One of my readers sent me an article that announced a fiber-to-the-home expansion in Nepal. The ISP for the country is Vianet Communications P. Ltd, which uses Nokia GPON technology. The company built the first FTTH in the country in 2011 and already serves 10 of the 75 districts. The current expansion will bring fiber to an additional 4 districts, with another district already scheduled after that. Vianet is a commercial ISP and began as a dial-up ISP in the capital of Kathmandu is now expanding across the country with fiber. By the end of the year the ISP will have 200,000 customers on fiber, providing a minimum customer speed of 100 Mbps.

You don’t have to look hard to see similar stories around the world. Romania now has the fastest broadband in Europe and is ranked as having the sixth fastest overall broadband in the world. Romania’s broadband success story is unique since the fiber networks have largely been built by small neighborhood ISPs that have strung up fiber. There are local news articles that joke about the country having fiber-to-the-tree. The country had almost no telecom infrastructure at the end of the cold war and local entrepreneurs and neighborhood groups have tackled the task of bringing the needed fiber infrastructure.

I’ve often heard it said that one of the reasons the rest of the world has more fiber than us is because the governments in those countries build the infrastructure. However, when you look closer at a lot of countries like Nepal and Romania, it’s commercial ISPs that are building fiber, not the government. Singapore has had the fastest broadband in the world for years and their fiber was built by three ISPs. There are similar stories everywhere you look.

If ISPs are able to build fiber in Nepal and Romania, why are they having such a hard time doing so here? There are a few key reasons.

Big ISPs in the US are driven by quarterly earnings expected by Wall Street. They get crucified for not maximizing profits and none of them can undertake any major expansion that would earn infrastructure returns of 7% – 12%. It doesn’t matter that the ISP business is a cash cow and spins off piles of cash once the business is mature – the big ISPs are structured such that they really can’t consider undertaking building fiber to residents.

Years ago Verizon took a hit for tackling FiOS, and even then the company was very disciplined and only built where construction costs were low. People are praising AT&T currently for passing over 10 million homes and businesses with fiber – but their network is the very definition of cherry picking where they serve a few homes here and few homes there, nearby to their existing fiber nodes.

There are plenty of smaller US ISPs that would love to build more fiber, but they have a hard time raising the money. Fifty years ago banks were the primary source of infrastructure lending, but over time for various reasons they no longer want to make the long term loans necessary to support a fiber network.  The big banks are also Wall Street driven, and banks make a significantly higher return on equity by churning shorter-term notes compared to tying up money for 20 – 30 years.

One only has to visit a FISPA convention, the association for fiber overbuilders, to find numerous companies that would gladly tackle more fiber projects if they could borrow the money. Just about every member of FISPA will tell you that borrowing money is their biggest challenge.

The countries building fiber have found ways to overcome these issues. The ISPs there are able to borrow money to expand fiber networks. Their banks love the guaranteed long-term steady returns from broadband. The countries I’ve mentioned have one natural advantages over many parts of the US since they have a higher population density. Nepal has 29 million people and is about the same size as Michigan. Romania is a little smaller than Oregon with a population of 19 million. However, they have other challenges. As you can see from the map accompanying this blog, Nepal has some of the most challenging topography in the world. Both countries are far poorer than the US and yet they are finding ways to get fiber built – because like everywhere, there is a big demand for broadband.

I’ve said many times in this blog that we need government help to build fiber in the rural parts of the country. That’s due simply due to the cost of a fiber network calculated per household, and the numbers don’t work in most rural places. However, I’ve created hundreds of fiber business plans and it generally looks feasible to build fiber in most other places in the country, and yet there is no flood of ISPs building fiber in our towns, cities and suburbs. Detractors of municipal fiber always say that our broadband problems ought to be solved by the private sector – but I look around, and in 95% of America the private sector hasn’t showed up.