Description

Overview

Background

SA demand is a function of both HDD unit growth and the number of SA’s required per drive.  In 1997, the average HDD had roughly 3.35 platters per drive with an SA on each side of each platter for a total of roughly 6.7 SA’s per drive.  In the following years, however, drive makers made enormous leaps improving areal density which is the amount of data that can be packed into a given amount of space.  Areal density improved so quickly during this time that drive makers were able to increase drive capacities while using progressively fewer platters and thus SA’s per drive.  The number of SA’s per drive thus fell from roughly 6.7 in 1997 down to roughly 2.3 by 2003.  Even though industry drive shipments roughly doubled over this period, the reduction in SA’s / drive led to a 30% decline in the SA market.

Storing data on each side of a platter allows you to double a drive’s capacity for little incremental cost compared to using only one side of the platter, so SA / drive bottomed around 2.3 in 2003 and then increased slightly in the following years.  This has allowed the SA market to more closely track the drive industry’s growth since that time.  There have been tentative indications that areal density improvements are today falling behind storage demand growth which could increase SA / drive going forward.  This would result in additional SA demand beyond what I am expecting in the coming years. 

Competitive Dynamics

One of HTCH’s competitive strengths is rapid product development which is valuable since hard drives are commodities, and HDD manufacturers thus want to develop and ramp their new drive programs quickly in order to maximize the short window of time in which they enjoy little competition.  Another of HTCH’s competitive strengths has been product consistency which is very important since the robots that assemble hard drives require exacting tolerances for the components that they are handling.  Higher consistency thus translates into higher yields for the HDD manufacturers.  The last of HTCH’s competitive strengths has been greater technological capabilities than its competitors.  This has allowed it to produce certain products that its competitors have struggled to produce and enabled it to source components internally instead of procuring them from outside vendors.  This vertical integration provided HTCH with a cost advantage since it could better optimize the overall cost of an SA. 

Importantly, HTCH’s ability to capitalize on these strengths has not been constant over time but has instead fluctuated with the pace of change and degree of technological advances that the drive makers needed from the SA vendors.  Over the last several years in particular, drive makers were able to improve their drives by advancing other components such as the heads and platters which took much of the edge off of HTCH’s value proposition relative to its competitors.  HTCH’s current development activities, however, indicate that drive makers once again need advances in the SA which should play to HTCH’s strengths.

2003-2005 were probably HTCH’s halcyon days because its technical abilities were strong competitive differentiators during that time.  The company’s fortunes turned dramatically in the following years as its marketshare fell from roughly 60% in 2006 (when there were four SA vendors) down to only 20% today (when there are three SA vendors).  There are a few reasons for this.  The first was the slowing technical advancement of SA’s that was mentioned above.  In the late 2000’s, SA technology slowed and drive programs began to last longer than they had previously.  These both allowed HTCH’s competitors to close some of the gap.  In addition, aggressive capital expenditures from 2005-2007 left the SA industry with significant excess capacity as it moved into the recession.  This was particularly problematic for HTCH since its two remaining competitors (MPT and NHK) were both subsidiaries of Japanese conglomerates which made them much less sensitive to earning a profit.  (For whatever reason, some Japanese conglomerates seem to be willing to subsidize unprofitable subsidiaries indefinitely.)  The result of all this was four brutal years of marketshare losses and steep ASP declines which led to dramatically lower revenue and operating losses.  To make matters worse, HTCH had a looming debt maturity in early 2013 which, combined with its weak financial performance caused customers to question its viability as a supplier.  This, of course, simply reinforced the company’s problems.  In mid-FY08, however, HTCH began its first steps in a multi-year process of dramatically improving its cost structure and balance sheet. 

Operational Restructuring

Up until the recession, HTCH had resisted its customers’ requests to produce suspensions in Asia near their HDD factories.  HTCH instead produced from three factories in the Midwest US, reasoning that its processes were already relatively automated and that the engineering capabilities underpinning its rapid product development and ramp speeds was not readily transferable to Asia.  I believe that refusing to move to Asia niggled HTCH’s customers, and the higher costs of US production would later prove to be material when price declines accelerated during the recession. 

