Supply and demand for money and the central role that banks play in it can be very confusing, so what follows is a longish preamble aimed at explaining the dynamics. The main point then follows which is how positive feedback operates and causes instability.
Before Keynes published his general Theory (Keynes 1936) supply of and demand for money was regarded as a relatively simple business. Supply came from those who wished to lend money and demand came from those who wished to borrow it. Money was a marketable commodity like any other, where the price - the interest rate - was set by the relative levels of supply and demand. If savers were more willing to lend than borrowers to borrow then the interest rate would fall, and vice versa.
The involvement of banks in the process didn't change the fundamental understanding, whether or not people understood that banks created money rather than merely acting as intermediaries between savers and borrowers. Those who understood the money creation process regarded banks rather than savers as providing the supply of money, and borrowers again as providing the demand.
However Keynes recognised that there was something wrong with this understanding. Marketable commodities are normally produced by market participants and sold to other market participants, who then consume them. Money isn't like that. It is neither produced by market participants nor consumed by other market participants; it is produced by banks (and a bit by the state) and then circulates round the economy enabling transactions to take place. If we consider a borrower of money then they initially provide a demand for money but then very quickly supply it to someone else when they buy whatever it is that they wanted the money for. However their buying something with the money doesn't count as a supply of money, but why not? Simply considering borrowers as demanders and banks or savers as suppliers misses all the other transactions. If we are to pursue this line of reasoning then we should consider every buyer of wealth as a seller (supplier) of money and every seller of wealth as a buyer (demander) of money. The only difference between these transactions and a borrower is that the borrower sells the bank a promise to repay - to sell wealth in the economy in the future to earn the money to repay the debt. It's still a transaction albeit based on future rather than immediate wealth exchange.
We quickly realise that pursuing this approach leads to confusion because money isn't a commodity, it has a different character and must be treated differently, but how?
What we have to do is to stand back and consider the economy as a whole, and money either being supplied to the circulating flow (by bank lending or by people spending money that had been held out of use) or taken from it (by repayment of debt or by holding money rather than spending it).
In his General Theory (Keynes 1936) Keynes explained that demand for money comes from people's wish to hold money rather than other interest-bearing or dividend-paying assets, so the demand for money comes from those who wish to take money out of the circulating economy, even if only temporarily - to hold it rather than spend or invest it. This is what Keynes termed Liquidity Preference (Keynes 1936 Chapter 13). Money is the most liquid of all assets. Keynes explained that liquidity preference has three components:
i. the transactions motive - having enough ready money to cater for day-to-day transactions;
ii. the precautionary motive - keeping money available as security in preference to volatile assets and to deal with the unexpected; and
iii. the speculative motive - keeping money available to take advantage of bargains.
These relate to the various money buffers that were mentioned in chapter 24. The transactions motive is what keeps money in people's pockets, wallets, shop tills and in current accounts awaiting use for transactions; the precautionary motive keeps money idle, often for long periods; and the speculative motive keeps money idle temporarily in investor trading accounts, financial investment holding accounts and elsewhere.
At times of uncertainty the precautionary motive comes to the fore. Assets may well be giving a good return, but if investors think their price might fall then they prefer to hold money because although the return on money is zero it isn't negative, as asset returns often are. Until Keynes explained it in this way people regarded interest as a reward for saving, but Keynes explained that it was compensation for foregoing liquidity - the higher the interest the more liquidity that people will forego. When people fear for the future their liquidity preference rises, and a much higher rate of interest must be offered to compensate them for foregoing liquidity.
If demand for money increases faster than money supply (higher liquidity preference), perhaps because people are fearful of a downturn and falling asset prices, then they try to sell their assets for money before the expected fall, but because the money supply is falling relative to demand the only way to induce people with money to buy the assets is to lower the price, often bringing about the very fall that was feared. As asset prices drop the corresponding returns rise - directly as interest for debt assets (e.g. bonds), and indirectly for equities (company shares) as an increase in dividend yield. If, conversely, supply increases faster than demand (lower liquidity preference), say because lots of people are borrowing money and spending it and banks are happy to create it for them, then people bid asset prices up, and as they do so the corresponding returns drop.
To make sense of events we must be clear about cause and effect. In the above discussion increasing or decreasing liquidity preference (demand with respect to supply) was the cause, and the effects were asset prices falling or rising respectively, and returns (interest or dividend yield) rising or falling respectively.
If banks want to create more money, perhaps because they are optimistic about the future and want to attract more borrowers so as to make more profit, then they lower the interest rate to make borrowing cheaper, and the normal result is more borrowers, a higher money supply, higher asset prices and a correspondingly higher liquidity preference - because all money ends up being held by someone. People's liquidity preference (demand for money) has risen to match the increased supply because the corresponding interest rate has fallen - they are compensated less so they forego less liquidity. Note that the changes in money supply, asset prices and returns differ depending on whether the change in liquidity preference itself causes them or is an effect of another cause. It sounds complex but it's no more than our old friend 'willingness to buy' that was discussed in chapter 2. If people want more money then willingness to 'buy' money (by selling loan agreements or assets) is higher than willingness to sell it and they must persuade banks or others to part with it, whereas if banks want to persuade people to borrow more money then willingness to sell money is higher than willingness to buy it, and the effects are quite different in each case.
If we just consider the bank lending market it is the interest rate that balances supply and demand, as in the traditional understanding, but there is a difference. The traditional view considers that demand comes from one sector of the market - borrowers, and supply comes from another sector - banks. Banks certainly have control over supply, but in normal times (i.e. when creditworthy people are willing to borrow money) banks also have control over demand, because demand (initially from borrowers but eventually from those who receive the money that the borrowers spend) responds directly to supply, as described above where the interest rate provides more or less compensation for foregoing liquidity. By means of the interest rate banks have control over both supply and demand for money.
