Power Integrations Inc (POWI) 2005 Q3 法說會逐字稿

Good afternoon and welcome to the third quarter 2005 earnings conference call for Power Integration. Today's call is being recorded. At this time I would like to turn the conference call over to Mr. Joe Shiffler. Mr. Shiffler, please go ahead.

Thank you and good afternoon. I'm Joe Shiffler, director of IR and corporate communication for Power Integrations. Thanks for joining us to discuss our quarterly results, which are outlined in a press release that we issued earlier this afternoon. The release has been e-mailed directly to those of you on our distribution list and is also available on the investor info section of our website, www.powerint.com.

With me on the call today are Balu Balakrishnan, president and CEO of Power Integrations, and John Cobb, our chief financial officer. Balu and John each have a set of prepared remarks after which we'll take your questions.

Before we begin, I'd like to caution you that our discussion today, including the Q&A session, will include forward-looking statements reflecting management's current forecast of certain aspects of the company's future business. Forward-looking statements are denoted by such words as "will," "would," "believe," "should," "expect," "outlook," "estimate," "anticipate," and similar expressions that look toward future events or performance. Forward-looking statements are based on current information that is, by its nature, dynamic and subject to rapid and even abrupt changes. Our forward-looking statements are subject to risks and uncertainties, which may cause actual results to differ materially from those projected or implied in our statements. Risks and uncertainties affecting our business, which could cause actual results to differ materially, are discussed in our most recent reports on forms 10-K and 10-Q filed with the SEC. With that I'll turn the call over to Balu.

Thanks, Joe, and good afternoon. Our third quarter results were solid overall, highlighted by a higher-than-expected gross margin, strong growth from our LinkSwitch product family and a significant reduction in inventories. We are also announcing a $25 million shared repurchase program reflecting not only our strong balance sheet but also a confidence in the near-term and longer-term future of our business.

Next year, in particular, should be very exciting for Power Integrations, and I'll spend a few minutes this afternoon outlining the key growth drivers for 2006.

Looking first at our third quarter results, total revenues were $36.5 million, up 4% sequentially, which was the low end of our guidance range. We did not meet the midpoint of our guidance due largely to a delayed production ramp on two high-volume Motorola charger designs that we announced last quarter. The delays were caused by issues at the charger manufacturer unrelated to our device. We expect these designs to begin ramping during the fourth quarter and reach full volume by the first quarter, which should a few million units per quarter.

Obviously, we are disappointed that these programs did not ramp as expected. This was the only negative in an otherwise very good quarter. In spite of the delayed cell phone program, the reported record revenues with an increase of 11% from a year ago, unit shipments were up nearly 20% year-over-year reflecting growth in our overall market share and significant growth from our lower-power, lower-ASP LinkSwitch products.

We continue to improve our gross margin, which was 49.5% for the quarter, better than expected, and up nearly 2 points from a year ago. This reflects not only ongoing cost reductions but also an ability to maintain relative price stability in a highly competitive market.

Our operating margins for the third quarter was 15% including a negative impact of about 5 points from litigation expenses.

Net income was $0.18 per share including a $0.05 impact from litigation. While the cost of litigation is high, our underlying profitability has improved significantly in the past year, which sets the stage for strong earnings growth, once litigation expenses begin to decline.

Looking ahead, while it's difficult to predict end-market demand given today's uncertain economic conditions, we believe we have the pieces in place to sustain strong growth through 2006 and beyond.

For the fourth quarter 2005 we are forecasting strong year-over-year revenue growth with gross margin improving again to approximately 50%. Next year is shaping up to be even better, and I would like to spend the next few minutes going over some of the key drivers that make 2006 a very promising year.

One major driver, as you all know, is the California Energy Standards on external power supplies, which take effect next July. We are marketing our solution very aggressively in response to these standards and design activity continues to build. Most of this activity is in the early stages, and any significant revenue impact is still a few quarters away. However, we did win three additional designs in the third quarter as a result of the standards.

All of these designs were in cordless phones, a market that previously used linear power supplies almost exclusively. The impact of this area continues to grow, as they are adopted in one geography after another. New York State has passed a law that will likely lead to a standard similar to California's, while Rhode Island recently became the fourth state to fully adopt California standards.

China, which currently has a warranty program similar to EnergyStar is now strongly considering mandatory regulations as well. It is clear that California standards are on the way to becoming a de facto global standard for external adapters and chargers, which accounts for roughly one-third of all AC to DC power supplies.

With our LinkSwitch and TinySwitch families, we have a complete portfolio of products to take advantage of this opportunity. We have ICS [ph] to support any kind of charger or adapter designed to meet the standards regardless of the power level or the desired output characteristics.

Earlier this week we announced the latest edition to our LinkSwitch family, the LinkSwitch LP. The LP has been specifically designed to replace unregulated linear transformers, which are typically used in very low-power applications such as trickle chargers where the low cost is the only criteria. Chargers designed in LinkSwitch LP typically need only 14 components, making them the simplest, lowest-cost electronic chargers in the world. No other solution comes close to achieving this level of simplicity. In fact, a comparable solution contains about 35 components, which gives us the bill-of-materials advantage as well as a significant advantage in terms of design and manufacturing costs.

Our solution is also superior in terms of safety and reliability with features such as auto-recovering, thermal shutdown included on our chip. LinkSwitch LP is therefore not only lower in terms of absolute cost, but it is also superior in terms of design cycle time, manufacturability, and performance. LP is a significant breakthrough, and it is timed perfectly to capture the opportunity created by the energy standards.

Energy efficiency will remain a key driver throughout our business, well beyond external power supplies. The world's energy resources are being stretched ever thinner by rapidly growing demand and by the recent supply shocks caused by natural disasters. Policymakers around the world are stepping up their efforts to reduce consumption, and electronic products continues to be a major focus.

