Pure Storage: The Great Price War That Wasn't

| About: Pure Storage (PSTG)
This article is now exclusive for PRO subscribers.


Pure Storage announced the results of its fiscal Q3 Wednesday afternoon.

Results were a significant beat on both revenues and earnings.

Expense management was a visible component of the beat.

The company forecasted moderate but not hyper revenue seasonality in the coming quarter.

The company is beginning its rollout of NVMe, a technology it believes will be as disruptive in the storage space as the introduction of flash.

Pure Storage - It's Valuation Shrinkage Time

Pure Storage (NYSE:PSTG) reported the results of its fiscal Q3 that ended 10/31/16. I guess I would describe the results and the guide as one of the unexciting beats. I was amused in reading one commentary this morning to find the author unhappy with results that diverged from his model by 80bps. While the language was temperate, the thoughts - quelle horreur. An 80-bps divergence from his forecast in non-GAAP product gross margins. Armageddon must soon be at hand.

The beat was basically seen by investors and analysts as to be expected. The shares were down around 10% today although that was partially a reflection of a very visible drubbing taken by the entire tech group based on very little substantive news…other than that age-old culprit of sector rotation.

Pure's beat was simply not large enough to satisfy investors. We have reached the stage where investors are worrying that beats might not be large enough going forward and are left handicapping the size of the expected over-attainment.

I first wrote about PSTG two months ago, and said it was a buy then at about this price. The share price has decreased marginally. The tech valuations are also marginally lower but the outlook for Pure is better than it was two months ago. I still like Pure Storage shares and think that they will outperform the tech index. Tech is being sold heavily today and it is possible that we are in one of those market phases where tech has a difficult hill to climb. That is a different subject. If tech reprises what happened in its wreck earlier this year, so will the shares of this company. I think the shares are cheap, but not cheap enough in that kind of environment to protect them from material share price devaluation.

I think that revenue growth will track higher than the current consensus and I think there is no potential price war on the horizon for the specific storage products that this company sells. The days of spinning disc are coming to an end, and this company remains a leader in the world of all-flash arrays (AFAs). There is a huge amount of potential operating leverage although the company is in the earliest days of making that happen. The cavils that I have seen regarding the shares, and there are more than a few, are of little importance when looked as part of valuation parameters.

Just for the record, PSTG announced a 50% growth in revenues, a beat of about 3% compared to guidance - a bit more on a constant currency basis. Gross margins were up 380 bps year on year and were exactly in the middle of the guidance range. Operating margins, particularly non-GAAP operating margins, improved as well, although GAAP opex spending growth slowed significantly both year on year and sequentially. Non-GAAP financial performance showed even stronger performance helped by the rapid rise in share-based comp., which basically doubled in this last quarter year on year.

The company is still a long way from cash flow break-even despite the high level of stock-based comp. The rapid growth in revenues is also driving rapid growth in receivables and deferred revenues, not really a major operational consideration, fell quite a bit year on year.

Guidance was pretty much a non-event if you consider non-events to be consistent with 48% revenue growth. The mid-point of the non-GAAP margin guidance produces an EPS loss a bit less than the current consensus but there have been complaints that leverage needs to be greater. Fact is that with the upcoming quarter bracketing the end of both the calendar year and the fiscal year, I would be surprised to see sequential growth of 11-15%. While that is approximately consistent with the percentage growth of revenues between Q3 and Q4 fiscal 2016, I expect that Q4 seasonality will be stronger than forecast. But the question is, will it be enough? I imagine that if revenue beats, then so too will earnings. There is plenty of leverage at scale available.

I initially wrote about Pure Storage on this site in an article of September 20th. The shares, in the wake of the share price movement this morning, are basically down a bit since that time. The software/tech index is also down a bit after today's drubbing. It has been a wasted few months in which to have been a Pure shareholder.

The current consensus rating from First Call for Pure shares is a tepid 2.2, holding between buy and hold and the latest coverage initiations were at neutral. That being said, the consensus analyst price target is $16.50, 27% higher than current quotes.

Price Wars and Storage - The two are not synonymous!

