Elliott Associates Letter to EMC [throwback]
October 8, 2014
The Board of Directors
176 South Street
Hopkinton, Massachusetts 01748
Attn: Joe Tucci, Chairman & CEO
Dear Joe and Members of the Board:
I am writing to you on behalf of Elliott Associates, L.P. and Elliott International, L.P. (collectively, “Elliott” or “we”), which collectively own 2.2% of the common stock and equivalents of EMC Corp. (the “Company” or “EMC”), making us one of your largest shareholders. We greatly appreciate the dialogue we have established with Joe and his team and we look forward to continuing it.
Since the publication of news reports detailing our position, we have received numerous calls from fellow shareholders requesting our views and sharing theirs. In addition, EMC management has spoken publicly about their view of the Company’s structure. The purpose of today’s letter is to share our thoughts on the right path forward. We hope this will help to inform Joe and the Board as part of EMC’s current review process regarding the long-term value-maximizing pathway for the Company.
The summary takeaways from our letter, which are more fully described below, are:
- EMC’s current structure – “the Federation” – obscures enormous value at EMC
- EMC should pursue pathways to recognize this value, including a separation of VMware from Core EMC and/or various M&A opportunities
Elliott is an investment firm founded in 1977 that today manages approximately $25 billion of capital for both institutional and individual investors. We are a multi-strategy firm based in New York, active in debt, equities, commodities, currencies and various other asset classes across a range of industries. Investing in the technology sector is one of our most active efforts at Elliott and one in which we have built a long track record.
Our approach to EMC is consistent with our approach to many of our current and previous technology investments. We have conducted extensive research to better understand the Company’s operations and strategy, including working with respected technology and management consultants to examine the broader storage, data center virtualization, security, big data analytics and end-user computing landscapes, and EMC’s position within those markets. We have also worked with engineers to examine and assess the capabilities and competitive positioning of EMC’s products and technologies across all of its offerings. Our efforts during the past year include a survey of over 580 EMC and VMware customers, enabling us to better understand the storage and data center virtualization landscape from a buyer’s perspective and to identify what factors are most important in driving purchasing behavior. We have also retained senior executives in the storage, data center virtualization and broader technology marketplaces to advise us on higher-level corporate considerations. We believe this time- and resource-intensive exercise has given us a strong understanding of the markets in which EMC participates, as well as a deep appreciation for the Company’s competitive strengths and challenges.
Our Thoughts on EMC
We have grouped our thoughts into the following three categories, which we discuss more expansively throughout the remainder of this letter:
1. The Federation
2. EMC and the Reality Today
3. The Right Course Forward
It is important that we convey that Joe, for whom Elliott has the highest respect, and his team deserve immense credit for assembling, integrating and building EMC into what it is today. The substantial value opportunity we describe in Section 3 is only possible due to the high quality of the Company and its individual businesses.
EMC has developed an incredible set of assets including EMC II, VMware, RSA and Pivotal. Each of these companies is comprised of acquired and organically developed product offerings and businesses that by themselves are frequently the best in the industry.
It is well known and has been broadly recognized by those who follow EMC and its markets that the structure of the Company is unusual: VMware, Pivotal, RSA and EMC II are run as independent companies. Each of VMware, EMC II and Pivotal is headed by a CEO who reports to Joe. RSA is headed by a CEO-level executive who reports to David Goulden, EMC II’s CEO (and, until last week, CFO of EMC). EMC management refers to this arrangement as “The Federation.” (Follow this link for a graphic illustration of the Federation: http://elliott-graphics.com/slide1.html).
The origin of this unorthodox structure is an important part of the story: EMC used to be only what is now called “EMC II,” the storage company. Over the years, however, EMC bought VMware (2004), RSA (2006), and the assets that comprise Pivotal (2009-13), among many other acquisitions. Joe had the foresight to allow VMware to develop independently as a separate and distinct company, paving the way for its extraordinary success in server virtualization and positioning it to enable the virtualization of other components of the data center.
Much has changed since 2004, though, when VMware was solely a server virtualization company and EMC was primarily selling high-end storage arrays. Indeed, the correct decision more than a decade ago is not necessarily the right one for today or for the future.
EMC and the Reality Today
The reality today is that the Federation structure, which may have served EMC well years ago, no longer does. We have compiled considerable detail to support this conclusion and summarize our thoughts briefly below:
A. Stock Price Underperformance: EMC’s stock price has underperformed its proxy peers and the market over all relevant timeframes while this structure has been in place.