Next month I will join policymakers, researchers, and other experts around the world at a conference in Seoul, Korea. The aim of the conference is to seek a global solution to the problem of standby power losses and to consider the development of a universal one-watch standard covering a broad range of devices. These discussions and others like them are leading in only one irreversible direction -- toward tighter standards on a wide range of devices. Tighter standards are nothing but good news for us and should continue to provide tailwind for a long time to come.

Another growth driver, going forward, will be the strong year we have had in terms of design wins. We had another strong quarter following the record level of designs in the second quarter. Design wins for TinySwitch included a wide variety of applications such as LCD TVs for LG and Sharp; PC standby designs for HP, IBM, Dell, and Sony; several DVD and MP3 players, and number of industrial applications such as power meters, motor controls, and emergency lighting. We also won high-volume designs for a digital camera and a mobility IGO charger.

Top CGX won multiple LCD TV designs at LG and server standby design for Dell, several set-top box designs including one for Motorola, and a number of LCD monitors and projectors among other applications. Our newer products, DPA-Switch and LinkSwitch grew nicely in the third quarter and should be significant contributors, going forward. Combined revenues from these two product families were up more than 50% sequentially and totaled 6% of revenues, up from 4% last quarter.

DPA-Switch, our DC-to-DC product family, won several power-over-Ethernet designs for products such as wireless access points and voice-over-IP phone equipment. DPA-Switch has gotten significant traction in PoE applications this year including sizable wireless [inaudible] phone design wins at Cisco and Nortel, as discussed on previous conference calls.

LinkSwitch, our low-powered product family designed specifically for cost-effective linear replacement continues to gain traction across a variety of applications. Third quarter design wins included a differentiator for LG, a PDHR for Palm, several small consumer appliances, and a number of industrial applications including power meters and lighting applications.

We believe we have the most competitive product portfolio in the market today. Furthermore, we have a full pipeline of exciting new products planned to exception over the next several quarters that will position us even better for 2006 and beyond. We will introduce next generation of some existing products making them even more cost-effective through a higher level of integration.

We will also introduce brand-new products that will improve our cost-effectiveness in a wide range of power supply applications. We also continue to expand our library of circuit designs in order to facilitate the use of our products in a wide range of applications.

For example, in August we introduced a design for an LED driver circuit featuring the LinkSwitch TN. We believe that solid-state lighting has tremendous potential for us as LEDs continue to produce applications such as traffic lights, emergency exit signs, and neon sign replacement and ultimately moving to general purpose home and office lighting.

LED lighting is a perfect application for our technology since one of its major benefits compared to the traditional lighting is energy efficiency. The LEDs are driven by inefficient power supplies, much of the efficiency benefit is squandered. Our new circuit design drives a dozen LEDs using only nine components at a high volume BOM cost of just $0.50 including our chip.

Overall, lighting is becoming an important new market for us that promises to be a significant contributor to our growth in 2006. In addition to LED applications, we are rapidly increasing our penetration in emergency lighting and all our industry lighting applications. In fact, we had approximately 20 design wins overall in lighting applications in the third quarter alone. And the revenue from lighting applications was up more than 40% from a year ago.

A very promising growth driver for 2006 will be our expanding presence in Japan. In the past, we have essentially taken a back-door approach to Japan by licensing our technology to MEI, one of our foundry partners. MEI sells parts under their own brand name to Japanese manufacturers paying royalties to Power Integration. They are competing in Japan with their own products, in some cases going against MEI but only to a limited extent. This has been a successful approach for us and has allowed us to focus our sales and resources on other markets. Now we believe we have a solution to directly address the Japanese market.

Our technology license agreement with MEI has expired, and MEI will not have access to any technology introduced by us after June of this year. They may continue to sell products based on legacy technology and will pay as royalties on those sales at a reduced rate over the next four years. However, with a strong pipeline of new products coming to the market over the near term, we will have a very strong competitive position in the Japanese market, and we expect good growth there in 2006 and beyond.

As we extend our geographic reach and address a wide range of applications, we are adding more sales and technical resources. By the end of this year, we expect that our sales and field and remaining staff, we'll have more than doubled in size from the end of 2003 with a majority of that growth occurring this year. We have been especially focused on adding people in China, which is fast becoming the most important market in the power supply industry. Expansion of our sales force has already paid off in the form of a higher level of design wins in recent quarters, and we expect this trend to be a major growth drive for us next year.

Another driver for us is our increasing focus on tier 2 and tier 3 customers over the last two years. We have implemented a wide variety of tactics to reach our customers and enable them to design with our products. This has helped us diversify in the markets such as consumer and industrial, which are more fragmented and less exposed to competition than other markets. This shift has been very healthy for us in our business, contributing to price and [inaudible] stability. While we will continue to focus on growing our tier 1 business, including penetration at lower-tier customers will be a priority in our marketing and sales efforts, going forward.

Before I turn it over to John, I would like to conclude by summarizing our key growth drivers heading into 2006. The replacement of linear transformers driven by efficiency standards is obviously and important opportunity for us, and we are already beginning to take advantage of it. However, this is far from being our only priority. We have strong design-win momentum across the board with newer products becoming significant contributors and more mature products still going strong as they near the peak of their lifecycle.

Next-generation products will soon follow, keeping us well ahead of competition in the mid-power and higher-power markets. We will also have new applications and markets available to us as a result of our new products and our new approach to the important Japanese market. We have dramatically increased the size and the reach of our sales force in order to capitalize on these opportunities, and we continue to use innovative marketing approaches to drive adoption of our products at smaller, more far-flung customers around the world.

Finally, we look forward to the successful conclusion of our patent litigation, which will not only end the burden of legal expenses, but also improve our competitive position by removing infringing products from the market. We believe all these drivers should add up to strong growth of our division in 2006 and beyond. With that, I'll turn it over to John for a review of third quarter financials. John?