HTCH undertook a small restructuring in 3Q08 followed by a major restructuring in FY09 that saw the closure of one of its three Midwest US factories and the construction a new factory in Thailand.  HTCH’s Thai operations ramped throughout FY10, and in FY11, the company began another major restructuring of its US operations.  The end result is that HTCH’s three original US factories have been whittled down to one US R&D center and one US factory where HTCH will use its US engineering teams to develop and ramp new programs up to volume.  Once up to volume, new programs will be transferred to HTCH’s low cost assembly operation in Thailand.  Results from the June and September 2011 quarters indicated that the company was finally on the path to recovery, having earned customer’s confidence in its Thai facility and new TSA+ production line, the latter of which is discussed below.  Just as the clouds were parting, however, October 2011’s flooding in Thailand swamped HTCH’s Thai factory.  That facility is now re-ramping and expected to reach pre-flood capacity by mid-FY13.  While the flood has pushed out the full realization of HTCH’s cost reductions, the company was still able to demonstrate to its customers that it has a credible path to reclaiming low-cost leadership.

Attaining the lowest-cost position is also being underpinned by rapidly growing adoption of a technology called “TSA+.”  In the late-1990’s, drive makers connected the head to the rest of the drive with a twisted pair of wires that ran across the top of the SA.  As head-sizes shrank, a smaller connection was needed, and HTCH developed the Trace Suspension Assembly (TSA) which used a subtractive process to remove conductive metal from the SA, leaving thin leads across the top of the suspension.  TSA eventually became the industry standard.  As heads continued to shrink and add functionality, drive makers saw the need to create even finer leads, and this could only be achieved through an “additive” process—adding conductive material to create the leads instead of removing material.  These additive flexures were originally only available from Nitto Denko which can be thought of as the 3M of Japan.  So at their customers’ request, HTCH undertook the expensive, multi-year endeavor of developing and scaling their own additive process, TSA+.  TSA+ adoption gathered steam in FY10, though HTCH experienced two significant setbacks at its TSA+ line during that fiscal year.  By FY11, HTCH was reliably producing TSA+ at high volumes as the line continued to ramp, and TSA+ is expected to be 80% of HTCH’s production by FYE12 (up from 55% in the March, 2012 quarter).  The additive flexure on a TSA+ suspension represents roughly half of the cost of the suspension, and once HTCH’s TSA+ line has scaled up, they expect to have a significant cost advantage over their competitors who will both have to source from Nitto Denko.  I believe that this new position as one of only two additive flexure suppliers in the world should improve HTCH’s strategic positioning.

Debt Restructuring

As mentioned above, HTCH previously had a significant amount of convertible debt that was putable to the company in January, 2013 which encouraged customers to limit their exposure to the company.  HTCH successfully restructured this debt through a series of three tender / exchange offers which were completed in February 2011, July 2011, and April 2012.  While each tender / exchange offer was different, the net effect has been to increase HTCH’s interest rate, decrease the strike price, reduce the total borrowings, and push the earliest put date on the new notes to January, 2015.  HTCH now has only $11.9MM of the original notes that are putable in January, 2013 which is quite manageable. 

While HTCH’s TSA+ and Thai facility efforts have been underway for years, they didn’t begin to prove themselves to customers until around mid-2011 when both of those operations had begun to establish track records of reliable high volume production.  In addition, HTCH’s balance sheet was a cause of concern until about two months ago.  I expect the resolution of these issues to improve HTCH’s ability to regain marketshare going forward. 

The NAND Boogieman

One of the greatest risks to HTCH and the rest of the industry has been the threat that HDD’s will be displaced by NAND (a.k.a. Solid State Drives or SSD’s).  This could occur either directly by replacing HDD’s in PC’s and servers or indirectly by tablets replacing PC’s since tablets currently only use NAND. 

SSD’s primary cost is the NAND “media” itself, so SSD’s total costs rise linearly with drive capacity.  Said another way, a SSD has essentially zero fix costs and its variable cost rise with the amount of media (i.e. capacity) placed into the drive.  A HDD, by contrast, has fixed costs for components such as the motor and actuator plus variable costs for the platters and suspensions.  Consequently, HDD’s have a higher fixed cost than SDD’s, but their variable cost are relatively small and increase in long shallow steps as drive capacity increases.  So from a cost perspective, SSD’s are ideal for devices such as MP3 players and smart phones which do not require very much storage because at those capacity levels, SSD’s total costs are lower than HDD’s.  Because SDD’s costs rise linearly with capacity while HDDs’ flatten out, SDD’s are roughly 10x more expensive for devices such as PC’s that require significant storage capacity, and this greatly increases the total cost of the device.

