Affymetrix Q2 2006 Earnings Conference Call Transcript (AFFX)

| About: Affymetrix, Inc. (AFFX)

Affymetrix, Inc. (NASDAQ:AFFX)

Q2 2006 Earnings Conference Call

July 31, 2006 5:00 pm ET

Executives

Stephen P.A. Fodor, Ph.D. - Chairman and Chief Executive Officer

Gregory T. Schiffman - Executive Vice President and Chief Financial Officer

Doug Farrell - Vice President of Investor Relations

Analysts

Quintin Lai - Robert W. Baird & Company, Inc.

Paul Knight - Thomas Weisel Partners

Derik De Bruin - UBS

Un Kwon - Infinium Capital

John Sullivan - Leerink Swann & Company

Ross Muken - Deutsche Bank

Tycho Peterson - JPMorgan Chase & Co.

Operator

Good afternoon. My name is Derek and I will be your conference operator. At this time, I would like to welcome everyone to the Affymetrix second quarter earnings conference call. All lines have been placed on mute to prevent any background noise. After the speaker’s remarks, there will be a question-and-answer session.

(Operator Instructions)

Doug Farrell

Thank you, Derek. Good afternoon, everyone and welcome to the conference call. At the close of the market today, we released our financial results for the second quarter of 2006.

The goal of this call is to expand on the contents of our earnings release. Participating in the call with me today Steve Fodor, our founder, Chairman, and Chief Executive Officer; and Greg Schiffman, Executive Vice President and Chief Financial Officer.

As a reminder, today’s call is being recorded and the audio from the call is being webcast over the Internet on our homepage at www.affymetrix.com.

During this call, we may make various remarks about the company’s future expectations, plans, and prospects that constitute forward-looking statements for purposes of Safe Harbor Provisions under the Private Securities Litigation Reform Act of 1995. Such statements are subject to risks and uncertainties that could cause actual results to differ materially for Affymetrix from those projected. These risks factors are discussed in Affymetrix’s Form 10-K for the year ending December 31, 2005, and other SEC reports, including our quarterly reports on Form 10-Q for subsequent periods. We encourage you to review these documents carefully. Forward-looking statements are made as of today’s date and we expressly disclaim any obligations to update this information.

With that introduction, let me turn the call over to Steven Fodor.

Stephen P.A. Fodor

Thanks, Doug. Good afternoon, everyone. This last year has been the single most-challenging one for Affymetrix in our 13-year history. We have had a series of internal and commercial challenges that have been compounded by increased competition, especially in the genotyping market.

Given these realities, we are focused on two issues: increasing our top-line revenue and reducing our operating cost structure to improve profitability.

Our plans to increase revenue include new product solutions for the genotyping market, new advances in the expression arena, and continued progress in the clinical products market.

First, I would like to address our cost-saving measures. Since we have entered into a period of operating loss, our general and administrative cost structure has become out of line for the revenue base of the company. During this quarter, we will reduce our expenses by restructuring corporate, general and administrative expenses while continuing a strong focus on sales, product R&D, and product launches.

These savings should bring the company into more profitable operations and position the company for a healthier expenses structure for 2007. Greg will provide you with more details in a moment.

In addition to reducing costs, we will focus on building revenue in all areas.

In genotyping, we have made progress on several fronts. We made substantial improvements in getting our customers up and running at production levels during the second quarter of this year. We improved manufacturing yields on our 500K product and added manufacturing capacity. We also worked closely with the genotyping community to develop and deploy second generation software to optimize 500K performance.

As we announced recently, we are now focused on introducing next generation genotyping products. Information we have gained in tens of thousands of experiments using our whole genome assays is being used to develop new products such as the 1 million-SNP product that we are developing in collaboration with the MIT Harvard Broad Institute.

To set the stage for our new product introductions, we have forward-priced the two-chip 500K product at around $250. Although this pricing will affect Q3 genotyping margins, it anticipates the price of the single-chip product that will be introduced in Q4 at attractive margins.

The 1 million-SNP product will feature an approximate four-fold increase in the number of snips per array. This level of information density will allow us to deliver SNP content at a price point that will enable large-scale genetics to become better statistically powered and ultimately more common-place. It will enable us to include a wider sampling of genetic diversity.

This is an aggressive plan -- a plan that depends on high manufacturing leverage, but it is a plan that we believe will fundamentally enable the most successful scientific approach for large-scale genetic studies.

Over the coming quarters, just as we have done in gene expression, we will leverage our manufacturing and technological infrastructure to pursue product offerings with higher and higher genetic content at aggressive price points.

In the expression market place, we continue to increase product breadth to include higher resolution exon and tiling representations of genome content. These products, with more comprehensive and more accurate views of the genome, require time for adoption, but there is no doubt that the field will evolve to these more information intensive assays. We are just beginning to see the leading indicators of adoption in the way of published scientific literature.

In Q4, we will be introducing a new, 96 well reader that will provide core laboratories the ability to transition to automated [inaudible] expression analysis. Over the coming quarters, many of the assays now provided in cartridge format will be adapted to the 96 well format.

I would like to end with a few comments about clinical diagnostics. We now have eight commercial partnerships with 18 cancer diagnostic products in development. We believe that these products will be important growth drivers over the coming years as we help bring complex genetic tests to the regulated diagnostic marketplace.