Pure shares are not highly valued in terms of EV/S. And while it is not quite yet objectively provable, I think a business like this with a gross margin above 65% is going to be quite profitable when it reaches some kind of steady-state growth. I will discuss valuation a bit later in this article, best as can be done. At the moment, the company has a market capitalization of $2.38 billion. Its cash balance is a bit greater than $500 million and so it has an enterprise value of less than $1.9 billion. The current estimate for sales in the 2018 fiscal year is just a bit above $1 billion - thus the EV/S is less than 2X. I do not want to try to probe the psyche or the logic of the 10 analysts who rate this name a hold. They clearly have different and much higher bars than I do regarding evaluating stocks. Or as is really the case, they believe that the growth trajectory for Pure is far more fraught than I do or that the path to profitability is littered with boulders and hidden landmines.

Basically, it all comes down to two things. Many analysts, looking at the results of the storage wars of the past few years, can bring themselves to believe that this is a different world with new players and new sets of key metrics. And other analysts can't believe that companies that sell products intended for corporate data centers can do so successfully in the context of the explosion of cloud computing. I suppose that is unfortunate - but when there are urban legends that animate share price discussion, it presents investors with opportunities that are normally not available. The fact that PTSG has a short interest of 16 million shares which is a substantial percentage of its current float is an indication of just how deep this misapprehension is in the minds of some investors/hedge funds. There are investors who often suggest that it is hard for them to go against a high short interest because the short interest "must know something." Sometimes it pays to go against that kind of thinking as the rewards can be far more significant than otherwise.

I do not want to spend a substantial amount of space in this article in reprising the history of the enterprise storage space. I have been involved with it one way and another since 1969. It has always been competitive, prices always fall and price elasticity has always been enormous. As the saying goes, plus ca change, meme ca chose. Flash, which is the subject under discussion here, must become cheaper if it is going to entirely supplant spinning discs. And it is becoming cheaper. A few days ago, I got a post from a reader who is a storage engineer. The reader was writing about an article I had published regarding the transition of NetApp (NASDAQ:NTAP) from its near-death experience to a company now becoming based on selling AFAs. He commented that his company was introducing the next generation of solid state drives (SSDs), the core ingredient of AFAs that are very advanced and incredibly cheap. That is how the storage market works. The components decline in price, the storage vendors incorporate the technology in their designs and cut prices and demand booms.

Oh, there is one difference here. Flash is competing against spinning discs. The cheaper flash becomes, the higher its market share will be and the faster it will obsolete spinning disc.

I expect that Pure will continue to lower prices for the foreseeable future and I expect that Pure will both continue to gain market share and maintain relatively high product gross margins indefinitely. (Concerns about the current prices of NANDs are way overdone.) Those investors waiting for some collapse in product gross margins have not only missed the bus - they are in the wrong bus. I have identified one link that discusses the growth of AFAs in the overall storage space; there are many, many others that suggest the structure of the market and its growth.

The AFA space is probably on track to see growth of 70%+ this year overall. Pure is in the third position in the market behind EMC/Dell including all of their different brands and NetApp whose market share has exploded thanks both to their own acquisition strategy and their focus on AFA.

Of the companies in the space, Pure and Nimble (NYSE:NMBL) have significantly higher gross margins than do their competitors suggesting they have significantly lower costs. Would it make sense for EMC or NetApp to start a price war when they have substantially greater costs than either Pure or Nimble? To say that it would be unusual is an understatement. Does it make any more sense for either Pure or Nimble to start a price war? Not really when both companies are achieving spectacular growth in all-flash and would only be cutting their own revenues.

Does Pure have a Secret Sauce and how does it keep the sauce secret?

One of the issues constantly cited by both potential investors and commentators regarding the enterprise storage space is that the products are commodities. Storage is storage is storage, or so they believe. I think that to be a substantial misapprehension that has a patina of verisimilitude and the substance of a bowl of mush. These days there are about six credible vendors of AFAs. But they do not all sell the same things. Pure has prospered because of its very high Net Promoter Scores, second only to those of Nimble and within hailing distance of those Nimble has received. (Net Promoter Scores are a measure of user satisfaction and depict the percentage of users who are willing to recommend a particular product or service. Pure has a Net Promoter Score of 83.5 with 100 being the highest on a scale of +100 to -100.)

