In Defense of the Greenspan (and Bernanke) Put

I support the "Greenspan put." If Bernanke shares the credit (or blame) for this, then I count it in his favor.

When I first heard about Greenspan's approach to "irrational exuberance," I thought it sounded just about right. I have not changed my mind. Of course, it is possible that I never understood him correctly.

First, I will explain what I understand the "Greenspan put" to be. Greenspan was open to the possibility that asset prices are sometimes driven by irrational exuberance. That is, stock prices might be higher than justified by the fundamentals of expected corporate earnings discounted by an appropriate interest rate. The same could be true of the prices of single family homes. The prices might be too high relative to expected future rents discounted by an appropriate interest rate. Those "rents" are mostly the evaluation of the housing services by owner occupiers.

The first part of the "Greenspan put" is that the Fed would not use monetary policy to "fix" the problem of excessive asset prices. I agreed and agree with that principle. Could the Fed use monetary policy to decrease asset prices? I think the answer is yes. Should it use monetary policy to decrease asset prices? The answer is surely NO!

Monetary policy should avoid monetary disequilibrium--most fundamentally any divergence between the quantity of money and the demand to hold money. If the Fed thought that stock prices or home prices were higher than justified by the fundamentals, the Fed could engineer an excess demand for money. Since the demand for money is generally growing, the Fed could simply restrict increases in the quantity of money.

It is possible that those short on money would sell off some of the particular assets whose prices are too high--stocks or houses--and that this would directly pop the bubble and return the prices to levels justified by the fundamentals. Those selling assets would intend to increase money holdings to match their demands to hold money, but the unintended consequence would be to bring asset prices closer to the fundamentals.

However, it is almost certain that those short on money would reduce expenditures on a variety of other assets, goods and services. In particular, as the growth of real expenditure on currently produced output slows, or perhaps even shrinks, firms are likely to respond by slowing or even reducing production and employment. The resulting recession would negatively impact real incomes and the ability of households and firms to purchase any number of assets, including the ones the Fed thought were inflated.

This policy would be destructive. The Fed would be disrupting the flow of production, income, and expenditure by engineering monetary disequilibrium. It would be doing so to manipulate asset prices--getting them to what what the Fed considers their proper levels. While having asset prices at the right levels would be a good thing, and perhaps "experts" at the Fed can determine what those might be better than market investors, it is wrong for the Fed to engineer monetary disequilibrium and disrupt real economic activity to achieve that goal.

An alternative way to describe this process uses neo-Wicksellian language. This approach assumes that one important short run effect of monetary disequilibrium is a shift of market interest rates away from the natural interest rate. The natural interest rate is the level of interest rates that keep saving and investment equal. Saving is that part of income not spent on consumer goods and services and investment is expenditures by firms on capital goods. While saving and investment can be defined in ways that make the Wicksellian framework ambiguous if not entirely meaningless, the sensible definition makes the natural interest rate simultaneously the level of interest rates that keeps aggregate real expenditures equal to the productive capacity of the economy.

From this neo-Wicksellian perspective, monetary policy involves manipulation of market interest rates by the Fed. The equivalent of avoiding monetary disequilibrium is to always keep the market interest rate equal to the natural interest rate. If asset prices, stock prices or home prices, are too high, the Fed could lower them by raising the market interest rate above the natural interest rate.

Interest rates are fundamentals that impact all asset prices, including those that are purportedly inflated. From this perspective, the fundamentals would include not only expected future corporate earnings (or expected future rents,) but also the natural interest rate that coordinates saving and investment. By raising market interest rates, the Fed could reduce all asset prices and while "ceteris paribus" this reduces those prices below their fundamental levels, if some particular asset prices were too high before, this is a move towards the "right" price.

Of course, raising the market interest rate above the natural interest rate also would result in saving being greater than investment and real expenditures growing more slowly than the productive capacity of the economy. As firms respond to reduced spending growth (or decreases in spending) by slower production and employment growth, the resulting recession will result in lower demand for all sorts of assets, including the ones that had purportedly been driven by irrational exuberance.

The Fed should never do this because it is failing to follow the key rule--keep the market interest rate equal to the natural interest rate. The Fed would be engineering monetary disequilibrium in a way that would disrupt the process of production, the earning of incomes, and the expenditure of those incomes on the output. Why? Because the fallout of that disruption will plausibly impact asset prices that the Fed determines are too high.

What about the other element of the "Greenspan put?" If an asset bubble burst, and stock or housing prices drop rapidly, the Fed will act to prevent this from having an adverse impact on the real economy. I thought and think that this is exactly correct.

The greater fools that purchase assets at the peak of a speculative bubble lose money. If they borrowed money to purchase those assets, the bankruptcy of the greater fools will show that their lenders were fools as well. It is never the role of monetary policy to seek to avoid such losses. The Fed should avoid monetary disequilibrium. The Fed should keep the market interest rate equal to the natural interest rate.