In normal times (i.e. other than when creditworthy borrowers can't be persuaded to borrow at any interest rate) there is an almost unlimited number of potential borrowers, ranging in creditworthiness from governments of mature well-established democracies at one extreme to people on low incomes and no security at the other. A cautious bank will lend only to a limited range of secure borrowers, and charge rates of interest that are competitive with other banks for each of this type of customer. A less cautious bank will lend to a wider range of borrowers of different levels of security, again charging competitive interest for borrowers in each range. In this way banks decide on the level of default risk that they are prepared to take, and achieve it by balancing the security of the borrower (the less secure the higher the risk), against the interest rate charged (the higher the interest the more defaulters the bank can cope with but the higher the chance of default), always taking account of rates charged by other competing banks. In this way the banking system as a whole provides a supply of money at different interest rates depending on the creditworthiness of borrowers. When banks have the borrowers the money is created at the stroke of a pen (or the click of a mouse) - money supply is simplicity itself.
In a nutshell both supply and demand for money are limited only by the risk that an individual bank is prepared to take - the more money they supply the higher their profits, but the higher the risk of defaulting borrowers overwhelming the bank's capital.
In abnormal times, as in severe recessions or depressions, banks' freedom of choice is much more limited because the demand for loans from secure borrowers is much lower. In fact people repay loans at a higher rate than new borrowers take on new loans at such times, so the money supply diminishes. The demand for loans from insecure borrowers remains high, but the risk of defaults in these circumstances is all the higher, so few banks will lend to such people.
Liquidity preference features again later in discussing Keynes' rejection of neoclassical economics in chapter 81.
The bank lending market is not normal because the bank and the borrower are not independent - dependence in general terms between market participants is discussed in chapter 34. The dependence is a subtle one, and it works through the information that banks use to assess creditworthiness. They do this in two ways:
· security of borrowers' income to be confident of repayment; and
· availability of collateral in case of default.
If there is spare capacity in the economy, then when banks lend money and at least some of it enters circulation, wealth creation increases, more workers are employed, and there is generally increasing prosperity all round. This provides positive feedback to the creditworthiness assessment in both of the above respects:
· the very fact that banks create money when there is spare capacity increases the number of creditworthy borrowers because more are employed and therefore more have a sufficient income to justify banks' lending to them; and
· with most bank lending being for asset purchase (property, equities etc.), asset prices rise thereby providing the secure collateral that the banks need in order to lend.
Here we see the effect of positive feedback - the activities of banks in supplying money increases the ability of borrowers to borrow it, so supplying money causes even more money to be supplied.
This process fuels an economic boom. People are spending the money they have borrowed in addition to their normal income, and the money they spend provides incomes for other people, so more wealth is created as spare capacity is absorbed. When that wealth is traded spending rises in the second and subsequent rounds in line with the income multiplier discussed in chapter 17. More people are now apparently more secure so they are better able to borrow, and in a growing economy there are good opportunities available to use borrowed money so more is lent (created) and borrowed, and the income multiplier does its work yet again and growth continues ever upwards at an increasing pace.
However there is a downside to borrowing. Borrowed money must be repaid, and as debts build so do the repayments. As banks create new money when they lend, they also destroy money when it is repaid (this is explained in more detail in chapters 39 and 44), and destroyed money is of course taken out of circulation. The increasing quantity of repayments acts as a brake on people's spending, but growth continues as long as more new money is created than old money is destroyed and spare capacity remains available. Unsurprisingly this situation can't go on forever, because eventually either the availability of new borrowers dries up before the spare capacity is absorbed, or the spare capacity is absorbed stopping the rise in new employment and overall income earned from employment, so again the availability of new borrowers dries up. At this point the money supply starts to shrink because the repayments of existing borrowers continue to take money out of circulation without sufficient new borrowing to replace it, and this precipitates the downturn.
Here the positive feedback loop (lending affecting creditworthiness) is still in operation, but now it works in reverse. As people continue to repay debts and borrow less, money is lost to circulation, spending is cut back, and less spending means falling incomes and higher unemployment. Falling incomes means that people sell assets in order to repay debts so as to keep their heads above water, and that causes falling asset prices. Lower incomes and lower asset prices means that people have poorer creditworthiness, so banks cut back on lending even to people who would otherwise still wish to take on debts (Dalio 2015 p13). Again we see the dependence loop - the activities of borrowers in borrowing less money reduces the ability (or in this case the inclination) of banks to supply it, so borrowing less money causes even less to be borrowed. Even worse for banks is the fact that falling incomes and poorer collateral means that there will be substantial numbers of defaulting borrowers, which seriously threatens banks' solvency.
Positive feedback often causes oscillations in dynamic systems in the absence of deliberate measures to control them (see chapter 34), and the system of bank lending and borrowing is no different. The severity of the oscillations depends on total debt - in the growth phase on the rate of spending from newly acquired debts as a proportion of total spending, and in the decline phase on the rate of reduction in spending due to repayment of existing debts as a proportion of total spending. The way it works is explained in more detail in the next chapter. These oscillations give rise to the business cycle and also to more severe booms and depressions.
The existence of positive feedback between banks and borrowers causes the lending market to oscillate between boom and bust, without ever finding a stable equilibrium state.
All this is explained very clearly in a video made by Ray Dalio available on Youtube.[192] I strongly recommend it - it makes a complex process much easier to understand.