Cloud Computing Today and Tomorrow

Cloud computing is a new buzzword (or is it buzzwords?). Everybody is talking about it and most people don't understand it. I will try to review current state of cloud computing from investment perspective and possible future developments in the area.

IT (information technology for those who spend last 20 years in a cave) expenses are huge in most of the companies. At least that's what any manager (with some exceptions among IT managers) will tell you. To reduce those expenses, companies adopted various business practices: cutting number of personal (usually with disastrous consequences later), appointing MBAs as IT department managers (easier to kill him/her than to explain what you are doing), buying software packages instead of in house development (not a bad idea) and, at last, outsourcing (with the whole specter of results, from awful to great). In most cases, IT costs ignored all heroic efforts of managers and continued to grow. They continued to grow for two reasons: IT does more work every year and most of efforts of managers are counterproductive, they really increase costs instead of cutting them.

Looks like managers see cloud computing as a new, great way to cut IT costs. They are both right and wrong. They are right, implemented properly, cloud computing can cut costs and/or increase productivity. They are wrong, nothing comes for free, and right implementation still costs a lot of time, efforts and money.

In the news, especially investment news, there are three different IT developments which are lumped together under name of "cloud computing". Below is a quick review of them.

Internal Cloud

Also called server farm, this is a new way of organizing computing infrastructure. Companies set up big server farms with thousands of individual servers. Servers belong to the company, although their management might be outsourced. The biggest plus of internal clouds is the fact that all data is kept on company's own hardware. Usual features of such farms: virtualization, automatic computer management, virtual networking. I am not going to explain all these terms, there is enough explanations in the Web. First server farms, as far as I know, appeared in Web oriented companies, such as Amazon.com (AMZN), Yahoo!(YHOO), Ebay (EBAY), Google (GOOG). But for investors, the most interesting companies are those which provide hardware and software solutions for internal clouds. The biggest of them are:
in hardware
Dell (DELL), HP (HPQ), IBM (IBM), Cisco (CSCO), EMC (EMC)
in software
Microsoft (MSFT), VmWare (VMW), Oracle (ORCL), which is also a hardware vendor after buying Sun Microsystems. There are also a lot of smaller players, but judging by latest earning report from Brocade (BRCD), competition is stiff and prospects are not certain.

External Cloud
If the idea of storing data on somebody else's hardware doesn't scare you, cloud itself can be outsourced. Currently, Amazon, Google, Microsoft, AT&T (T) and a lot of smaller companies provide this kind of services. I think that Amazon and Google have advantage here, both because they are better at managing relationships and have better hardware/software combinations. Microsoft's policy of using exclusively Windows operating system is a drag on performance, when Amazon and Google reliance on Linux is a plus. AT&T is at disadvantage here as well, because its problems with customer service are not restricted to mobile phones area. Of course, companies provide virtual machines to their customers, with operating system of customer's choice. But Linux is a better base for virtualization than Windows. Unfortunately for investors, it doesn't look like a significant piece of business for any of these companies or any other big companies which might get into it. Possible candidates are IBM, Ebay, Yahoo!, Dell, HP, Oracle. Of specialized companies, I only found Rackspace Hosting (RAX) and Enomaly, which is not public (yet?). I don't know if specialized companies have any chance inside of the herd of elephants, but Rackspace is on my watch list.

Software As Service

I don't really know why is it often called "cloud computing", it has nothing in common with the other two. These are suites of applications provided to business online, usually through web browser interface. True, companies providing applications might use internal or even external computer clouds, but business model is completely different. From my point of view, this is a very interesting development. There is only one problem for the companies here: data is kept on devices which belong to different company. But in this case companies don't need expensive IT departments to run the application. It's not a big help to big companies, which use hundreds of different application, including a lot of custom built. But for a small company, which needs less than a dozen application, this is a very interesting proposition. Current competitors in this area: Salesforce.com (CRM), Oracle ORCL), Rightnow Technologies (RNOW). Salesforce.com is a leader, and any independent company is a possible acquisition target for Oracle and SAP (SAP). There is a possibility that Microsoft might get into this business, using acquisitions or internal developments, but so far I don't see any indication.

Of above mentioned, software as a service is the most interesting investing area. I'm looking at Salesforce.com often, but stratospheric P/E scares me every time. I might be wrong and P/E might be justified. For internal clouds, software companies look like the best bet, VMW being the leader. I don't see any investing possibilities in the external clouds yet. I am long GOOG for different reasons and I think that AMZN is a great company, for other reasons as well.

Full disclosure: at the time of publication author had long positions in GOOG, VMW and BRCD and no positions in other stocks mentioned. Positions can change any time.