Its 2001, I am sitting in my office in Leeds, UK wondering if this was the biggest scam in the world. Spectrum or wireless air waves were being acquired for obscene amounts of money and to the layman it would seem as if the industry had dished out way too much money for using “air”? Was that right? Why were telco’s bending over backwards to spend billions? What really was being unleashed here?
Lets start with the basics, what really is spectrum?
Air waves are the invisible pipes used to transmit information for wireless networks- be it Wi-Fi, Bluetooth, cellular or any other technology. A government appointed regulatory body defines the type of applications which can be used on the allocated spectrum in each country: TV, Radio, Mobile, DoD; each getting slots. With over 7 BN mobile devices (and only 4 BN tooth brushes) in the world and data usage going through the roof, the demand for the pipes is increasing exponentially. An alternative to having more of the spectrum is to put more electronic infrastructure, on tens of thousands of cell sites- which is certainly a lot more cost prohibitive.
What is licensed vs unlicensed spectrum?
Licensed spectrum is the use of radio spectrum designated by the governing regulatory body reserved for organizations that have been granted licenses. Though licensed spectrum is regulated and costly-it keeps out electronic noise/rift raft and allows the licence holder to cut down on potential conflicts and source of interference. Unlicensed spectrum is free and open to everyone, as long as the follow some basic ground rules. It’s cheaper, more accessible and generally has a much lower Quality of Service.
Is Radio Spectrum a scare resource?
Theoretically, spectrum is not scarce- there is an abundance of it. However, useful spectrum is ‘scarce’.
· Radio waves have different properties and not all waves are suitable for communications.
· Lower frequency bands are the “beach-front property”, as they have better penetration (buildings, walls etc)and wider range ie lesser cellsites needed to be deployed to cover rural population.
· Availability of technology for different applications.
How much spectrum do we need?
Data traffic is going through the roof. A decade back, an average customer used 30–40Meg/month today its nearly 100x of it, and the tsunami of bandwidth hogging applications is going to accentuate the problem over time. Case in point being AT&T, which saw a 5,000% increase in its mobile data traffic in the past three years; forcing them to invest billions of dollars to augment capacity.
How much have operators paid for Licensed Spectrum?
An absolutely mind boggling number! The reason is the economic value and return it is creating. Now that we have debunked that spectrum valuation is a scam, lets look at some investment dollars in this space:
- At the height of the dotcom boom in 2001, the UK telecom operators dished out £22.5BN to the government.
- To cite more recent auctions, the Indian govt. raised $23.5BN for 3G and broadband spectrum in 2010
- Not too long ago (2006); US telecom operators paid over $15BN for the AWS airwaves
The unit used to compare spectrum valuations is MHz-POP; this indicates how much it cost per megahertz to cover a person. The following table compares and contrasts some of the historical spectrum auctions in North America:
So how do we actually put an economic value to spectrum?
Well, two approaches used generally:
- “The Comparables approach”: In simple terms look for a precedence ie how much have others paid for it in comparable markets
- “Income and Cost Reduction Approach” — Cost avoided in terms of rolling out additional infrastructure (small cells, sectorization, cell splitting etc), which could be 10–100x of the cost of spectrum in certain geographies. At the end of the day investing top dollars on spectrum is a strategic decision to limit capital outflows and also to keep competitors at bay from key markets. That’s one reason why spectrum prices deliver unexpected bonanzas to governments, going upwards of 10 times the amount predicted by analysts.
Apart from pure financial valuation models there’s a lot more that goes into defining the intrinsic value of the spectrum; namely Regulatory environment, Number of bidders, Current Spectrum Holdings & future demand, Potential of Future Availability of spectrum, Spectrum band maturity, geographic size of licensing area etc.
Having built spectrum valuation models for PE firms, global and local operators- I can vouch for it being one of the most intriguing topics to have previously worked on. It’s a lot of science and more so art in double guessing competitors & applying game theories.