微晶片科技 (MCHP) 2005 Q3 法說會逐字稿

Good day everyone and welcome to the Microchip Technology third quarter financial results conference call.

As a reminder, today's call is being recorded.

At this time, I would like to turn the call over to Microchip's Chief Financial Officer, Gordon Parnell.

Please go ahead, sir.

Thanks.

Good afternoon, everyone.

During the course of this conference call, we will be making projections and other forward-looking statements regarding future events or the future performance of the Company.

We wish to caution that you such statements are predictions and that actual events or events may differ materially.

We refer you to our press release of today, as well as our 10(K) for the fiscal year ended March 31, 2004, and our 8(K) current reports that we have filed with the SEC that identify important risk factors that may impact Microchip's business and results of operation.

In attendance with me today is Steve Sanghi, Microchip's President and CEO.

I will comment on our third quarter fiscal year 2005 performance, giving information by geography and product segment.

And Steve will then give his comments on the December quarter, discuss Micro control of product line performance, outline our guidance on the March quarter, and discuss our factory and inventory strategy going forward.

We will then be available to respond to specific investor and analyst questions at the end of these statements.

Our net sales for the December quarter were 205.4 million, up approximately 15.4 percent from sales of 178 million in the prior year's third quarter and down approximately 6.9 percent from sales of 220.7 million in the immediately preceding quarter.

GAAP net income for the quarter was 53.1 million or $0.25 per diluted share, an increase of 30.1 percent from GAAP net income of 40.8 million, or $0.19 per share -- per diluted share from the prior year's third quarter and a decrease of 12.1 percent from GAAP net income of 60.4 million, or $0.29 per diluted share from the immediately preceding quarter.

Relative to non-GAAP net income of 40.4 million in the December quarter of last year, the current year's net income was up 31.6 percent.

Geographically all three geographies were down sequentially.

Europe and the Americas were affected more due to the holiday shutdowns in those regions in the December period.

Europe was down 11.2 percent.

Americas was down 6.6 percent.

And Asia was down by 4.5 percent.

All product areas were also down sequentially.

Analog was down approximately 18.1 percent.

Memory approximately 7.8 percent and microcontrollers down 5.8 percent.

Our gross margins for the December quarter were 56.9 percent, modestly lower than gross margins of 57.3 percent in the September period.

The tax rate for the December quarter was in line with our guidance of 24 percent and we are anticipating continuing at this rate for the current quarter.

Operating expenses were 24.9 percent of sales in the December quarter compared to 23.3 percent in the prior period.

Research and development costs were 23.1 million, representing 11.2 percent of sales.

Sales and general administrative expenses were 28.1 million, representing 13.7 percent of sales.

Microchip's total inventory position at December 31 was approximately 97.8 million, an increase of approximately 7.8 million from the prior quarter level.

Inventory turns were at 3.6 with days of inventory representing 101 days.

These levels compared to inventory days of 87 as of the end of September.

Our distribution inventory is continuing at very moderate levels in all geographies.

And as of the end of December, our distributors held about 2.5 months of inventory, the same as they did at the end of last quarter.

At December 31 our receivable days were 47 and this was 2 days higher than the receivable days at the end of last quarter.

The overall receivable balance decreased approximately 15.3 million quarter-over-quarter, or 12.7 percent.

As of December 31 our cash balance, our cash and short term investment position was 695.8 million, with 44.3 million of debt on the balance sheet.

During the quarter, Microchip generated free cash flow from the business after capital costs, working capital costs, dividend payment and treasury activities, which resulted in adding $90 million to our overall treasury balance.

During the December quarter we did invest 2 million through our stock buy back program -- in our stock buy back program using the leverage on our short term investments to borrow to the debt levels that I mentioned a moment or two ago.

Our capital spending was approximately 21.3 million for the December quarter.

Depreciation expense for the December quarter was 30.2 million versus 27.4 million for the same quarter last fiscal year and 30.8 million in the September quarter.

Capital spending for the fiscal year is expected to be approximately $65 million and depreciation for the year will be about $120 million.

Microchip shipped 11,380 new application development systems to customers during the quarter.

The continued strength of our development sales is an excellent indicator of our new product strength and the design cycle that we are -- we feel is part of our business at this stage.

Our total number of development systems shipped now stands at over 345,000.

I wanted to give you a heads up on two areas that we currently have under analysis here at Microchip and (indiscernible) where we have yet to reach a final conclusion.

The first relates to cash repatriation from offshore locations under the terms of the American Jobs Creation Act where we have the potential to bring cash back to the U.S. under the terms of the act.

