The conventional Fisher-Friedman approach to the determination of nominal spending and income focuses on the equilibrium demand and supply of money through the quantity theory:
(1) PY = MV(i)
If the economy’s money-supply process has produced “too little” money for the current level of spending, businesses and households attempt to shift their portfolios and accumulate more money by (i) trying to sell some of their other financial assets for cash, and (ii) cutting back on their spending. Thus the flow of spending—and income, and production—falls until PY has fallen by enough to make households and businesses no longer seek to increase their money holdings. Combine this focus on money supply and money demand equilibrium with some model of price theory and price adjustment, and the result is a theory of the monetary business cycle.
At the zero-interest rate lower nominal bound, this Fisher-Friedman framework breaks down. With no opportunity cost to holding wealth in money rather than in other short-duration safe nominal assets, there is no reason for changes in the money stock to induce any changes in the flow of spending at all. Anybody seeking to model nominal and real income determination must then find another, alternative equilibrium condition to focus on.
The conventional theoretical macroeconomic step to take is to focus on the intertemporal Euler equation of a representative household with rational expectations: is it on an optimal consumption-savings path? But focusing on the Euler equation requires that there be (a) no disagreements between investors, and (b) a meaningful representative household, with (c ) correct rational expectations. If we are unwilling to make those assumptions, we are then driven back to the equilibrium condition from which, given the existence of no disagreements, a representative household, and rational expectations, the Euler equation was derived.
This equilibrium condition--appropriate for a more general theory—-is the Wicksell-Hicks flow-of-funds through financial markets condition: the supply of savings vehicles must be equal to the demand for savings vehicles. If the demand for savings vehicles is greater than the supply, people will cut back on spending their savings on currently-produced goods and services as they try to boost their holdings of savings vehicles. Spending, production, and incomes will fall until people feel so poor that they no longer wish to boost their holdings of savings vehicles. When the supply of savings vehicles is greater than demand, people will increase their spending as they try to turn their excess asset holdings into current consumption. Spending, production, and incomes will rise until the richer society is happy holding the existing supply of savings vehicles.
Thus theories about the effectiveness or ineffectiveness of any kind of policy—monetary, fiscal, banking, whatever—at the zero nominal interest rate lower bound are ways of deploying Wicksell-Hicks flow-of-funds equilibrium condition. They are theories about:
(2) B + S = B + I + (G - T)
about the demand for savings vehicles—the sum of the current stock of financial assets in the economy B and desired additions to that stock through savings S—and the supply of savings vehicles—the sum of the current supply of financial assets B plus business investment I plus government debt issuance G - T.
Note that if you wish to assume a (a) representative agent with (b) rational expectations you are still committed to an analysis via (2). It is a consequence of your intertemporal Euler equation. If and only if (2) holds, then your representative agent with rational expectations is on its optimal consumption-savings path.
Equation (2) gives us insight into the potential effectiveness of monetary policy at the zero lower bound. Such non-fiscal policies do not, by definition, change the supply of savings vehicles: they simply swap one savings vehicle in private portfolios for another—cash for Treasuries, short Treasuries for GSEs, private obligations subject to bankruptcy risk for private obligations backstopped by a government guarantee. For them to boost the economy therefore, they too must work via their effects on reducing private savings S or boosting private investment I.
How can they do this?
Mostly, they do this by convincing financiers that the path of real interest rates will be lower in the future than they had expected. Even though at the zero nominal lower bound the current size of the money stock is irrelevant because the opportunity cost of holding money is zero, this will not always be the case. Someday the size of the money stock will matter again. And if the central bank now takes steps that credibly promise such a larger money stock in the future when it matters, then expectations of lower nominal interest rates and higher price levels in the future which should boost investment and other real interest-sensitive components of spending now.
What are these policies?
Jawboning to reduce anticipated real interest rates: The central bank can claim that it will maintain interest rates in the future at a lower level and the money stock at a higher level than that of its normal policy reaction function. Thus, the central bank would claim, inflation will be higher in the future than forecasts based on normal reaction functions would allow. That means that real interest rates will be lower—and that would push down savings and push up investment.
The potential difficulty is that open-market operations can be unwound. That the Federal Reserve has raised the money stock this year does not necessarily mean that it will keep the money stock five years hence above the level called for by its standard reaction function. Pure jawboning is the ultimate in cheap talk. Jawboning backed by quantitative easing at the short end of the term structure is not quite pure cheap talk: the central bank is taking action. The problem is that the action is easily reversible.