This fall, we will open the doors in our CLIA testing laboratory in West Sacramento with plans to offer dozens of different molecular tests. Stay tuned for updates and announcements in this important strategic market.

Finally, we have opened a routine search for a senior level commercial operations leader. Our objective is to recruit experienced talent to help transition our established markets into well-run, long-term profitable growth opportunities.

In closing, although this past year has been challenging, we continue to reinvigorate our product line to drive our top-line growth and institute operational cost-savings in order to improve profitability for the company.

Let me now turn the call over to Greg.

Gregory T. Schiffman

Thanks, Steve. Good afternoon, everyone. Before I review our operating results for the second quarter, I would like to give you some details on a restructuring that Steve mentioned.

The primary focus is on getting our operating expenses back in line and improving our profitability. Over the last five years, we have gained significant cost leverage utilizing a financial model in which our operating expenses grew at half or less the rate of our revenue growth.

That model served us well through the middle of last year, when operational challenges impacted our top-line growth. During our first quarter conference call, we told you that we were taking steps to reign in our expenses. Based on our second quarter results, we are taking a more aggressive approach to restructuring to bring our operating expenses in line with the business.

Our primary focus is in the general and administrative functions and will include rationalizing our facilities. We intend to take a charge in the third quarter as a result. We will be providing additional details on the scope of this restructuring and the charges when we complete our analysis.

Although we have made significant progress in addressing the operational and technical challenges that we have faced over the last several quarters, improvements in financial results lag operational improvements.

On a GAAP basis, the company reported a net loss of approximately $10.1 million, or $0.15 per diluted share in the second quarter of 2006, as compared to net income of $7.8 million, or $0.12 per diluted share in the second quarter of 2005.

Excluding the impact of FAS-123R, the company reported a non-GAAP net loss of approximately $6.2 million, or $0.09 per diluted share in the second quarter of 2006. This loss includes $5.5 million in taxes, or $0.08 per share.

Product and product-related revenue in Q2, which excludes sales to Perlegen, was $75.5 million. Product and product-related revenue included array and service revenue of $44.8 million, of which around $4.5 million was genotyping services, reagent sales of $9.6 million, and instrument sales of $11.2 million. This is compared to $45.9 million in array revenue, $9.8 million in reagent revenue, and $13.5 million in instrument revenue in Q2 2005.

During the second quarter, we achieved total consumable sales of roughly $56.6 million, which includes arrays, reagents, and genotyping services, as well as $2.2 million in products sold to Perlegen, as compared to $57.6 million in Q2 2005, which included $1.9 million in sales to Perlegen. The overall mix of consumables for the second quarter was about 70% from RNA analysis products and about 30% from DNA analysis products.

Our gene expression business was in line with our initial guidance, with the exception of our exon and tiling products, where the rate of adoption was slightly lower than expected. However, we see strong interest in customers and are aware of peer review publications that are in the works.

Our instrument revenue in the second quarter include 39 new GCS 3000 systems, consistent with the prior year. This brings our cumulative systems shipped to about 1,445 units. The difference in instrument revenue over the prior year was driven by reduced sales of system upgrades and automation.

Royalties and other revenue were $2.4 million in the quarter. Royalties consist primarily of the amortization of up-front license payments and minimum royalty obligations under existing license agreements. Other revenue is principally comprised of research revenue earned from government grants.

Cost of product in product-related revenue, which excludes cost of product sold to Perlegen, totaled $27.1 million for the quarter, generating a product and product-related gross margin of 64%. This included charges incurred while ramping up our new plant in Singapore. We have nearly completed the final validation of the plant and expect to begin shipping commercial product from Singapore during the second half of the year.

We have many customers that we are in dialog with on large genotyping studies which will be carried out over the next several years. In order to enable these projects on our platform today, we are offering forward pricing on our current 500K. Using the manufacturing leverage that we have demonstrated over the last decade, our costs will be reduced in half next quarter, allowing us to preserve margin while expanding the market.

Our decision to reduce the price of our two-chip 500K mapping set during the third quarter, ahead of our scheduled release of a single-chip 500K in Q4, will contribute to a lower gross margin on genotyping products in Q3. We expect to be exiting Q4 with more traditional chip margins for genotyping products.

Perlegen revenue was $2.2 million during the second quarter, with a cost of product sold to Perlegen of $1.1 million.

SG&A costs were $39.5 million for the quarter, as compared to $32.5 million in Q2 2005. This growth includes approximately $5.2 million associated with stock-option expensing under FAS 123R and ongoing SG&A costs associated with the acquisition of ParAllele. The remaining delta over the prior year was driven by increased litigation expenses.

On July 17, Judge John Sprizo of the U.S. District Court for the Southern District of New York ruled on the constructions of 15 disputed claims involving eight patents. The court adopted the key interpretations proposed by Affymetrix and six codefendants relating to eight [enzo] patents covered by the order. We are pleased with the decision and look forward to the successful conclusion of this litigation.

With regard to our ongoing patent litigation with Alumina, we are awaiting the [Markman] decision from the honorable district court Judge Farnan of the District of Delaware.

