A long-standing controversy in monetary theory and macroeconomics involves the question, "Is it money, or is it credit?" In most of the macro theory of the past thirty years, the focus is on money. This is nice because you get to write down M and stick it into equations. Thinking about credit does not lend itself to equations. But if equations are the proverbial lamp post, and we've lost our watch somewhere else, then we should dare to look elsewhere than under the lamp post.
Suppose we have a butcher, a baker, and a candlestick maker.
The butcher says to the baker, "I'd like a cake. I'll pay you for it when I get paid for my hamburger, which should be soon."
The baker says to the candlestick maker, "I'd like a candlestick. I'll pay you for it when I get paid for my cake, which should be soon."
The candlestick maker says to the butcher, "I'd like some hamburger. I'll pay you for it when I get paid for my candlestick, which should be soon."
If they are all willing to extend credit to one another, you can see how this economy will work. The baker will bake a cake and sell it to the butcher, etc. Everybody gets what they want, and everybody gets paid.
In general, though, trade does not take place this way. Instead, trade is facilitated by an institution that is trusted by all parties. That institution might be currency. Or it might be banks.
Suppose we have an economy that has come to depend on banks for supplying trade credit. You can see that if the banks are very generous, lots of businesses can operate on the basis of accounts receivable. Maybe some of these businesses even operate unsoundly--perhaps their capital expansion plans are based on overly optimistic assumptions.
Suppose it turns out that the baker bought too many ovens, issued bonds to pay for them, and it turns out that he cannot repay the bonds. His bank no longer wants to advance him funds against receivables, because the bank is afraid that the receivables might not actually materialize or, if they do materialize, the bondholders will claim them. So the baker's bank credit dries up.
You can see that if bank credit dries up for a lot of businesses, economic activity can decline. The butcher can no longer get bread, the baker can no longer get candlesticks, and the candlestick maker can no longer get hamburger.
What are the policy implications? If the banks are still lending but the baker has too many ovens, then it seems as though the best thing to do is just let the baker and his creditors suffer. The economy's capital needs to be re-deployed, and there is nothing to be done about it. People are not as wealthy as they thought they were.
If the baker has not over-extended, but bank credit dries up for no good reason, then government can improve the outcome by stepping in to replace the banks. If the government provides credit, or it induces the banks to expand credit, economic activity can be restored.
In real life, a slowdown in credit can reflect both factors. That is, borrowers may have over-extended, and there are adjustments and wealth losses to be endured. But in addition, banks may have become exceedingly cautious, denying credit in ways that produce a needless contraction in economic activity. The ideal government policy then involves letting failed firms fall by the wayside while promoting credit expansion for viable economic activity. But how does the government know what is viable economic activity and what is the desperate flailing of insolvent firms?
The challenge is difficult enough for a government that is motivated solely by economic considerations. In practice, one can be certain that the political bias will favor bailouts of failed firms. This is the opposite of the correct economic approach, since it keeps capital misallocated.