All the reliable data I have seen shows the opposite.

Note, returns very a lot by company the industry average is ~6%, but that includes a lot of poorly run lemons.

Also, there was a huge bubble that put in a lot of dark fiber, but that glut has passed.

PS: Yes, those returns look low compared to say MS. But, software company's have larger issues borrowing money at very low rates.

The Chattanooga network is illustrative: http://www.nyls.edu/advanced-communications-law-and-policy-i....

Revenues are about $100 million, with operating expenses being $85 million. That's about $15 million profit, on about $400 million invested, or under 4% ROIC. The system was funded with bonds issued at 4.5%, meaning it would never break even without being subsidized by electric ratepayers.

As to Chattanooga network, economy's of scale are huge in this business and those 4.5% bonds are well above the industry average right now. Still, a city breaking even is not bad if their debt is growing slower than inflation. It's a question of deprecation of assets not rate of loan repayment.

PS: They city may also not be trying to make much in the way of direct profit, it would be interesting to take a closer look at their books.

Your is assertions are incorrect. EPB Fiber is profitable, is servicing it's debt in full, pays less than 4.5% in interest in their bonds, is not subsidized by electric ratepayers and would have broke even even without the grants it has received.

Not only are electric rate payers not subsidizing the fiber network, the fiber network is subsidizing the electric side!

Looking at the finances, about $400M was invested, but EPB fiber only has ~$260M in debt left due to a $111M grant for smart meters. Furthermore the $15M profit is after debt service. EPB paid $19M in debt service (FY14).

Source: https://static.epb.com/annual-reports/2015/wp-content/upload...