Thanks, Balu, and good afternoon. As Balu noted, our revenues for the third quarter were at the lower end of our guidance primarily due to a delayed production ramp on two major Motorola programs. However, in spite of these delays we showed strong year-over-year growth in revenues in units shipped, and we continued the recent upward trend in our gross margin. We also demonstrated strong underlying earnings growth and operating margin improvement aside from the impact of patent litigation expenses.

Turning to the income statement detail, net revenues for the quarter totaled $36.5 million up 4% compared to the prior quarter. As you may recall, our second quarter revenues included a $900,000 benefit from the reversal of reserves for sales returns. Our sequential revenue growth excluding this item was 7%. Year-over-year revenue growth was 11% driven by nearly 20% growth in unit shipments. ASP was $0.44, down from $0.47 a year ago, driven primarily by pricing but also by product mix as lower-power products such as LinkSwitch have become a bigger portion of our revenue over the past year.

Looking at some of the revenue detail, turns orders were 77% of revenue, the highest percentage in our history. Distributors accounted for 61% of revenue in the quarter including Memec and Synnex at 19% and 18%, respectively. As a reminder, we recognize all of our distributor revenues on sell through.

Revenue mix by end market in the second quarter was 31% consumer; 28% communications; 23% computer; 11% industrial; and 7% other. The computer and industrial segments drove our sequential revenue growth for the quarter, up 13% and 16% respectively, while consumer and communications revenues were each down slightly. All segments contributed to year-over-year revenue growth with industrial up about 60%, computer up 15%, communications up 6%, and consumer up 1%. Gross margins for the quarter was 49.5%, up more than a half a point from the prior quarter and nearly 2 points from a year ago. Again, this is in spite of a pricing environment that remains highly competitive with both discretes and competing integrated parts being priced very aggressively.

Operating expenses for the quarter were $12.6 million, up 12% from the prior quarter. Patent litigation expenses were the primary driver of the increase, totaling $1.8 million, up from $800,000 in the second quarter. Litigation expenses will continue to be elevated over the next several quarters as we progress through our cases against Fairchild and System General. We expect litigation expenses in the fourth quarter to total about $2 million bringing the total for 2005 to about $5 million, or $0.13 per share. It is difficult to predict at this time how much we will need to spend in 2006, however, we do expect litigation expenses to decline next year compared to 2005 with the bulk of the spending coming in the first three quarters of the year.

Returning to the income statement, income from operations in the third quarter was $5.5 million, giving us an operating margin of 15% including a negative impact of about 5 points from litigation expenses.

Other income was $939,000, a slight increase from the prior quarter due to higher interest rates. We have adjusted our effective tax rate for the year to 21% from our previous estimate of 24% due to a higher percentage of profits from lower tax jurisdictions. Our third quarter effective tax rate was 12%, which brings our year-to-date tax rate to the new 2% level.

Net income was $5.7 million or $0.18 per share, the lower-than-expected tax rate contributed $0.02 while as mentioned earlier, litigation expenses reduced earnings by $0.05 per share, about $0.01 more than we had expected when we gave guidance for the quarter.

Turning to the balance sheet, we ended the quarter with $122 million in cash and investments, down $22 million from a year ago due to the $40 million repurchase that we completed in June of this year. Cash flow from operations was $6.1 million in the third quarter. As Balu mentioned, our board has authorized the repurchase of another $25 million of our shares.

Inventory decreased by $2 million during the quarter to $22.3 million, giving us inventory turns of 3.4, up from 3 in the prior quarter. Accounts receivable were $14.7 million, up $1.8 million. DSO was 36 days, up from 30 days in the prior quarter.

Moving to the outlook, visibility remains limited, as usual, and the demand environment is highly uncertain, given current economic conditions. Power supply manufacturers are managing their inventories very tightly, and we expect the turns component of our revenues to remain in the mid-70s in the fourth quarter. We expect sequential revenue growth in the range of 2% to 6%, which is better than our historical fourth quarter average of approximately flat revenue. This equates to year-over-year growth in a range of 11% to 15%.

We expect our gross margin to improve once again to approximately 50%. Operating expenses should grow by 1% to 3% sequentially with patent litigation expenses totaling around $2 million. As mentioned earlier, we anticipate a tax rate of 21% while earnings per share should be in a range of $0.16 to $0.19. The midpoint of this range would be an increase of 25% from the fourth quarter of 2004 despite an expected impact of $0.05 from litigation expenses.

Our fundamental underlying business is performing well and, as Balu explained, we are very excited about the opportunities in front of us as we head into 2006. And now I'll turn it back to Joe. Joe?

Thanks, John. Before we take questions, I'll do a quick rundown of our very busy fourth quarter conference calendar. We will be at the Deutsche Bank Global Semiconductor Conference in Las Vegas on November 3rd; the AEA Classic in San Diego on November 7th and 8th; the Thomas Weisel Innovations in Power Conference in New York on November 9th; the Raymond James IT Supply Chain Conference on November 15th, also in New York; the B of A Technology Conference in San Francisco on November 16th; and the First Albany Growth Conference in New York on December 7th. And now, Operator, would you please open the line for questions.

[OPERATOR INSTRUCTIONS]

Mr. Shawn Slayton with SG Cowen and Company.

In the scripted comments, if you mentioned it, I didn't catch it -- a trend in average selling price?

The ASP was $0.44 down from $0.47 the year before and, I believe, down a penny from the quarter before.

Could you speak to that just a little bit and also help us reconcile the expansion in gross margin with that ASP trend? Thanks.

The expansion in gross margin is really a function of cost reduction. The ASP came down, as we mentioned, both a function of price reductions on a part-by-part basis but also a mixed shift, as we started to shift more of LinkSwitch and also a shift to lower-power TinySwitch product, it reduces the overall ASP. However, the margin we've been able to improve through cost reduction, and we've discussed on prior conference calls, we focus on all areas of cost reduction -- wafer prices, packaging prices, test costs, and we've been continually successful in reducing our cost, obviously, at a faster rate than the ASP reduction.

