DCS: Are the finance houses still keen to invest in the data centre (DC) market, or has the credit crunch had an impact on the financing options?
KT: We've recently seen an increased desire and opportunity to invest in the DC market, both from a debt and equity perspective. Some of the recent debt financings raised have been directed to existing debt refinancing.
The credit crunch, although affecting some companies' ability to raise debt, was more relevant in 2008 and 2009.
DCS: And has the credit crunch had an impact on your plans to invest in new DC facilities?
KT: The credit crunch didn't affect us negatively – we had ample cash on our balance sheet, and we were generating positive discretionary free cash flow. The credit crunch allowed us to expand when others were more focused on managing their existing debt and operations.
DCS: Put another way, the market statistics suggested a slight slowing down in DC activity, but no significant impact of the credit crunch – is that a fair assessment?
KT: I believe the impact from the credit crunch was more of a 2008 and 2009 phenomenon. Although, today, more capital is sitting on the sidelines waiting to be deployed into the industry, given it is such a significant capital investment, I don't believe we'll see reckless spending moving into the industry. Even when money was at its cheapest, being 2005 – 2007, there wasn't over-investment in the industry. That said, this is a strong industry and many want to participate in the upside.
DCS: And with new DC builds seemingly announced most days of every week, is the DC market buoyant again?
KT: The DC market isn't too buoyant and although there have been may announcements, some are speculative and conditioned on customers signing pre-leasing arrangements. The large deployments are far and few between – with Equinix, AT&T and Digital Realty making the largest investments in the space.
DCS: Are you able to share with us a typical new build/expansion project in terms of the financing lifecycle – ie plans, seeking investment, starting the build etc.?
KT: The average project life cycle is 18-24 months assuming you have yet to determine the location of the DC. Once determined, depending on size and permitting, the time to deliver is 9-18 months. In part, this is why you've seen Equinix taking down large amounts of space and then phasing the capital deployment. Once the initial decision is made, our ability to put up the next phase is much quicker and typically a lower cost per sellable unit of capacity.
DCS: Do you tend to work with the same set of VCs and investment banks, or each time you go to market do you have to ‘re-sell' your company and the specific new DC build?
KT: At this stage of our existence, we typically would not work with a VC but we do spend plenty of time with our investment bank advisors on capital raising efforts and inorganic market opportunities.
DCS: Do the investors understand the DC market, or are they simply interested in the attractiveness of the sums?!
KT: Our investors very much understand our industry, and in particular, our company. Being the global leader in data center services, with a differentiated story plays well to our investors and debt holders.
DCS: How competitive is the DC build market in terms of accessing financing?
KT: The DC market can access capital fairly easily – the question is at what cost? Digital Realty is an investment grade REIT that can raise cheaper money than any other major player in the market, in part given the length of their customer contract, and in part due to the ownership of many of their assets. Other companies, both REITS and managed service providers, have a more difficult time raising capital at a reasonable cost.
Equinix is fortunate to have recently raised $750 million of capital at 8.125% with an 8 year maturity – the benefit we have, although not investment grade, is our global leadership position in data centre services and long term operation performance.
Our cost of funds is about 200-250BPS higher than Digital but lower than many of the other providers in the market. Also, our debt is unsecured with no guarantees from our subsidiaries – this gives us great flexibility as we look forward in time.
DCS: And how competitive in terms of the capacity being built or planned?
KT: We are very competitive with our capacity planning, not to mention our ability to offer services across 35 markets in the US, Asia-Pacific and Europe. We look at each market prior to making an incremental investment decision – if a market gets sloppy, or we can't differentiate our service offering, it may cause us to invest our capital in the other markets.
DCS: Many DC ‘openings' are announced before they are built – presumably to attract customers, which then pleases the investors – what is your company's approach to this issue?
KT: In many ways that is why some feel the market is getting over built. Companies announce but don't start the investment until they get a customer to sign. We build based upon our market intelligence and customer demand as up to 80% of each quarter's incremental business comes from our installed base. Although I would state we don't speculatively build, we do commit capital prior to a formal customer commitment.
DCS: Put another way, do you announce a new build, then hope to attract customers, or start a new build because customers are requesting it – or a mixture of both approaches?