Pure has a more reliable platform than much of its competition with six nines of availability. These days, the company is focusing its development on something called NVMe (Non-Volatile Memory) technology. I am not qualified to state as does Scotty Dietzen, the CEO of this company that "NVMe offers the potential for a 100-fold increase in performance to processing many requests simultaneously (Parallel) rather than one at a time…We believe all-NVMe is going to prove (to be a) bigger disruptor than all-Flash." Based on what I have read, this seems likely to be the case and its wide adoption by Pure will constitute a significant technology moat behind which it can grow and prosper…until some competitor is able to replicate the capability. NVMe is being used as part of the Pure offering currently but it has been a development focus for the last several years. At some point in the near future, NVMe will become a major factor in the Pure offering and it will become a major differentiator in terms of storage performance. Essentially, it is the difference between sequential and parallel read points. (I apologize if I have over-simplified the technology but so far as it goes, it is suffice to say that it will be a part of the secret sauce for Pure, and an important one at that, and one that should be appreciated by investors in terms of how they might value the shares of this company.) To what extent NVMe will be a differentiator and exactly how much of lead Pure has in the technology is not totally knowable at this point. I do think it is fair to say that it is something that has already helped the company win enterprise deployments and it is likely to become even more of a competitive factor the closer Pure gets to a major NVMe launch.

Pure is also in the midst of launching FlashBlade, which is a product line for unstructured and Big Data. This has been an under-served market for years now since EMC absorbed and later eviscerated Isilon. (Isilon still exists in some truncated state for EMC/Dell but investments have not kept up with the pace of technology change in that market.)

Again, many hedge funds and some individual investors continually expect to see revenue growth for this company revert to the mean, whatever the mean might actually be. But the storage market is not a commodity and all storage simply is not fungible. Again, I do not want to attempt to define the features of FlashBlade that facilitate the performance and capacity of the solution. Here is a link to Pure's marketing blog regarding FlashBlade. While it will not produce a great deal of revenue this year, it obviously has the potential of being a significant revenue generator in fiscal 2018 and beyond and is one of the factors that should allow this company to expand its market share within the flash storage market.

What is the current competitive state of play in the storage market?

I would never suggest that the storage market is not competitive. Pure doesn't go out and win deals as a sole source. One area that some believe to be an area of potential competition that has emerged recently is the offerings of hyper-converged products. These offerings appear to have befuddled investors and analysts and in particular, those offered by Nutanix (NASDAQ:NTNX) are thought to be Pure's competitors. I have written favorably about Nutanix, but hyper-converged technology is better suited to remote and branch offices for its workloads and, at this point, does not often compete for workloads in primary data centers. Hyper-converged is very attractive because of the automation it offers in running data centers, and its decreased footprint, but at the moment, it is not intended to compete against AFA for most applications.

When it comes to what I might call basic workloads, Pure has developed a form of storage efficiency in the flash world. This is akin to the advantage that NetApp used to enjoy when it was able to use storage efficiency to charge more per unit of storage but still was considered to be a lower cost competitor because it used less storage than competitors to provide the same throughput. Pure makes a similar claim and it would seem to bear out in the fact that gross margins for this company are significantly higher than those of the traditional competitors who now sell flash. Those competitors simply have to provide more capacity while matching the Pure price in order to sell.

EMC remains and is likely to remain the principal storage competitor and it is also the largest vendor of flash. Anecdotally, there is still a fair amount of confusion and sales force dysfunction in the new enterprise - but over time, that is likely to fade. The CEO of Pure said that EMC offers nine different flavors of flash and while I imagine that it does not offer nine products to one customer at one time, the confusion of the non-integrated product line affords competitive opportunities for a company such as this.

At the moment, Pure maintains that when deals get to the Proof of Concept phase, it has a two-thirds win rate and given its technology advantages, that is not terribly surprising. Pure has a meaningful partnership with Cisco (NASDAQ:CSCO) which is a significant competitive differentiator and Cisco partners have begun to migrate from EMC/Dell to Pure. In order to develop offerings that include Cisco compute, Cisco networking and Pure Storage. I believe that over time, that will become a meaningful market share driver for Pure.

