Drowning in data

Opinions expressed whether in general or in both on the performance of individual investments and in a wider economic context represent the views of the contributor at the time of preparation.

Viewed from the outside, the building your author recently visited as part of his research for this piece looks most uninspiring. Located a brief walk from the Canary Wharf financial district of London, it was built around 20 years ago out of concrete and glass, and stands five floors high. Yet inside, this is where all of Amazon’s and Facebook’s European internet traffic is exchanged. It also acts as a perfect metaphor for the ‘data deluge,’ a phrase we first coined when we discussed the topic in March 20111. In other words, data is ubiquitous, whether we realise it or not. Demand trends look irrepressible and there seems no reason to expect this phenomenon to reverse any time soon. Many related investment opportunities are inevitably arising as a result, some of which are profiled below.

According to my guide for the tour I undertook of the above data centre, “what we build [in terms of new capacity], we can fill.” Most of the markets in which they operate are now “capacity constrained” and when asked what constituted the biggest execution risk for the company in question, the response was simply being able to develop new centres quickly enough, overcoming planning requirements and other logistical hurdles such as negotiating with power suppliers. Readers of this piece should be far from surprised by these comments given the rapidity with which data is being created. IBM estimates that 2.5 quintillion (i.e. 2.5 x 10 to the power of 18) bytes of new data come into existence every day. By way of easier explanation, consider that every minute of the day, more than 200m emails are sent, over 2m search queries are made on Google, greater than 600,000 pieces of content are shared on Facebook and some 50,000 Apps are downloaded globally.

It is not just consumers, however, that are being caught up in the data deluge, but businesses too. By way of example, take Wal-Mart, the world’s largest retailer: it handles over 1m customer transactions every hour. These are imported into its databases, which are estimated to contain over 2.5 petabytes of data, equivalent to approximately 50m filing cabinets worth of data in the ‘real’ world. For Wal-Mart, or indeed any other organisation, irrespective of size, analysing this data can help drive optimisation – from supply chains to customer relationships – in order to establish competitive advantages. A recent McKinsey study found that retailers analysing their data could have the potential to increase their operating margins by over 50%. Elsewhere in the report, McKinsey calculated that the US government could generate up to $300bn of annualised cost savings through the analysis of healthcare data, while most European healthcare administrations could make reductions in the region of $100-150bn annually.

Two clear trends are at work here: first, the growth of data and second, the need to store, secure, retrieve and analyse the data in question. Cisco expects overall internet traffic to quadruple over the next four years and for growth in data to run at an annualised rate of over 30% through to 2015. The spread (and speed) of broadband, mobile, digitisation, social media and the cloud constitute the key drivers, underpinned by improvements in processing power (Moore’s Law) and storage costs (which continue to fall – by some 30% in the last five years according to some estimates).

To take just one example from the above list, approximately 4bn people, or 60% of the world’s population currently use mobile phones. Of these, around 12% are smartphones, a category that is currently expanding at about 20% a year. Average smartphone data generation is currently running at 150 megabytes a month, almost triple 2010 levels. The growing use of tablets such as the iPad is also generating additional data demand. Some 70m were sold in 2011 and shipments are presently expanding at a rate of more than 40% per year. As a consequence, according to Cisco, by the end of this year, the number of mobile connected devices will exceed the number of people on earth, and by 2015 the company forecasts that over 10bn such devices will exist globally.

Such growth inevitably creates both opportunities and challenges. Most new data currently being created is referred to as ‘unstructured.’ In other words, items such as text messages, images, social media feeds, video and much machine data does not fit into conventional tabular form and so requires sophisticated tools in order to store and retrieve it. Industry researcher IDC estimates that unstructured data accounts for around 90% of total data volumes and is growing at a rate three times that of structured data. Viewed from another perspective, McKinsey calculates that only about 20% of all data currently in existence is being analysed. A separate (but increasingly important) issue is also how securely data is stored, should it be compromised via a cyber-attack or something similar.

Against this background, it would seem highly logical – and more importantly, necessary – for corporates to be increasing their investment in data analytics and storage. The dangers of failing to implement such a strategy and running the risk of losing ground relative to the competition is significant. Amazon, for example, says that over 25% of its sales are now generated by its recommendation engine (“you may like…”), a product it has been investing in for a number of years. Chief Information Officers cite data analytics as being their number-one investment priority for next year and in a recent study undertaken by Morgan Stanley of 300 IT purchasing managers, spend on technologies that more efficiently access and utilise data are set to expand to 28% of average IT spend over the next three years (versus 24% at present), creating an incremental $115bn spending opportunity.

How and where this money will be spent is still unclear. The market is complex and also heterogeneous. Investment is required not only in software and hardware, but also in bandwidth and connectivity. Protection and security will also become increasingly important factors. An additional challenge is constituted by the speed at which the industry is developing, implying that established vendors often lack next-generation technologies, while new entrants have been quicker to embrace the opportunity.
From an investor’s perspective, the over-exuberance witnessed at the time of the tech bubble is still fresh in the minds of many, implying the need for caution, particularly when a number of companies present in the space seem to feel it incumbent to enunciate a data or cloud strategy even if it may ultimately resemble the emperor’s new clothes. Understanding the most essential constituents of the value chain is therefore crucial in terms of identifying potential winners, irrespective of the strength of the underlying theme.

One starting point is to consider that any data strategy will likely necessitate an expansion in services to move and store data. While around 80% of corporate data centres are currently in-house, these are typically inefficient (with either too much or too little spare space, implying wasted operational costs or increased future capital expenditure requirements). Data centres enable customers to warehouse and easily scale up their data strategies and such centres also bring the benefit of carrier-neutrality, allowing clients to transfer their data quickly and via a range of different service providers. External storage lowers the risk of enterprises misallocating costs.

Leading players within the field include Telecity and Equinix, who have delivered annualised returns of 23% and 12% respectively over the last five years. Telecity has eight data centres in London alone (including the one Helicon visited) that offer a total of 24,000m2 of colocation and hosting space in total. The industry has structurally high entry barriers, resulting in high recurring revenue streams and correspondingly attractive margins for the businesses present.

Within the hardware arena, those companies that have been most proactive in embracing and acquiring new technologies look to be best-placed. The boundaries between hardware and software are also becoming increasingly blurred, with the former typically expanding into the latter’s preserve rather than vice-versa. EMC and Cisco are among those that have established positions of strength. EMC is the global market leader in data-centre services (enterprise storage systems, software and networks) and should benefit from data growth, virtualisation and cloud services.

Meanwhile, Cisco has made data centres one of its five priority areas and reported over 90% year-on-year growth in this area at its most recent results. In addition, Cisco plans to return over 50% of its free-cashflow to shareholders going forward, while the highly cash-generative nature of EMC’s business could also allow it to embark on a similar strategy. In order to remain at the forefront of developments we should also expect continued M&A from both these companies (and others – such as IBM, SAP and Oracle), particularly in the area of security, which is likely to become increasingly crucial.

Even if the likes of Telecity and Equinix, EMC and Cisco do not emerge as among the ultimate winners from the data deluge, what does remain clear is that demand for their and other related services is likely to remain unabated for the foreseeable future. Correspondingly, for corporates of all size and across all industries, it is not an option not to spend on espousing and incorporating the relevant technologies. The deluge is upon us.

Alexander Gunz, Fund Manager