KT: Per the above but recognizing that once we commit to a project, we work with our sales and marketing organisations to roll out the deployment with the objective of pre-selling within six months of opening one of the advantages Equinix has is that we tether all IBXs back to the network dense primary IBX, an advantage others don't have.
DCS: In simple terms, what factors influence the decision to build a new DC/expand an existing facility?
KT: We have many factors affecting our decision – market size and opportunity, customer need and demand, market vertical opportunity, strength from a pricing perspective, cost to build and the competitive landscape.
DCS: Do you have a typical ROI model for the DC build/operate cycle, or is each project treated as a one-off?
KT: Equinix has a hurdle rate whereby we target IRR returns over a 10 year period of 40% but that said, each project is unique, and to the extent we need to deploy capital at rates below this target, we'll look at the key factors and then make that decision.
We work to drive down our cost of capital, which will give the company increased opportunity to meet its financial targets under a different economic scenario. Also, we're trying to maintain as much flexibility in our terms of financing to allow the company to maneuver as efficiently as we can given the changing landscape in the Internet infrastructure industry.
DCS: Which market sectors do you see as healthy for expansion right now – whether geographical or industry-sector driven?
KT: Many of the market opportunities are healthy today whether it be network, electronic financial trading, mobility, cloud computing, software-as-a-service or enterprise. We have teams focused on each segment allowing us to diversify our revenue base and create strong opportunities for expansion in each of our three regions.
DCS: And have you begun to try and factor in the likely impact of The Cloud in to your DC build/expansion programme?
KT: Consistent with the above comments, as we continue to grow our verticals, including cloud, we've assessed the potential market opportunity for Equinix.
As one can imagine, this opportunity is large, and only highlights the potential for the DC service providers.
DCS: Indulging in a little more crystal ball gazing (!), do you see steady, or even rapid, growth in the DC market for the next few years (esp. as older DCs are upgraded and/or replaced)?
KT: Equinix continues to be optimistic about the opportunities for growth, and since we've been growing at a multiple of the market growth, this gives us continued confidence in our build opportunities for the foreseeable future.
We believe we have good market and trend data for the next 24-36 months but thereafter, it is more of a crystal ball gaze than anything else. Given the market supply and demand still appear to be at an imbalance, we've shared our perspective on our last earnings call suggesting that our next internal target will be $2 billion in revenues, effectively doubling the size of the organisation over the not too distant future.
DCS: What factors influence the decision to build from new as opposed to upgrading an existing DC facility?
KT: The company generally focuses on building new opportunities versus retrofitting older data centres. This is not to suggest that someday down the road it may become economically opportunistic to retrofit older data centres.
But today, given the increase in energy consumption at the server level, it would cause a company to make a much larger investment
to upgrade the older data centres to meet today's needs. I see this as another opportunity to limit the level of new investments in the DC industry.
DCS: Although not strictly part of the investment cycle, how does the price and availability of power impact on your DC build programme?
KT: Actually, both price and power availability do factor into our investment decisions. Access to power, given power and cooling infrastructure are generally limiting factors today, not physical space, ultimately determines the size of the investment in a data centre. Clearly, the larger the centre, the better the economies of scale. This consideration also causes the company to consider average consumption per sellable unit, which determines our cost to build. The other side of the equation is operating costs, which includes power per KWH, as this impacts our ability to price and sell in a given market.
Clearly, if a customer needs to be close to their servers, they need to recognise there are some lower and some higher costing environments. We take that into consideration when we build.
Silicon Valley is a perfect example – some build in Santa Clara, as the power is delivered by Silicon Valley Power, a cheaper source of power than San Jose, whereby this power is delivered by Pacific Gas and Electric.
DCS: Do you see the entrance of the telcos and the large IT corporations into the DC market as a threat, or confirmation that it's good market to be in, and that, as an established player, you're in the ‘box' seat?
KT: Today, we already compete against the large Telcos and the large IT shops, or System Integrators. In fact, depending on what their end customer needs may be, both of these services providers could also be our customers. An example of this is IBM – they are our largest customer represent approximately 4% of our global revenues.
Additionally, AT&T also represents a large customer of Equinix although this would typically relate to a component of their organisation that is more networks focused.
Tags: Hosting & Colocation, Applications, Design & Facilities Management, Green Tech