Research and development expenses for the quarter were $21.6 million as compared to $20.8 million in the second quarter of 2005. Our current R&D priorities include next generation genotyping products, which Steve described, and 96 well plates for automation. The $1 million growth in R&D was driven by stock option expensing under FAS 123R.

For the quarter, the company recorded a net interest and other income of $3.3 million, which included currency gains of around $800,000.

As a reminder, this is our first year booking a fully-burdened tax rate for GAAP reporting. Depending on the level of profitability and the jurisdiction in which it is generated, we will see significant variability in our reported tax rate from quarter to quarter. For example, during the second quarter 2006, we had an income tax expense of $4 million on a GAAP basis, despite having a loss for the quarter. The bottom line is that on a cash basis, we expect to pay approximately $2 million for the year.

As discussed in today’s press release, the company has been engaged in an internal review, performed on the direction of the audit committee of the board of directors of its historical stock option grant practices from January 1, 1997 through May 31, 2006.

This internal review has revealed certain documentation lapses in the period 1997 to 1999, including one instance where the company has determined that the option grant date should have been recorded differently. The review does not indicate that there was any pattern or practice of inappropriately identifying grant dates with hindsight in order to provide discounted or in-the-money grants.

Based on the company’s review to date, and given the dates of the grants in question, the company does not anticipate any material adjustment to the financial results of operations included in this press release.

Let me take a moment to summarize our balance sheet. During the second quarter, our accounts receivable DSO improved to 76 days, in line with our target range. Inventory for the second quarter of 2006 was $45.1 million, an increase of $4.4 million or 10.8% over the prior quarter. This was primarily due to a build-up of consumables as we prepared to transition some manufacturing capacity to Singapore. We expect to see inventory turns improve during the second half of the year.

For the quarter, the company was cash-flow positive, generating about $19.5 million from operating activities. Our cash and marketable securities balance decreased by about $7.6 million to $269.2 million, including the investment of about $28 million in capital expenditures associated with our manufacturing expansion.

While the last year has been a painful one, we believe that we are turning the corner and expect to emerge as a much-stronger company in 2007. We have a tremendous market opportunity in front of us. We have a robust product pipeline and a highly scalable technology that will continue to substantially differentiate Affymetrix over the coming quarters.

We believe that this market is one of the most exciting opportunities in health care today, and that we are uniquely positioned to deliver comprehensive solutions at the right price point to accelerate market growth.

At this point, we would like to open the call for questions.

Question-and-Answer Session

Operator

(Operator Instructions)

Your first question comes from Quintin Lai with Robert W. Baird.

Quintin Lai - Robert W. Baird & Company, Inc.

The 39 instruments that you placed during the quarter, can you break that out to customers using it for gene expression versus genotyping?

Gregory T. Schiffman

Quintin, that is a very difficult one because actually a lot of the customers that we sell to are utilizing both technologies and really solving the problems that they are dealing with. If we look in particular at the area of cancer research, which is a large percentage of our customer base, they are applying both the expression products to understand the nature of the disease they are looking at in genotyping to understand how individuals are responding to treatment, or how it is affecting, and so from that standpoint, I do not think that we can give a clear indication one way or the other.

Quintin Lai - Robert W. Baird & Company, Inc.

Maybe a question for Steve. Your competitor has announced several large genotyping contracts over the same time period that you have also improved your manufacturing yield and gotten your next generation algorithm. How do you think that 1 million product that you are going to launch later this year is going to win over the new customers to come in and adopt the Affy technology?

Stephen P.A. Fodor

Quintin, that is a great question. I think there is no question that as we look retrospectively, we did launch the first products into this area, and I think did a pretty good job of opening up the field. Unfortunately, a lot of our operational mistakes I think created an opportunity for competition to come in and launch their own products.

As we looked at this, we are certainly recovering on the application of the 500K. I think there was a time period there where the reputation of the 500K suffered. However, I think that it has improved dramatically. As we started to look at this, we really thought we could incrementally approach this business area or we could really try to drive the whole field to a new level, and that is what we decided to do.

Part of this is this forward pricing of the 500K immediately to make the whole genome studies of 500K much more cost-effective for the marketplace, and second of all now, using a lot of the real world experience that we got in commercializing the 500K as feedback now, technical feedback. We are incorporating all those changes to actually drive to the million level.

Taking 1 million-SNP’s, which will just have superb genetic coverage of the genome and then putting that out in a very high-leveraged and I think very price-competitive way, I think will change the paradigm once again in genetics because it will allow people to do two and three times as many samples for the same sort of cost structure that they could have done in the past with that level of genetic coverage.

Time will tell on this, but that is going to be our strategy, is one is very aggressively pricing the 500K, and then two, launching this next generation 1 million-SNP type product in order to really push the fields forward.

Quintin Lai - Robert W. Baird & Company, Inc.

Last question, and then I will jump back in. A lot of these genotyping contracts are very high-volume -- lots of samples being processed. Are you going to take that 1 million-SNP ship and then go out of the cartridge onto the 96 well plate format?

Stephen P.A. Fodor

We are going to do both. We will initially launch it into the cartridge format, and even in the cartridge format today, with the carousels and so on, it has pretty good throughput in those systems, but again, as we mentioned, we do have a new 96 well reader that is being launched in Q4 of this year, and we will be porting all of these assays onto that format.

Quintin Lai - Robert W. Baird & Company, Inc.

