Northern Sky Research

Capping One’s Opex-tations

Oct 27th, 2016 by Jose Del Rosario   More from this Analyst | Profile

Redirected to subject matter Research Director - original author no longer with NSR

The balancing act currently being undertaken by satellite operators and other industry players involves launching more capacity, entering new markets, migrating customers to different frequency bands, and in some instances doing all this while managing a pile of debt. While NSR has discussed previously the evolution of CAPEX and OPEX in this new market dynamic, our Satellite Operator Financial Analysis, 6th Edition covers this discussion in more depth.

The first major takeaway from this discussion is a trend apparent for some time now—GEO-HTS is, from a raw CAPEX per Mbps perspective, quite a bit cheaper than GEO-FSS. With CAPEX increasing by a factor of <2, operators have seen throughput increase by up to, or exceeding, 10 times. As the chart below shows, even using fairly conservative figures for capacity, GEO-HTS sees its CAPEX cost per Mbps falling by over 70% compared to GEO-FSS. While the capacity is sometimes not a “one-for-one” conversion—that is, the spot-beam nature of GEO-HTS Ka-band capacity, for instance, makes it more difficult to achieve a very high fill rate—it is nonetheless clear that there are, from a CAPEX perspective, some significant advantages in launching GEO-HTS as compared to GEO-FSS capacity, with satellites of many times the size being launched for 1-2 times the cost. However, when taking into account metrics such as EBITDA, the situation becomes somewhat murkier.


EBITDA in GEO-HTS: The Thaicom Example

The EBITDA environment for GEO-HTS is somewhat more difficult to quantify, given the lack of data, the lack of GEO-HTS systems that have truly become “established”, and the fact that EBITDA is oftentimes harder to pinpoint/assign to a particular function than CAPEX. With that being said, there does exist some data that, when coupled with assumptions, can provide insights as to the EBITDA per Mbps equation in GEO-HTS as compared to GEO-FSS. Having launched IPSTAR in 2005, Thaicom is in possession of what many would consider an “advanced” GEO-HTS business model, in the sense that it has now had a decade to become established and build a customer base. Likewise, Thaicom does publically report figures such as top-line satellite revenues/EBITDA, IPSTAR revenues, and conventional satellite revenues. When taking the above data and making assumptions for EBITDA margin for Thaicom’s raw satellite leasing arm (conventional satellite), we can make assumptions for EBITDA, and subsequent EBITDA contribution per Mbps leased for both GEO-HTS and GEO-FSS satellite capacity.

*NSR Estimate                                                                                                                                   Source NSR

When looking at the numbers in this way, we see that the average leased Mbps of traditional FSS capacity in 2015 contributed around $21,600 in EBITDA over the course of a year. Conversely, when looking at capacity leased on IPSTAR, NSR sees an EBITDA per Mbps of around $2,200 for 2015, a figure that was down somewhat from 2014 due to Thaicom having increased IPSTAR capacity leased during that time by ~20%. When looking at these two figures, one might be surprised at the stark differences—i.e. FSS being higher by a factor of 10—but it is important to consider that this includes Thaicom’s video contracts (which would be generally witness a higher per Mbps/MHz price), and that in general, applications on GEO-HTS will tend to see higher bits per hertz conversion ratios, which means that in per MHz terms, the difference is somewhat less.

What other meaning can we take from these EBITDA figures? First, as NSR has noted before, it is clear that selling data services in the Gbps will not be a 75% EBITDA margin business. While operators have for some time been able to sell capacity at a very high EBITDA margin due in part to CAPEX/D&A being such a large percentage of the overall cost structure, this is changing very quickly. One need only look at Thaicom’s reported EBITDA margin above, which dipped as low as 33.4% in 2010, to understand how OPEX-intensive the efforts can be to set up a distribution network for a newly launched GEO-HTS.


Bottom Line

The age of GEO-HTS will certainly cause significant disruptions to the way that satellite operator business models function. With more capacity being launched in ways that prove more cumbersome to optimize, operators will need to expect to see a spike in OPEX coming at a time when CAPEX relief from GEO-HTS is forcing a rebalance of spend. As we have seen with programs such as IPSTAR, when an operator pivots to put one foot into the telco industry, they can expect significant upfront OPEX just as a “buy-in”, another challenge to add to the myriad variables to be taken into account when entering the brave new world of HTS.