Final decisions are obviously subject to our board's approval and would not be implemented before next fiscal year, if at all at this point.

The second area relates to the consideration of vesting of underwater options and the effects of implementing the expensing of stock options after June 15 of this year.

Many companies have implemented changes to their option accruals -- semi-conductor companies, as well as others in different industries.

We are currently analyzing our option pool and will determine if any changes should be considered in this area, again in conjunction with our board.

I will now ask Steve to discuss the performance of the business, last quarter and current conditions, our guidance for the March quarter and outline the factory and inventory strategy we intend to utilize.

Steve?

Thank you, Gordon, and good afternoon, everyone.

First I would like to reflect on the results of the December quarter, then comment on our Microcontroller product line, discuss the guidance for the March 2005 quarter, and discuss our factory and inventory strategy going forward.

So let's begin.

I am disappointed that overall industry conditions have resulted in sequentially down revenue for Microchip.

However, I am very encouraged and proud that our business model has held up once again during a very difficult period for the semi-conductor industry.

Our gross and operating margin percentage remain among the best in the industry and far ahead of our closest competitors.

We have always managed our business through industry cycles based on a longer term focus on overall results, resulting in higher highs and higher lows in operating performance.

We certainly believe that to be the case in the current business environment.

I shall comment on this further as I get to the guidance for the next few quarters.

Before I do that, I shall talk about the Microcontroller product line.

Microcontroller business performed best out of our 3 business units because it was down the least.

Microcontroller business was down 5.8 percent sequentially, but it was up 17.5 percent from a year ago quarter.

On calendar year basis, our Microcontroller business was up a healthy 24.1 percent in calendar year 2004 over calendar year 2003.

Now at this point in time I always hate to bring Freescale into the discussion, but whenever I don't you, the investors and analysts, always bring it up in your questions, so I might as well provide you the comparison now.

Freescale's transportation and standard products group in the last quarter was up 3.3 percent versus a year ago quarter, versus Microchip's Microcontroller business up 17.5 percent versus a year ago quarter.

On a full calendar year basis, Freescale's TSPG segment was up 7.9 percent from calendar 2003 to calendar 2004, while Microchip's Microcontroller business was up 24.1 percent from calendar 2003 to calendar 2004.

Therefore, we are continuing to gain market share.

We are continuing to see a very robust design activity with our 8-bit flash Microcontrollers and we are seeing a lot of new applications emerging which are utilizing our products.

We are very well-positioned to continue to gain market share in this segment.

By the way, our Flash Microcontroller business was up even during last quarter.

Flash Microcontrollers grew approximately 2 percent sequentially and were up 50 percent over a year ago quarter.

On a calendar year basis, our Flash Microcontrollers were up 67 percent from calendar 2003 to calendar 2004, substantially out-performing the industry.

And now I shall discuss our guidance for the March 2005 quarter.

As we looked at the March quarter, we took several factors into account.

First the entire semi-conductor industry is experiencing weak market conditions due to inventory correction.

We also believe that inventory correction that began nearly 6 months ago has almost run its course.

Microchip's direct customer backlog -- this is a very important data point by the way.

Microchip's direct customer backlog which started out about $2 million lower on January 1 versus that on October 1, with 3.5 weeks of January behind us now, direct backlog is now slightly higher than at the same time last quarter.

Which really means that terms have been strong.

So we started on January 1, $2 million lower.

But in 3.5 weeks, we have filled that all up and the backlog is slightly higher than it was at the same point last quarter.

The December quarter is also seasonally front end loaded quarter; versus the current March quarter to be seasonally back end loaded quarter.

Therefore, stronger direct backlog as of now, sort of bodes well for the quarter.

Having said this, March quarter is also our seasonally weakest quarter of the year through the lunar new year coming in February, which results into less shipping days in Asia.

At the same time, there are more shipping days in U.S. and Europe because of lack of Thanksgiving and Christmas holidays.

So putting this all together is more of an art than science, but after taking all these factors into account and after checking expectations from our direct as well as distributor channels, we expect net sales and net earnings in the December quarter to be about flat with the December quarter.

Now I wish to remind the investors and analysts that Microchip historically experiences only one sequentially down quarter followed by a flattish quarter and then followed by growth.

The only exception we had was in the tech bust of 2001, when we had 2 down quarters.

So this time after having a sequentially down quarter in December 2004, and now flattish quarter expected in March 2005, we expect to return to growth coinciding with our seasonally strong June quarter.