Okay, and on the competitive side of things, no one else there -- since we talked three months ago -- no one else there is being any more or less aggressive on the pricing or on their efforts to take market share in this space?

The competition remains extremely aggressive.

Okay. You gave a little bit of detail but not a lot on the out-year legal expenses. You said, if I understand, you'd give $5 million in lawyer costs this year, and it's going to be something less than than in absolute dollars next year, is that correct?

That's correct. And then, as we said, it will be less overall for the year but most of those -- the spending will be in the first three quarters. We don't expect very much in the fourth quarter because by then most of the actions should be somewhat complete.

Okay. Maybe for Balu -- so -- and I know you guys give 90 days of guidance. Q1 of this year, you guys actually had a sequentially up year, and I know there was -- largely, that because of the dislocation in the second half of '04 that caused your Q1 to be up. Are you seeing any change in the seasonality of your business? Maybe as you have more exposure to the consumer and the handset end market as a percentage of your business? And obviously what I'm trying to do here is get my arms around what the trajectory should be for Q1. Thanks.

Well, there are two components to it. The end market itself -- it's not clear to me what is going to happen in the future, especially with the cost of gasoline and heating expenses. So there is some uncertainty there. But if you look at our company, we have some differential advantages, mainly the leader to LED efficiency. As energy becomes more expensive, there is certainly more focus globally on improving energy efficiency. So that will help us have a differential advantage.

In terms of seasonality, I am not sure whether there is any significant change in seasonality, because there have been lots of unusual things that happened last year, which made last year different from our historical seasonality, and this year has been a little bit closer to historical seasonality because we have had effectively a 7% increase in Q3. It would have been larger but for the two programs that got delayed. But this year looks more normal to me but, as we mentioned, Q4 should grow reasonably from Q3 because we didn't do as well as we expected originally in Q3.

Okay, so now normal seasonality implies a modestly down Q1 as opposed to 2005 Q1, if I understand you.

Yes. Usually, the Q1, and again if I take normal seasonality -- I don't know what's normal anymore -- is flat to down 5% is the range.

Okay, that's helpful. And, Balu, can you just speak a little bit about -- you've talked a little bit about the traction that you're getting related to these governmental energy efficiency initiatives -- what is your estimation on who is going to be -- what manufacturers most likely -- maybe you can speak to Asia and Europe -- what manufacturers are going to be later in adopting and realizing that this is a legitimate mandate that impacts their business, and who is embracing this notion now and maybe starting the design process?

To me, I guess, I'm kind of speculating that maybe the Asia manufacturers will be -- will wait until the last minute to accomplish some of this work. Can you speak to that?

Well, if you look at the large consumer manufacturers, whether it is Sony or Panasonic or Samsung, LG -- they all know about the standard by now, and they are all taking steps to make sure they can meet the standards. In the consumer business, there are not too many small companies.

Now, there are -- I mean -- I should say, having said that, there are smaller companies which sell things like security systems and sprinkler systems and so on. I'm not so sure how quickly they will adapt and how much aware they are, because we don't go touch those customers. We totally deal with tier 1 customers directly; the rest of them come through distribution. So we really have to wait and see what happens, but in terms of tier 1 customers, almost all of the customers are aware of it, and they are all taking steps to follow through. Certainly, some of them are more practiced than others.

Okay, just one last item, and I'll give up the floor. Just on the balance sheet, John, you guys had a kind of a meaningful spike in other assets.

What we did during the quarter is we made a strategic loan to one of our vendors for purposes of implementing new technology to continue to get cost reductions. And it was a $10 million loan.

Shawn Connor with Waterstone.

A couple of quick questions -- first of all, on the inventory -- do you expect to bring that down again this quarter like you did last quarter?

We expect our inventories to be relatively flat this quarter sequentially.

And can you talk a little bit about inventories within the distribution channel?

At the end of the quarter, we had 3.6 weeks of our inventory in distribution, which is kind of at the lower end of the average.

And just to remind you, we only count sales after the distribution has sold. It's only on sell-through.

Right, and then within comm, can you talk or give a little bit more color about what you're seeing in the handset side of the business?

Our view into the handset business is really through Samsung, Motorola, and then, to a lesser extent, LG and the Chinese. So we had a quarter equivalent with what Samsung announced a couple of days ago. We had strong growth in our China business, and LG was down slightly. Motorola was down predominantly for the reasons that we discussed -- that the ramping of the new design win was much slower and delayed from what we had expected, so they continue to ship an older design.

Samsung currently represents about 50% of our revenue, so at this time they represent our largest window into the cell phone market.

Andrew Huang with American Technology Research.

Just a couple of housekeeping items first -- what kind of a tax rate should we be using for Q4 of this year, the current quarter?

Twenty-one percent.

Twenty-one percent? And then for 2006?

We haven't done an estimated tax rate for 2006, but our expectation is that it wouldn't be higher than where it is and probably not significantly lower but within the same ballpark.

In your prepared remarks you kind of talked about the ASP going down from $0.45 in the last quarter to $0.44 this quarter, and part of it was due to mixed shift, you know, because of LinkSwitch, I assume. But then also part of it was because of price reductions you commented. So I'm just kind of curious, you know, are there any areas and market where you're being more aggressive on price, and have you done any kinds of sensitive analysis to kind of figure out how much incremental volume you could get if you were a lot more aggressive on price?

Certainly, in markets where there are large tier 1 customers, we have the highest pressure mainly because of the discrete and integrated competitors targeting those markets much moreso than tier 2, tier 3-type customers. The two big markets are characterized by tier 1 customers. The first one is cell phones and communications; the second one in computers, because most customers tend to be very large.

