down to '02?

Yes, I'll go ahead and answer that. Listen, we do have one major case and one minor case, so even though the Linear case is minor at this stage, it's still money. The main thing on these cases is you need to spend your money on all the expert witnesses and so forth. That's where patent cases get expensive -- before April. You have to have your case done by February. You have to be ready to try and file your briefs and all that, so you spend your money between October and February, so we'll spend large chunks of it. We've also made sure we have a very strong legal team. We've added advisers and so forth this quarter to make sure we have a very strong team. This is a critical case. We feel very good about tour chances, but we're not going to short-sight this case by making sure we spend what we need to. And a lot of it is on these very expensive outside experts.

Seems like a wise economic decision in the long run.

That's right. We take this matter very seriously and we make sure we spend the money wisely, now. And we look at this, it's a long-term investment.

Perfect. Thank you guys.

And your next question comes from the line of Rick Schafer with CIBC. Please proceed.

Yes, hi, thanks guys. A couple of questions. First one's just a follow-up to Ross's on when he's talking about legal expenses. I mean, by my math, you guys have -- well, first of all, you guys have settled really three of the four original lawsuits, and by my math I think the total cost to settle them all was I think less than $5 million combined. First, is that pretty much accurate? And, second, have you guys considered pushing forward with settlements like with O2, let's say, on this case you're getting ready for right now versus going ahead and litigating it? Or is that even an option?

Well, in terms of your math, in terms of total settlement cost, that sounds about right. Relative to settlement, as always, we would welcome the opportunity to settle the case on reasonable terms and we'll certainly always pursue that. We have in the past and would pursue the same thing in the O2 case. However, both parties have to be willing to be reasonable and frankly in the case that has not been the situation with O2Micro.

Okay, and then on Linear -- I know a couple guys brought it up, but Linear Tech, the timing or is there any way to size that potential liability and give us any kind of idea of timing on when something might happen there?

I'm sorry. The Linear case is actually -- it's very early stage. I don't really have a timing at all. We're just in the preliminary parts of evaluating their claims and deciding what to do about it. So it involves a single product, our MP 1543. It's a relatively new product. So it's difficult to quantify -- our revenues are pretty small so far, so because it's a new product and It's also a new case, we're evaluating it.

I can't really give you any dates. Sometimes these things take a year or two before they go to trial at all, so that's the only thing I can think of.

Okay, and then, just a follow-up, on the CCFL business, I guess we've -- I know you talked about it, you guys expect it to be basically flat in the fourth quarter. I mean, you have seem Maxim and Microsemi I think started to make some headway in that business, particularly at Dell right now, which ha been a bigger customer for you guys. Can you describe, like, the impact that you guys are seeing right now? Do you expect that we've sort of seen that business stabilize at these levels going forward, or how should we think about CCFL when we look into '07?

Well, we believe MPS, the product remains superior in the notebook applications. We're still the only company that offers single-chip solutions for the notebook backlight. But our customers -- because of the litigation, our customers do look for alternative solutions and a company like Micrel, a company like Microsemi and the Maxims. So if this April trial is a result of that matter, so I think that our product is very competitive in the market.

Okay, and so to grow that business next year, if you guys were assuming any growth in CCFL next year, does that include pushing forward, maybe pushing your CCFL solution into new areas, where you haven't been historically strong, like LCD TV, FPD, something like that?

Yes. It has a lot to do with the litigation outcome.

Okay, thanks a lot.

And your next question comes from the line of Quinn Bolton with Needham & Company. Please proceed.

Afternoon, guys. Was wondering if you could just repeat about what you said in the 129 patent in the O2 case?

Yes, we basically -- O2 dismissed the 129 portion of their case, again, in the Northern District Court of Oakland. There was a 722 patent and a 129 patent case pending, and O2 withdrew the 129 patent case recently.

Okay, and then I understand that there might have been a preliminary decision here in the last week, they might have dismissed some of the claims of the 722 patient?

I don't think that's correct. There was a hearing on Friday when there's various motions in front of the judge and that's what went on last Friday.

But in terms of 722?

Oh, okay, you're talking about -- this was another case. There was recent press about the court's refusal to dismiss MPS's inequitable conduct claims. Is that the one you're talking about? Basically, in the 722 case, there's a discussion about the 722 patent in the case coming up and because O2 failed to disclose material prior [ARC] in its 722 patent applications, we believe that their patients should not be filed. Thus, we filed what is referred to in legal terms as an inequitable conduct claim.

O2 contested this, but the drug recently agreed with MPS that this claim -- this claim for invalidity -- should be heard in the case. So it was more of a technical matter. It basically said our claim about invalidity will be heard. That's all it is.