At this point in time I would like bring up Microchip's performance through the industry cycles to be higher highs and higher lows in gross and operating margins.

We expect our low of gross margins during this down cycle of the industry to be within 80 basis points of the all time high of 57.3 percent, which in itself would be the remarkable.

We expect March quarter gross margins to be between 56.5 to 57 percent.

Next subject that I would like to talk about is inventory.

Microchip's inventory over our corporate life in the last 80 years has fluctuated between 74 days and 134 days.

So in the high inventory environment during the down cycles, several times over the last few years we have gone in the range of 130 days, 134 days.

In June of 2003, our inventory was 134 days.

Because of the long life cycle of our products, we have demonstrated through several cycles that we do not take large inventory write-offs like many in the semi-conductor industry.

Our current inventory days is 101 days, which we do not consider high since its much lower than the high number and pretty close to really the band of inventory we would like.

At the same time, the distributors are getting moderate levels of inventory at about 2.5 months.

So therefore, Microchip is planning to continue to run our factories at the current production rate even though that rate is slightly higher than our current revenue guidance.

This will increase inventory somewhat.

We expect inventory to peak out at about 115 days or so, depending on the growth in June and September quarters and also the distributor appetite for inventory, which we are currently assuming to be flat at 2.5 months of inventory.

So running of factories at the current production rate, which is slightly higher than our current revenue rate, accomplishes several things.

Number 1.

It maintains gross margin percentage in the current range plus or minus 50 basis points for product mix and ASP variances.

Number 2.

It maintains high equipment utilization and builds inventory actually quite inexpensively.

And number 3.

It lowers the capital expenditures over the next 2 years by utilizing the inventory build by manufacturing higher than the current sales.

So when you look at it over a year or sometime frame, in fact it generates a higher amount of cash because it builds inventory when the factory is running slightly idle and then it lowers the capital expenditures later, saving a substantial amount of cash.

So utilizing this strategy, we currently expect to generate about $730 million of free cash flow over the next 2 years that we have modeled.

We will share that cash with our investors aggressively by increasing dividend as evidenced by a 35 percent sequential dividend increase announced just today.

So let me summarize a few key points.

Number 1.

We see the bulk of inventory correction to be behind us.

We believe that December quarter was a (indiscernible) in revenues.

After flattish revenue in the March quarter, we expect growth to resume.

Number 2.

We expect our Flash Microcontrollers to continue to grow and outperform the industry.

Number 3.

We expect gross margins to be sustainable within 40 basis points of where they are now and within 80 basis points of our all time high margins of 57.3 percent.

Number 4.

We expect inventory to grow in the next few quarters, peaking at about 115 days or so, well below the 134 days peak of the last cycle.

No inventory corrective actions are required and we will continue to run the factories at the current run rate.

Number 5.

We continue to generate a very substantial amount of free cash flow after a relatively low capital expenditure going forward.

And we will continue to share this cash with our shareholders with aggressive increases in dividends going forward.

With that, Angela would you please poll for questions?

(OPERATOR INSTRUCTIONS) First, Adam Parker with Sanford Bernstein.

Hi.

Just a couple quick questions.

Why are you confident that the bulk of the inventory correction is behind you?

Can you share any extra thoughts there, given you said your own days, you're going to grow for a couple of quarters here?

The inventory correction is largely behind us because we see that pattern in our bookings and turns.

As I mentioned on the direct bookings and turns where our direct book-to-bill was near parity last quarter.

We started the current quarter $2 million less in backlog and in 3.5 weeks. we have filled that plus some more and the turns and bookings are stronger in January than they were actually in any of the last 3 months.

So this pick up -- the recent pick up is making you more confident?

Recent pick up plus, if you look at the larger customers who were not buying as much product into October and November because they had some inventory, returned back to buying it.

And seeing our distribution where last quarter, our total distribution inventory in dollars actually declined substantially and the months of inventory and distribution was flat, even though the revenue was down.

So that is really -- does not need any correction.

So we really think that it's largely behind us.

Is that the same thought process that fuels you to believe that you will be back up to mid single-digit growth on a sequential basis in the June quarter?

It's the historical patterns in this recent trend or is there any other color you can throw on to your comments, that you expect to be back up to mid single -- you know, under a normal economic recovery here, up to mid-single growth by June?

Well, you know, during September quarter is our historically stronger quarter.

If you go back in the history of the Company, very rarely we go down in the June and September quarter.

I think the exception was the tech bust of 2001.

And then right after SARS, there was high inventory in Asia and we were still up in that June quarter, I believe.

