Saturday, December 19, 2009

"Tight Money" and Interest on Money

Scott Sumner has again claimed that economists talk about "tight money" without having any idea what it means. I know, I know, he is only counting important economists, and gives me a pass anyway....


"Tight money" is an excess demand for money--the quantity of money is less than the amount of money households and firms want to hold. But because "tight money" can impact the economy, and those changes in the economy impact the amount of money people want to hold, "tight money" must be defined in terms of an imbalance between the quantity of money and what the demand for money should be.

What does that mean? So far, I have argued that tight money is an imbalance between the quantity of money and the amount of money people want to hold when "the" interest rate is equal to the natural interest rate, coordinating saving and investment, and real income is equal to the potential income, the productive capacity of the economy.


What about the interest rate paid on money? While hand-to-hand currency--paper money and coins--generally pays no interest, most money takes the form of deposits. Using the MZM measure of the money supply, zero-maturity deposits make up 90 percent of the quantity of money. While there have sometimes been regulations prohibiting the payment of interest on some types of deposits, most money takes the form of deposits that can pay interest.

Conceptually, the interest rate on money could fall enough to prevent "tight money." Other things being equal, paying less interest on money should reduce the demand to hold money. If the interest rates on deposits used as money are lowered enough, it is difficult to see how "tight money" could continue to be a problem.

From the point of view of the banks issuing the money, paying less interest on deposits reduces their costs and enhances profits (or reduces losses.) There should never be a problem with banks being unable to afford to lower the interest rates paid on money. Further, it is superficially attractive for each individual bank to pay lower interest rates on money, reducing costs, and adding to profits.

There could be problems in the opposite situation of "loose money," an excess supply of money. If the quantity of money were greater than the demand to hold money, banks could pay higher interest rates on money, raising the demand for money to match the existing quantity of money. This would, however, raise bank costs, reduce bank profits, and perhaps result in heavy losses. Perhaps banks could not afford to raise the interest rates paid on money enough to clear up an excess supply of money. Further, banks are less motivated to take an action that at best reduces profits.

Of course, if we assume that the interest rate on money aways adjusts to keep the amount of money people want to hold equal to the quantity of money, and add that competing banks are able to adjust the quantity of money in response to profit and loss signals, then the result is similar to any competitive market. Prices and quantities adjust to maintain a balance between quantity supplied and demanded.

If "tight money" results in lower interest rates on money, lowering bank costs and enhancing bank profits, then the banks will be motivated to expand their balance sheets, funding the purchase of additional earning assets with the issue of more deposits that serve as money. If bank lending plays a large role in credit markets, then the interest rates on earning assets will fall. As the quantity of bank issued money rises, the interest rates that must be paid on those deposits to motivate people to hold them rises. In equilibrium, the difference between the interest rates paid on money and the interest rates on the nonmonetary assets banks hold will shrink to match the cost of providing intermediation services.

These considerations would be more important if there were "loose money." If banks are assumed to raise the interest rates paid on money enough to increase money demand to meet the existing quantity, and they can reduce the quantity of money by shrinking their balance sheets, then there is really no problem. If higher interest rates on money result in losses, then banks can contract their balances sheets. Again, if the banking system is large relative to credit markets, then interest rates on monetary assets should rise as banks shrink their balance sheets. As the quantity of money falls and the interest rate on money needed to motivate people to hold that quantity of money falls, the difference between the interest rates on nonmonetary assets and money grows until it matches the cost of providing intermediation services.

In their "A Laissez-Faire Approach to Monetary Stability," Leland Yeager and Robert Greenfield describe what they call the Black-Fama-Hall Payments System. They briefly describe a supply and demand approach to monetary equilibrium. They even have a graphical and mathematical treatment in an appendix. Black-Fama-Hall refers to papers by Fischer Black, Eugene Fama, and Robert Hall. The Black and Fama citations refer to papers where similar processes are emphasized. The yields on what serves as the medium of exchange are assumed to continually adjust to avoid any imbalance between supply and demand.

