Is the Storage Space Investable?
Since the advent of the cloud and its continued hyper-growth, it has proved to be difficult to invest in storage companies or even their adjuncts such as Commvault (NASDAQ:CVLT). Some of the companies within the space have been consolidated. Many other companies have been through a phase of implosion. Most lately, the overall decline in external storage revenues appears to have abated. And now, as in the recent past, the migration to flash storage has continued at an unprecedented pace.
But valuations remain dramatically compressed, and most recently, Nimble Storage (NYSE:NMBL), a company whose growth rate has been 30%, was sold to HPE for less than 3X EV/S. The most prominent flash storage vendor these days is Pure Storage (NYSE:PSTG). I first wrote about Pure back in September 2016. The shares have depreciated 25% since that time and 16% since my second outing on the name. The issue hasn't been operational performance. That has proven to be strong. Both of the earnings releases since my first article have been beats and raises. That is usually thought to be a recipe for share price appreciation, but these days, storage companies need something more to see share price appreciation. By now, Pure's EV/S ratio has dropped to 1.55X and its enterprise value is just greater than $1.5 billion. Anomalies such as these most often are followed by consolidation.
At the start of this month, Pure released the results of its fiscal Q4 that ended 1/31/17. The company, having not had a great reception to beat and raise quarters, tried something different. It beat, but it noticeably cut expectations for fiscal Q1 (ends 4/30) and simply bracketed expectations for the current fiscal year in terms of its forecast for revenues and EPS. That has actually worked a bit better. The shares are only down 10% since the start of the month. One writer on this site has opined that the guidance appears aggressive/worrisome for the full year, rather than the current quarter. He also suggests that much of the questioning during the conference call reflected skepticism about the company's forecast and particularly as it applies to a new product called FlashBlade. There is no question that many analysts both focused on and expressed skepticism about the ramp for the new product.
Is it reasonable to be a contrarian here if that is what it is to recommend Pure's shares? I am not sure if shareholders would necessarily have been happier with worse guidance. And I am not too concerned about the tone of analysts' questions. Analysts are going to be unhappy with guidance cuts even if they are for one quarter and even if they have more to do with seasonality than anything else.
By tweaking guidance to the extent, it has for Q1 and by reducing the percentage growth rates it was willing to support, management has significantly de-risked investing in the shares. Far from guidance appearing aggressive, for the second half or for the full year, it appears to be ultra-conservative with far more potential for upside than downside. This is not a company with classical compressed valuation metrics. It doesn't earn money or generate cash. But for the leading independent company in what is, after all, still a high growth space, its EV/S at 1.55 is remarkable, remarkably low I think.
As it is, opinions of analysts are more or less evenly divided between buy and hold, and no one thought the quarter that was announced was of significant enough performance to change their rating. So, it isn't quite being a contrarian to recommend this name as an investment. Can investors invest in Pure. I believe they can and should, and I find the concerns about the introduction of FlashBlade and its ability to bring in $80 + million of orders this current fiscal year to be much concern about very little.
The question at the start of this section is relevant in that the predisposition of many observers in terms of investing in the storage space appears to be negative. There have been many years of disappointing operational performance on the part of many storage vendors. The largest of them is now private as part of Dell. The second largest, NetApp (NASDAQ:NTAP) even after a terrific rally, is still trading for less than it did five years ago. The balance of the article presents the investment case for Pure at its current valuation and tries to look at moats, the kind of competitive barriers that a storage company can build and not the kind that are built surrounding castles.
Some Lessons in Seasonality, Guidance and Profitability
One of the principal reasons why Pure's shares are valued as they are has to do with the guide the company provided for its fiscal Q1. Simply put, guidance for a pronounced Q1 pull-back in revenues was seen as a significant concern and the company's forecast for a sharp rebound in Q2 and beyond was thought by some to be less than credible.
Many observers find it surprising that IT vendors have significant quarterly sales perturbation based on seasonality. The most commonly observed seasonality in the IT world has to do with the spike in demand seen for companies that have significant sales to the Federal government. Federal procurement is marked by a profound budget flush that inevitably makes quarters that include September 30th particularly strong as government agencies and departments hasten to spend all allocated funds before the end of the fiscal year. Enterprises, while not as rigid, tend to flush their budgets at the end of their own fiscal year and that, more often than not, ends in December. Contra wise, the quarters after December and the quarters that span the summer months, particularly August, are known to be unusually slow, especially so for those companies with lots of exposure to European procurement patterns. There is nothing either mysterious or malignant regarding these trends. They almost inevitably are seen in the operations of all IT vendors that reach some scale, and it is not particularly concerning that they have made themselves seen in Pure's results.