So June quarter is our seasonally strong quarter and with the inventory correction behind, we believe the stage is set for us to grow in June quarter.

Okay, great.

Thanks, Steve.

You're welcome.

Next we will go to Michael Masdea with Credit Suisse First Boston.

Hi, guys. (Inaudible) here for Michael.

Can you tell us, Steve, either today or at the beginning of the quarter what your backlog coverage was -- the guidance you have?

Yeah, go ahead, Gordon.

The overall backlog coverage at the beginning of the quarter was about 50 percent.

It's about 64 percent as we speak.

And again if you look at it at the beginning of last quarter, it was about 59 percent entering and was about 70 percent at the same point in the prior quarter.

The rate of fill acceleration there is important, but also the fact that we are in the quarter where we have the effect of the lunar new year.

If you look at the complexion of the backlog we have on us -- and again the backlog in terms we are speaking of here are on a POP basis, not on a POS basis.

So that is what our distributors as well as our director customers have placed orders on us.

And with the effects in Asia of the lunar new year being in the second week of February, we do see a tendency for the Asian Disti to lower their inventories -- to be more conservative in placing orders at this time relative to what their habits would be in the prior quarter.

And when you look at the complexion of our visibility and our available backlog, that is one key ingredient that needs to be kept into consideration here.

That's helpful.

Gordon, with lunar new year second quarter can you compare linearity of this quarter and the last quarter?

Is it similar?

You know, it just varies by each of the geographies.

Obviously February will be impacted in Asia.

Linearity in Americas and Europe don't have too many significant holidays.

So if you look at their progression through a 13 week quarter, it would be more linear and perhaps driven by some of the features that, as Steve mentioned -- with, as the inventory correction is improving in each of these geographies, some acceleration of revenue as you go through the periods.

In the December quarter, people are shutting down factories at the end of the year.

Significant, almost two weeks, holidays in Europe and about at least a week holiday in U.S. at the end of December.

So that phenomenon does not take place in March.

When you do the math on all that, the December quarter becomes slightly front end loaded and the March quarter becomes slightly back end loaded and the Chinese new year just makes it that much more in that direction.

Great.

That's helpful.

Thank you very much.

Not to be repetitive here but I promised myself that I would really talk about the turns, to a degree here.

So hopefully everyone on the call can listen and perhaps understand.

I honestly feel that it's a misunderstood statistic, so I would repeat again that this is based on POP.

It's the orders we have from our distributors, as well as the direct customers.

And I often hear comments back to a strict correlation to our revenue target.

Obviously we recognize revenue, as we've said, on distribution on sell out -- not on sell to the Disti channels.

So that makes a tremendous difference.

So I would just offer that for consideration as people look at these statistics.

It does give us an indication of backlog visibility and certainly we are talking today about the fact that we are seeing an improvement in that trend period-over-period.

But that's what it does give us is (indiscernible) business rather than absolute elements.

Thank you.

Next we will go to Mark Edelstone with Morgan Stanley.

Good afternoon and nice job in running the business.

Two questions if I could.

One is -- can you give a sense now as to how large the flash controller business is, as a percentage of your controller business overall, and then I have a question on gross margins.

The Flash Microcontroller is about half of the total Microcontroller business.

Very close, slightly less than half, but it's in that range.

So it's very, very sizeable business.

I believe the total Microcontroller was about $167 million or so last quarter.

So approximately half of that.

It's a very, very sizeable business.

And largest of any of the companies that we know.

Thanks, Steve.

Then just to get a sense for what the margin trends might look like once you start ramping the Fabs back up again.

Can you give a sense as to what the gross margin degradation would look like here in the next quarter or so if that you were actually throttling wafers back enough so that you were not building inventory?

I can't do that analysis on the fly and I -- that's really not the strategy we run the business on.

If you throttle the wafers starts right now -- see, if you look at Microchip's products they have very, very long life cycle.

Our products do not follow (indiscernible) has spoken eloquently about it in the past.

We are still shipping products in volume that we produced 9, 10 years ago.

So if a business model is such that our products have very, very long life time and during these cycles we do not right off the inventory, and if you have -- you build a slightly 10 days of inventory let's say from here and you deplete them over the next couple of quarters, it would absolutely be shameful for us to not use that attribute of our business model in trying to really manage the overall health of the P&L and cash over the cycles.

Therefore our strategy is we are going to run the factories slightly ahead of the current sales level, but we are not changing the factory, we are not lowering it, we are not increasing it, we are leaving it flat.