Unfortunately, interest rates on monetary liabilities are unlikely to directly adjust to meet the demand to hold them. Why not?

To the degree we are considering the existing monetary regime, convertibility between bank deposits and base money fundamentally changes the process. (Greenfield, Yeager, Black, Fama, and Hall were all writing about alternative monetary regimes which have no base money.) The interest rates paid on deposits are often an avenue by which banks respond to excess supplies or demands for base money--currency and reserves--rather than to excess supplies or demands to hold the money they issue.

For example, suppose there is an excess demand for currency. The likely result is a withdrawal of currency from banks. The banks respond to their reserve deficiencies in the usual way, by selling off assets and contracting loans. An individual bank can obtains more reserves in that fashion, but only by taking reserves from other banks. While the banking system doesn't obtain more reserves, as banks shrink their balance sheets, the monetary liabilities they use to fund earning assets should shrink. If the banking system is large relative to credit markets, then the reduction in bank lending and sales of securities increase the interest rates on earning assets.

Each bank is also motivated to pay higher interest rates on deposits, including deposits that serve as money. By attracting deposits from other banks, an individual bank attracts more reserves. While this is largely at the expense of other banks, they may even dampen or reverse the initial currency drain. Note, however, that the higher interest rates on deposits is both reinforced and reinforces the changes in the interest rates on bank earning assets. With higher interest rates on deposits, banks must earn higher interest rates to avoid losses. And the higher interest rates on loans and securities allow banks to pay higher interest rates on deposits.

However, this process is not correcting the "tight money." The initial excess demand for currency has resulting in a decrease in the quantity of deposits. Rather than the interest rates on those deposits falling to limit the demand to hold those deposits to the now more limited quantity, the reverse has occurred. The banks raised the interest rates on deposits, exacerbating "tight money."

Suppose that the initial problem wasn't an excess demand for currency, but rather an excess demand for checkable deposits. Do the banks immediately respond to this shortage of their product by lowering the interest rates they pay? It is very unlikely because people short on money are unlikely to show up at the bank asking for additional deposits even if they want to hold larger balances.

Those short on deposits will instead obtain more by selling nonmonetary assets and leaving the funds received in payment in their checkable deposits. (They might also reduce their purchases of goods and services or nonmonetary assets out of current income.) While the sellers' banks receive increased deposits and favorable net clearings, those are exactly offset by the adverse clearings faced by the buyers' banks and the decreases in the deposit balances of the buyers. There is no change in the quantity of money, and so the initial excess demand for money remains.

The sale of nonmonetary assets should lower their prices and raise their yields. Given the interest rate on deposits, this increases the opportunity cost of holding money. The amount of money people choose to hold declines to meet the existing quantity of money. This is, of course, the liquidity effect that plays a key role in Keynesian monetary economics.

How do banks respond to the higher interest rates on nonmonetary assets? The most likely scenario is that they raise the interest rates paid on deposits. The individual bank can earn more on its asset portfolio. To expand its balance sheet it needs to obtain more deposits. Those deposits used for monetary purposes serve this purpose as well as any others. So, the interest rate paid on money can easily rise in the face of an excess demand for money.

To the degree that the higher interest rates on deposits attracts increased deposits of currency into the banking system, the quantity of money created by the banks expands, which does help correct the shortage of money. The primary impact, however, is to make the "liquidity effect" disappear. As banks adjust the interest rates they pay to the interest rates they earn, the opportunity cost of holding money adjusts to the cost of providing intermediation services and no longer causes the demand to hold money to adjust to the existing quantity of money.

While the liquidity effect may disappear, the excess demand for money still exists. Something else, presumably lower real income or a lower price level, will be required to reduce the nominal demand for money to again match the existing quantity.

Are there no scenarios where changes in the interest rates on deposits tends to correct "tight money?" On the contrary, there are many scenarios where changes in deposit interest rates prevent or dampen monetary disequilibrium.