Thank you.

Operator

Your next question comes from Paul Knight with Thomas Weisel.

Paul Knight - Thomas Weisel Partners

What was your year over year array revenue? What was the Q2 revenue versus last year?

Gregory T. Schiffman

Sure. So if you look at year over year, Q2 last year was $45.9 million. This year, we are at $44.8 million when you take array and services, which is predominantly arrays that are being processed at Affy on behalf of other customers, so that is what we have referenced before as the comparable metric.

Paul Knight - Thomas Weisel Partners

What do you get when you back out service?

Gregory T. Schiffman

The actual just service only is not a component that we provide. It is $4.6 million was the total services, but a large component of that is [REA’s] and reagents.

Paul Knight - Thomas Weisel Partners

Why with the large ramp-up in Sacramento capacity are you showing this type of flat or even down year-over-year revenue? Are you pricing down in Q2?

Gregory T. Schiffman

The capacity expansion that we have put in place actually was one that we had been running through last year. We are expanding in Singapore, and Singapore is a pretty strategic expansion, one I think long term in terms of needs of the company as well as very tax efficient means of operations as we moved, looking long-term at getting our tax rates down.

From the standpoint of Sacramento, I think we have the yields and the capacity in place. I think we are well-positioned, but I do not think it really relates to volumes this quarter.

Doug Farrell

Paul, this is Doug. Maybe I can help out a little bit too. Our second generation algorithm, that was not widely available until we were into the second quarter, so there certainly were customers as we had flagged back in February in an 8-K filing that some of the reorder rates were running lower than expected. Getting the new algorithm out to people was critical for that. That happened some way into the second quarter, depending on the customer, so it was kind of an April-May timeframe, so there certainly were some projects that were stalled in the first-half of the quarter while people got their software updated.

Paul Knight - Thomas Weisel Partners

Any booking trend that is positive? Order trend?

Gregory T. Schiffman

I do not think we can comment on specific orders, but I will say that we have been getting an awful lot of feedback, both from customers that we had not spoken to at all previously that I do not think had been looking at running any genotyping experiments that said the price point where it is at is not going to enable them. We have also certainly been having discussions with a lot of other participants in the marketplace today and how this might affect their decisions on experiments going forward.

So all the feedback has been very positive on the announcement we put out recently, but I do not think we would be commenting on specific booking trends because it is very early first month of a quarter.

Paul Knight - Thomas Weisel Partners

How is your pricing now on a genotyping project of a thousand arrays compared to any competitor that might be out there?

Stephen P.A. Fodor

It depends on the products, Paul. Certainly it depends on what assay somebody might be using, but with the current reduction on the 500K at $250 or so for 500,000 SNP’s, we believe that is best-in-class pricing out there. Obviously one of the key things that makes this genotyping market so much different than expression is an expression, you can run a small number of arrays and see trends and things that were consistent. In genotyping, because everyone is 99.99% alike to begin with, you have to do very large sample sizes to understand the significance of these subtle differences. That I think is the fundamental change in the marketplace.

Paul Knight - Thomas Weisel Partners

Thank you.

Operator

Your next question comes from Derik De Bruin with UBS.

Derik De Bruin - UBS

Good afternoon. Could you walk me through the price volume calculations that gives you confidence that you can cut price without critically wounding your gross margin over the long-term? In other words, how much additional volume do you need to pull through in order to compensate for the price cut?

Gregory T. Schiffman

Actually, Derik, I guess on this one, we look at it a little different. This is similar to what we did in expression, where we went from five chips to two chips to one. The concept we sell today is on two chips. We have essentially cut the price in half of where we were selling it. We will be providing it on a single chip next quarter, so our cost cut in half and the margins will be consistent with what we have been seeing back at the early part of the year.

From that standpoint, you do not require any real volume growth to fundamentally hit it. I guess there is some spreading of fixed costs or leverage there, but the primary drivers on it are the fact that it goes to a single chip instead of two.

Derik De Bruin - UBS

I was curious about your comments on the tiling and the exon arrays. Could you tell me, give us a little more color on what has been the hold-up in getting some of those pulled through?

Stephen P.A. Fodor

There has been pretty good scientific up-take in the leading -- I guess you would say the thought leading labs, but there is still -- even when we first introduced our expression products in the beginning, it takes a little bit of time for the first papers to come out and for transitions to go on during the product evolution.

There is clearly, the large majority of the marketplace today uses things like the U133 or the gene level representations of the genome as the primary product. That is the big majority of users.

The tiling arrays and the exon arrays are just now being introduced into a number of different centers where people are putting them through the ropes and dealing with the information issues and analysis issues as they learn to use them.

Unlike what happened in the first introduction of expression, they are being used even in -- particularly in things like biomarker discovery in cancer research and so on, and the people that are using them uniformly come back and say this is much more powerful, I get much better markers, and I am seeing things that I did not see before. It just takes -- all our expectations aside, it takes a while for it to be seriously adopted I think by the marketplace.

Derik De Bruin - UBS

Some of that has been at your core gene expression business, it is still running about mid-single digits growth year over year?

Gregory T. Schiffman

Yes, I would say, Derik, the core gene expression business is low- to mid-digit growth, single-digit growth.

