Fractional reserve banking is the idea that you can loan more money than you have, because the obligations of the people with the loans to repay you are just as good as money.
This is one of those powerful indirect civilization ideas; it is especially powerful because the value of the repaid loan is coming from the future, and combination of due caution on the part of bankers and the assumption of good value means this is a way to make it much, much easier to innovate. If you can convince a banker that your idea will let you repay the loan, you can get the capital to make your idea real. (This has lots of practical drawbacks, starting with the scant knowledge of fields not their own typical of bankers, but it's ever so much better than the situation where, no matter how good the idea, it was at best a serious political problem to get access to capital, and generally your good idea was going to die with you.)
It's not an accident that this idea got into practise in the Renaissance, or that everyone who has real trouble with the basic truth that money is fundamentally socialist, an agreed accounting system for value that isn't directly related to anything real has some serious issues with the whole idea.
But, anyway; "risk" in this sense is a measure of how likely the value from the future is to show up, and interest is a way to distribute that risk—since some of that value isn't going to make it—over everyone involved.
Because that value from the future takes awhile to show up, the obvious trick is to convince a banker to accept something as valuable when it isn't; you then get real money in return for fake, since the money you get is right now, and convertible into stuff, and the discovery of the non-value happens later, after you've had a chance to abscond with your gains.
These days, there aren't very many places to abscond to, and the rate of communication is really fast, so the traditional forms of parasitism on fractional reserve banking don't work.
What obviously has worked is to create financial instruments, concoct explanations for why they are less risky than common sense would indicate, and let everybody involved in treating those instruments as having less risk than they actually do in on the scam, so they have a chance to take real money out in return for the fake future value.
The problem now is three things:
- The people running this con were and are bankers
- No one knows how much of the value currently being counted is fake
- No one know what potential real value is supposed to be paid for by something fake
It's going to take a ground-up rebuild for the basis of trust, which, as a devil-in-the-details task goes, doesn't strike me as highly tractable, even if all financial institutions are required to run entirely open, public books henceforward.
Doing that while the bits of value that are lies evaporate, dropping the stuff that depended on them, well.
The core fix is obvious—it's the usual choice between concentrating wealth and expanding access to choice—but the availability of bankers who are able to think in terms of expanding access to choice may be seriously limited, and the political solutions will only be as good as the advice they're getting. If there's no one saying "the view of the people who run the banking system that it's for concentrating wealth is part of the problem", there can't be a policy fix for the problem.
I don't think this is going to be a quick fix.