B. Core EMC is Deeply Undervalued: The Federation structure has led to a widely-recognized undervaluation of “Core EMC” (EMC excluding VMware).
C. EMC and VMware now compete: Both EMC and VMware have grown and are now competing against one another, confusing customers, employees, Street analysts and shareholders.
D. EMC II and VMware hinder one another: Though separate companies, they are still “together,” a fact that inhibits each business’s ability to partner.
E. The Federation and Joe’s Tenure: Joe is the architect and manager of the Federation. It is not viable to keep the structure as it stands today beyond his tenure as CEO.
Stock Price Underperformance
Though EMC is a leader in numerous markets with great products, EMC’s stock price has deeply underperformed its proxy peers and the market over all relevant time periods. It is important to keep in mind that the current structure (EMC + VMware) has existed over the entire timeframe illustrated below. The below chart highlights EMC’s stark underperformance as of July 18, 2014, the trading day prior to the public disclosure of Elliott’s position in EMC.
On the right side of the above chart, we include underperformance since the VMware acquisition. These numbers are astonishing – EMC’s stock has materially underperformed its peers and the market since the VMware acquisition, despite that purchase bringing EMC over $32 billion of value (VMware has a $40 billion market cap and EMC owns 80% of it). The driver of this underperformance is EMC’s unusual corporate structure and the damaging valuation impact it has on Core EMC.
The Federation structure also adversely impacts VMware, a leader in server virtualization and a would-be leader in virtualizing storage. Many in the industry believe VMware should have created and dominated the software-defined storage market by this point had it not been owned by a storage hardware company.
Core EMC is Deeply Undervalued
Elliott is hardly alone in pointing out that Core EMC is significantly undervalued. This problem has been widely reported and discussed, yet it has gone unaddressed for years.
The undervaluation is staggering: Prior to the disclosure of Elliott’s position in EMC on July 21, 2014, Core EMC traded at just ~3x 2015E EBITDA and ~8x 2015E P/E. Those metrics are incredibly cheap on any basis but especially so when you consider that NetApp, Core EMC’s closest peer, currently trades at ~6x 2015E EBITDA and ~13x 2015E P/E, a premium to Core EMC of ~80% and ~70%, respectively. This is even despite the fact that Core EMC is the market leader in storage with such attractive assets as VNX, Data Domain, Isilon, XtremIO, RSA and Pivotal (all of which are included in Core EMC). Core EMC, if it traded as a standalone company, would undoubtedly trade at a premium to NetApp.
The discount serves no one’s benefit – not shareholders, not employee-shareholders and not the executives trying to run the company. Here’s a brief sampling of what independent observers have said:
“In fact, at current price we believe investors are getting core EMC ($16B sales, 20% Op margins, $1.35 EPS) for FREE as EMC’s price reflects value of: VMW ($15.85), Emerging storage ($6.24), RSA ($2.99), Pivotal ($1.71) and Cash on hand ex-VMW ($1.21). Should core EMC get valued close to NTAP, it would add an incremental ~$10 to EMC's stock” – RBC Capital Markets (10/7/14)
“The case for a VMware spinoff is straightforward, more compelling if combined with cash. Based on our SOTP, core EMC remains undervalued, trading on P/E of 9x versus the IT Enterprise/Hardware sector of 12x, despite superior growth and technology positioning potential” – Credit Suisse (7/24/14)
“[O]ngoing lackluster relative share performance should leave us / investors to question (or pressure) EMC’s willingness to do something else to unlock value […] (e.g., consideration for any distribution of VMware shares to EMC shareholders). […] [W]e believe the current valuation warrants EMC management to provide investors with a better understanding of how the company plans to unlock what appears to be a very discounted EMC Information Infrastructure valuation” – Stifel Nicolaus (1/12/14)
“Core EMC’ getting no respect. 4 years with no gains? – ‘Core EMC’ has been essentially flattish over a 4 year horizon, suggesting all gains in EMC shares over the last 4 years have come from appreciation in VMW stock”—ISI (1/29/14)
EMC II and VMware Now Compete
Management continually states that EMC II and VMware are “better together.” But this sentiment is at odds with the reality that the Federation keeps the two companies in a sort of limbo where they are “together but not really.” The Federation has actually created the worst of both worlds: EMC II and VMware are not together enough to have completely joint infrastructure, development and go-to-market capabilities, or to eliminate confusion among customers, employees and investors. However, they are together enough to create damaging competitive dynamics and to hinder essential third-party partnerships.
