Sunday, January 24, 2010

Financial Health of the Colocation Industry

I would like to preface this post with the statement that I am very much a novice when it comes to stock analysis and any form of insightful opinion into the valuation of a publicly held company. However, I think it is interesting, I'm learning, and it allows me to really dig into the inner workings of a company.

I decided to do a little analysis of the colocation industry and the specific stocks I was interested in
  • Equinix
  • Digital Realty Trust
  • Savvis
  • Terremark
  • Dupont Fabros
  • Rackspace
  • Internap
I also like to track merger and acquisition trends and predict how the landscape will change in coming months. With acquisition activity picking up in 2010 I started by looking at the cash and cash equivalents for these companies. Combined they have $810.57 million. That is a little shy of the $842 million that Google's peak quarterly Capex was in 1Q 2008. Out of curiosity I also analyzed the cash equivalents of the big guys - looking at Intel, IBM, Cisco, HP, Dell, and EMC. Their combined cash of $50.01 billion is slightly more than the 2008 Gross Domestic Product for the country of Lithuania.

Key figures
The average price to earnings ratio for my colocation group is 24.1, with a few troublesome negative p/e's for Savvis (-60) and Terremark (-16). Another interesting one to look at was the debt-to-equity ratio. For the most part, the group stayed under 2.0, but Terremark's 6.18 and 3.61 for Savvis is an indicator of how they have been financing their growth. In the case of Terremark however, building such intense data centers like they do requires some serious capital outlay up front.

An interesting stock comparison number to look at is the revenue per employee. Digital Realty and Dupont Fabros aren't really in the same category as the rest in this group, and their revenue per employee figures reflected as much ($2.9 million and $2.6 million respectively). Otherwise Equinix and Internap were the highest with $745,000 and $597,500 respectively.

I ran across a few 'insights' from the industry on Equinix and Internap. The Motley Fool website included Equinix in their '5 Deathbed stocks?' article, talking about companies where revenues dry up, margins contract and profit evaporates. Apparently Equinix has an Altman Z-Score of 1.7. This score is used to predict the probability that a firm will go into bankruptcy within two years, utilizing multiple income and balance sheet values to measure the financial health of a company. I'm sure numbers don't lie, but I think I believe this one about as much as I believe in Jim Cramer's knowledge of the industry (in other words, I don't!). An article at Smartrend notes the uptrend for Internap by showing that they are currently above their 50 day moving average of $4.22 and above their 200 day average of $3.36.

Finally, I looked at Morningstar to review financial statistics on these companies. The morningstar 'grades' were interesting, as they analyzed growth, profitability and financial health of the company. Equinix received a B for both growth and financial health, but a F for profitability. Rackspace received the best grades with a C for growth and a B for both profitability and financial health.

If anyone is interested -- here is the Google spreadsheet where I tracked the data.

Overall I think the industry is doing well and set for a positive 2010. The toughest part may be just defining 'what' the industry really is. Many if not all of the financial sites I visited had not attempted to categorize the industry that these companies are in, or put them in odd categories with peers that really weren't industry equivalents at all. I think some of the smaller data center companies may be acquired and some interesting things will certainly happen in the related telecommunications sector.


chuck goolsbee said...

It is hard to use publicly traded companies to gauge the health of the colocation industry as a whole because very few "pure colo" companies are public. Of the companies you list, only Equinix can boast colocation as it's primary business. The others are either Network Operators (who do colo on the side) or Real Estate companies whose customers are partially made up of colocation companies. Finally, Rackspace is really a dedicated server (or maybe classified as "managed hosting"?) provider. Try to buy colocation from them… they can't/won't sell it to you. The vast majority of colocation providers have never gone public. This is because of the 2001 downturn and the very slow recovery from it. Had that event not happened I imagine there would be a lot of colo stocks to gauge the health of the sector from.

Given that the 1999/2000 facility-building boom left us with a huge surplus of empty datacenter/colocation space, and that space has finally been all used up and new facilities are once again being built, I'd say the colocation industry is doing pretty well. If there were more colo companies being traded publicly then I'd imagine the sector would be viewed as a leading indicator of economic recovery. As it stands right now though it is largely hidden from the view of Wall Street analysts, hence the ridiculous statements we've heard from some of them.

Dave said...

I would tend to agree with Chuck's comments here.

Some of the organizations you've list, I would either consider Real Estate companies as opposed to Colocation providers. In the case of Internap, they are more of a network and route optimization technologies provider.

Granted, you can get colocation service from most of them, but I wouldn't condier it their core business.

In the above case, are you realy analizing the health of the colocation industry?

Unknown said...

Both Chuck and Dave make good comments to qualify the pool of companies you are comparing. DRT and DF are certainly direct comparisons, arguably they should be compared to other REIT's be they office/industrial, retail etc. I beleive Terremark might be considered the hybrid, with substantial investment in traditional bricks and mortar real estate, as well as colo operations infrastructure. As Chuck pointed out the 99/00 building boom left a surplus of empty data center/colo space, but that should be qualified that it left not just empty but "un-built or first generation" space (infrastructure needs). The mix of real estate and operating companies in the pool is somewhat a reflection of how that infrastructure is financed now versus the 99/00 triple net lease approach by both landlords and colo operators.

John said...

All excellent points - I think I'll rename the list "John's misc. companies to watch" :)
I think the stock trackers need some help also though -- here is how they categorize and list competitors for Equinix:

Morningstar - telecom Services industry
AT&T, Qwest, Level 3, Colt Telecom Group, Savvis

Google Finance - Communications Services
Switch & Data, Terremark, Savvis, Level 3, Verizon, Internap, AT&T, Qwest, Dupont Fabros, Rackspace

Yahoo Finance - Telecommunications Services - Domestic
AT&T, Verizon, BCE Inc., Centurytel, Qwest

CNN Money - Computer Services
IBM, Wipro, Infosys Technologies

Bloomberg - Web Hosting/Design - Internet Services
AsiaINFO holdings, Switch and Data, J Global Comm

chuck goolsbee said...

Based on those listings John, I suspect that Wall Street is essentially ignorant of the colocation market as an independent sector.

Corelink Data Centers: Colocation said...

Thanks for the interesting and informative post. I look forward to more in the future.

Paolo Gorgo said...

Great post.

I agree with some other comments: DLR and DFT are a different category (REIT).

Equinix is the only pure colo player of the list. The others have a more complicated business model.

Having checked this metric (among many other) for quite a long time, I think I can say that EQIX has always achieved a very good number, much better, for example, than SDXC, the only real listed pure player (almost 50% better, by heart). However, a good reason is that Switch and Data has a network of smaller data centers, and scale is in favour of Equinix.

With the merger, EQIX will be "hit" (in the short term), if you just look at this metric.

A while ago, Ike Elliot made an interesting comparison in the Telecom sector:

As to your Equinix number, I make a slightly different calculation (4 times latest quarter), and usually take it as an indication rather than an absolute number - given the fact that EQIX keeps expanding, they hire people (for the new data centers) well before the revenues are generated. A good reason why this number will look better once they stabilize (not in the short term, I hope...).

Internap is an example of the opposite: stable revenues, and they just made a 10% haircut to work force. The number sounds much better than a while ago, but it's just one of many indicators of a Company's performance.

I also agree Wall Street has a few problems understanding the sector...