The inventory might grow from about 101 days to 115 days like I said, over the next couple of quarters.

And they may not get that high, depending on how much the growth is, which is really not very much at all.

And then in the following 2 to 3 quarters, the inventory will slowly come back to these levels.

And we believe that is the best way to use Microchip's assets in building the product at the lowest cost and generating the most amount of cash, which we then can return to the shareholders.

Yeah, I certainly agree with the strategy there, Steve.

I guess what I was trying to get a sense for is the impact that you would have now if you were -- as you are building inventory, so that once that inventory build stops and you start depleting the inventory as the business grows again later in the year, I was hoping to get a sense for what the incremental margins would look like as you are re-ramping the fabs, if you will, and holding the inventories flat once again at lower levels at a time when the business is growing.

I understand the question and I think it's a good one and the best analogy I could give you probably is to go back in our last 7, 8 quarters when we were ramping Fab 4 and gross margin was improving.

Just from there, try to get an incremental gross margin over sales and maybe use that as a proxy or something.

But otherwise, we haven't done that analysis.

I don't want to do it on the fly and you will have also decide which products to build and whether you build a Fab 2 or Fab 4.

What do you do with the wafers?

Do you finish them to the front -- to back end or do you leave them in die?

So the analysis you have to actually make some decisions to really get to that.

The easiest way to look at it is the last time you are growing what happened.

Thanks a lot, guys.

Next we will go to Tore Svanberg with Piper Jaffray.

Good afternoon.

This is actually Jeremy calling for Tore.

Just a question first on your capacity utilization.

Can you just provide an update as to what it is now and also look at the Gresham facility itself and where that is along the ramp.

It really hasn't changed much in the last period as we said.

The Tempe facility is at approximately -- close to 100 percent utilization and Gresham continues to be at very moderate levels of utilization.

So we haven't given out specific guidance on where Gresham is, but obviously, at this stage, although it's no an option that we have, we have tremendous upside capability in that Gresham facility that will return lower manufacturing costs and help to drive our margins when we return to a situation where we are growing the fabs again.

Great.

And maybe just asking it a slightly different way as well.

With forecasts for capital expenditures coming down quite a bit or staying low, how long can you maintain this low level of expenditures?

Well, you know, to figure out the capital expenditure over longer periods of time, you first have to start from the revenue over a longer period of time.

Which we can't guess right now.

Having said that we have been seeing for quite awhile that with the large amount of investment in Fab 4, we could keep capital expenditures in the $75 million range for years to come -- for several years.

And probably much less than that in the immediate fiscal year that's coming up.

And part of the reason is the inventory strategy we are utilizing.

If we were to throttle back the factories today, we would have to turn on more capital and really take more depreciation expense in the September time frame, when we have depleted the inventory and we need to grow more, maybe even earlier, maybe even in June.

By building the inventory now and running our Fab full out, our inventory may grow by 10 days or so but then we will ship that inventory into the growth in June and September quarters and hence actually generating large amount of free cash flow out of no incremental expense, which would we really think is the best way to do it.

That effect is creating a situation where total capital expenditure in the next year would be very low.

Gordon, do we have a number?

We don't have a number as yet.

That's still something that we are looking at.

But suffice it to say, this year was 65 million.

I think it's less than that.

It will be less than that.

I think it's less than that.

It is very much driven through our wafer Fabs.

Great.

Thank you.

Just one last question on your dsPIC.

Is there -- can you provide an update there -- I think you mentioned you entered volume production.

Are there any initial applications driving this production?

Yeah.

At this point in time we have 16 dsPIC products in production.

The original plan that we have laid out for investors was to have 20 dsPIC products in this family and our first product went to production about a year ago.

So in basically one year we have filled out the family.

We have 4 products left which should be done in the next 3, 4 months or so.

So we have populated the family very, very quickly.

We have several thousand development systems that have been sold in the market and a large number of tools are getting sold every month.

We have a large number of design ens, a good number of customers are in volume production.

We already -- our design engineering groups are already moving forward to define the second and third family of products in the dsPIC family.

And there is a road map for 20 products or so in each of the second and third family, which designers are already starting to work on it.

So the family is doing very nicely.

It's very well-positioned.

We are winning designs.

Having said that, we have chosen not to share at all exactly where we are winning because that would be very competitive information and competition could really just come and block those areas -- bring competitive focus in those particular applications.

Great.

Thank you very much.

You're welcome.

(OPERATOR INSTRUCTIONS) Next we will go to Ambrish Srivastava with Harris Nesbitt.

Good afternoon.