At the time of VMware’s acquisition and for years afterward, EMC II, composed primarily of EMC’s storage business, and VMware competed in different worlds: EMC II sold sophisticated storage hardware and VMware sold software that virtualized servers.
Over time, however, as VMware saturated the server virtualization market, it began to consider its next growth driver. Storage virtualization was a clear choice. At the same time, EMC II, a vendor of storage hardware platforms (and the software that runs on top of them), began developing its own software-defined storage platforms.
Today, EMC II and VMware are completely at odds. In fact, we have heard from many in the industry (including former EMC and VMware employees) that they would have been at odds long ago but that VMware was slow to enter the storage market because it was owned by a storage hardware company. Regardless, they are now competing, a fact that EMC management concedes:
“[T]here's aspects of software-defined storage that Pat's [VMware’s] doing, that David's [EMC II’s] duplicating, or David's doing that Pat's duplicating, depending on how you want to look at it” – Joe Tucci, EMC CEO and Chairman
“So the confusion between [ViPR and vSAN] is the ViPR block data service, which is based on ScaleIO, and vSAN, those 2 are storage systems which do similar things” – Jeremy Burton, EMC President of Products & Marketing
“We [EMC and VMware] both partner and compete in SDS [software-defined storage]” – Chad Sakac, EMC SVP
Even VMware’s recent 10-K acknowledges this growing competition: “EMC could assert control over us [VMware] in a manner which could impede our growth or our ability to enter new markets or otherwise adversely affect our business“ and “[t]here can be no assurance that EMC will not engage in increased competition with us in the future.”
EMC’s response to this obvious problem has been to say that they like the competition and overlap:
“Will there be some overlap [between EMC and VMware]? I hope so. I hope so. Because if you leave seams, that's where our competitors go through” – Joe Tucci, EMC CEO and Chairman
“But, anyway, there's not as much joint development as you think [between EMC and VMware]. As a matter of fact, it's very little. I would rather see a little bit of overlap than trying to get these seams perfect” – Joe Tucci, EMC CEO and Chairman
Unfortunately, this “no seams” rationalization isn’t working and is causing confusion among employees, customers, Street analysts and shareholders. The marketing messages are different and the products are competitive – it’s unclear and is helping neither company:
“They should not be developing competing products, since they are, they should be separate so they can develop the best for their consumers” – EMC & VMware customer
“Originally, it made sense to keep VMware and EMC together. But now there are competing products. It is time to separate the two companies” – Former EMC employee
“We are in phase 2 of the VMware life cycle. Tucci's decision not to spin it completely off 7 years ago was appropriate at the time in order to continue to help VMware grow in server virtualization. Now there are competing products…it’s time to spin off” – Former EMC employee
Our view is that over time vSAN will create some challenges for traditional storage vendors, including EMC. We believe that as the product matures there is a risk of conflict with many of VMware’s traditional storage partners. In particular, it will be interesting how EMC manages this conflict as there will clearly be sales overlap and competition between EMC and VMware. In fact, we believe that the EMC parent may have two sales forces competing against each other. – Barclays (7/21/14)
As part of our extensive diligence on EMC, Elliott commissioned a leading management consulting firm to conduct a survey of 287 EMC customers and 298 VMware customers. Follow this link for the results: http://elliott-graphics.com/slide2.html.
“They should not be developing competing products, since they are, they should be separate so they can develop the best for their customers” – EMC & VMware customer
“Both companies could benefit from this divorce” – EMC & VMware customer
The decision to keep VMware affiliated but still separate as part of the Federation was right for its time. However, it is abundantly clear that this is no longer the case. As time passes, this untenable situation is going to get worse.
EMC II and VMware Hinder One Another
This “together but not really” structure has damaging consequences far beyond just their competition in the storage market – the structure is hindering critical relationships that Core EMC and VMware need to maintain to advance their presence in their markets, as well.
For example, service providers are one of EMC II’s fastest growing verticals and EMC II’s relationship with these customers is critical, particularly as the storage customer base consolidates with more traditional enterprise customers turning to these telcos for cloud storage offerings. However, VMware is directly competing with those customers through its vCloud Air offering, a hybrid cloud service built, owned, operated and supported by VMware through eight data centers globally. We have heard repeatedly that this new source of competition has recently become an issue for EMC II’s telco customers, who now have their storage vendor’s affiliate competing against them in cloud services. This competition can be damaging to EMC II, as the CEO of NetApp pointed out, “[W]e actually looked at should NetApp become a service provider… and realized that a large number of these customers, these cloud providers, were our customers. So we felt like it was much more effective for our business to be a pure play. We have no intent to become a service provider ourselves.”