It is perhaps relevant in assessing guidance to take a look at the quarter Pure actually reported. Pure's quarter was certainly within the range of expected results. Revenues grew more than 50% year on year and almost 16% sequentially. Gross margins were consistent with past levels despite some component shortages and component pricing increases. The company showed a significant improvement in operating expense ratios both sequentially and year on year, both GAAP and non-GAAP.
The company had an incentive program to focus its sales teams on new name accounts. The incentive worked - perhaps too well as there were 450 new logos closed during the quarter. Needless to say, these tended to be smaller enterprises than a more typical quarter has seen for Pure and that, in turn, has led to lower levels of potential "expand" business in the current quarter. It is nothing that is either unusual to see or particularly worrisome for a company like this still in its adolescence.
The company provided guidance for the current quarter that was at significant variance with the prior First Call consensus. Revenue growth year on year is projected to contract to 27% and revenues are projected to decline 22% sequentially. The prior year, revenues declined by 7% sequentially and then rose by 16% the following quarter. All that says is that seasonality is not a novelty for this company, although the current projection may represent an extreme example.
I am not sure that the chart that the company presented as part of its earnings release depicting its expectations for seasonality will prove to be quite correct, and I am not too sure how accurate the guide (22% decline in revenues from the sequential quarter) will prove to have been when the quarter closes. But there is nothing unusual in expecting this company to see a sequential decline in revenues in this current quarter followed by a significant sequential growth in fiscal Q2 at its scale.
And given the anticipated seasonality, it is not terribly surprising to see operating margins deteriorate substantially in fiscal Q1 as well. Many IT vendors front-end load investments in sales and marketing and with sales projected to slip, and sales and marketing expenditures projected to rise, I think the results to be expected are or ought to be pretty obvious.
Perhaps of equal interest, the company is forecasting gross margin stability. That is a bit surprising given the shortages and price jumps in terms of NAND and DRAM that other storage vendors are dealing with. Management talked about the fact that its products are between two and five times more efficient in terms of NAND usage due to a form of storage compression. The company also suggests that its software-based products have allowed it to mix different kind of flash as well as consumer-grade MLC. I have linked to an article that relates to how Pure uses this cheaper kind of flash without compromising reliability. I am not qualified to evaluate the arguments against this strategy. All I can say is that if the proof is in the eating, Pure has not been on any visible diet lately.
Gross margin differences are very visible. Both HP (NYSE:HPE) and NetApp talked about supply constraints coupled with higher component costs in their forecasts. It is hard to gainsay the double-digit advantages in gross margins that this company both has achieved and continued to forecast. There is some myth that all storage is fungible. There is a myth that all-flash arrays, the kind of storage that this company sells, are fungible. Corn and wheat are fungible, more or less although there are obviously different types of those commodities.
Storage is simply not fungible because of the significant software content that all arrays offer. I have heard the argument about commoditization since the days of NetApp's storage efficiency solutions. Back then it was ONTAP that allowed the company to achieve better product pricing than its much larger rival EMC. As is often said and is self-evidently true at this time in the flash storage market, the devil is in the details. A bit of study of how the storage industry has evolved will probably put paid to the idea that there is not significant product differentiation even within the context of all-flash arrays and that Pure has been able to leverage its technology successfully.
Flash Blade - "Lightning in a Bottle?"
Lightning in a Bottle is said to be California's premier summer camp-out festival featuring music, art, yoga and education. It apparently includes sustainability, social cohesion, personal health and creative expression. A song has been named "Lightning in a Bottle." Given my musical tastes, I imagine I would find the festival and the song more than a bit painful and a bit of cacophony. But my taste in music is not universally admired either. The company's COO, Dave Hatfield, described Pure's FlashBlade product as the equivalent of lightning in a bottle, presumably with malice aforethought.
Presumably for many of us who are not quite involved in the new age exemplified by the festival, the quote is supposed to mean that the product will be a very strong contributor to the company's revenue growth and will be a significant factor in establishing more rapid growth in the second half of the just started fiscal year.