Derik De Bruin - UBS

Finally, one more question, and be patient with me, because I am not an accountant. Could you walk me through what is going on with your taxes?

Gregory T. Schiffman

I will say the GAAP reporting of taxes is a very confusing area. What we have taking place right now is we have incurred losses, and when you incur losses domestically, and you also have overseas operations, which you are starting up that are in a slight loss position at this point because we have not begun shipping product, you are not able to offset any -- you basically do not get the chance to be able to use any of the losses overseas to offset profit that you are making in your international locations other than Singapore, so you end up reflecting a tax rate that you are paying, not being able to take out the losses in the low-tax jurisdictions.

As we look at next year, we should see a lot more stability in our tax rate and we should see it moving down towards the low 30%. The challenge we have this year is really as you start looking at the one-time charges in Q3 and the losses this quarter, coupled with starting up the offshore operations, we are going to see very unusual rates reflected through, but on a cash basis, we will be paying a very [diminished] AMT rate.

Derik De Bruin - UBS

That is actually very helpful, but then you could say that on a non-GAAP basis with the $0.06 in stock option expenses, then it was basically a $0.15 loss -- is that how I am readying this?

Gregory T. Schiffman

A $0.15 loss on GAAP -- with the taxes in there from a GAAP reporting, that is correct.

Derik De Bruin - UBS

I will call you afterwards. Thank you.

Operator

Your next question comes from Un Kwon with Infinium.

Un Kwon - Infinium Capital

Could you give us an idea of what percentage of your installed base has the MegAllele software upgrade?

Stephen P.A. Fodor

Off the top of my head, I do not know the exact amount. It is a relatively small proportion that are running the MegAllele assay relative to the whole genome genotype end products at this point. We can get you a better breakout on that, but it is a relatively small proportion that are running it, relative to the 100K and 500K type applications.

Un Kwon - Infinium Capital

Do you see any pick-up in demand in the second-half of this year for those targeted genotyping products, or is that more of a 2007…?

Stephen P.A. Fodor

There are a number of very specific panels that are being released for targeted genotyping. I think one of the things that will end up playing out long-term here, and we will have to see how this shakes out, I think Greg mentioned that we had some very positive comments on this pricing structure from some of our customers. We have had some customers that have said “I wanted to do a targeted genotyping” but with this sort of cost structure, I might as well just run the whole genome. So this is exactly what we saw happen in expression as we moved everything to higher and higher density and higher and higher content, although we still get some requests for custom chips and flexible content in expression analysis these days, that is a very much smaller piece of the entire marketplace.

Genotyping will probably be a little different. It will be a combination I think of where people want to do validation studies. With this type of pricing, they will be able to just run the whole genome assays again, which will then actually increase the statistical power of all the studies that they have done, and so that is one piece of it.

On the other hand, there are a number of -- a series, both in human and other bio-organisms, as well as infectious disease, where people want to use very targeted genotyping for very specific questions.

This is a very volatile marketplace right now, with a lot of change. I think we will just have to see how it shakes out over the next year-and-a-half.

Un Kwon - Infinium Capital

Secondly, can you describe the SNP selection and content for the 1 million-SNP chip? Will this include a lot of hapmap SNP’s or does it allow you to skip SNP’s throughout the genome or will it be restricted by your restriction enzyme base assay?

Stephen P.A. Fodor

It is a combination of a lot of things. There are about -- there are with the 500K product today, I think there are around -- forgive me if I do not have the numbers exactly right, but around 300,000 plus or minus SNP’s that are in fact in the hapmap and another 150,000 or 200,000 that are actually not even include in the hapmap studies that are in fact in the genome.

I believe the numbers are right now, we are screening another 1.3 million, 1.2 million SNP’s. I think around 800,000 of those were actually included in the hapmap analysis, and another 500,000 or so were not.

So this will be a combination product that includes SNP’s that have been included in the analysis of the hapmap, but also a number of different SNP’s that in fact are not even present in the analysis of the hapmap sample that will be using the enzyme [fractions] are similar to what we are doing with the 500K, but they will be expanded.

Un Kwon - Infinium Capital

Lastly, would you be able to walk us through how some of these contracts are structured for these large genotyping studies? If a customer wanted to do a 2,000 patient sample and they order all the chips up front, is there flexibility where they can sort of shift vendor, or are they locked into the contract? Is there a follow-up on the discount that you are going to be offering on the 500K, will that flow through to the customers that have already ordered, that are actually in the middle of their studies?

Gregory T. Schiffman

I think we can only comment really on the contracts that we have done as a company. In general, as we have signed customers up for large studies, and if there are commitments that they make when they placed orders and we have shipped product, and typically we would -- these things run over a couple of years, so we do not ship all the product in a single shipment. Once products are shipped, it really has become the customers. We are not able to take products back on that basis.

I do think customers throughout an experiment would always have the potential, if they have not placed orders, of running whatever product they would choose. In general, I do not think they are obligated on orders that they have not placed with a company going forward. At least, I think that is how we have seen contracts that we have dealt with historically.

Un Kwon - Infinium Capital

What is the average order size?

Gregory T. Schiffman

That really does vary in a wide range when you look at these studies. I think most of the people look to try to stock up somewhere between three to six months of supply of consumables on hand, so if they are able to run the studies and keep it moving, it is probably in that type of a range.