I want to close this section by addressing what many investors think to be some kind of long-term disease that Pure has to deal with and of course, that is competition from the cloud. I want to constrain the length of this article, at least to some degree, but the fact is that to many users, cloud storage is cheaper than, far cheaper than any kind of on-premise storage. A few weeks ago, Amazon (NASDAQ:AMZN) took an axe to its pricing, and in particular, to its storage pricing, which was cut by as much as 28% for one kind of storage in one region. (Amazon has a multiplicity of offerings such that it is not possible to talk about the overall impact of its pricing changes in the storage area. There are apps that have been developed to help AWS users understand their costs - and even with those apps it is hard to speak about any kind of knowable pricing for a typical workload).

It is obvious that over the past several years, the emergence of the cloud has dramatically inhibited the usage of external storage and it is just recently that shipments of external storage have stopped falling in dollar terms. Many of those institutions and investors that have short positions in Pure shares believe that the phenomenon that happened in the past will happen again going forward and that Pure's growth will be inhibited materially by competition with steadily lower cost cloud storage alternatives.

But things really do not stand still in technology. I have to once again talk about NVMe, something I am more than a bit loath to do. But the architecture of NVMe, which has thousands of parallel points of connection, is ideally suited to both Big Data/Analytics and unstructured data applications. By this point, I think, most investors recognize that the world of the next several years will be one of hybrid clouds and it is fair to say that the advent of NVMes will allow users to cheaply and efficiently develop applications in their own clouds that make use of this technology. At some point, I anticipate, those institutions with their major short interest in Pure will come to the realization that the technology is far more complex and nuanced than they have realized and will throw in the towel and allow valuations to rise to levels commensurate with the company's rocketing growth.

Costs, Valuation and a restatement of the investment case for all who may have missed it

I understand that in an attempt to be thorough I sometimes befuddle. I don't like to write about specific new technology trends and base an investment thesis on that - but the fact is that shareholders and those thinking of investing in this name are really buying into very specific technologies that are being brought to market by Pure's team. There is no hidden cash flow or margin generator here. If readers believe that the technologies, described as best my poor fingertips and keys are capable of doing, are for real, then the numbers will align and the investment will work. And the opposite is true. It is lots more than just flash. It is flash, it is unstructured data. It is storage efficiency and it is the advent and proliferation of NVMe. Is all of that worth an expectation of 30%+ growth for several years? I think it is.

So, the question becomes one of profitability. I am not going to get involved in debating the course of NAND pricing, or what the ramp will be on FlashBlade gross margins. Over time, NANDs will get cheaper. Over time, volumes on FlashBlade will reach levels consistent with gross margins in the range of this company's model.

Overall, GAAP operating expenses grew by 52% sequentially. Included in that amount was a one-time $30 million legal settlement. Absent the settlement, GAAP opex grew by 30% year on year, substantially below the 50% growth of revenues and demonstrating significant operating leverage. I am puzzled that such a level of expense discipline did not receive more and more favorable commentary. For nine months, and excluding the legal settlement, GAAP expenses grew by 46% while total revenues grew by 72%.

Sequentially, revenues grew by 21% reflecting some of the seasonality of the end of the fiscal year of the Federal Government. Sequential operating expense was only up by 6%, again excluding the legal settlement. These are all GAAP expenses.

During the course of the call, the company's CFO, Tim Riitters explained the seasonality of expenses. At this point, the company has diminished hiring although it will always have room to add engineering talent. Hiring will re-accelerate in Q1 and Q2 of the new fiscal year. But I think the takeaway here is that the path to profitability is very visible here. The company says it will reach sustained positive cash flow in the second half of next year; the results this quarter do show a company exercising a fair amount of expense discipline.

One can tinker with numbers and models every which way. I think that based on the business case I have laid out, it is reasonable to believe that even with an external storage market basically not growing, flash will continue to grow rapidly for years to come and this company is likely to grow faster than the flash market as a whole.

I also think that the company can readily achieve its margin goals over several years. It is entirely reasonable to put together a model that shows non-GAAP earnings per share of $2.25-2.50 in fiscal 2020 for this company and free cash flow should be roughly equal to non-GAAP earnings. What is the worth of $400-450 million in non-GAAP earnings/free cash flow in 2020, growing at perhaps 15%? Again, the answer depends on the discount rate that is used, but if the variables fall into place, then a reasonable valuation is several times the current enterprise value of $2 billion.

Disclosure: I am/we are long PSTG.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.