April 25, 2017 - 8:12am EST by
2017 2018
Price: 44.84 EPS 2.64 2.83
Shares Out. (in M): 4,115 P/E 17.0 15.8
Market Cap (in $M): 184,502 P/FCF 20.8 17.9
Net Debt (in $M): -6,157 EBIT 11,243 13,293
TEV ($): 178,345 TEV/EBIT 15.9 13.4

Sign up for free guest access to view investment idea with a 45 days delay.

  • Technology


As one of our portfolio managers was building his software business in the 1990s, he witnessed first-hand the speed with which the Oracle database overtook its competitors and became the solution of choice for most of the new industrial strength data-driven applications being custom built at that time. As investors, we kicked the tyres on Oracle a few years ago but were put off from digging deeper by widely reported risks to this dominant position in enterprise databases from the open source movement and SAP’s new database product, HANA.
In 2015 and 2016, we undertook industry-wide research looking at software application providers. The catalyst for this research was the very positive impact on long-term cashflow and margins which can be achieved by these businesses switching from traditional ‘on premise’ business models to the new paradigm of delivering products via the ‘cloud’. This research led us back to Oracle’s door. We were struck by the fact that the risks which had concerned us five years ago had not materialised. Digging deeper, we also realised that Oracle presented a rare opportunity to invest at an attractive price (12x free cashflow on initial purchase) in a technology industry giant that is well positioned to benefit from a coming disruptive wave. 
At Oracle’s 2016 investor day, co-founder and Chairman, Larry Ellison, rebuffed the doommongers with characteristic swagger:
“You can argue Oracle hasn’t been competing in database for at least 10 years… there have been so many stories of the next thing that was putting us out of business… I've heard this for as long as I've been here.”
Management bluster is nothing new in Silicon Valley, nor the corporate world at large. Accordingly, we examine such claims with due scepticism. Yet Ellison’s boasts stand up to scrutiny. Oracle’s control of the RDBMS market has persisted for two decades, with the company maintaining at least a 2x RMS for the last ten years. We have tested hard three potential threats to Oracle’s dominance in the enterprise database market:
The threat from Open Source:
“Free” Open Source alternatives have been available for 15+ years but have failed to make serious in-roads into Oracle’s installed base of enterprise customers. An expert with one of the top consultancies explained to us:
“Open Source products like PostgreSQL are good for small environments where you have a very technical development team; however, it lacks the resilience and scalability of the industrial strength database products like Oracle and SQL Server. Enterprises are unlikely to take this risk for their mission critical apps”.
The threat from HANA
It is estimated that 50% of SAP’s Enterprise Resource Planning ‘ERP’ implementations run on an Oracle database. SAP has been targeting the database part of its ERP stack since 2010, when it launched its own database solution, HANA. Initial deployments of HANA were largely focused on BI (Business Information or management reports) but for the last few years, SAP has been pushing its customers hard to consider HANA as the solution for their core transactional database driving their ERP system. SAP claims that making the switch from Oracle to HANA is a largely automated procedure. Oracle is characteristically dismissive of this threat, claiming that SAP implementations represent a “tiny” proportion of Oracle database maintenance revenues. We are sceptical of this claim but even if this figure was as high as 25%, this only represents 7% of Oracle’s total revenue. If 100% of these customers switch to HANA in the next fifteen years, then it represents a 0.5% of revenue headwind per annum. We think this scenario is very unlikely. One of our industry interviewees told us:
“Moving SAP to HANA is really easy if all you’ve done is just configure your SAP application. However, I don’t know of a single SAP instance where there isn’t a significantly customised code-base. To make the migration easy you first have to de-customise your ERP and then shift it to HANA. This is not a trivial process. Anyone who tells you it’s simple either doesn’t know what they’re talking about or is being disingenuous.”
Added to this is the fact that many SAP customers will have other applications running off the same Oracle database so a shift to HANA would require all of those applications to be upgraded and in many cases rewritten.
Threat from NoSQL
The latest alleged challenge to Oracle’s database crown comes in the form of “NoSQL” databases, which deviate from the traditional tabular structure of RDBMSs and are a good fit for handling big data, geospatial data, real-time web applications and the like. Nonetheless, RDBMS systems remain better suited for use cases with high intolerance for data loss or inconsistency. For the foreseeable future, the RDBMS vs NoSQL debate looks to be a case of “horses for courses”; systems that hold critical financial transactions at their heart look highly unlikely to switch over from traditional SQL databases any time soon. The market for RDBMS is currently more than 10x that of NoSQL. Nevertheless, the rise of NoSQL is an area we will watch closely.
The relational database is typically integral to the day-to-day functioning of an enterprise and deeply ingrained within the information architecture of a business. This dependency makes Oracle’s core product exceptionally sticky. One technology director we interviewed opined that “moving off Oracle would be a three-year project for 20 people” – far from a trivial switching cost! Another technology professional with over 25 years’ experience described the process of switching database
provider as trying to “change the engine while the bus is moving”. The product is made even more sticky by Oracle’s ecosystem of trained and certified database administrators (“Oracle DBAs”), a complementary network of legions of professionals who earn a living off the back of Oracle’s database and therefore share an interest in its endurance.
For all of these reasons, we are inclined to agree with Ellison Oracle won the database wars a decade ago and its moat remains wide.
Oracle’s application business provides Enterprise Resource Planning (ERP), Human Capital Management (HCM) and Customer Relationship Management (CRM), and other software products to businesses and organisations across the world. In contrast to database, Oracle's applications have typically lagged in second or third place in every major category behind rival SAP. The applications market is just as sticky as the database market, and for similar reasons; enterprise applications are revenue-critical systems which are risky and painful to switch. SAP's c.97% customer renewal rate bears testament to this as does Oracle's failure to surpass SAP in on-premise applications.