EMC II also needs to maintain strong partnerships with networking providers like Cisco, with whom they partner to sell converged infrastructure offerings (integrated solution of storage, networking and servers). The Cisco relationship, for example, was nearly fatally damaged when VMware entered the networking market through its acquisition of Nicira. EMC II did not benefit from VMware’s Nicira acquisition but did suffer significant damage to its Cisco partnership as a result.
VMware has substantial potential in storage and networking virtualization, attractive markets that are complementary to its pioneering success in server virtualization. We believe VMware is the natural player to dominate these opportunities, but both established and upstart competitors are also vying for share. In order to achieve the same level of ubiquitous success it has in server virtualization, VMware needs to be the partner of choice at IBM, HP, Cisco, Dell and others. Being “together but not really” with EMC II, a direct competitor to many of those companies, is harmful to these critical efforts.
Though the Federation strategy for EMC and VMware does not work and cannot be continued, the two companies can easily continue their partnership after a separation. Indeed, VMware need not be owned by EMC in order for the companies to work together. For example, we note that there is a reseller relationship between EMC and VMware, which, per VMware’s 10-K, brought in $141 million last year (less than 3% of VMware’s revenue). This relationship and any revenue generated by EMC for VMware could certainly be maintained through a close reseller and technology partnership (as often occurs among enterprise IT companies). Other non-revenue generative strategic benefits could also be achieved through technology sharing and partnerships, as also happens frequently in the IT industry.
The Federation and Joe’s Tenure
Elliott has enormous respect for Joe and looks forward to continuing our dialogue with him, his team and the Board as we convey the views of EMC’s shareholder base on this issue. As illustrated above, Joe is the CEO of EMC Corp., the holding company for the Federation’s assets. Reporting to Joe are David Goulden, Pat Gelsinger and Paul Maritz. Art Coviello, the head of RSA, reports to David Goulden, because RSA is part of EMC II. These men are well-respected, CEO and CEO-caliber executives who are paid like CEOs, who are industry leaders, who can leave to go elsewhere, and, yet, who still report to another CEO.
This highly unique and tenuous structure is only tenable under Joe’s active leadership as CEO. Joe is the Federation’s architect and its leader: There is only one Joe and he is not substitutable.
As Joe and the rest of the Board evaluate the path forward, we believe it is critical for the Board to fully appreciate that it cannot retain a structure that doesn’t work without one specific individual. The most important thing a Board does is succession planning and preparation – to retain a bespoke structure dependent upon one very unique individual is not the correct decision. It may hang together while Joe is still the active CEO, but it will not work beyond him.
The Right Course Forward
It is clear to us that the Federation is not the right structure for EMC and that from shareholder-value, operational and financial perspectives, there are better alternatives. We believe the alternatives fall into two main categories:
A. Spin-off of VMware from EMC
B. M&A Opportunities for EMC
Spin-off of VMware from EMC
A tax-free spin-off of all of VMware from EMC would result in EMC distributing its VMware shares to its current EMC shareholders. Post-transaction, current EMC shareholders would own both their existing EMC stock and stock in a newly independent VMware. For example, a shareholder who currently holds 100 shares of EMC would own the following after a spin-off of VMware: i) their same 100 shares of EMC (now Core EMC) and ii) approximately 17 shares of VMware. Thus, this shareholder would receive continued benefits from both Core EMC as well as from VMware. Immediately and over the longer-term, we believe such a separation creates financial and operational benefits that are extremely meaningful.
EMC owns great assets, but shareholders and management are not realizing the value they should from the highly attractive businesses within EMC’s holding company structure. The potential for a separation has often been discussed, and the benefits from a value standpoint are enormous. The sum-of-the-parts valuation is meaningfully higher than EMC’s current stock price and represents a significant opportunity to drive a substantial increase to shareholder value by adjusting the Federation structure.
We have modeled our assumptions below and discuss them here. First, we believe it is beyond debate that EMC II would trade at a premium to NetApp. EMC II is the market leader in storage and has coveted assets such as Data Domain, Isilon, RSA and XtremIO that NetApp does not. Second, we believe that VMware, at its current valuation of 14-15x Enterprise Value/2015E Unlevered Free Cash Flow, is trading inexpensively given the company’s growth profile. Though we don’t model it below, we believe that as an independent company with the opportunity to more freely partner and more aggressively invest in software-defined storage, VMware would have the opportunity to grow faster and be worth more.