SSD’s have some other real and illusory advantages over HDD’s.  SSD’s are much faster at “random reads” of data, meaning data that has to be searched for as opposed to data is read sequentially (like playing a movie).  This is part of why NAND allows devices to power up so quickly compared to HDD’s which require a longer boot cycle if the device is turned off.  SSD’s also have lower power consumption, though this is less of an advantage in PC’s since the drives do not constitute a large part of the power budget.  SSD’s are also perceived to be more reliable since they lack moving parts, but I have read reports that the cells that store data within a SDD can only be rewritten over a certain number of times before they fail.  This will lead to reliability issues if the cells are rewritten over extended periods of use.  I believe that this reliability problem gets worse as the cells get smaller which they have to do for NAND to get cheaper.  NAND proponents cite these real and perceived advantages as reasons that SDD’s will eventually displace HDD’s.  SSD’s have obviously already displaced HDD’s in many Apple laptops, but those are premium PC’s as illustrated by Apple’s mere 5% share of the global PC market.  The preponderance of the market has proven to be extremely price sensitive and unwilling to pay several hundred dollars more for a PC with a relatively small amount of storage. 

I’ve been grappling with the NAND risk for years now, and I don’t believe that SSD’s will live up to their hype or extend beyond the premium segments of the market.  For most of the market, the most likely outcome seems to be that PC’s will eventually use hybrid architectures that store applications on NAND and data on HDD’s.  This could be achieved through a “hybrid drive” or by using standalone NAND and an HDD components.  Such an architecture would provide the best of both worlds since it’d provides features such as “instant-on” and allows applications to load quickly, but provide those features at a much lower cost since you’d only be using the premium storage where it can improve the user experience.  If the NAND reliability issues that I have read about are correct, this architecture would alleviate them since the data would be stored on an HDD.

With respect to tablets, there seems to be a growing admission that tablets are best suited to consuming data, but that you still need a PC to get work done, not to mention store your data.

Lastly, I’d mention that NAND proponents usually point to NAND’s falling costs as a reason for adoption, but they ignore the fact that HDD costs are falling as well.  In addition, building enough NAND production capacity to displace a large part of the HDD market would require $10BB’s of capital spending. 

Industry Consolidation

Over the last two years, the HDD industry has consolidated dramatically with STX acquiring Samsung’s drive business and WDC acquiring Hitachi’s.  This leaves only STX, WDC, as well as Toshiba who has a relatively small marketshare.  HTCH has historically had a very good relationship with WDC, but they did not supply much content to Hitachi, so there is the opportunity to gain share at the combined WDC / Hitachi over time.  HTCH lost most of its business with STX in early 2009 after STX announced that it would concentrate its SA orders with a single supplier and HTCH was underbid.  STX was ultimately unable to leave HTCH completely because HTCH has unique capabilities for enterprise drives, but HTCH’s business with STX has been very limited since then. 

What is very encouraging is that the drive makers are currently preparing to launch new drive programs that involve technology transitions, and this has opened up new opportunities for HTCH across the board, including at STX.  One of the new technologies being adopted is Dual Stage Actuation (DSA) which adds a piezoelectric motor to the SA in order to facilitate more precise control of the head’s positioning.  HTCH is the only SA vendor that has shipped DSA in volume in the past, and increasing adoption of DSA, and TSA+ along with drive maker’s upcoming technology transitions should help the company regain some marketshare.  During the last two calls, management indicated that they have been qualifying on new drive programs across the board, and they hope to begin regaining marketshare in the next 12-18 months.  They had previously indicated that they should be able to get their marketshare back to at least 30% which would be a 50% increase from current levels on a larger TAM since industry HDD shipments continue to rise. 

InSpectra

HTCH has a fledgling Biomedical division that is based on metrology technology that they developed for its SA manufacturing processes.  Their product, InSpectra, uses infrared light to monitor how much oxygen is present in capillaries (a.k.a. tissue perfusion).  This has proven useful in a number of medical applications, but HTCH has focused on trauma applications since trauma wards lose money on many of their patients and are thus motivated to adopt a diagnostic tool that will help them avoid unnecessary costs.  For example, declining tissue perfusion has  proven to be a good indicator that body is going into shock which is very useful since the vital signs are not helpful in detecting this.  Knowing whether a trauma patient is going into shock allows physicians to know whether they have time to use radiology to better diagnose the problem or whether they are running out of time and need to begin surgery. 