This is William Tell (ph) calling for Ambrish.

I was wondering if you could comment a little bit in terms of the overall demands and environment for -- across geography, I guess more specifically within China and Asia since we are still hearing that continues to be a problem area for a lot of the companies.

Thanks.

Well, our revenue was sequentially down in China last quarter and in our flat guidance of the current quarter, we are modeling China to be sequentially down again.

Particularly because of the facts of the lunar new year.

Particularly they will be down a week or two because of the lunar new year and the general weakness that everybody hears about in China.

But this is a very, very good strong growth quarter in Europe because there are no holidays and it's a reasonable quarter in the U.S.

When you put it all together, then you still come up with that flat number.

So in general in terms of distributor inventory in Asia are you seeing that normal -- higher than what you are normally prefer versus other region?

Not at all.

They typically run lower inventories than Europe and Americas.

And there's no difference in the profile that we are seeing at this point.

We do not believe we have excess distribution inventory anywhere in the world.

The average is 2.5 months which has been there for several quarters, which is historically, if you look at it over the many cycles, that would be considered moderately lower end of the inventory spectrum.

The Asian inventory is less than that 2.5 days.

On the average, they hold a lower amount of inventory and over the next one month, they will get even lower because they tend to deplete the inventory going into the Chinese new year, which is the middle of February.

And then when they come back, they quickly run by again for a very strong March.

So at the end of the March, we expect currently that the inventory would be flat.

And one last question.

I was wondering if you could share with us the turns for the December quarter and also what are the turns requirement for the current quarter.

Thank you.

We have answered that maybe in the interest of time, you could listen to the conference call again -- record it.

Next we will go to Paul Leming with Soleil Securities.

Hi, just one question on clarification, then a broader question.

Could you give me the depreciation expense for the December quarter again?

Certainly.

Depreciation was 30.2 million in the December quarter.

Thank you.

The broader question I would be interested in you talking about is you clearly have been incredibly aggressive on increasing the dividend over the last 2 years since you first instituted one.

Can you talk a little bit about 2 issues related to returning cash to the shareholders?

The first issue is just, philosophically, how you trade-off the decision on share repurchases versus dividends.

How you try and strike a balance between them, just in fact how you stack them up against one another.

And then secondly can you talk a little bit about how high you are willing to let the level of cash on the balance sheet build?

Is there some level at which you won't let cash build beyond that point?

I will try to answer that.

It's difficult to be totally complete in giving the answer to this question because it's a tough balancing act.

As far as a dividend increase versus stock purchase is concerned, if I look at stock buybacks for large number of companies in the history over the last several years, my general perception is that the investors believe companies have largely squandered a way cash on stock buybacks.

That may or may not be your view.

I haven't talked to you personally.

In general that is the prevailing view.

So if you look at back in the height of 2000, when company's stock was in many cases 3, 4 times higher than where their stocks are today, they were buying stocks all the way down as it went from 100 to 10.

And then they are still sitting at 10 or 15.

So companies in general haven't been very, very wise in buying their stocks.

I tend to say our record has been better.

In general we have -- our record is better because we only by stock when there are those moments when we believe that investors have thrown the baby with the bath water, if you want me to use that term.

That happened at the end of the SARS fiasco in March, April of 2003, where stock went down to 16, $17, down almost 40, 50 percent and people were still saying, sell this baby.

We bought a large amount of our stock at that time and the stock subsequently doubled during the recovery.

We have a strong business model that tends to do good in good times and bad times.

However, when the times are bad, we could take a sequential hit for a quarter or so like we did last quarter.

We always recover.

We recover very strongly.

So our stock buyback strategy has been very opportunistic, which I have said in these terms.

If investors don't like the stock and they are not buying it, it's down lower than we think it should, then we by a large amount of stocks only in those opportunistic times.

Which doesn't give us a lot of opportunity to continuously use our stock in depleting cash.

Which leaves the -- 2 of revenues.

One is acquisitions and second is dividend.

Microchip management team, in general, do not like acquisitions.

We haven't done very many over our history.

We've done a few small ones here and there to try key pieces of technology.

With don't do large acquisitions.

We don't like them.

They are hard to acquire.

They are hard to integrate.

Cultural differences are too large and usually takes years of diversion to digest an acquisition.

Which leaves us with the third option is to aggressively return the stock to shareholders at dividend.

Right now with our discussion with the board, we have not identified any limits to how far we think dividend percentage can go or how high we will let the cash build.

We have studied S&P 500 and DOW and the dividends -- the average dividend of S&P 500 I believe is 2 percent and DOW is about the same.

