Sunday, August 5, 2012

Comparing Cash Flows at Equinix and Rackspace

Superficially, Equinix $EQIX and Rackspace $RAX have a lot in common. Both are high growth plays in the cloud computing sector. They also seem to be in similar stages of development in their industry life cycle. In other words, they are generating high revenue growth, but little cash flow. However, I would argue that they have different business models and, going forward, we could see divergence in some key metrics for the relative businesses.

In summary, Equinix builds and owns state of the art data centers which provide collocation and infrastructural services to public and private clouds. It is the ultimate ‘picks and shovels’ play on cloud computing. I think its current cash flow situation is masking very strong underlying cash flow and, providing the industry doesn’t create overcapacity, investors can look forward to significant improvements in the future. Equinix’s capital expenditures are in order to build data centers which typically take around four years to reach peak demand.

Rackspace mainly provides managed hosting services although, it also provides data center services as part of the mix. The key difference is that Rackspace offers a high level of ongoing support and, its business model involves significant capital outlays on customer gear in order to support top line growth. I don’t think its cash flow profile is as attractive and investors need to keep an eye on how this develops.

How Rackspace Works

Essentially, Rackspace allows its customers to outsource their hosting IT requirements. This is wonderful for clients because, it allows them to focus time and resources on their core business. It also lets them shift fixed long term costs of their own IT expenditure, into the variable cost of outsourcing to Rackspace.

Naturally, this is giving Rackspace very high growth rates but at what cost? It is all very well for Rackspace to generate growth, but there are some indications that it is being paid for by increased capital expenditures. If so, then the investment thesis for Rackspace is not a strong one in the long term. Moreover, the company’s high level of service -the so called ‘fanatical support' engendered by its ‘rackers’- demands management focus and resources.

Here is a table that shows how capital expenditures are evolving over the years. The numbers for 2012 are based on estimates.

Note how the sharp increases in revenues have not caused the proportion spent on capital expenditures to tail off. In fact, it appears to be increasing in recent years. This implies that Rackspace is continually spending on capital expenditures in order to service growth. In addition, whilst EBIT margins are going up (hitting 12.3% in the last quarter) this has not causing operating cash flow margins to deviate much from the 31-33% they have been in for the last few years. In other words, free cash flow generation appears to be declining. This is a worrying sign.

Supporters of the company will point out that, technological advancements, particularly in servers, will see costs and expenditures decrease proportionally in future as Rackspace generates more bang for its buck. I’m sympathetic, but I prefer to see it happening now, rather than rely on an abstract argument.

Similarly, the bulls would argue that the capital expenditure requirements are creating significant capacity, which will be used up in years to come. However, if this was the case, then surely we would have already seen capex/revenue dropping by now? In addition, Rackspace i clear on the issue of its capital expenditure requirements being necessary to support current growth.
'As in prior periods, we expect the bulk of capital expenditures to be on customer gear, which is predominantly success-based spending to support revenue growth.'
I’m not convinced by the case for Rackspace.  But investors who are less skeptical, might be inclined to watch this metric carefully.

A Similar Story with Equinix?

I think Equinix’s underlying picture is actually better. Despite the negative free cash flow, Equinix genuinely is investing in long term capacity. The key consideration for this company is not so much the business model, but rather the timing of industry wide capacity expansion.

Equinix provides data centers, and the big plus is that the growth in internet traffic is broad based across sectors and has proved to be recession resistant. This looks like a structural growth story and will only increase as cloud computing and data traffic increases. Whilst, superficially, there is no moat in data center provision, it is, in fact, a mission critical application which requires substantial planning and trust on behalf of the clients.

The potential downside here is that this ramping of capacity (the likes of Equinix, Telecity and Interxion are expanding aggressively) will cause a supply glut which will lead to falling margins just as the data centers need cash flow in order to pay back the debt needed to expand capacity.
No matter, right now, this doesn’t look like being a problem and investors will see the early signs when Equinix, Telecity, Interxion $INXN or the Digital Realty Trust $DLR start reporting slipping EBITDA margins. The growth in internet traffic is progressing at an exponential rate and, increases in cloud computing will only strengthen this growth.

A quick look at gross margins across the industry reveals that margins are holding up well, so there is no sign yet, that there is overcapacity in the industry.

Data center customers tend to be very ‘sticky’ and it is not uncommon to see 95% of the previous year’s revenues reoccur in the next year. Therefore, companies like Equinix can generate significant organic growth simply by servicing its existing customers’ expanding data center needs.

What Investors Need to Know About Equinix

The key thing with Equinix is to delineate its maintenance and expansionary capital expenditures. For example, in 2012, Equinix is forecasting revenues greater than $1,890m and operating cash flow of around $655m. These numbers tally well with historical cash flow conversion which is around 35% of revenues. As for capital expenditures, the forecast is for a total of $700-800m with $135m coming from maintenance and $565-665m from expansionary capital expenditures. In other words, the discretionary free cash flow forecast is for $520m or 5.4% of the current Enterprise Value.

Whilst this evaluation is very attractive for a stock with this kind of growth rate, investors will need to keep an eye on utilization and, as discussed earlier, gross margins. For now, all the key metrics are indicating that there isn’t overcapacity in the industry so investors have a right to be positive. Equinix’s current churn rate is a very low 2.4%, gross margins are rising and, despite the strong expansion, utilization rates are rising sequentially across all three geographic regions.
At some point, these metrics may well turn, but for now market conditions remain bright for Equinix.