Un Kwon - Infinium Capital

Thank you.

Operator

Your next question comes from John Sullivan with Leerink Swann.

John Sullivan - Leerink Swann & Company

A couple of quick ones -- can you tell us what the return rate is running at right now for the 500K two-chip set? I am hearing numbers in the 5% area. Is that the right neighborhood?

Gregory T. Schiffman

You are referencing with regard to warranty returns?

John Sullivan - Leerink Swann & Company

Chips coming back to you, right.

Gregory T. Schiffman

That probably sounds like it might be about right, John. I do not know that I have that -- it has been a number which has come down dramatically over the past few months, and I do not know that I have the latest one at hand, but it would not surprise me that the numbers you are talking to are consistent with where I believe it is at it. We can certainly get back and confirm on that one.

John Sullivan - Leerink Swann & Company

Can I ask you this -- in the near-term, how low do you think -- what is a reasonable target for that quality control measurement? Can you get it down to 1%? Can you get it down to 2%?

Gregory T. Schiffman

Traditionally, we have seen returns at a 1% to 2% range, of which I will say the returns really vary between cases where products were bad to issues where customers may not have processed, done their target prep and processed the chip correctly. In all cases, we tend to take the returns back, but it is in that 1% to 2% range.

Just as we looked at improving yields in the 500K and getting it to the 90%, I think we would expect over time in genotyping to see the return rates and failure rates equally move that direction.

John Sullivan - Leerink Swann & Company

Can you just talk about as you -- in the near-term as you lower price for the two-chip 500K set drives more volumes, how much capacity do you have? How much ability do you have to meet increased demand at the lower price point?

Gregory T. Schiffman

As we indicated earlier, by the end of this year, we expect to have increased our throughput capacity in wafers by essentially 100% from where we exited last year. So it is a fairly substantial leverage that we have in terms of being able to run additional chips. I think one that is important is you look at the volume growth here, and we are certainly targeting that this is going to expand the market. We would not be doing this if we thought that the volumes were constant. It is going from two to one chips, and so fundamentally, we manufacture half the number of chips for the current experiment, so it is one that we are looking to see the market expand quite a bit. I think we have heard a lot of positivity across the board in this marketplace. We are just looking to really get our share of the market as we go forward.

John Sullivan - Leerink Swann & Company

Last question -- as you work through the improvement process for the 500K and fair rate and return rate, are there things that have been learned of a global nature that you will be able to apply to the proposed 1 million-SNP product? Or should investors worry that you will see the same sort of evolution from a QC standpoint in the 1 million-SNP?

Gregory T. Schiffman

Let me start, and then we will go to Steve. One, and I think it is an important one, is to understand that the 1 million-SNP is really the same feature size, same kind of an assay that we are working with today. So I think we have learned a lot in the process, and those will leverage. I will turn it over to Steve for that, but this is not a large technical shift. It really is an expansion of what we have already been doing.

Stephen P.A. Fodor

I think your question was do we have to re-learn from the lessons all over again, or is this truly something that we learned and used the advances from the previous experience. That is definitely the case.

There are a lot of things involved in it. One is, of course, as we shipped out the volume of chips that we did last year, we have instituted different measures of chip quality and chip validation as we shipped them out, and actually have instituted some very strong functional testing of our chip manufacturing runs, which made dramatic improvements in what we were shipping.

Second of all, the original products that we sent out had a massive number of probes on them that looked at the SNP’s in different ways. Now, what has happened is we have literally run billions of genotypes and customers have run billions of genotypes off of these chips, and so now we know which of the chip designs are most efficient now in actually capturing the SNP information, and that is all going to be implemented on this next generation.

It is going to be a pretty aggressive program, but the evolution from the two-chip set to a single chip is relatively straightforward and very, very low risk. As we go to 1 million, we think that it is an aggressive program but it does not have the same risk profile as the initial introduction of half-a-million SNP’s into the general marketplace.

John Sullivan - Leerink Swann & Company

Could you just comment briefly, can you describe for us how the collaboration with the Broad Institute will work? As far as I know, you folks have mostly developed your own products. You are calling out the Broad Institute as a partner in the development of the 1 million-SNP product. How will that work as a practical matter?

Stephen P.A. Fodor

As a practical matter, we are going to work with them in order to take the lessons, as you know, or as you may know, the Broad I think is up at a very, very high level of -- what is it, about 3000 -- 2500, 3000 samples a week or something? I have forgotten now, but they have done an enormous amount of processing on this technology.

Again, the knowledge base of the product performance and the central features of what we need to extract to go to the next level really reside within them, and they have an absolutely spectacular set of experimentalists as well as bio-informatics guys that we are drawing on in order to design the next generation.

We actually are working very collaboratively with them to bring that knowledge into this entire field. The idea is, as I mentioned earlier, not to stop incrementally around half-a-million and stops around too much there, but to really push the field so that we can offer very, very much increased capacity at really, really attractive price points in order so that the statistical genetics can be done well.

This product will have a significant number of SNP’s, obviously a million. It will have a significant representation of both the hapmap as well as SNP’s that are not even in the hapmap, and this will power, given that we will be able to do this in price points that will be very attractive, the population sizes that can be run for these diseases will be powered all the more better.