As mentioned earlier, I have written about this company in the past and have mentioned in those articles that FlashBlade would be a key to the ongoing growth of Pure. That remains the case. FlashBlade is designed for workloads that require very fast access to unstructured data and is used with FlashArray which is basically for primary data storage. The product was released earlier this year in general availability and a recent tweet talked about a FlashBlade system that replaced a 1,008 rack unit installation, filled with 13,000 disc drives that had come from NetApp. Obviously replacing 1,000 racks with a single unit is an impressive achievement.
Needless to say, the FlashBlade installation that replaced the NetApp disc cost millions less and dramatically outperformed the existing disc array because flash is far faster and because the I/O load is automatically processed across all blades (parallel vs. serial if you like analogies).
Mr. Hatfield said that the way the forecast for Pure has been constructed is that it takes between one and two FlashBlade systems per selling team for the current fiscal year to beat the company's internal plan. Overall, less than 10% of Pure's installed base needs to convert over to FlashBlade in order for the company to achieve its guidance. Summing up, Mr. Hatfield declared, "We're super enthusiastic and very confident in the FlashBlade launch."
I am not really sure why it might have been that the negative writer on this site would have found that kind of assertion less than credible. Later on, during the course of the conference call, another questioner asked about the ability the company would have to sell $86 million of FlashBlade hardware this current year. Management again reaffirmed the forecast. At that point, one would either have to believe that management is somehow being economical with the truth or that it is completely unaware of the prospects for its own business. Given the record of this company since it has been public, it is simply hard for me to find any specific substantiation for such a serious allegation.
The fact is that the products that FlashBlade competes against are really super-annulated, or as the product VP put, there has been a dearth of innovation in that specific market. EMC paid a lot of money to buy Isilon, but after the first few years of ownership, it has clearly under-invested in trying to improve the performance and functionality of what it had bought. NetApp has made some smaller acquisitions in the realm of unstructured data over the years. It has never been a major focus for the company and it really is not a major focus currently. It would not be surprising if Pure found it easy to both leverage its installed base and to take market share in selling FlashBlade, especially given the current state of sales execution plaguing Dell/EMC.
I do not, to be sure, have some crystal ball of my own that tells me how successful FlashBlade is going to be. Sales execution is always important in product launches no matter what they are and who is expected to buy the products. It certainly helps that much of the expected success of this product is going to come from sales within the installed base. But selling $86 million of a hot new product within a total forecast of $800 million of hardware sales this current year just doesn't seem like a particular stretch. Best as I can determine, FlashBlade probably does have an edge in terms of its capabilities compared to competitive offerings - at least for the moment and that is about as good as anyone might expect in the storage hardware space.
During the course of the call, management said that it needed to achieve 70% of its bookings from its installed base to reach the yearly revenue target. This is a quintessential land and expand company and hence 70% is really not some outlier in terms of prior performance. The CFO declared that the numbers for repeat customers "are in the same zip code. Over time, we've seen that rate continue to climb up. Now that we've got the history (in front of us) and repeat business, we've got a lot of confidence and how those (numbers) are going to curve out over the course of the next year."
While clearly some of the analysts had a bit of trouble wrapping their minds around the forecast, for all of that, it hardly seems incredible or stretched. Management spoke about the statistical analysis it had done on its various customer and salesforce cohorts in putting together its forecast. A forecast by definition involves things that haven't yet happened. There will always be expectations which aren't achieved - or for that matter are over-achieved. There are many ways of putting together forecasts, but the statistical analysis used in the process by this company is certainly a reasonable technique and one that is really not producing outlier results, for the most part. I do think that the statistics that suggest the rather sharp Q1 seasonality may well prove to be challenged, but that is really a good thing.
Competition, Evolution and Species Die-Off
I am not going to try to answer all of the questions that arise regarding the future of Pure's market share. I do not have the resources of a market research organization, and the space required to try to be objective and thorough is far greater than is necessary to draw conclusions about Pure as an investment.
One of the issues that some analysts have had with Pure is that to grow, it seems to need to gain share consistently in what has been a marketplace populated by giants. There is some validity to the terms in which the argument is framed. The AFA market is said to have a CAGR of just over 20% in the latest analysis put together by IDC. Pure is in fourth place with an 11.5% share compared to EMC/Dell with a share of 40%.