SAP has, however, been surpassed in CRM by upstart software as a service (SaaS) company,  Salesforce was founded by Marc Benioff, a former Oracle employee and protégé of Larry Ellison.  Salesforce’s stratospheric rise over the last decade or so is a classic case study in technological disruption.  Both SAP and Oracle shared a time-tested model for selling on-premise applications: customers purchase an upfront licence fee, followed by an ongoing annual maintenance fee of c.20% of the initial licence cost.  Salesforce opted for an entirely different model.  It delivered its CRM applications as a service over the internet and eliminated expensive upfront licence fees.  This shifted the cost of CRM from capex to opex – customers would continue to have use of Salesforce’s software so long as they paid their ongoing subscription. 

Using this SaaS model, commonly referred to as the “public cloud” approach, Salesforce unseated the industry incumbents and took pole position in overall CRM market share.  In the process, Salesforce became a $60bn market cap company, albeit with a formidable 300+x trailing P/E multiple.


In the meantime, Oracle’s top line has remained broadly flat since 2012 as the company has worked on its own public cloud strategy.  Alongside currency headwinds, management cites the switch from the front-loaded, capex business model to the subscription-based, opex model as a key reason for the slowdown in revenue growth.  Management teams are, of course, often proficient at explaining away less than flattering financial numbers.  Accordingly, we have closely tested the switch-over effect of Oracle’s new cloud-based business in our modelling.  The substance of the effect is best illustrated by a simple table:  


The loss of the upfront licence fee payment in year one is felt for nearly four years, which goes some way to explaining Oracle’s multi-year sales slowdown.  The above logic applies to both the database and the applications side of Oracle’s business.  In the example given, the on-premises licence that produces $20 of revenue annually can be said to have an “installed base value” of $100.  In our modelling, we have also considered the simulated “installed base value” of Oracle’s cloud customer base, i.e. the equivalent value that the cloud subscriptions would represent if they were treated as licensed products.  Interestingly, this modelling exercise indicates organic growth in Oracle’s installed base value of 6-7% during the last four years (after adjusting for currency effects and the impact of acquisitions), in line with the average growth of the business over the last 15 years.  In spite of scepticism from the market, we believe Oracle’s underlying business continues to grow. 