However, it is possible that a sharp decrease in some asset prices could indirectly result in an unusually large increase in the demand for money. It is also possible that a sharp drop in asset prices would result in a decrease in the natural interest rate.

Other things being equal, a decrease in asset prices reduces net worth. The reduction in wealth provides an incentive for households to increase saving to rebuild wealth. It is even possible that a sharp drop in the prices of assets, even financial assets, might reduce the willingness of firms to invest. An increase in saving and/or a decrease in investment reduces the natural interest rate. If interest rates fall to the new natural interest rate, then saving and investment are again equal. What happens to the level of saving and investment is ambiguous.

An alternative way to describe this same process would be that a decrease in asset prices reduces wealth and consumption expenditures. Investment expenditures may fall as well. These two effects decrease aggregate real expenditures below the productive capacity of the economy. The natural interest rate has fallen. A lower level of interest rates is necessary to keep aggregate real expenditures equal to the productive capacity of the economy. If interest rates fall to the new natural interest rate, saving and investment are equal and total real expenditures equal the productive capacity of the economy.

If the natural interest rate falls and the Fed were to leave its target for interest rates unchanged, it would create or allow monetary disequilibrium to develop. The "Greenspan put" simply requires that if the aftermath of an asset bubble results in a decrease in the natural interest rate, the Fed will lower its target for market interest rates so that saving remains equal to investment and aggregate real expenditures remain equal to the productive capacity of the economy.

Similarly, it is possible that a sharp decrease in asset prices and net worth may result in an increase in the demand to hold money. In the context of a growing economy, the real demand for money is generally growing. It is possible that heightened uncertainty associated with a popped asset bubble would cause people to choose to increase their money holdings more than usual. Further, if market interest rates are falling with the natural interest rate, the lower opportunity cost of holding money may generate a liquidity effect--indirectly leading to an increase in money demand.

To avoid monetary disequilibrium, the Fed must increase the quantity of money more than usual to meet the temporarily increased demand to hold money. A sharp drop in interest rates and an extra large increase in the quantity of money would be the consequence of the Fed following the proper policy--avoiding monetary disequilibrium and keeping the market interest rate equal to the natural interest rate, in the context of a sharp increase in the demand to hold money and a decrease in the natural interest rate.

Further, suppose the Fed failed to follow the proper policy and allowed an increase in the demand for money to generate monetary disequilibrium. Equivalently, suppose the Fed kept the market interest rate unchanged despite a lower natural interest rate. The result of these policies would be reduced real expenditures, reduced output, and reduced employment--in other words, a recession.

A recession would certainly reduce corporate earnings and presumably the demand for homes. To the degree that stock prices or home prices are reduced because of expectations of a recession, this is an expectation that the Fed will fail to respond to changing conditions in a way that will avoid monetary disequilibrium. Equivalently, it is an expectation that the Fed will fail to respond to changing conditions by keeping the market interest rate equal to a changing natural interest rate.

It is entirely appropriate for the Fed to explain that it will avoid monetary disequilibrium that might result from the aftermath of an asset bubble. The Fed should be committed to avoiding monetary disequilibrium at all times, and this is just one situation that might result in monetary disequilibrium. Again, another way of stating this commitment would be for the Fed to explain that it will adjust market interest rates to any change in natural interest rates that might result from the aftermath of an asset bubble.

The arguments regarding the market and natural interest rates when a bubble pop are symmetrical. If irrational exuberance results in increases in the price of some asset, the resulting increase in net worth can be expected to result in more consumption and less saving. Perhaps there could be some positive impact on investment by firms. A decrease in saving and increase in investment result in an increase in the natural interest rate. If the Fed is keeping the market interest rate equal to the natural interest rate, then it will raise market interest rates.

Further, the increase in consumption and perhaps investment due to an increase in asset prices would involve an increase in real expenditure beyond the productive capacity of the economy. The proper response of the Fed is to raise market interest rates to keep interest rate at a level that will keep real expenditure growing no faster than the productive capacity of the economy.

However, this has no necessary connection with irrational exuberance or an asset bubble. If asset prices were rising because of a correct assessment of some positive change in the fundamentals--say corporate earnings really will be higher in the future, then the result would be a decrease in saving and increase in investment. The natural interest rate has increased. And the Fed should increase the market interest rate so that saving remains equal to investment and total real expenditures grow with the productive capacity of the economy.