And it may [perceive] as our favor, but that's only a hearing. It's not the final judgment.

Okay, but in the -- within the last week or two, there's been no claims on the 722 patient dismissed?

No, that's only a hearing and a judge indicated one way or the other, but that's not the final judgment.

Yes, the only thing that came out was more of a legal motion thing to make sure -- and it ended up our argument will be heard in court. That's all that says.

Okay, and then moving on just with Chengdu ramping up now to pretty much full capacity, I know you said that you're still doing some R&D in Los Gatos, but are there further costs that can be taken out of the business now that Los Gatos is effectively only an R&D site, or have already taken out most of the cost?

I think we've really seen most of the benefit in the third quarter. The fact that our ongoing operating costs are a little lower in Chengdu, that's good, but there isn't going to be much more taken out. We've reduced the headcount and so forth, or transferred them to R&D.

Okay, and then lastly, in past calls, Rick, I thought you had mentioned sort of tax rate potentially could go down towards maybe the low 20s and 2007, you had mentioned a mid 20 number, which I think you said for Q4. So I'm just trying to clarify. Do you think the tax rate in '07 can get to low to mid 20s, still? Or do you think mid 20s is now a better number for '07?

Yes, actually, let me clarify that. Originally, we thought we would get some impact from the tax structure earlier. We won't in '06, '06 will be similar to what it is now. But I think in '07 we can go into the mid 20s. That's how I'd like to put the outlook, now.

Okay, thank you.

And your next question comes from the line of Craig Berger with Wedbush Morgan Securities. Please proceed.

Good afternoon. Thanks for taking my question. I want to ask about the white LED drivers. How big are they still and what's the impact on Q4?

Hi, Craig, thanks. As we said, the range of white LED revenues is generally about 10 to 15% of the DC-to-DC number. This quarter, it was more like the 10 part of it, closer to the 10. It would probably be that or lower in fourth quarter, simply because of the -- some of the standard inductor-type products having very low prices and we're not trying to do that business, to that's affected our revenue.

Overall, a white LED driver for handsets, the price has become very commoditized and the product can be replaced very soon by other competitors. And most of the companies in the business, they all have a [ping to ping] compatible device and they can replace it. And so it's really a commodity business. MPS would like to get out of that.

Yes, some of the margins we hear is 25 to 35%, even, on these parts.

Yes.

Okay, so less than 10% in Q4?

It could be. It's probably declining.

We actually said in the late last year we will decline. This portion of the business will be declined.

Right. Okay, Rick, with respect to the Q4-specific tax rate, did you say what that should be.

It's probably the same. The book tax rate, I mean, our effective tax rate year to date, is 81%. That's what we recorded in Q3. The non-GAAP are is about 36%. It will probably stay 36%, which is pretty much statutory rate right now.

Okay, thank you. With respect to Q1, I realize you guys aren't providing any guidance, but could you discuss any potential new product-related revenues that may offset some of the normal seasonality that you guys would typically see? Are you going to be getting any new product revenues in Q1?

Well, I only can talk about the designing and also possible revenue [work again] in Q1 '07. And we introduced the battery chargers in the beginning of this year. We see a healthy growth. And that will be also in the Q1 revenues. And other products, like, we released some DC-to-DC product in the middle of year, saying that we will see some [risk] back to that.

Who are the battery charge customers?

We don't really disclose any particular customers and mostly in China and some other ones. We have a design win in Japan.

Great. And then with respect to the CCFL business, $7 million a quarter for the last couple of quarters. Is that kind of a consistent rate, or run rate, that we can use for 2007 at this point? Or do you -- asked differently, do you still see any major CCFL customers moving away from you that haven't already moved away from you?

In the last conference call, we took a very conservative approach and we forecast our CCFL revenue lower than actuals. And so our customers obviously remained, stayed with us. And in the future, the outcomes really depend on the litigations, and it's difficult to comment on it.

Right. With respect to litigation, you guys are guiding Q4 to 3.5 to 4 million. Is there any reason to believe that Q1 would be less than that?

If it goes to trial on April 30th, it would probably be a similar range. That's been our experience when we take major trials to court. Again, this one relies heavily on a lot of outside expert opinion. So you get to spend twice, your lawyers and then their experts and so forth. So it would probably be -- can't really say it would be 3.5 to four, but it would probably be more than three.

Well, a lot of work will be done in this quarter and the next quarters. We can't really foresee that, so these are litigations and the lawyers [are] on their own costs, so we believe this is the best investment for the company in the future.

Great. One more for me. With respect to the BCD competitive landscape, what are you seeing out there among your competitors? Any new entrants, and what are you seeing from AATI?

