The Earth's moon is one of the largest natural resources of our planet. My question: how much would it cost to build an internationally accessible commercial facility (moonbase) on the Moon?
Since we don't have any moonbases yet, this is going to be difficult to answer. So the following will guesswork with a few bits of data in between.
The only large-scale construction we've built in space is the ISS. Building the ISS cost on the order of $100 billion. The ISS can house about 10 people, which is reasonable for a scientific outpost, but really small for a commercial venture.
If we assume mining is the goal of this venture: a large gold mine on Earth employs ~500 people, and uses thousands of tons of equipment to move 15 million tons of rock. It produces 28 tons of gold/year, sold for $420 million/year.
Transporting the moonbase and industrial equipment to the Moon isn't going to be easy. A Saturn V could only transport about 15 tons to the Moon's surface (and about 5 tons of that would take off again).
Today, going to the Moon is just about feasible for scientific missions. Thousand-ton commercial ventures are way out of reach. Even the most precious materials mined on the Moon would be orders of magnitude more expensive than their counterparts mined on Earth.
Projects like the SpaceX ITS aim to change that, with a projected per-launch cost of $60 million for 300 tons to LEO. That's 2 times the payload of the Saturn V, so it might be able to land a single ISS module (weight around 20 tons for the large modules) in one piece. Let's try transporting 1000 tons of equipment to the Moon this way. That's 50 launches, or 3 billion in launch cost. Just transporting the materials to the Moon costs more than you can possibly make mining expensive stuff like gold.
The dream of sending humans back to the moon could become a reality for the relatively low price of \$10 billion — only about 10 percent of one previous cost estimate — as long as NASA is willing to buy from private spaceflight companies, a new report shows.
In 2005, NASA estimated that returning humans to the moon would cost 100 billion Dollars (approximately \$122 billion in today's dollars). But if the success of private spaceflight companies like SpaceX and Orbital Sciences continues, NASA could send humans back to the lunar surface in as little as five to seven years, at a highly reduced cost, the new report shows.
That's not all: 10 to 12 years after that first commercial moon trip, NASA could develop a permanent base on the moon for about \$40 billion in today's dollars, the report said. The proposed permanent moon base would be used to convert lunar ice into hydrogen propellant that could be sold for use by other spacecraft, including missions headed to Mars.
Here is an additional link that seems to support the findings of the above or at the very least be inline with the above (as provided by call2voyage's comment to the OP.)
As much as I would like to offer you a finite answer to cost, you're going to need to accept a bracket indicating the startup cost. \$10 Billion USD – \$40 Billion USD. Once the startup phase is concluded, the startup cost is over and the project is profitable from that point on.
The cost will be the cost of shipping people there and a startup habitat and supplies.
After that everything will be built from lunar resources by people living on the Moon.
Delivery to Earth is a piece of cake. Luna colony builds transfer vehicles from lunar resources and fuels them from lunar resources. Transfer to low Earth orbit from low lunar orbit (large structures) can be done by nuclear engines (see Project Rover, NERVA), or laser powered, solar wind powered, all of which are magnitudes cheaper than any Earth OR low Earth orbit launch cost.
Profitability (measured sans startup cost) is instant once production of space assets, fuels and transfer vehicles are started. Earth would, given today's propulsion systems, be incapable of competing in any aspect.
The basic equation is 1/6th G (the Moon) which is infinitely greater gravity than the microgravity of the ISS vs 1G (the Earth). Mining … that is not what you asked but it also will be magnitudes cheaper and … shall we say be a Lunar Colony decision. With the added bonus of nuclear engines to capture and return near Earth orbit objects for processing or processed in place and the processed products are shipped where needed.
The greatest problem with space is shedding being Humans: A Planetary Species to become Humans: A Space-faring Species, as in a fundamental change in thinking and how we solve problems. We start by discarding the perception that everything is a problem needing an Earth Hammer solution.
Habitats: lunar lava tubes are our Habitats, they are already made. We need to seal them, pressurize them and make them habitable.