The numbers are as high as 5 percent.

Would we go that high?

We don't know.

But certainly we have large amount of cash availability and with the extreme amount of cash we are going to be generating over the next several years, our internal plans are to aggressively return the cash to shareholders.

And percentage somewhat is also depending on stock price so we can't really project that.

That's what I was looking for.

Thanks, Steve.

You're welcome.

Simona Jankowski with Goldman Sachs.

Thank you very much.

Just wanted to follow up on an earlier question where you talked about your Flash-based Microcontrollers about being half of your Microcontroller sales.

Do you have any expectation you can share with us on how much that would comprise by the end of the calendar year?

No, we do not.

We do not.

Really most of the growth of Microchip is really coming from Flash Microcontrollers.

So if you have a growth estimate for next year for Microcontrollers, you can do the math.

But, no, we do not have any longer term guidance.

Okay.

Second question then is, for your inventory increase both in the December quarter and what you are expecting in March, can you comment on the composition of that inventory being treated perhaps by your 3 segments?

Gordon?

Well, certainly with 80 percent continuing to be in micro, and there in product areas where we are very comfortable with the long life of those products, a large piece of the composition will be in that area.

But we can certainly look to keep products in Analog and Memory, where again we are comfortable with the applications and customer penetration that we have.

Overall, we have a balanced approach to the growth in inventory.

Okay.

I know you talked about the long product cycles.

Are we to conclude -- and it sounds like a reasonable conclusion, that there is not really a scenario you can foresee in the next couple of quarters where you have to have a significant write down of the inventory.

Absolutely not or we wouldn't be investing in this strategy.

And history has shown that to be the case.

There is not been a situation where Microchip has had to absorb a significant obsolesce write down in our inventory.

And we have supported through this -- the business model we have here of taking advantage of the longer life cycles and building inventory, effectively low costs.

Plus the inventory is not very high either at Microchip nor is it high at distribution.

And Microchip's products don't really go through a record product transition where a graphics product, one is coming down rapidly because it only sells to 3 customers and they are all going away to graphics product 2.

If I were to take an analogy of a different company and if you have a large amount of graphics product 1 in inventory, nobody wants it after 3 months, that is not Microchip's business.

We sell every one of our products for 10 years plus.

So there is really no issue here.

There is no graphics transition going on of any kind and every product is really going along and if there is a slightly tend to high inventory, that's really nothing.

Thank you.

Next we will go to Jeff Rosenberg with William Blair.

Hi, how are you?

Let's see, first I wanted to ask if you could give us some feel for what memory pricing did during the quarter and maybe some color commentary on the overall market conditions there?

From the perspective of pricing?

Yeah.

Your -- just relative to your competitors, I know you tend to be opportunistic and if you are seeing -- what you are seeing there.

You know, the overall conditions are getting back to more normal conditions and flat to down, low single digits in pricing in December and probably in -- somewhere in the same range for January.

We are certainly seeing the inventory correction take effect in this product area and so we are -- we see the industry conditions as improving in the memory area as we described in the overall business.

So would you say it's been more benign than in some of the more fast growing bigger memory markets where pricing got pretty rough?

Absolutely.

The comment here to add here, Jeff, and for everybody else to note this -- that the analog product as well as memory products were down a larger amount sequentially and actually were the first to recover of our inventory.

Those two products, when you look at various internal charts at Microchip, you can say more strongly that the inventory correction on those is over.

Microcontroller never was high inventory -- it's doing well.

But memory and analog products actually recovered first.

You can actually see from our gross margins here, obviously the 2 components of gross margin are utilization rates in the factories and ESPs.

And we've talked to utilization.

You can see that overall margins are essentially the same.

There are other issues of mix, et cetera, that reverberate around the business model.

So when you look at our business versus these other companies who have much stronger pressure on their ESPs, you can see that the basic kind of differences in those business models which are not inherent in our own.

And I also wanted to ask you on the gross margin -- are you still getting a benefit from mix relative to the Flash growing as a percentage of sales or is that no longer as much of -- I think you felt like that was a pretty nice advantage early on when it was ramping.

I wouldn't break that out as a significant difference and right now it's such a large portion of the business that it would just be average.

The last thing I wanted to ask, Steve, if there was any color you are willing to provide on the comment you made in your prepared remarks about your seeing penetration in new applications?

I guess I'm wondering if that's being driven predominantly by new products like the 610 product and the dsPIC.

Or any kind of color there in terms of where you are seeing success in penetrating new end market applications.