AATI, I don't see AATI as a competitor and we don't have any -- they don't compete with us in the higher-voltage area, anything about 12 volts, let's say, and, frankly, we're the only -- there are very few companies that offer solutions that have more than 16 volts and with the higher current capabilities, and we don't see a lot of competitors. But the real competitor is the discrete. People making the controllers and other companies make power MOSFETs. We are replacing those solutions.

Thanks for your insight. Good luck.

Thank you, Craig.

And your next question comes from the line of Michael Davies with Next Generation. Please proceed.

Yes, thanks, guys, for taking my call. I just have a few items here. A clarification, with the BCD Plus and the dropped litigation with Micrel. I mean, this allows you to pursue what you're doing in the BCD Plus kind of uninhibited? Can you just clarify that?

No, it's regardless of Micrel, the patent lawsuit settlements or not, we're not violating Micrel's patent, so now, with the settlement, it's strengthening the MPS process technology patent portfolios, and BCD Plus, we developed totally based on MPS and MPS inventions, and this is a very unique, very -- has excellent, superior capabilities.

Okay, so taking this a step forward here, you're free to pursue your own directions and your own developments and products uninhibited from what they were challenging you with?

Yes, and we can't disclose the full details of the settlements, but in essence the claims they made have been settled and we can just continue doing what we're ding.

Okay, thanks. A couple other points.

Okay.

Turns business, what was that in the business?

Well, we don't really disclose that. We're a typical analog company that a good chunk of our business is churns business. We don't really talk about it. We're no different than most people. People don't book a year ahead for our kind of parts, so we always have a fair amount of turns business, but we've never disclosed the percentage.

On the audio segment, you're up nicely and those have pretty decent margins. Can we expect that this kind of level can be sustained?

Well, yes, as you know, that's one of the things I mentioned. Part of our gross margin performance this quarter, there's some cost impact from changing it, but it was also the mix issue, and you picked up on it. The increase in audio gives us a little better margin. It's one of the better products. So, in fact, our current outlook for audio is to continue to hopefully hold these levels, not slip back.

That Q2 number of 400K was a bit of a timing issue, with new product designs not kicking in on time, but we think we can keep the $1 million growing, based on the sales force outlook, so it's looking pretty good. We've introduced a couple of good new products, Michael.

I agree. And just one follow-up there. Looking at your end markets, anything to read into it, I mean, given what we've heard from a lot of the analog players out there, any weakness in any one market such as notebooks or so forth?

Michael Hsing will answer your question, Michael, sorry.

Yes, I think since you mentioned the patent life, by the way, I forgot to mention, $3 million settlement with Micrel is just getting my own patent. That was my patent I granted to Micrel at the time. So [inaudible]?

Just any -- from what we heard from some of the analog players, we've had some maybe systemic weakness in perhaps the notebook segment. Just wondering if you had any color to add to that?

We are -- in the current product family, we only have a CCFL exposure in the notebook side, but in the future products, definitely we're targeting in the notebook application. Is I can't add a lot of color to the market conditions for you.

All right. Thank you guys.

And your final question comes from the line of Steve Smigie with Raymond James. Please proceed.

Great, thank you. Just to follow-up, with the litigation, say you get to the April period, does that spending level continue into June and would we expect it to drop off pretty significantly into the back half of the year?

Yes, Steve, actually, at that point the trial is done -- the major trial is done, so we should know one way or the other. So I would assume it would drop off. That would be our projection, given that we wouldn't have any major trials at that point, but it's always hard to say with litigation, but that's probably a hard thing to assume. I don't know how much, but that's the likely outcome.

And then on the spending for building out the sales force in R&D, does that continue to ramp on a dollar basis throughout '07?

Yes, we're working on the plan now. As I said, quarter over quarter, just to give you an idea, SG&A was up about 255,000, and only about 70,000 of it was G&A and about 180,000 of it was sales and marketing. So a couple hundred K a quarter right now in sales and marketing is kind of the increased rate. The increase in R&D this quarter was more than it would normally be. We wouldn't normally, again, go up a couple hundred K a quarter from investing in new people, but we're trying to get a ton of new products out, so we've got [mass] costs and a lot of one-time costs and layout costs and things like that. So we had some usual charges, spending.

It's also a good thing. We're obviously getting out a lot of new products, so we wouldn't expect R&D to go up as much after this next quarter or two.

Okay, great. Thank you very much.

Ladies and gentlemen, this now concludes the question and answer session. At this time, I will turn the call over to Mr. Neely for closing remarks.

Well, thank you all for listening to our conference call. We look forward to talking to you further in the upcoming months. Thank you.

Thank you for your participation in today's conference. Ladies and gentlemen, this concludes the presentation. You may all disconnect and have a good day.