InSpectra comes in a “SpotCheck” version that takes an instant reading to determine whether a patient is a candidate for perfusion monitoring as well as a traditional monitor version that will monitor perfusion on an ongoing basis which is required for using it as a diagnostic tool.  The monitor uses single-use sensors which are sold under a razor / razor-blade model.

Clinical results have been encouraging over the years, though InSpectra has yet to see broad adoption.  HTCH is conducting a large multi-site, outcomes-based study that was supposed to conclude during the March, 2012 quarter, but management decided to expand the study by including additional patients.  They have not said when the study is now expected to conclude, but I’d expect it to be finished within the next 6-12 months.  I am not ascribing any value to InSpectra, but it would radically transform the financial profile of the company if it gained traction in the market.  This division was steadily growing revenues off of a small base before HTCH’s restructuring efforts led them to begin selling through distributors. 

Upside in nearly every scenario

If you remove the portion of HTCH’s convertible debt that is classified as equity, HTCH’s TBV at the end of the March quarter was $7.99.  This compares to a current stock price of $1.46.  Because STX and WDC now control such a large portion of the HDD market, it would make sense for one of them to vertically integrate into SA’s.  (SA’s are one of the only parts of the HDD supply chain, if not the only part, that is still only served by merchant vendors.)  HTCH’s unique position as one of only two additive flexure sources would make such an acquisition interesting since it would force the other HDD vendors to sole source from Nitto Denko.  To be fair, the SA is a small portion of a HDD’s total cost, but it’s still be strategically undesirable to have Nitto Denko as a sole source.

So how much would a HDD maker be willing to pay to vertically integrate into suspensions and acquire one of the only two companies in the world that can make additive flexures?  50% of TBV seems like a really good deal for the HDD maker, and that would represent a stock price of $4.00, or 174% higher than the current price.  What if they were only willing to pay 30% of TBV?  That’d be $2.40, or 64% over the current price.  As you can see, HTCH’s price is profoundly low which creates enormous upside in nearly any scenario. 

Now that the restructurings have been completed, HTCH has a well defined cost structure that can be used to estimate earnings and FCF with simple assumptions about SA shipments and ASP’s.  HTCH is currently incurring additional costs to produce in the U.S. while it re-ramps its Thai facility, but the company has provided excellent visibility into what it’s cost structure will be once this is completed by mid-FY13.  I have estimated that FY13 EPS and FCF will be $0.22 and $1.14 respectively.  (FCF is higher than EPS because D&A is much higher than Cap. Ex.  HTCH also reports significant non-cash interest expense associated with its convertible debt.)  This implies P/E and P/FCF multiples of 6.7x and 1.3x, respectively.  I don’t expect HTCH to ever receive a high multiple, but even if it traded at a 4x FCF multiple, the price would be $4.56, or 212% higher than it is today.  30% of TBV and 4x FCF are silly multiples, but they illustrate how wide the margin of safety is at today’s prices. 

I ultimately believe that HTCH could reach $8.00 or more in the coming years depending on how successful they are regaining marketshare.  The HDD market is expected to continue growing, so moving from a 20% share to 30-35% share would result in enormous revenue gains and operating leverage.  As explained above, impending technology transitions at HTCH’s customers are playing to HTCH’s strengths which has created openings, and they believe they are performing well on the underlying development programs. 

Risks

I have followed HTCH for about seven years now and endured serial disappointments with their suspension and biomeasurement businesses.  This gives me a natural concern that I could simply be wrong about the prospects for recovery, though it is difficult to imagine the drive makers allowing HTCH to fail and their supply chain to further consolidate.  This is especially true given HTCH’s new position as a reliable TSA+ supplier.  I believe that the largest risk to my thesis is timing and that it could take them a year or two longer to regain share than I am expecting.  Consequently, it’s quite possible that my FY13 estimates won’t be realized until FY14.  As illustrated by the return figures I’ve shared above, however, the margin for safety is so enormous that the annualized returns would remain excellent even if it takes a couple more years to realize my price targets than I’m expecting. 

Catalyst