Quantitative easing at the short end to reduce anticipated real interest rates: The central bank can claim that it will maintain interest rates in the future at a lower level and the money stock at a higher level than that of its normal policy reaction function—and back up those claims by expanding the money stock now by continuing to buy short-term government bonds for cash even though this is simply a swap of one zero-yielding short term safe nominal asset for another. The idea is that the Federal Reserve is not just claiming it will expand the money stock in the future—it has already expanded the money stock.
The potential difficulty is that open-market operations can be unwound. That the Federal Reserve has raised the money stock this year does not necessarily mean that it will keep the money stock five years hence above the level called for by its standard reaction function. Pure jawboning is the ultimate in cheap talk. Jawboning backed by quantitative easing at the short end of the term structure is not quite pure cheap talk: the central bank is taking action. The problem is that the action is easily reversible.
Quantitative easing at the long end to reduce anticipated real interest rates: The central bank can claim that it will maintain interest rates in the future at a lower level and the money stock at a higher level than that of its normal policy reaction function—and back up those claims by expanding the money stock now by to buying long-term government, agency, and private bonds for cash and committing to holding them to maturity. The idea is that the Federal Reserve is not just claiming it will expand the money stock in the future—it has already expanded the money stock. And such transactions are not or at least are not as easily unwound. Price pressure effects mean that the Federal Reserve will, embarrassingly, probably lose money on the round trip if it breaks its commitment to hold its purchased securities to maturity.
Quantitative easing at the long end may have another significant effect as well. By taking duration and, in the case of private bonds, default risk onto its balance sheet the Federal Reserve transfers that risk from the marginal investor to the marginal taxpayer. If the marginal taxpayer is at a corner solution in their financial market holdings or has a higher rate of time discount than the marginal investor or simply does not see through the government’s balance sheet, one consequence of quantitative easing at the long end is that investors will then have unused risk-bearing capacity. Duration and default risk spreads should then fall, and this fall in spreads should turn more investment projects into positive NPV ones. Quantitative easing at the long end thus has not only its potential principle effect on private investment and other forms of interest-sensitive spending via expectations of inflation and thus of real safe interest rates but also a secondary effect on investment via the government’s assumption of a role as a risk-bearing partner in private enterprise.
Back-of-the-envelope calculations based on standard finance principles suggest that this portfolio-composition effect should be insignificantly small. But similar arguments based on standard finance principles have long suggested that standard open-market operations in normal times should not be able to shake the intertemporal price system out for thirty years either, and there is a lot of evidence that standard open market operations in normal times do in fact have substantial effects.
Quantitative easing accompanied by declarations that the central bank has not changed its long-run inflation and price level targets: I do not understand why a central ban would pursue such a policy.
But that is the policy that the Federal Reserve appears to have decided to pursue.
Ryan Avent:
Monetary policy: The Fed's communications problem: As I've written before, the commitment to allow higher inflation in the future is one of the key methods through which the central bank can have a positive effect on an economy stuck at the zero lower bound. The Fed's efforts to clarify and push out the date at which it is likely to raise rates strikes me as a means to try and commit itself to higher inflation in the future. But the Fed's communications efforts in this regard run up against a serious obstacle in the form of the Fed's long-term inflation forecast, which is 2%. The Fed can't force future central banks to keep to any policy path. If the Fed were to project a long-run inflation rate above 2% then, as Mr Bini Smaghi says, markets might suppose that monetary tightening lay ahead, whatever the fine print says.
This is not an unsolvable problem but is, I think, one of the tight spots in which the Fed finds itself as it transitions from a framework that wasn't very good at boosting the economy at the ZLB to one that might be. One way to get around the problem would be to change the target, to 3% inflation or to something else, like a price or nominal GDP level, that implies future inflation above currently acceptable levels. The Fed may get there eventually, but probably not soon enough to have a meaningful impact on this recovery.
An alternative might be to bring the point at which future inflation is tolerated a bit closer to the present. That is, the Fed doesn't necessarily run into problems of inconsistency if it projects inflation above 2% 1 or 2 years from now—a timeframe over which markets readily understand this group of policymakers to have control—while maintaining the long-run 2% goal. Achieving that would require the Fed to give itself a framework within which it's acceptable to have inflation above 2% (and even to try to generate inflation above 2%), and as I wrote last week, I thought the Fed took a big step in that direction at its latest meeting. But one then has to choose to act within that framework. I suspect that what that will take is a near-term projection of inflation above 2% combined with action—asset purchases—designed to demonstrate that, yes, the Fed is actually trying to create a little catch-up inflation. At the last press conference, Ben Bernanke all but admitted that that would be a sensible thing to do. Now we just need to excise the "all but".
2025 words
Comments