It is possible to "bootstrap" a business and you will be able to eventually grow based solely on the cash you started out with and the money you make. Sometimes, this is a great way to run things.

However, if you want to build something revolutionary and do it fast, it's better to start borrowing money. It turns out that controlled borrowing is a net-win for everybody, allowing a pool of money to do more work than it would otherwise.

The way financing works is that you make an educated guess about how likely somebody is to pay you back and make sure that on the average you are right. So say I have a hundred bucks and I give ten people ten bucks and ask for twelve bucks back the next year. Maybe all of them will give me the money back. It's a good year. Maybe one or two of them won't be able to pay me back, so it won't be as good of a year. Maybe one year three or four of them will go under, which means I'm going to be much more careful next year.

Now, my options when you can't pay me back also influence things. If you can walk away without any penalty from not paying me back, you are going to be a lot more likely to forget to pay me back, so I'm going to be very careful with who gets my money and I'm going to charge more interest. If you are borrowing the money to get a shovel and I can take the shovel back, then I'll be a little less careful because I can always sell the shovel and get some of my money back.

This part is the same if you are a person or a business. However, eventually you get bigger than the smallest bank. Or maybe even bigger than a big bank. And there's always a limit to how much any single bank will let you get out of them. Eventually, you need to create a way to finance massive amounts of money but in such a way that you aren't limited to talking to a few huge banks. Enter securities.

So, what happens is that if a company needs a truly huge amount of money, they break the huge amount into reasonable sized chunks and sell the little chunks separately. This worked out quite well, so the financial folks have been cooking up new and different ways to bundle and "securitize" stuff.

The traditional thing you'd think of is a "Bond". There are a few different categories, depending on who issues them (companies, towns, states, governments, etc), when you get interest (coupon bonds means that you get your initial investment back when the bond comes due and you get periodic payments the rest of the time), and if the bond can be paid off early (Callable).

Just like you have a credit report, so do companies. So depending on the rating of the bond, the interest rate changes. And you can actually sell these things at any point in time and there's some fairly complicated equations used to model how much value your bond has at any given point in time.

Now, most bonds are designed to last for years and used to fund huge things like new factories. It's pretty much like a loan, just done up so that a bunch of people can participate.

But companies also need other sorts of money to fund more immediate things. For example, if I know that another big company is going to pay me in 60 days, I can write a short term security called "paper" on a market called the "money market" and see the money now. This means that I get to shorten my cash cycle and then another person gets to take a little bit of that money at very low risk.

The most important part of this market is that all of the participants in it need to be fairly large and the "paper" needs to be have very little risk.

Now, what happens if you can't pay up on a piece of paper or a loan? Certain very large companies, governments, and municipalities can put up only their reputation as a guarantee. Everybody else has to put up collateral. So if I'm selling bonds for a factory, if I can't pay the bondholders back, they can come and take the factory.

Now, both "paper" and "bonds" are standardized things. You can't sell bonds with weird conditions associated to them because that means that somebody has to read the contracts and maybe adjust their mathematical models. It's like making screws that turn the opposite direction... people just won't accept it.

After a while, folks discovered that this works in reverse. If you can take a million bucks and break it into lots of five thousand, why not take a bunch of little five thousand bucks chunks and turn them into a million buck blob? You can then break it down again into your five thousand buck lots, but this way, if one of those five thousand buck chunks you bundled up goes bad, everybody's out a little bit instead of one person being screwed over. Which is really the subject for another article...

Anyway, the important lesson to remember that's applicable for the economic crisis right now is that everything depends on you getting your estimates right. If you misjudge the likelihood of any of these things to default, the consequences are huge.