For Rackspace, Growth Comes at a Price

• Report this Comment On November 11, 2009, at 5:51 PM, TMFMileHigh wrote:

  Hey Eric,

  Thanks for the mention, but I think your analysis is missing two big issues.

  >>Based on 2009 revenue forecasts, that number should approach a full 30% of sales. As a reference, telecom industry capex levels are traditionally in the 15% to 20% range.

  This comparison doesn't make sense. Rackspace isn't a telco, and its data center build out is necessary to support increased demand. What's more, return on capital is above the average cost of debt. Expansion isn't bad if (a) demand is present, and (b) it's paid for equitably.

  >>Since IT gear tends to depreciate in value pretty quickly, you can count on these investments amounting to a large chunk of Rackspace's revenues for the forseeable future, and keeping down earnings and free cash flow along the way.

  Of course they do. But again, how is this bad if the useful life of a server is at least four years and a multi-tenant architecture allows for increasingly higher returns on capital?

  There's no doubting that Rackspace faces unknowns. You're right to question whether this business is sustainable, and whether its capex outlay is justified. And yet I think there's too much alarm, with too little explanation in your analysis.

  Investors should remember that this is a business whose legacy is single-tenant, customized architectures. The higher capex recognizes that Rackspace is doing more business in the multi-tenant world of the cloud. Fewer houses, more hotels, and ever-higher returns on invested capital.

  That's why I'm thrilled to have Rackspace as a Rule Breakers pick.

  FWIW and Foolish best,

  Tim (TMFMileHigh and @milehighfool on Twitter)

• Report this Comment On November 11, 2009, at 7:15 PM, TMFNomad wrote:

  Hey Tim,

  You're right that Rackspace isn't a telco, and that it has a positive ROIC, but the high capex/revenue ratio is putting a limit on the kind of returns that the company's getting, as shown by its FCF numbers.

  Also, given that the company's capex is directly tied to customer demand, it doesn't follow that a positive ROIC thus far means that issuing more debt to finance additional capex will deliver the same results. At least not unless the company is badly supply-constrained.

  Regarding the depreciation issue, the key point is that IT gear carries much higher depreciation rates than for a data center build that a colocation company like Equinix might do. Servers need to be replaced every few years, whereas a new data center will only need incremental investments after it's been built.

  I agree that the ramping of the cloud business should lead to more efficient hardware use thanks to multi-tenant server usage. But as of last quarter, the cloud business was still less than 10% of revenues. In addition, its average revenue per cloud customer is much lower than that of the managed hosting business, leading to some legitimate margin concerns.

  I do think RAX's business is sustainable - the company has turned FCF-positive, and if growth continues, I can see it generally staying that way. I'm just skeptical about the company generating the kind of cash needed to justify its lofty valuation, given how much capex is tied to investments that need to be incrementally made as new customers are added, and which involve products with pretty short lifecycles.

  Thanks for the feedback and constructive criticism.

  Eric (TMF Nomad)

• Report this Comment On November 11, 2009, at 8:21 PM, TMFMileHigh wrote:

  Hey Eric,

  Good stuff. Three more clarifying points if you'll allow me:

  >>I'm just skeptical about the company generating the kind of cash needed to justify its lofty valuation, given how much capex is tied to investments that need to be incrementally made as new customers are added, and which involve products with pretty short lifecycles.

  I think you need to be careful here. The "product" isn't what has the short lifecycle. This is a service-driven business with multi-year SLAs in the cases of its managed hosting customers.We don't know exactly what the equipment replacement cycle looks like. Nor do we have a perfect view of maintenance versus growth capex.

  Still, you raise a good point and I'll be sure to do the math on the average cost of incremental server investment based on what data RAX publishes.

  Second, we can probably agree that RAX isn't going to be adding 10,000 new mostly cloud customers per quarter for much longer. So while growth and capex may correlate right now, they're likely to decouple long-term. When they do, older investments will earn higher returns.

  Finally, while it's not happening now, there's nothing preventing RAX from adding customization features to its cloud offerings, and customization is what creates margin in this business. I like RAX's chances of earning incremental margin and cash flow gains.

  Thanks for the extra comments,

  Tim (TMFMileHigh and @milehighfool on Twitter)

• Report this Comment On November 13, 2009, at 1:18 AM, TyrantBone wrote:

  Gotta love this site, constructive criticism on this subject is a win for us readers!

• Report this Comment On November 18, 2009, at 11:20 AM, RaulChapin wrote:

  To add to the list of happy "prospective" customers.

  Our company has a small rather static website (with a blog that gets updated daily at best)

  We had a hosting provider that was charging us $200 a month for "99.9% uptime" until it went bankrupt and left us with no site for a few days!!

  So our manager thought nevermind this hosting service from out of country (Canada here) lets go local. The local people wanted $700 per month (they are our ISP)

  I thought why not mirror our site on a cheap webhosting and then maybe take advantage of having better access speed etc (I admit my actual knowledge of Internet infrastructure is a bit outdated)

  Long story short, from an article at the Fool on Akamai and its competitors (Rackspace) i decided to have a look.

  Well i can get the same space on their cloud for $100 a month... that means i can have a top level provider for half of what i was paying our now defunct host and get our page served from top of the line servers from them.

  I am still looking, but if there is not something obiously wrong... we will be with them for years and years to come.

  Why is this relevant? because a) they have space to play a bit with ther prices (we were willing to pay 7 times more for likely lower service... due to our scare with the cheaper smaller player going bust on us)

  b) the pain of switching is not one that many companies want (except the ones that are heavily involved in IT) meaning a lot of people that are small to medium once they find how cheap you can get top service for will not want to change later.

  c) if they go bankrupt i am in sooooo much trouble LOL (Actually we will pay $150, $100 to them and $50 to a small supplier myhosting.com to keep a mirror of our site up at all times but not actually available to the public... so in the very remote possibility that Rackspace goes bust... we just change our DNS information immediately and wait for 24 hours for propagation.

  So, even with a silly added security of having two sites up at all times while only using one, we still save $50 a month over our difunct provider and $550 over the option we were considering... That is $6600 a year in savings for a small to medium company...

  Regards and thank you for providing way more information on webhosting that i found in google LOL (You know i am an accountant when i look in the Fool for IT help ROFL)

  I hope my anecdote will be of use to the Fool community !

Add your comment.