Suppose households and firms decide that they are "overleveraged" and have too much debt. As they restrict new borrowing and seek to pay off existing loans, banks and other lenders respond to the lower credit demand by reducing the quantity of funds lent and reducing the interest rates they charge.

The reduction in the quantity of new lending and the repayment of existing loans shrinks the asset side of the banks' balance sheets. The deposits of those repaying the loans will be decreased, reducing the quantity of money, and creating an excess demand for money.

Worse, given the interest rate on money, the decrease in the interest rates on nonmonetary assets reduces the opportunity cost of holding money. The result is an increase in the demand to hold money. Combined with the decrease in the quantity of money, the shortage of money is exacerbated.

If banks respond to the reduction in the interest rates they earn by lowing the interest rates they pay on monetary deposits, then the the reduction in costs makes the expansion of bank balance sheets more profitable. Further, a lower interest rate on money increases the opportunity cost of holding money.

It is possible that the interconvertibility between base money--reserves and currency--and deposits is equilibrating in this scenario. If the demand for currency, reserves, or both are positively related to the supply of bank deposits, and the quantity of base money is fixed, then a decrease in the quantity of deposits results in an excess supply of base money. As each bank attempts to rid itself of excess reserves, it expands loans and increases deposits in the usual way. Lowering deposit interest rates is another way to reduce excess reserves.

However, if the lower interest rates result in an increase in the demand for base money, either an increase in the demand for reserves because their opportunity cost falls as the return on earning assets falls, or else an increased demand for currency because its opportunity cost falls with the lower interest rate on deposits, then adjustments in the yields on bank issued money will fail to prevent monetary disequilibrium.

In the limit, lower interest rates on earning assets will push interest rates on deposits below zero. At some point, storing currency is cheaper than financial intermediation. By far the cheapest way to store currency is for banks to operate as 100 percent reserve institutions--storing currency for depositors. A sufficiently low demand for credit and low interest rates on money could greatly raise the demand for base money.

So, how does the interest rate on money relate to the concept of "tight money?" Ideally, the interest rate on money would adjust to keep the demand for money equal to the existing quantity. If that were to happen, there would be no monetary disequilibrium, even if bank balance sheets are not at the optimal size to maximize profits.

As an analogy, consider the market price of gasoline. Suppose the market price of gasoline is high and gasoline sellers are making high profits, and will expand the production of gasoline. Clearly, the gasoline market isn't in full equilibrium. But as long as the gasoline market clears, and there is no shortage of gasoline, the plans of the buyers and sellers are consistent.

The problem created by "tight money" is just this sort of inconsistency of plans, but with economy wide consequences. Instead of gasoline buyers being frustrated by shortages and lines, people throughout the economy are frustrated by the inability to sell goods and services, resulting in falling output, income, and employment.

However, the interest rates on money are unlikely to adjust to keep the demand for money equal to the quantity of money. Instead, any adjustments are much more likely to be movements relative to the interest rates on nonmonetary assets, particularly the earning assets held by banks. Any movement is likely to be towards the long run equilibrium where the difference between the interest rates on nonmonetary assets and the interest rate on money is equal to the cost of providing intermediation services.

Again, working by analogy to a market for a single good where excess supply or demand depends on the current market price of the good, the definition of tight money should based upon the existing "price" or rather, nominal interest rate, on money.

"Tight money" exists when the quantity of money is less than what the quantity of money demanded would be given the current nominal interest rate on money, the level of real interest rates on nonmonetary assets consistent with the natural interest rate, and a level of real income equal to potential income.

There are two more factors that need to be considered in understanding tight money, perhaps the most important--the price level and its expected rate of change. What should the price level be now? What should it be in the future? Perhaps the answer is whatever is necessary to keep the real quantity of money equal to the demand. But it really depends on the nature of the nominal anchor of the monetary system. More later...

No comments:

Post a Comment