But other than that differentiation, I would say across the board our margin is relatively same for the same volume for the product. So we have pricing pressure from discretes and integrated competition, and we have had severe pricing pressure for quite a while now. It continues to be that way.

I would just like to add, as we've commented in the past, for us to reduce our prices, that would not automatically equate to new volume. And, in fact, to be quite blunt, we've actually tried it on several occasions in the past, and it doesn't work because what happens is the competitors just drop their prices, and all that happens is the prices overall in the market collapse. So we try to neutralize price so that by using our device versus some other design, the system bill of material is equivalent, and then we win the business on the ease of design, ease of manufacturing, or quality, reliability, delivery -- all those factors that have helped us gain share over the years.

But just reducing price would not lead to an increase in revenue and, in fact, it would probably lead to a decrease in revenue because we'd probably get the same business, we would just get it at a lower price.

Got it, okay. The next question I have is you also commented on kind of being a little bit more aggressive on the tier 2 and tier 3 customers, and I would think that normally that's the type of customer you go after through your distribution channel. So my question is can you give me an idea of maybe what your operating margin would be selling to a tier 2 or tier 3 customer through distribution and then what it would be if you sold it -- your part -- to those customers directly?

Well, your first statement was correct. I mean, most of the smaller customers will be through distribution. We can't really afford to do business with them directly. So when you say -- you said "aggressively," but aggressively doesn't mean lower prices. Actually, tier 2, tier 3, is higher prices and higher margins, but we are implementing strategies that will allow us to leverage our resources to help get design into these customers.

One example of that is the software program we've developed that will allow tier 2, tier 3 customers to design a power supply by themselves rather than depend on our SAEs like the tier 1 customers generally do. So I can't give you exact numbers on what the profit margins are, the gross margins are on tier 2, tier 3 customers. Obviously, the lower the volume the higher the margin, and so it helps us offset lower margins at tier 1 customers.

Okay, so just so I'm clear -- when you said that you are trying to more aggressively go after the tier 2 and tier 3 customers, it would still be through distribution but you're just trying to go at them with maybe incremental spend on marketing?

Yes. It is basically using different strategies to expand our tier 2, tier 3 market. When you're a smaller company, you tend to focus on tier 1 customers, because that's the faster turn on the investment. But as you grow, you have enough bandwidth and enough resources that you can come up with tools and also Web-based applications notes and seminars and a whole lot of other strategies that we implement, including ways to get technical questions answered through the Web and so on. And by doing so, we are able to address a lot of smaller customers, and we actually have seen a significant increase in tier 2, tier 3 revenue since we started implementing this two years ago, and we are very, very pleased to see that, because that's not only higher margin, it's a lot more stable business, and it is similar to the type of business Maxim and Linear have. They have much larger tier 2, tier 3 content. Of course, we still continue to grow the tier 1. We are not in any way neglecting the tier 1.

And we are also adding distributors specifically to help us go after the tier 2 and tier 3.

Tore Svanberg of Piper Jaffray.

This is actually Jeremy calling for Tore. Balu, I think you mentioned the MP3 player being one of the important drivers to design wins. Can you tell us how big this business is currently and possibly where this could go next year and any notable customers you might have in this segment?

I'm afraid I don't know the exact volume of this design, and Joe is trying to --

It's a medium-volume design win.

Does that answer your question?

And how big is the percent of sales today and where do you think this type of business can grow?

The total MP3 business?

Yes.

At this point, it's less than 2% or less than 1 percent.

Yes, I think it's less than 1 percent.

Okay, and in terms of Samsung, I think, you mentioned is a very important customer -- with the number of new phones coming out, can you help us understand how this could help your charger business?

It doesn't have much of an impact on the charger business. They have a few models of chargers that service all of their phones, so I would like to say they have, like, three or four models of chargers, and two of those models are the highest-volume models, and they just choose one of those chargers to go with the phone. So there's not a direct connection between the phones and the chargers.

Great, and one last question -- in terms of your relationship with MEI, can you help us understand, also, how by entering the Japanese market on your own, does this affect your relationship on the manufacturing side for the middle?

Not at all, in fact we have signed our five-year agreement for the next five years with MEI. So we had two separate agreements. We had a wafer fabrication agreement and a technology agreement, and the last five-year technology agreement expired in June, and we have chosen not to -- mutually chosen -- not to continue it mainly because we think we are now big enough in size to be able to enter this Japanese market on our own, which, of course, has advantage that we get the full revenue and margins rather than just getting royalties. And so that gives them opportunity to grow that market as an additional growth driver for 2006 and beyond.

Vernon Essi with Janney Montgomery Scott.

Wondering -- just a point of clarification on the segments that you break out -- is cordless phone going to be slotted as a communications-based revenue or consumer?

Communications.

Okay, and then in the prepared comments, you discussed a little bit of the drivers behind the computing side. What were they again for the revenue growth sequentially?

Well, the main drivers were the LCD monitors, was that pretty significantly quarter over quarter? And, in fact, every category was up. LCD monitors was up the most, but PC standy, server standby, PDAs, were all up sequentially.

Okay, and then just to revisit a prior question, can you give us a little more detail on the other asset line? You discussed a -- it sounded like a technology loan? I didn't grasp all the stuff you were saying there.

Yes, we made a $10 million loan to one of our vendors to help them fund some technological development that will lead to lower cost over time.

When you say "vendor," you're talking about -- would this be a distribution partner or a --?

It's one of our vendors -- or supplier.

Supplier, yes.

All right.

A manufacturing partners.

Sumit Dhanda with Banc of America Securities.

A few questions here -- first, in terms of the delay that you talked about for the high-volume design ramps for Motorola -- could you perhaps quantify the extent of the revenue impact? Are we talking 1% or 2%? Are we talking 4% or 5%? And what kind of unit volume are you talking about here?

