The gold of the Knowledge Economy is dark fiber. Those that have it have something of great value. Those that don't are in trouble. It is really a currency of sorts, but much better than gold. Gold's value fluctuates up and down, but fiber's value will continue to go up indefinitely. How? As the network electronics that light dark fiber improve, you can expect to be able to continually increase the capacity of dark fiber strands, creating value each time.
We are already well into a transition away from single wavelength fiber use, where you need two fibers to complete a circuit--one fiber transmits down to a home or business connection, and one fiber transmits back up in the other direction. You can already buy off the shelf equipment that puts two wavelengths on a single fiber, with each wavelength providing one half of the circuit; this technique is called wave division multiplexing. But DWDM (Dense Wave Division Multiplexing) equipment that puts dozens of wavelengths on a single fiber is already in use on long haul, high capacity fiber routes. And eventually, fiber to the home and business will also be DWDM-based, increasing the capacity of a single fiber by orders of magnitude--try that with gold.
As bandwidth hungry applications are beginning to choke current copper-based systems, fiber is finally coming into its own, and I have been getting calls lately asking about how to price dark fiber. The simple answer is, you don't. Dark, unlit fiber is a core asset of any network, and you absolutely don't want to sign long term leases handing over dark fiber on your network to some third party. Once you do, you are foregoing the revenue potential of that fiber permanently (or at least as long as the lease, which are often twenty years or more).
If you have dark fiber and are getting requests to buy or lease it, you probably don't want to do that. Instead, you want to sell bandwidth, because you can always "make" more bandwidth by upgrading your network electronics. Once you give the fiber away, you lose that capacity.