I need to expand here, on a subsidiary, but very important factor. That of fractional reserve banking (FRB). FRB is a financial process whereby if I go to a bank and borrow money from them I must pay it back with the monetary fruits of my own future productive activities. The money that the domestic bank lent to me, they also borrowed from a commercial bank. The commercial bank lent that money into existence. However, they only did so on the back of knowing that they can go to the central bank, after-the-fact and make good their lendings. Thus, ultimately, central banks/commercial lenders decide how much is lent into existence to you and me based on an assumption of how much productive activity will be happening in the future such that the lending can be repaid. Of course, the reality of how the credit trickles down through the system is far more tortuous and indirect than the schematic, above. but, in principle, this is how our FRB "money" supply grows.
A few points in response to the above para.
1. I'm not sure that there's really a conceptual distinction between a "bank" and a "commercial bank" -- they're all commercial. The important distinction is between commercial banks and Central Banks. Any bank that lends you money can go directly to their Central Bank to obtain funds needed to balance their books, it's a privilege that goes with having a banking licence. They'd usually prefer to raise the money on the commercial debt markets, but they can go to the Central Bank if necessary.
2. When a commercial bank seeks funding from a Central Bank,
it has to provide collateral so this ability of theirs is not, in theory, an unlimited credit-expansion machine. The Bank of England (BoE) traditionally demands collateral of the highest quality, i.e. government bonds (gilts). The Bank faced criticism for not relaxing this requirement when UK banks such as Northern Rock needed funding, but only had mortgage-backed junk to offer as collateral.
3. Other Central Banks, including the Fed, are less rigorous about the kind of collateral they will accept. The great thing about insisting on gilts is that banks can't manufacture them; there's a real cap on the amount of credit that commercial banks can obtain from the CB. As soon as the CB starts accepting other kinds of bonds, that limitation is removed, and the vulnerability of the process to feedback from asset prices back to debt-raising ability (i.e. bubbles) is obvious.
4. So how did the UK bubble happen, if the BoE wasn't handing out money willy-nilly? I suggest (based entirely on circumstantial evidence) that it was money being recycled (lent back to us) from countries with which we ran large, persistent trade deficits. So, UK commercial banks who made loans to us, were able to raise the funds from foreign commercial banks, who were glad to find a home for all the pounds they'd collected by running a trade surplus with us. Again, the dangers of feedback are obvious, since the more we borrowed, the more we spent; and the more we spent, the richer our (exporting) trading partners felt and the more money they had to lend back to us.
So ... I'm not persuaded that the debt-based banking system itself inevitably leads to disaster. All the matters I've discussed above are matters of policy, somewhere or other, and amenable to regulation (but not necessarily in a democracy).