Well, we mentioned that the overall -- once the program is ramped to their full volume, it would be a few million. And to quantify it, as I think we said in our comments, that was the reason that we didn't achieve the midpoint of our guidance. So if you calculate the midpoint of our guidance, that will get you pretty close to the amount of revenue that we were expecting but that got pushed out.

Okay, so the way to think about a fourth quarter guidance, then, at midpoint is really up 1%, and you're going to make up the 3% that you sort of fell short of versus the midpoint fourth quarter?

Well, the delay has impact on two quarters. One, we expected the ramp to start in the third quarter so we would get revenue in the third quarter, and then we would be pretty much at full volume in the fourth quarter. What's happened is the ramp has been delayed now to the fourth quarter, so we lost all of that revenue in the third quarter but now it won't achieve full volume until the first quarter. So it has a negative impact also on the fourth quarter.

Was this a design issue at your partner or what exactly was the issue that caused the delay again?

In the interest of confidentiality, we would prefer not to disclose what it is. All I can say is it is not related to our device.

Let me just move on to something different -- three additional designs in the cordless phone market -- could you give us an idea of what kind of volumes we're talking about for product?

They are smaller designs; they are less than a million units each a year. But we also have additional designs that are currently in progress, in fact, quite a few on cordless phones. These were significant because this is the first cordless phone win that we've had, and as we mentioned in the remarks, it was related specifically to the CEC regulations.

I am assuming, based on your guidance, your backlog was up slightly heading into the fourth quarter?

Correct, it was up about $1 million.

Two more questions -- on the litigation front, are you suggesting that next year, on a full-year basis, it will be less than what we saw in '05? So then is it fair to assume that $2 million, that run rate, will mark, roughly, the peak terms of spending?

It's very difficult to say. As we said earlier and in the comments also, we expect the overall spending for the year to be less than it was in '05, so we also expect the spending to be predominantly in the first three quarters. So we could end up having some spending in one of the quarters that would be higher than $2 million, but it's very difficult at this point to be much more specific than that.

And any reason why you're litigation exceeded your forecast for Q3 -- 1.8 versus the 1.5 that you forecast?

The main reason is the discovery has taken more time and obviously cost more in legal fees than what was estimated by our attorneys back in July. So as they received the documents and went through the documents, they just had much more work than what they had expected back in July.

I guess what I'm driving at is how confident are you that, on a full-year basis for '06, your litigation expenses will, in fact, be lower than what you saw in '05 given what happened with the third quarter litigation forecast?

I think we are fairly confident. I think the amount that we expect less than the $5 million is sufficient that there would have to be a pretty significant change for that not to be the case.

Just two more, and then I will go. Anything you can tell us about how the bookings shaped up through the course of the quarter? Whether one month was weaker or stronger versus the other one, et cetera?

July was a very weak month and, in fact, August was somewhat weaker, but then it picked up. We had very high turns in the month of September. In fact, we had a record level of turns for the month of September -- almost 60% of our revenue for the month of September was from turns orders. So just in the month we had nearly 60% turns orders.

So for whatever reason, customers were a bit cautious early on, and then they flooded us with orders in September, and the orders have remained strong through the first part of October.

And then the final question -- your share count up fairly substantially here. Number one, what's the reason for that; number two, how do we model this, going forward?

Our share count up from where?

The previous quarter?

No, I don't think so. Last quarter we had 30.9 million shares, and this quarter we have 30.7. So it's --

My error, I apologize. Thank you that's all I have.

Matthew Robison with Ferris Baker Watts.

This is actually Mark Downey [ph] for Matt Robison. The first question I have is what was the ending headcount for the quarter?

Three hundred thirty-six.

Three hundred thirty-six. Was that mostly driven by the increase in field?

In field people?

Yeah.

Yes.

Yeah? Okay. Can you speak to the unit volume in ASP specific to DPA-Switch and the DC-to-DC opportunity and the expectations for growth there?

Well, as we mentioned, LinkSwitch and DPA combined were about 6%, and the majority of that is LinkSwitch. So DPA-Switch still represents around 1% of our revenue, and so we expect that to grow. As Balu mentioned in his comments, we have lots of design activity and design wins that are ramping. So over the next several years, we expect it to grow as a percentage of our revenue.

Is there any way to rationalize the unit volume in that area? I'm just trying to find out, you know --?

Our average ASP is about $1, so I think the math will get you there.

Okay, thanks. And last question -- was there any progress above the 30% outsourcing of the IC tests -- I'm just trying to find out, is this an event that's going to happen? I know you spoke that would probably happen next year, last quarter. Is it event-driven or is it just a slow transition?

It's a transition. I'm not sure I want to call it slow, but we're transitioning. This quarter we had about one-third of the testing done offshore, and by late next year we'd like to have about 70% of it offshore, and then with an ultimate goal to have somewhere between 80% to 90%.

So we've made some good progress, but, frankly, we expect to double the percent offshore by the end of next year. So next year will be a key year in the transition.

Doug Jones with Lucent.

Hello?

-- your guidance?

I'm sorry, I didn't hear the question.

In terms of your guidance for the next few quarters, you're talking flat growth -- I'm sorry, if you apply the numbers you gave, and if you take the midpoint for Q4, it will get you to somewhere around $38 million in revenue. You're saying we get to the first quarter, we're looking for flat to down 5%, which is going back to historical seasonal trend, however --?

On the first quarter that was not our guidance. We haven't given any guidance on the first quarter. The question that Balu was answering is what the typical seasonality has been, but that was not our guidance.

Okay, okay. And then, well, let me ask the question this way, then -- in terms of the Motorola business, could we expect that couple of million dollar ramp up to -- would we expect that to come in the first quarter?

As I mentioned in earlier comments, we expect the ramp to start this quarter and then to be more or less fully ramped in the first quarter. So between now and the first quarter, that will definitely contribute to revenue growth.