The implications for the demand to hold money and so the proper level of the quantity of money are a bit more puzzling. It is possible that an expanding asset bubble might motivate people to hold less money in order to profit from the bubble. If that were true, then the quantity of money should fall to match the lower demand to hold money. However, it is possible that higher asset prices would result in an increased demand to hold money. Perhaps the greater net worth generated by the higher asset prices motivates people to hold more of that wealth in the form of money. Further, some assets change hands, and at higher prices, it takes more money for those transactions. This could lead to a higher demand for money. So, if the Fed is committed to avoiding monetary disequilibrium and rising asset prices lead to an increase in the demand to hold money, an increase in the quantity of money is the proper response.

This is a puzzling result because if the Fed failed to accommodate this increase in money demand, then the resulting shortage of money could possibly motivate people too sell overpriced assets and so dampen the bubble. However, there is no particular reason why reduced expenditure should be centered on the particular assets that are overpriced. More importantly, if the assets have increased in price because of a correct assessment of the fundamentals--for example, future corporate earnings really will be higher, then failing to accommodate growing money demand not only is likely to impact other markets along with the one with the run up in prices, lowering those prices are pushing them away from fundamental values.

So, why all the complaints about the "Greenspan put?" Part of the problem is an inappropriate application of ceteris paribus thought experiments.

Suppose the natural interest rate and the demand to hold money are given. Any increase in the quantity of money is by assumption an excess supply of money. Any decrease in market interest rates are by assumption a reduction of the market interest rate below the natural interest rate. If the plausible institutional assumption is added that newly created money is lent out by the Fed and the banking system, then an increase in the quantity of money is an excess supply of money and this excess supply of money is matched by an excess supply of credit that plausibly results in the market interest rate being pushed below the natural interest rate.

It is certainly possible that the short run effect of the excess supply of money and equivalent excess supply of credit is additional demand for some particular type of asset and an increase in its price. The arguments given above about how higher asset prices raise money demand are similar to the better known liquidity effect that can temporarily absorb an excess supply of money.

Further, one of the fundamentals that impact asset prices is the rate used to discount expected future earnings. If the lower market interest rate is used as a discount rate, then asset prices should rise.

Of course, the long run effect of an increase in the quantity of money is for the prices of goods and services to be higher than they would have been. With the market interest rate below the natural interest rate, real expenditures outstrip the productive capacity of the economy. Shortages of final goods and the resources needed to produce those goods results in higher prices, including the prices of resources like wages.

In the end, the real quantity of money falls again to meet the demand to hold money. And at the same time, the real supply of credit falls so that the market interest rate again rises to meet the natural interest rate. While nominal asset prices should be higher along with all other prices, there is no real excess supply of money or credit.

The story is plausible enough, assuming given money demand and a given natural interest rate. And while one could say that the Fed shouldn't create asset bubbles in this fashion, a much more sensible approach is to just point out that the Fed should always avoid monetary disequilibrium. It should not create excess supplies of money. The market interest rate should not be pushed below the natural interest rate.

Further, in the real world, the demand for money and the natural interest rates are not given. And so, not all increases in the quantity of money are excess supplies of money and not all decreases in the market interest rate involve it falling below the natural interest rate. And so, observing an increase in the quantity of money or lower market interest rates along with rapid increases in the price of some asset do not show that monetary disequilibrium is at fault.

As explained above, if there is an asset bubble and it pops, it is very possible that the natural interest rate will fall and the demand for money will rise. Lower market interest rates and a rapid increase in the quantity of money is not creating a new bubble.

Won't a policy of preventing sharp decreases in home or stock prices from causing recessions lead to higher asset prices? Of course it will. If the Fed keeps market interest rates equal to the natural rate and adjusts the quantity of money to meet the demand to hold money, then recessions from that source will be avoided. If those recessions are avoided, then they won't result in depressed corporate earnings. Similarly, lower income during recessions won't depress housing demand and prices. Better macroeconomic performance is a fundamental! For the Fed to commit to following a sensible policy is a factor that should raise asset prices.

For example, if the Fed were to use a roulette wheel to randomly pick years to reduce the quantity of money by 30 percent, then this would be a new risk for the market. This risk would tend to deter investors from purchasing assets, and would be a factor that would keep asset prices from rising "too much." It would also be foolish policy. If the Fed stopped that policy, investors would stop worrying about it, and so buy more assets and bid up their prices. That would be a problem?

The "Greenspan put" didn't promise to prevent greater fools from taking a loss from an asset bubble. It doesn't even promise to prevent "undershooting" where an asset price falls below its fundamental value. It simply promises to avoid monetary disequilibrium--to keep the quantity of money equal to the changing demand for money. Equivalently, it promises to keep the market interest rate equal to the changing natural interest rate. The goal is not to keep stock prices or home prices always rising. It is to keep saving equal to investment and real expenditure equal to the productive capacity of the economy. It is to keep monetary disequilibrium, perhaps from the aftermath of an asset bubble, from disrupting the process of production, income, and expenditure.