Last week, Twitter raised $1.8 billion in a convertible bond offering. It originally planned to raise $1.3 billion; the offering was made up of five-year notes, which received a 0.25 percent interest rate, and seven-year notes, which received a 1 percent interest rate. Those are very, very low interest rates, typically indicating a very low risk.

At the same time, the ratings agency Standard & Poor's released an unsolicited rating of Twitter's debt, and rated it as junk, pointing out that Twitter is a new, unprofitable company in an untested market.

What explains the dissonance?

First, there's simple publicity. Ratings firms have to make a living like everyone else, and they crave PR like everyone else. Many of the ratings agencies' sovereign bond ratings are little more than public relations exercises — while they might help a small developing country to raise money, the U.S., Germany, Britain, and Japan don't need them. Ratings agencies provide ratings because it gets them press coverage. Here, the same factor is at play. Twitter didn't ask S&P to rate its debt, S&P just decided to publish a report. Why? "S&P Rates Twitter Debt Junk" — the press release, and the blog headlines, write themselves.

Second, there are macroeconomic factors. Low central bank interest rates — a response to continued sluggishness in the global economy — are only part of the story. The bigger story is that we live on a very rich planet (lots of capital) with subpar economic growth (not enough places to invest all of that capital). This means that whenever there's an opportunity, capital will chase it for all it is worth. Hence, companies like Twitter can sell bonds at very low interest rates. Does that mean we're in a bubble? Not really. There are still very few technology IPOs, and very few of the companies getting funded are silly. But we do live in an easy money world, and that redounds to Twitter's benefit.

Third, and more important, there is the bizarre disconnect between perceptions of the internet business and the reality of the internet business. This is not about "This time it's different" or "trading bricks for clicks" or any other tired cliché from the dot-com era. It's about the fact that there are now 2 billion people online, which means we can see the true dynamics of internet business play out.

And the key to understanding why so many people are bullish on Twitter, despite its being young, hyped up, and unprofitable, is the following concept: network effects.

A network effect is what happens when a product or service becomes more valuable as more people use it. The classic example is the telephone. If no one has a telephone, it makes no sense to own a telephone, since you can't call anyone and no one can call you. On the flip side, if everyone has a telephone, then it becomes less feasible not to own a telephone. The more people there are on the telephone network, the more valuable the network, and the more valuable a telephone becomes for the next marginal consumer to get one. (We can see this 19th-century trend playing out right now with mobile in Africa.)

Businesses with network effects are very different from other kinds of businesses, and it just happens that the internet — being itself a network — lends itself very well to creating network effects. Setting up a network is typically very expensive and hard, since the value of the product will be very low for the first users. But once the network reaches a critical mass of users, it typically snowballs and grows very quickly because of a virtuous cycle: Now that the network is valuable, it attracts new users, which makes it more valuable, which attracts new users, and so on.

This means that over their lifetime, businesses reliant on network effects can be very profitable. Once you have spent the fixed costs of building up the network, the variable costs of maintaining it are typically low. Furthermore, these businesses are often natural monopolies, since usually there will be only one network of a particular type (Facebook is here, but no one remembers MySpace). That means the network can charge a lot of money for ads and other services.

To return to Twitter. Is it unprofitable? Yes. But how many users does it have? North of 200 million, and growing. Think that's a lot? The global internet population is 2 billion; pretty soon the global smartphone-owning population will be 4 billion. So Twitter is still in the early stages. And does Twitter have network effects? Yes it does, like all social networks. The more people are on Twitter, the more sense it makes to be on Twitter. Since Twitter has network effects, everything suggests that it should be unprofitable now that it's still growing, but that over the long run it will be very big and very profitable.

In other words, economic logic reaches the same conclusion as the market: it's a very good bet that Twitter will be able to pay back its debt.