Kevin Clinton Winter 2005 Lecture notes 10 Monetary policy: nominal anchor for the system 1. Monetary stability objective Monetary policy was a 20 th century invention Wicksell, Fisher, Keynes advocated price stability objective, and hence active monetary management. Pre-1914 Gold Standard. Bank of Canada not established until 1934. Rationale for price stability objective inflation and especially deflation costly. In the absence of a serious problem of this kind, outside the short run money should be approximately neutral. That is, not affect real variables. The best that monetary policy can do for the economy is to provide an environment of price stability where thoughts about inflation or deflation do not pre-occupy people. In economic jargon, there has to be a nominal anchor for the system to solve, and it does not make sense to have an unstable anchor. Moreover, confidence and credibility are core issues in maintaining the value of assets that have no intrinsic value. Price stability has unique credibility as an objective. A common approach historically for small economies has been a fixed exchange rate or currency board. The hope is to import the credibility of a major central bank. Whether this will work well or at all depends on a number of factors, including the very credibility of the fixed rate. There is a debate on this issue. Central banking issues independence instrument versus goal delegating monetary power in a democracy Bank of Canada monetary policy governance transparency and communications Defining price stability What does it mean in practice? agreement that a low positive rate of change in the price level is an appropriate objective o avoid deflation like the plague o measurement error in price indexes but is the objective to limit over time the variance of o the price level o or its rate of change
There is a technical difference: non-stochastic versus stochastic trend. In practice, at low rates of inflation, this issue has not been important. 2. Instrument of monetary control Focus here on independent monetary policy. Current debates are about what price stability should mean, in theory and practice the institutions, strategies and tactics to achieve this goal. The operating instrument of the central bank is a short-term interest rate, usually overnight. Technical framework Reserve requirements are no longer a key feature of monetary control they are zero in Canada. Neither are the text-book stories about bank reserves and money multipliers. The BofC supplies approximately zero reserves (to be precise, usually a small positive) to the banking system each day, using open-market operations. The bank controls the overnight rate through the terms on which it provides these reserves, or settlement balances, not through the quantity. Banks have to settle interbank payments daily, using their settlement accounts at the central bank. The BofC levies an interest charge on settlement account deficits (bank rate) and pays a deposit rate (bank rate minus 59 basis points). These official rates provide a ceiling and a floor to overnight rates in the interbank money market. They also provide an incentive for banks to keep their balances at zero use of the BofC credit or deposit facilities is in costly relative to the overnight market. The BofC has no objection to temporary random borrowing by banks on settlement accounts. Digression: doesn t the central bank control money? For a while in the 1970s the set of ideas known as monetarism, or the new quantity theory, gained ground. Central to these were the ideas that the demand function for some monetary aggregate was stable: maybe narrow money, M1 (currency plus demand deposits); maybe broad money, M2 (M1 plus other bank deposits); or in the end maybe one of dozens of definitions the central bank could control one or more of the favoured aggregates the central bank should abandon discretion and adopt a rule of constant money supply growth Milton Freedman would replace the central bank with a computer The absence of stable empirical relationships quickly did this in. Such instability is inevitable because of financial innovation. The quantity theory may be useful in situations where monetary policy is out of control and the central bank wants to re-assert itself. That is, where changes in the money stock are so large that they dominate monetary relationships. For example, rapid inflation double digits or more. 2