In order to realize the intrinsic sum-of-the-parts value of EMC, a tax-free spin-off could be initiated immediately. However, the opportunity for shareholder value creation does not stop there. Core EMC is a slower-growth but extremely stable company, and its capital structure should reflect this reality. Once VMware is separated, we believe Core EMC can and should add leverage to repurchase stock. Core EMC could easily repurchase 20% of its stock while retaining sufficient cash to do growth-additive M&A. In the below analysis, we show the highly positive impact that a buyback would have on Core EMC’s stock price, post-spin, even after its valuation multiple has meaningfully increased from its implied level today – as you can see, the value accretion is highly attractive:
(1) Funded by $4.8B of new debt (pro forma leverage of 2.0x), $2.7B of US cash and $1.5B repaid intercompany loan. Core EMC will generate over $3.3B of free cash flow per year with the additional interest expense.
(2) Repurchase Price represents 6.75x 2015E EMC II EBITDA of $5,605M plus $1.3B valuation for Core EMC's 62% interest in Pivotal based on a 7x 2015E Revenue valuation and includes repayment of $1.5B intercompany loan in Core EMC cash balance.
(3) Pro forma for buyback. Assumes interest rate on cash of 0.5% and on debt of 3.75%.
(4) In line with NetApp's current P/E multiple of 13x as of 10/6/14 for conservatism.
(5) Represents Core EMC's 62% interest in Pivotal at 7x 2015E revenue, which is a conservative assumption given Pivotal could IPO at a 10x forward revenue multiple based on certain comparables.
(6) Additional upside potential associated with sale of Core EMC to strategic buyer after spin-off.
(7) Represents market value of EMC's 80% interest in VMware as of 10/6/14.
(8) Uses treasury stock method based on Total Value to EMC Shareholders share price.
Elliott recognizes that there are other mechanisms to accomplish a separation of Core EMC and VMware which also drive meaningful stock price upside. One of these mechanisms is a split-off whereby EMC shareholders are given the option to exchange their shares for VMware shares at an attractive exchange ratio (which effectively enables the Company to use VMware stock as currency to repurchase EMC shares in a tax-free fashion). The above analyses contemplate a spin-off, but the benefits of a split-off are also compelling.
Whatever the mechanism used for a separation, both Core EMC and VMware would retain their significant strategic value. Core EMC contains spectacular assets, making the company highly attractive to numerous strategic buyers. After a tax-free spin or split of VMware, Core EMC could be acquired on a tax-free basis, an outcome that would result in even greater value to shareholders than the value illustrated in the analysis above. We believe this is the likely outcome as we have a well-founded understanding that Core EMC is a strategically attractive asset to a number of parties.
M&A Opportunities for EMC
As has been long speculated, EMC and its assets are attractive to a number of parties in the broader enterprise IT landscape. Since the announcement of Elliott’s position in EMC, we have learned of acquisition interest in EMC’s assets on the part of several large companies that make strategic sense and which involve parties that could both afford to acquire the EMC Federation as a whole or distinct assets within the Federation, and could also successfully integrate these acquisitions. An acquisition of EMC by any of these buyers would create the leading enterprise IT company in the world, selling every component of the data center and a dominant converged infrastructure offering. It would be a highly attractive outcome for both the acquiring company and EMC, as well as for the combined company’s shareholders.
We believe that it makes sense for EMC to fully explore acquisition interest (and to do so in a way that preserves the option of a tax-free spin-off of VMware) and we understand that much of that work is ongoing. Elliott would be supportive of a value-creating transaction that would allow current EMC shareholders to continue to benefit from the upside of a combined EMC and acquirer company. Additionally, we believe it may be advantageous for EMC to pursue a spin-off of VMware while evaluating these other options. Announcing and pursuing a spin-off does not preclude an M&A pathway.
Now is a Great Time
We believe now is not only a great time but the optimal time for EMC to establish a future structure that makes financial and strategic sense for the long term. Whether through a tax-free spin-off of VMware or through M&A, the options are compelling due to the incredible quality of EMC’s assets. Regardless of which path is chosen, the conclusion is clear to us – as it is to many – that the current Federation structure is not right for EMC or its shareholders.
We hope that you find these perspectives and observations useful. Elliott has spent a considerable amount of time diligencing and developing these ideas, and I personally believe in their merits.
I want to reiterate the tremendous amount of respect we have for Joe and his team and for what the Company has accomplished.
Thank you very much for the time and consideration.