That is the strategy that we are taking, and we are working with Broad mainly because they are really good geneticists and we are going to push this along.

Operator

Your next question comes from Ross Muken with Deutsche Bank.

Ross Muken - Deutsche Bank

Clearly I think the strategy going forward seems to be as if you feel you can be a cost leader and thus a price leader. Can you talk about the sensitivity to price you think your customers have from an academic versus a commercial perspective, and really how important price is you think in the mind of the purchase for both of those customer groups?

Stephen P.A. Fodor

Again, you need to think about -- it is not quite as black and white as price and cost leader, as you put out.

The issue when talking about statistical genetics is to get the number of samples up and run so that you can actually get statistical power out of these studies. I think that this is a new and very volatile marketplace, and we saw a couple of years ago, we really did not even have an offering in this area. Last year, we blossomed pretty strongly into this area. This year, we are facing some competition, and so we are going to drive this to the next level.

It is about, can you push this field to a point where the statistical genetics can be done not just at a few places but can be done very, very broadly? So there is obviously premium pricing to be had when we are just looking at a very limited technology base with a limited set of customers who can pay. The question though is if we are really going to enable human genetics in the broad sense as well as other genetics, can we enable the true statistical power of this, which means we have to look at both how many markers can we spread across the genome at what sort of price points in order to really enable these large studies.

We have had some feedback. Like I said, a couple of initial feedbacks, although we do not want to make a big deal out of this. I think the next few quarters will show, but we do get comments back from people now that say “I never thought I could afford to do this, and now this is getting into a point where I can actually play this game”. The second type of comment that we have got, I mentioned earlier, is that some people that are thinking about doing focused panels to look at a certain area because they thought that is all they could afford to do are now saying “well, shoot, I can get much better genetic power now by just including the entire genome and increasing my data set”.

Again, I think time will tell over the next few quarters about how elastic this marketplace is, and in terms of how many new studies do you open up and how many new customers do you bring in by offering this sort of genetics at this price point.

Gregory T. Schiffman

I think one thing that is important with it also is to recognize that I do not think a lot of the strategies that we are doing have really been different than what we have done historically. I mean, if you look at the expression market where it was five chips and we dropped it down to one, we have gone from providing 60,000 transcripts to a million exons on a single chip, and the price point again is pretty similar.

We are doing the same in genotyping. I think one thing that really has differentiated Affymetrix historically, and one area that we do feel we demonstrated for a long time an awful lot of leverage is the manufacturing of photolithography. Fundamentally, we are not shrinking feature size at this point. We are actually just enhancing the design of the products, and over time will be reducing features and continuing to gain leverage and get less and less expensive products. That has continually opened up the marketplace if we look at the history of the company.

So I think there is an awful lot being raised. The reality is, we are doing a little forward pricing because historically, we have not had people that are signing up contracts that tend to be multi-year projects, and to not get that information out where you are a quarter away we felt was not going to be doing the right thing by our shareholders or by us as a company.

I think that is a little different, maybe being a little more aggressive, but the strategy in what we are executing to, this is a ten-year history of Affy and I do not think we see that changing.

Stephen P.A. Fodor

It is actually a good point too. Initially, in gene expression, if you will remember the criticisms of Affymetrix back in the mid- to late-90’s, people would say “Yeah, but it cost $2,000 or $1,500 or $2,000 a sample to run a gene expression experiment” and how expensive is that, and so on.

I think here we are just in a slightly different dynamic. We came out with the first whole genomic genotyping product. We are being now -- we have some competition in this area, and so therefore we need to [fight] competition looking to drive the market to the next level. It is pretty much that simple.

Ross Muken - Deutsche Bank

Just quickly on the cost side, you discussed the restructuring, more focused corporate G&A -- is this a long-term solution or more of a short-term solution? One would think that the crux behind the whole price versus volume tradeoff is that you believe over time you will be able to capture a bigger piece of the pie, and thus you will see recovering sales and gross margins that will be more in line with where you were historically, but would that not also sort of note that you would also have -- you would need a similar cost structure to where you thought you would be, given that growth trajectory you had been on, given a recovery.

Where is this coming from? Is there a little bit of discongruence there?

Gregory T. Schiffman

I think that as a company, if you just looked over the last several years, we have realized that our cost structure in terms of what we were spending, we were spending outside of a lot of the industry norms, and we put a model in place that we are looking to grow operating expenses half the rate of revenue.

As we talked with investors and if we looked at it as a company, we made some investments, if we go back about a year-and-a-half ago, in infrastructure, really targeted at top-line revenue growth and areas where we were driving the market.

I think that as we look at where we are at right now, growing into our cost structure in terms of the G&A, it makes sense to structurally just make the changes now, get us into a position where we are in a position of strength and growing from there. We are very focused in terms of funding the sales efforts, with feet on the street and all of the efforts associated with growing the market, as well as development activities and getting products out there.

I think this is just one that it would not be prudent as a management team not to address it in a little quicker fashion than we would have addressed this if we had been growing the top-line revenue.

Ross Muken - Deutsche Bank

Just one other thing from what you had mentioned on the call, so basically from a management perspective, you guys are looking to add someone at the -- sort of a chief operating officer, is that right?