Even then, the use of M1 as an intermediate target in Canada 1975-82 did not help to contain a surge of inflation. Interposing an intermediate objective between instrument and goal proved to be a distraction. In situations of moderate or low inflation i.e. the typical situation of the large economies the correlation between M and P has been quite unreliable, over any horizon, as a basis for inflation control. There is still a view that monetary aggregates may be nonetheless as information variables about future output and inflation but this is an inconclusive, and highly technical, empirical argument that we do not need to get into. 3. Monetary neutrality Long-run neutrality of money is one of the oldest propositions in economics. An equal proportionate change in all nominal magnitudes money supply (any definition), prices, etc., leave the real equilibrium unchanged. Under a metallic standard, large exogenous increases in the money stock would in the long run result in roughly proportionate increases in the price level. (It has long been acknowledged, however, that in the short run, a monetary stimulus could boost output.) In a modern monetary economy, this cause-effect story is no longer relevant because the money supply is endogenous. But one can still believe in long-run monetary neutrality. Charts showing correlations between M and P over widely differing inflation environments (either cross-section over countries or time-series) confirm monetary neutrality but say nothing about causality, or the appropriate way to maintain price stability. Impulses from central bank policy actions i.e. changes to short-term interest rates show up in other variables the yield curve, the exchange rate, output and prices more or less simultaneously with their effects on M. And these variables are directly relevant to our welfare. 4. Inflation targeting Strategic choices in Canada price index CPI and core CPI point target (2%) or band (1-3%) defining a quantitative rate is 2% too high? too low? just right? control horizon 6-8 qaurters contingent exemptions volatile items, indirect taxes 3
Monetary policy rule Taylor rule contingent rule nominal anchor target deviations the horizon for inflation control should the output gap be in the rule? also the exchange rate, and asset prices? trade-offs or efficient strategies i n T = r + p + γ ( p p ) + φ ygap (1) where i = nominal short-term interest rate p = inflation rate p T = target inflation rate 2% in Canada r n = equilibrium interest rate (set by world rate in small open economy) γ = φ = 0.5 = rule parameters Said to be a fair approximation of Fed behaviour in the 1980s and 1990s. For every point increase in inflation, the Fed raises the interest rate 1.5 points. For every point increase in the output gap, 0.5 points. This is an inflation control rule. The output gap is in the rule as an indicator of future inflation. It is not there to trade output for inflation. In the model the equilibrium value of the gap is zero regardless of the target rate of inflation (no LR trade-off). It is not a fixed rule in the Friedman sense, but a contingent rule. Not necessarily an optimal control rule either. Tests have been done on: the appropriate size of the coefficients Taylor s initial stab works quite well o a large value of g means that the central bank puts a high price on deviations from target o BofC uses an 18 month horizon, uses a forward-looking model, rather than Taylor rule per se whether the output gap is helpful especially given the measurement error whether the exchange rate should be included this is less an issue in the US than in Canada Ball makes a case (c.f. BofC MCI) whether asset prices should be included e.g. stocks, real estate. 4
Policy rule as an a nominal anchor In the standard modern 4-equation macro model, the policy rule provides the nominal anchor to close the system. The rule provides feedback that will eliminate deviations of inflation from a target rate. In the real world, whether this will work depends on credibility: The target has to be credible, p T cannot be just any old inflation rate it has to be a low one, and have some permanence. The rule itself has to be credible, and followed consistently. If the central bank has a good reputation, credibility may allow it to divert from the rule in the short run to deal with financial disturbances. 5. Monetary policy and stabilizing output If the economy starts at a full-employment, low-inflation equilibrium, monetary policy will generally be counter-cyclical. It does not need an objective of output stability the key objective for the central bank is price stability or steady low inflation. Since deviations from full employment excess demand or deficient demand would move the price level from target, the actions to stabilize prices are the same as those to maintain full employment. Example: the current easing by the Bank of Canada, in the face of inflation below 2%, and slow GDP growth. If the economy is experiencing chronic high inflation the central bank may have to impose a short-run recession on the economy to get back to price stability. However, once there, monetary policy would be in counter-cyclical mode. Maintaining low inflation does not require below-par economic growth. Most examples of strong economic expansion in the post WW2 era involve low rates of inflation 1950s, 1960s, 1990s US, etc. References http://www.bankofcanada.ca/en/inflation/index.htm Inflation-Control Target Inflation-Control Target: Definitions and Notes 5