If Oracle will earn more revenue from cloud software in the long-run, that must, of course, mean that the customer is willing to pay more money to Oracle for that software. We always seek to test whether customers are gaining real value from the products or services offered by our portfolio companies. A piece of software that does the same thing it did before, but now does it in the cloud at an increased price, does not seem like a strong value proposition. After talking with several customers, however, it became clear to us that this was a misleading characterisation of the cloud offering. The unseen cost of on-premise software lies in paying for customisation, data centre maintenance and expensive consultants to manage upgrade cycles every six years or so. Hence, cloud software can strip out enough ancillary cost that its long-term total cost of ownership is actually lower for customers, despite the higher on-going subscription charge: a winwin for Oracle and customer alike. Cloud applications are also quite likely to provide a better user experience for the simple reason that responsibility for running the software becomes aligned with the people responsible for writing it. Moreover, customers always receive seamless upgrades to the latest version.
The economics of cloud applications look attractive per se, but the switch to the cloud also offers an attractive opportunity for Oracle to gain share in the applications market. SAP has consistently outdone Oracle in on-premise applications, but early signs suggest that Oracle has stolen a nontrivial lead on SAP in the cloud. Cloud CRM continues to be led by and cloud HCM is strongly contested by another SaaS contender, Workday. Cloud ERP, however, is currently
dominated by Oracle; the data suggests that Oracle has more than 10x as many customers as Workday, its nearest cloud ERP competitor. Moreover, ERP is the largest total enterprise applications market at over $30bn.
To meet more Oracle cloud customers, we attended Oracle’s “Modern Business Experience” conference at London’s ExCeL centre in early February. The conference was impressively well attended, with over 3,000 customers, prospective customers and business partners congregating to hear from Oracle co-CEO, Safra Catz, among other senior executives. The event was a timely opportunity for us to speak to a broad range of clients and gain a sense of the corporate appetite for Oracle’s cloud offering. We say Oracle’s cloud offering, because cloud software was unambiguously the focus of the day, as it now is in every major Oracle presentation and earnings call. We entered the conference cautiously optimistic about Oracle’s cloud prospects. We left the conference rather more bullish on those prospects and impressed by the number of large clients transitioning to Oracle’s cloud or considering doing so. Oracle co-CEO, Mark Hurd, has recently boasted that over half of Oracle’s cloud ERP customers are “net new to Oracle”. Our experience supports this claim.
Our conversations with customers did however uncover a customer advocacy risk: in two separate cases (one a FTSE 100 company and the other a FTSE 250 company), they confirmed that it would be very (too) expensive to transition away from their existing Oracle installations, but they also told us that Oracle was a very aggressive vendor always looking to extract additional fees where it could. In both cases, this meant that they would not consider using Oracle for new applications. This is certainly a risk that we will watch.
Technological disruption has proven the demise of many a large industry incumbent over time. On its current multiple, the market seems wary that Oracle could follow a similar path. Following our research, we feel far more confident in Oracle’s direction. Oracle’s total customer base is over 400,000 strong yet the company maintains a 95%+ customer retention rate, a level of stickiness that is hard to come by in any industry. This existing customer base remains highly cash generative. At
the same time, Oracle will benefit from any cannibalisation that its cloud business achieves and the company has demonstrated innovation and business execution in taking the lead in one of the largest potential cloud applications markets.
From talking to people who have met Ellison, and reading extensively around the subject, including the excellent business biography, SoftWar, we have built up a strong and largely positive impression of the competitive and brilliant co-founder of Oracle. Ellison is often described as the best  technologist in Oracle even at the age of 72. His passion has always been more in the technology but his attempts at delegating the operational management of the business in the 80s and 90s were
not pretty; the business nearly went bankrupt in 1990 as a result of some aggressive accounting by the Oracle sales team. More recently, as the business has matured, Ellison has become more interested in the intellectual challenge of making a business more operationally efficient. There is evidence that this is working: of its largest enterprise software application peers, Oracle has for the last few years vied with Microsoft for the top slot on margins. The current management team
consists of a CEO duo: Safra Catz, who has been one of his more durable appointments, joining the Board of Oracle in 2001 and Mark Hurd, the highly-regarded ex CEO of HP. Ellison has in theory stepped back to focus on his role as CTO, however when the current structure was in put in place, Catz was quick to dispel concerns: “There will actually be no changes, no changes whatsoever.” Ellison remains a very visible part of this management team (when he is not winning the Americas
Cup) and despite his age, appears to have lost none of his competitive spirit. Ellison retains a personal stake in the business worth $50bn and has stated that “I’m stuck here for the duration; there’s no way I can stop until I know how this story ends”. As co-investors in his business, we hope (and believe) that the story has a long way to go.


I hold a position with the issuer such as employment, directorship, or consultancy.
I and/or others I advise hold a material investment in the issuer's securities.


Transition from on premise to cloud.

    show   sort by    
      Back to top