Setting a Value on the US 2.3 GHz WCS Band

When the WCS band was first auctioned off in 1997, 30 MHz of spectrum in 2.3 GHz was sold for $14.8 million. At the time, it seems reasonable to pay that much for spectrum with stringent regulatory requirements on RF out of band emissions to protect the adjacent satellite DARS band (Digital Audio Radio Service). Now, the picture is completely different: AT&T acquisition of Nextwave’s WCS and some AWS assets for $650 sets a new price for the WCS band. It also clearly demonstrates how in spectrum limited market with high demand for mobile broadband services, the price for higher frequencies can increase significantly. So what would the price of the WCS band be and are there any lessons to draw from it?

To start, let’s take out the AWS spectrum out of the total price. Nextwave bought 154 AWS licenses for 115.5m in 2006 ($0.13/MHz-PoP), then sold about 67% of it (92 of 154 licenses; 599 MHz-PoP) to T-Mobile and MetroPCS for $150.1m. This leaves Nextwave’s remaining AWS spectrum valued at around $74m. Therefore, Nextwave’s WCS spectrum would be around $576m (based on $650m total value of AT&T’s acquisition).

My estimate is that Nextwave’s WCS assets are about 29% of the total MHz-PoP, with many licenses in the A, C, and D blocks in many of the prime markets. This places a value of at least $1.96 billion on the 30 MHz WCS band, or $0.21 per MHz-PoP. This is 133x the price paid in 1997! Not a bad return on investment.

It was also reported that AT&T filed with the FCC to acquire Horizon Wi-Com’s and Comcast’s WCS assets which in my opinion are very valuable as they include the A and B blocks, respectively, in many prime markets. I estimate Comcast’s holdings at 11% and Horizon’s holding at 8% of the total WCS band (in MHz-PoP), putting a potential value of $210 and $151 million on Comcast’s and Horizon’s spectrum, respectively. In fact, I think these values may well be on the low side considering they are mainly A & B blocks and that they provide AT&T access to a large percentage of prime markets. Sprint’s WCS assets (7%) should be valued at least at $142m.

Comparing to other markets, the US WCS band is quite pricy. This is no surprise considering the high demand for mobile broadband services in the US which exceeds that of other countries. This is the prime driver for ballooning spectrum costs. For instance, Canada’s WCS band was auctioned off at $6.1m (CAD0.018/MHz-PoP). A few years ago there were no bidders in Hong Kong’s 2.3 GHz band auction (although a repeat earlier this year fetched $61m for 90 MHz).

When it comes to spectrum valuation, demand for broadband service is one driver of valuation. Competition, GDP, quality of spectrum and its properties and regulatory laws are some of the other drivers. But when we compare the WCS spectrum price to that of other markets and regions, I think that geography and population density is an important factor. There’s a heavy reliance in the US on building macro-cells which makes sense given the vast expanse and distribution of population in the US. Spectrum becomes important in this case as macro-cellular architectures offer lower spectral efficiency than an architecture that leverages indoor systems and small cells. In heavily dense urban markets particularly in Asia such as in Hong Kong, Singapore and Tokyo, there’s a more entrenched tradition of using smaller cells in addition to extending coverage into buildings, subways, and other such areas with distributed antenna systems. US spectrum valuation continues to be comparatively high for many of the factors above. The factor of geography and population density should not be ignored.