Stephen P.A. Fodor

Well, I think we are going to keep some flexibility here, but the question is, sort of level of a president of the company, or the extent of what this person should oversee will depend a lot on the talent that we bring in and the qualifications that the person brings in.

I just wanted to add to what Greg said. We have begun to really separate out a little bit more about what our objectives are from a life sciences perspective and how we are managing that in terms of the principal customer segments in the pharmaceutical industry as well as the academic segment, and then beginning to really manage to a slightly different drum what we are doing in the clinical perspective and in diagnostics. So long-term, I think part of this is Affymetrix has grown up into this level. This is a good opportunity for the company to try to get its cost structures and its organizational structures in a little bit better sense.

Ross Muken - Deutsche Bank

Thank you.

Operator

Your final question comes from Tycho Peterson with JPMorgan.

Tycho Peterson - JPMorgan Chase & Co.

Thanks for taking the call. A number of the questions have been answered, but I guess following up on the last question on the cost structure, Greg, did you mention there was going to be some facility rationalization in there as well?

Gregory T. Schiffman

One of the activities underway, and I think it is one that, in light of current market conditions, is just understanding facilities we have, what makes sense, and how do we want to move forward from where we are at.

I think that over the next month, month-and-a-half, two months, we will be able to provide more updates to you with regards to what we will be doing, sort of the size, magnitude and charges that would be flowing through.

Tycho Peterson - JPMorgan Chase & Co.

Second, on the -- as I look back over the inventories over the past year, year-and-a-half, turns have obviously come now and data has gone up. What is a reasonable way to think about turns over the next two to three quarters? What is a good goal to have?

Gregory T. Schiffman

There are two things that have driven the inventory turns, if we looked at it over the past year-and-a-half. One of them has been a shift to some of the instrument manufacturing where we are working with third party OEM providers, and so you are no longer working with raw material that you are purchasing, but actually inventorying entire units with an OEM profit built into them.

I think we saw our turns fundamentally shift a bit when we introduced a couple of those products.

With the transition to Singapore, we are building some inventory to consumables just to assure that we do not have any shortage of supply. We have essentially completed qualification of plant. We are bringing products over and have seen good results from them, but we just do not want to have any risk there.

I think as you look at turns going forward, I think we traditionally talked somewhere in the three-and-a-half to four-and-a-half times, and that is probably still in a range with four to four-and-a-half the most probable as we look in the near future.

Tycho Peterson - JPMorgan Chase & Co.

Finally, on the Roche collaboration, I know you typically like to defer a lot of the questions to Roche, but can you just give us an update if there is anything we should think about near-term in terms of launches, and then what the status is right now, the 450, how that is doing in the market?

Stephen P.A. Fodor

You are absolutely right. We will refer you to Roche for any updates on their products. What I will say is that the interaction with Roche is very positive and we see a lot of progress for the future there.

Gregory T. Schiffman

There are several products in the pipeline, but I think specific status, they have talked about some specific status, and we really do have to refer you back to Roche.

Tycho Peterson - JPMorgan Chase & Co.

Thank you.

Operator

Please proceed with your presentation or any closing remarks.

Doug Farrell

Thanks for taking the time to join us in the call today. If you did miss any portion of the call, a phone replay will be available for the next seven days beginning at around 5:00 p.m. specific time today. To access the replay, domestic callers should dial 800-642-1687 and international callers should use 706-645-9291. The pass code for both is the same, it is 2926144. Alternatively, an audio replay is available under the investors section of our website at www.affymetrix.com.

Thanks again for joining us and have a great day.

Operator

This concludes the Affymetrix second quarter earnings conference call. You may now disconnect.

Copyright policy: All transcripts on this site are the copyright of Seeking Alpha. However, we view them as an important resource for bloggers and journalists, and are excited to contribute to the democratization of financial information on the Internet. (Until now investors have had to pay thousands of dollars in subscription fees for transcripts.) So our reproduction policy is as follows: You may quote up to 400 words of any transcript on the condition that you attribute the transcript to Seeking Alpha and either link to the original transcript or to www.SeekingAlpha.com. All other use is prohibited.

THE INFORMATION CONTAINED HERE IS A TEXTUAL REPRESENTATION OF THE APPLICABLE COMPANY'S CONFERENCE CALL, CONFERENCE PRESENTATION OR OTHER AUDIO PRESENTATION, AND WHILE EFFORTS ARE MADE TO PROVIDE AN ACCURATE TRANSCRIPTION, THERE MAY BE MATERIAL ERRORS, OMISSIONS, OR INACCURACIES IN THE REPORTING OF THE SUBSTANCE OF THE AUDIO PRESENTATIONS. IN NO WAY DOES SEEKING ALPHA ASSUME ANY RESPONSIBILITY FOR ANY INVESTMENT OR OTHER DECISIONS MADE BASED UPON THE INFORMATION PROVIDED ON THIS WEB SITE OR IN ANY TRANSCRIPT. USERS ARE ADVISED TO REVIEW THE APPLICABLE COMPANY'S AUDIO PRESENTATION ITSELF AND THE APPLICABLE COMPANY'S SEC FILINGS BEFORE MAKING ANY INVESTMENT OR OTHER DECISIONS.

If you have any additional questions about our online transcripts, please contact us at: transcripts@seekingalpha.com. Thank you!