Steve Smigie with Raymond James.

Hi, this is Ethan Binn [ph] for Steve Smigie. I believe previously we talked a little about Samsung, and I think it was mentioned that Samsung is at 50% off the cell phone revenue, correct me if I'm wrong.

That's correct.

Okay, and that's cell phone revenue. So what percentage of POWI's revenue is from Samsung, if I may ask? And then what percentage of Samsung's revenue is direct from Samsung versus from subcontractors?

Right. Our total shipments related to Samsung products is typically in the mid-teens, and I would -- we haven't added it up, but I would expect that's about where it was this last quarter. Of that, about half of it goes -- or less than half of it goes directly to Samsung, and then the other half is shipped to their subcontractors who build either power supplies or products on behalf of Samsung.

Right. Okay, and I have one more question -- I was just wondering if any of your -- Power Integration's -- components is designed in on the Motorola Razr phones?

The way Motorola uses their chargers is they don't have specific chargers for specific phones unless they have a specialized phone, like a Nextel phone is an example. So we have a percentage of their business, and it's our assumption that our IC would be in potentially any model that they would have that a consumer might buy.

All right, so this is similar to what you were referring to before about Samsung phone chargers as well?

Correct.

Auguste Richard with First Albany Corporation.

Hi, guys, real quick -- when I try to back out the impact of the reserve reversal in Q2 and then try to factor in the delay in ramp from Motorola, it looks like your sequential growth in that scenario would have been about 9%, which is sort of at the low end of your historic seasonal growth in Q3. And I was just wondering if you could talk about any other factors that may have slowed the sequential growth in the current quarter -- or -- Q3, rather?

I think just, you know, the overall economic environment, if you look at prior years that might have been stronger in the overall environment, but I think doing the math that you did, and I would agree in terms of your calculation and the results from it, I think, on a relative basis versus the market and versus other semiconductor companies, it's pretty strong. Other years, we've done better but so have other companies, and the end markets have been stronger. But I think relative to the end markets and the overall environment, it was pretty good.

Okay, and then tax rate for next year?

We haven't calculated it yet. As I mentioned earlier, I wouldn't expect it to be significantly higher and probably not significantly lower. So for the fourth quarter, we're expecting 21%. So it should be relatively close to that.

Got it, and then at 50%, you're bumping up against sort of the higher end of your gross margin range. Do you think you can hold that gross margin, going forward? Or will you sort of give some of your cost savings on to your customers?

It depends on a number of factors, including how aggressive the pricing is in the marketplace and also, to some extent, on the end rate. But if we assume everything is relatively the same -- meaning that the price doesn't get anymore aggressive than it is, we should be able to maintain similar margins, but the only thing that's affected is the external factors, because we have a lot of cost reductions that should offset a normal price reduction. But, again, the timing will not be perfect in terms of exactly how that happens. But unless there is some significant change in the marketplace, we should be -- I can't say exactly how much, but we should be in the same range.

Okay, and then just for MOSFET market, has pricing been fairly stable?

No, actually, all of the components that are discrete -- MOSFETS, diodes, transistors, capacitors, or integrated solutions -- they have been going down at a significant rate, and that's been the case for quite a while now. There was a small period of relative stability in the first half of last year. After that, it's taken the same normal price erosion down.

Okay, and roughly 10% per year?

It a little bit varies from component to component, but, yeah, in that kind of range.

Craig Ellis with Citigroup Smith Barney.

Hi, it's Terence Weil in for Craig Ellis, thanks for taking my question. I was wondering if you could give a little bit more color on the cost reduction in the quarter that led to the gross margin improvement? Did it come through improved wafer pricing or a shrink or was it better pricing on packaging? If you could add some color, thanks.

Well, we have a number of programs going on in [inaudible]. Certainly part of that was wafer pricing, probably more significant part was the shrunk dies that we are using right now because of smaller lithography. That's probably the most significant factor in the last, let's say, one or two quarters.

Okay, great, and then a follow-up if I could -- Balu, I think you mentioned that you thought some of the CEC-related product traction -- you see a significant impact in a few quarters, I believe that's what you said. How would that track to your expectations a couple of quarters ago? Thanks.

It is pretty much in line with that expectation. We expect that customers start doing the redesigns at this stage so that they'll be ready to switch over in the middle of next year. So we see a lot of activity, but we don't expect to see the design wins until probably first or second quarter of next year.

Okay, great, and then, lastly, and related to that, it looks like your LinkSwitch and DPA line might come out to be about 5 percent of sales this year. Where do you see that going next year in kind of a moderate CEC-related scenario? Will that segment of your business of LinkSwitch and DPA, will that approach 10% or surpass 10% of sales next year? Thanks.

That's very possible. Again, we will give you more guidance as time goes on, but LinkSwitch and DPA-Switch should grow quite significantly next year based on what we see right now in terms of design activity.

Todd Cooper with Stephens, Incorporated.

Yes, I'm kind of tagging on to the last question, but can you define for us again what qualifies as a design win for Power Integrations?

We consider it a design win when we get a pre-production order. That's mainly because we go through power supply vendors who then supply to the OEM. So just because we got designed in doesn't mean that they are going to win the designers OEM. So we wait until they place a pre-production order before we consider that a design win.

Balu, I asked that because I'm, frankly, not overly impressed with three design wins related to the CEC, but is that indicative of the design activity or is the design activity associated with the CEC much more active than the design wins you are able to announce?

Well, it's a lot more active than the design wins. In fact, I am surprised we are getting design wins so early. We got a couple of them last quarter, which was a big surprise to me, and we got three of them that we can directly attribute to CEC. There are some others, which are influenced by CEC, but it's hard to directly attribute it to CEC. But most of the design wins, we would see probably in the late and the first quarter or in the second quarter of next year. Because of the way we define design win, it will be relatively close to production.