First there are a large number of areas -- we introduced over 15 new products last year, maybe even more than that.

And with the breadth of the product portfolio filled, a lot of the new products we now announce are all generating incrementally serve available market because they are doing things we didn't used to do before.

So there are a very large number of new applications emerging everywhere that we are going into.

Secondly I think we have gotten increasingly conservative to not talk about that information because, over the years, when out the bullseye, everybody is looking at what is Microchip doing.

Why do they grow 24 percent when the competition is growing 7?

And I think that information we believe is damaging to us so we don't provide it.

Okay.

Out of the 50 products how many of those were analog, roughly?

That does not include analog.

These are 50 plus products of Microcontrollers only.

They are all highly integrated analog Microcontrollers, which means there is a large amount of analog on the Microcontroller.

None of the Microcontrollers being produced are not pure digital.

They all have a very high amount of analog, but they are still called Microcontrollers.

In addition to that, we have a large amount of analog products by themselves.

I think I may have a number, hold on a second.

In the last year we introduced 50 Microcontrollers, 42 new analog products, and 16 new dsPIC products.

Great.

Thanks.

Sumit Dhanda with Banc of America Securities.

Good afternoon, guys.

I just wanted to follow up again on the gross margin question.

Steve, if I look at what happened in early 2003 with a similar situation you had, one down quarter and then you started to see a resumption of growth, but your margin recovery took about 4 quarters to materialize and your inventories were at roughly similar levels on a days basis.

Is that how we should think about getting back to peak margins for Microchip going forward?

I think that would be a reasonable proxy.

I'm not sure about the exact time without modeling it.

The reason it takes longer for the margins is because, if you are building inventory high like we are trying to raise inventory, and then you bring it down, so during all that time you are not growing the factory.

This time around, the inventory does not go as high as it went right after SARS.

So the timing of it is slightly different.

But in general, your thinking the right way.

Okay.

Thank you.

Gil Alexander (ph) with (indiscernible) Associates.

My question has been answered.

Thank you very much.

Ruben Roy with Pacific Crest Securities.

Yes, thank you, Steve.

Sorry if you talked about this already, but the direct business improvement that you are seeing, was that specific to any end markets or can you provide a little more detail on that?

Thank you.

The end market commentary on Microchip hasn't really changed.

We just looked at numbers -- looking at our last calendar year versus the numbers we have been providing for the prior calendar year.

And it really almost was no change.

About 34 percent of our business was consumer, 18 percent was automotive, I believe 17 each was office and industrial, and about 13 or 14 percent in the communications.

That's correct.

So quarter after quarter and year over year, these numbers really tend to not change.

We are serving close to 46,000 plus customers right now and, at any given point, over a short period of time, our business does not get driven by either one market or one customer or one application.

Okay.

Thank you for that.

How big a percentage is distribution right now?

About 65 percent of our business is through distribution.

Thank you, Gordon.

Craig Ellis with Smith Barney.

Thank you for taking the question.

I just had a clarification on the prepared comment and some of the Q&A with regard to revenues through the cycle.

You said a flat quarter followed by a quarter of growth.

What would be, looking back over the Company's history, what would be the path to seasonally normal growth?

Would it be a quarter, 2 quarters?

Can you just help us understand what trajectory you think you might be on as we look at getting back to seasonally normal growth quarter to quarter?

One thing nobody knows is where the industries growth rate is going forward in absolute terms.

If you give me a number, I will divide that for you seasonally.

Do you have a number?

You want to divide 10 or 20 or 25 or 15 or what number would you like to divide?

That's one thing that we do not want to provide longer term guidance.

But if you give me a number, I will break that up seasonally for you.

Is that fair?

I see what you are saying, Steve.

The point is that you think you can get back to growth, you're not sure if it's going to be seasonally normal growth, but growth nonetheless in the calendar year.

I think it gets back to seasonally normal growth.

June and September quarter are 2 stronger quarters, and December is weak and then March is also weak.

So that's a pattern.

In terms of trying to put it numerically, I would be giving you guidance for the next 4 quarters, which I do not want to give.

Got you..

Thank you.

You're welcome.

There are no further questions at this time.

Mr. Parnell, I would like to turn the conference back to you for closing or additional remarks.

Thanks.

We appreciate everyone's attendance today.

We know it's a very busy day with many, many companies reporting earnings.

So thank you for hanging in here with us.

Steve and I will both be available for any follow on questions for the rest of the day.

Take care.

Thank you.

That does conclude today's teleconference.

We thank you for your participation and have a wonderful day.