One thing to note is that the legacy vendors these days are simply hoping to hang on and not grow a great deal. NetApp is hoping that the culmination of its re-imagination and product line rejuvenation is that it can grow mid-single digits. I doubt that EMC/Dell even has that level of aspirational goal.
Data reduction continues to be a major factor in the Pure competitive story. Again, data reduction is all about software and architecture, and having a more modern architecture has allowed this company to exploit its software advantages to provide users with better performance regardless of the specific components it uses.
The company has a significant partnership with Cisco (NASDAQ:CSCO) in terms of developing a converged product offering. Just how successful the offering which is called FlashStack has actually been is difficult to determine. But it has certainly helped Pure develop relationships with many Cisco resellers that historically worked with the other storage vendors. Going back some years, Cisco had alliances in its data center offerings with both EMC and NetApp. Over time, it seems likely that FlashStack offerings will supplant and replace the installed base that Cisco still has that use older technologies from its two former partners.
There are a couple of other opportunities that are likely to be significant going forward in allowing Pure to continue to gain market share. One of these is a concept known as synchronous replication. It is a technique that storage vendors use to enhance the ability of their users to replicate data so that it can be safely copied and stored in disaster recovery architectures. Most very large users require that their storage partner have this capability. Pure has been shut out of the highest end of the market because it hasn't had this capability. The company is now in the process of preparing a roll-out of its version of the technology. It is a capability that is one of the factors that is perhaps amongst the reasons behind the significant sales ramp the company is forecasting beyond seasonality and FlashBlade.
The final concept that should probably be noted is the migration to NVMe technology. I tried in a previous article that was at the limit of my understanding to talk about the transition/revolution to NVMe and the opportunities it presented for Pure. Pure maintains that its product evolution will be different than those of its competitors. It claims it will be a non-disruptive evergreen transition which is another part of the company's mantra. And it claims that it will introduce this technology in a way that leverages current software. Since none of this has happened yet, one has to take claims on faith or on a willing suspension of disbelief. But it does seem, at least possible, if not quite probable, that the advent of NVMe technology will allow Pure advantages that should enable the company to continue to grow its market share for some period of time to come.
I think a more important competitive issue for this company remains how its solutions compare in terms of performance to workloads run on the public cloud. That is a horse that has been ridden near to death, and I doubt I will be able to shed any new light on the subject. The best guess of analysts is that regardless of the cloud alternative, the market for AFAs has a CAGR of above 20% for the next few years and that kind of opportunity provides plenty of runway for Pure.
Pure reported that it had 201 million shares outstanding at the end of its fiscal year. That represents an increase of 7.3% year on year. The company continues to use significant levels of stock-based comp, although the rate of increase has started to slow a bit. In the last quarter, stock-based comp represented 16% of revenues. As of last evening, the company had a share price of $10.38 and thus had a market capitalization of a bit less than $2.1 billion. The company had a cash balance of about $550 million, takings its enterprise value to $1.55 billion. The EV/S based on company guidance and the analyst consensus as derived from First Call is for sales this current year of $990 million. That results in an EV/S of just 1.55X. An unusual number for a growth company, to say the least.
The company should reach non-GAAP profitability in the second half of the year, and it has made progress in some expense ratios, even on a GAAP basis. But no rational investor is going to be buying this name for earnings in the next couple of years.
The company generated a modest amount of cash flow from operations in Q4, but it has not yet succeeded in generating CFFO for the full year. The decline in Q4 CFFO was a function of the slower growth in deferred revenues in the quarter. That in turn was a function of the company emphasizing new accounts/nameplates in the last quarter. These smaller businesses typically do not sign the kind of multi-year procurement agreements and services contracts that has in the past led to strong levels of deferred revenue increase. While management spoke to having sustained positive CFFO in the second half of the current fiscal year, it seems unlikely that the company will generate enough CFFO to do much more than pay for CapEx. It seems likely that the company will most likely produce a level of free cash in fiscal 2019 that would allow for a valuation using that metric.
Pure is not yet a profitable company although it has a readily discernible path to profitability. I think at these levels, Pure's shares represent a significant value and one that can produce significant positive alpha for investors.
Disclosure: I/we have no positions in any stocks mentioned, but may initiate a long position in PSTG over the next 72 hours.
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.