Shawn Slayton with SG Cowen and Company.

Hi, guys, just a quick follow-up -- these legal expenses for next year, this Fairchild thing, it's supposed to go to the jury sometime in the later summer, is that still the goal?

That's right. They haven't fixed the date yet, but the best estimate is third quarter of next year.

Okay, so after this thing goes to trial, the legal expenses, God willing, drops off to approximately zero for the Fairchild-related litigation?

Yes, it will drop off quite significantly, yes.

And the spending on the System General case should be -- actually, it will start dropping off after the end of this quarter. There will be some left in Q1, but after Q1, there should be very little.

Andrew Huang with American Technology Research.

Thanks, I was wondering if you could just elaborate a little bit more into the increase in the other assets? I think you mentioned a loan to a customer?

During the quarter we made a $10 million loan to one of our suppliers -- suppliers or vendors, whichever term you prefer. And the purpose of the loan was to help fund technological development that would lead to future cost reductions of the product that they supply to us.

Okay, and do you feel comfortable about the liability of this supplier not going away and all that kind of stuff?

If we made the loan, and we didn't feel comfortable, I probably wouldn't be here today.

We did all the due diligence expected to do, and it had to get board approval and so on. So it's not --

Obviously, it's a supplier/vendor that we know very well, but as Balu mentioned, we did lots of due diligence on it, and the purpose of the loan was for strategic cost reduction and clearly before we made the loan, we wanted to obtain some assurances that they would be able to repay it.

So when you talk about, like, technology development or strategic cost reductions, is it on the wafer side or the packaging side or test side?

It was one of our suppliers or vendors, and that's as far as we'll go.

Larry Freitag with North Star Capital Funds.

I just wanted to follow-up on a couple of comments you made earlier in this call. Earlier in the call you said that the competitive environment certainly is very aggressive, remains aggressive. So under the assumption that pricing will remain aggressive going forward, your recent comment was that you obtained cost reductions through a die shrink the last couple of quarters. So if pricing were to continue to be aggressive, where do you think future cost reductions would come from?

That's an excellent question. The biggest factor in maintaining our prices is innovation. We continue to come out with new products that have a higher level of innovation so that we can maintain margins even though the competitive solutions are going down in price. So if you look at our actual price erosion, it's actually significantly less than our competitive solutions, and that is directly related to the innovations we do.

But in addition to that, we have manufacturing cost reductions that are ongoing, and we talked about the shrinks, which is related to a new technology that is smaller lithiography, and we are spreading that across more and more products although it is not a one-time thing. It will happen over a period of time. It's not like we took all the products and did it at one instance. So there is an ongoing reduction that's coming from that.

What kind of yields are you getting on those shrinks as you put them on the new products?

It is -- I can't tell you -- we don't discuss yields publicly, but I can tell you there is no yield difference between the shrunk and the unshrunk products.

Okay.

Okay? And the other cost reduction that is ongoing is the test cost reduction, which is coming through the overseas movement of test. We are doing more and more tests overseas, which is less expensive. The third one is packaging. We are doing a lot of work in packaging to reduce packaging costs in terms of using metrics, tools, in terms of going to lower-cost packages, lower-cost manufacturing methods, and the last one is test time. We are constantly improving our test time so that our total test costs come down. So we are looking at all areas of manufacturing and -- the other one I didn't mention, of course, is the ongoing production of silicon cost, the cost we pay for wafers. So if I look at all of these cost reduction for the foreseeable future, the next two to three years, maybe even four years, we have cost reductions that are in place that will happen gradually, over time. And to a rough extent, to a rough approximation, they should offset the price reduction -- price erosion that we see, which is in the range of 5% to 7% for our products over the next several years. So if the price reduction remains where it is, we should, to a large extent, be able to offset that.

Okay, and then one last question, just a follow-up on the competitive environment specifically. You made the general statement that's competitive, but any particular vendor, whether it's STM or someone else that you're seeing more aggressively than in the past?

I would say discrete is probably the one that is most aggressive because many of these discrete components are made in China, and the Chinese manufacturers tend to be extremely aggressive, willing to take almost no margin on these products, if necessary, in some cases. But I would also say, our integrated competitors -- whether it's Fairchild or SV, are also very aggressive in price reductions.

Are you finding that for you to keep certain business that you're having to lower your cost as well just to keep the business, which might be putting pressure on your margins?

Well, as I said, we try always to use newer products, which have a higher level of integration as a way to stay competitive, because that allows us to keep the margin. So that's our number-one strategy. But we also have to give price reduction -- annual price reduction even though we don't give as much as other people do, but we do have to stay competitive.

Okay, Operator, could we take one last question, please?

Yes, our last question comes from Mr. Shawn Connor with Waterstone.

Just one quick follow-up on your receivables, obviously, they were up this quarter and then, as you mentioned, it was a very back-end loaded quarter. Should we expect the day sales -- or the receivables to come back down in Q4, so it probably wouldn't think it would be as back-end loaded?

I would expect them to come down but probably a couple of days. Not to 30 days but maybe to 34 or something like that.

And the last thing on the share repurchase -- will you begin that this quarter or have you already begun it?

The plan was approved by the board, and it will be put in place this quarter. Whether we actually purchase shares, that will be determined at a later date and in terms of how many we repurchase. But the plan will be in place this quarter.

Ladies and gentlemen, that does conclude today's question-and-answer session. At this time, I would like to turn the conference call back over to Mr. Shiffler. Mr. Shiffler, please go ahead.

Okay, thanks, everyone, for listening. A replay will be available on the investor info section of our website, www.powerint.com. Or you can access a telephone replay for one week. The number for that is 877-660-6853 from within the U.S., or 201-612-7415 from outside the U.S. And the replay account number that you'll need is 3055 with a conference ID number of 171536. Thanks for listening and good afternoon.

This concludes today's teleconference. Thank you for your participation.