Emmanuel Tumusiime-Mutebile: The elements of a modern monetary policy framework
The views expressed in this speech are those of the speaker and not the view of the BIS.
Introduction
The Bank of Uganda has now been implementing its inflation targeting lite (ITL) monetary policy framework for almost three years. I believe that our experience during this period demonstrates that ITL can be an effective framework for monetary policy, even in low income countries with relatively shallow financial markets. In my remarks I want to focus on what I believe are the most important lessons of Uganda's experience in terms of the institutional arrangements and the policy objectives of monetary policy.
Institutional arrangements
A feasible ITL framework requires a supportive institutional framework, of which two elements are particularly important. The first is that the central bank must have operational independence to determine its monetary policy. In practice this means that it must be allowed to set the policy interest rate to best achieve its monetary policy objectives (which I will discuss shortly) without interference from other institutions or persons. The importance of operational independence arises because in most countries the central bank is the only public institution which has a mandate to promote price stability. A central bank which does not have operational independence would risk having its monetary policy influenced by other institutions or persons with objectives other than price stability, which would obviously weaken the central bank's ability to control inflation.
The operational independence of the central bank should be enshrined in legislation, as is the case in Uganda where the Constitution states that the BOU shall not be subject to the direction or control of any person or authority. However, statutory independence is a necessary but not sufficient condition for the operational independence of the central bank. To be able to implement monetary policy in an independent manner, the central bank must also have sufficient instruments at its disposal, which it can use without having to seek the permission of other authorities. Hence there must be some arrangement for providing the central bank with marketable instruments (preferably Government securities as they are the most widely traded and liquid instruments in the domestic financial market) which can be used for monetary policy operations at the discretion of the central bank. Ideally the central bank should hold its own stock of Government securities. One possible channel through which Government securities could be transferred to the central bank is through the capitalization of the central bank, as has occurred in Uganda.
A second aspect of the institutional arrangements for monetary policy which is critical for the efficacy of an ITL framework is the absence of fiscal dominance. What does "fiscal dominance" mean in this context? For an inflation targeting central bank it is imperative that the Government does not systematically borrow from it to finance the budget, because this risks undermining monetary policy. Hence if Government is to support the introduction of ITL, it must be prepared to fully fund its domestic borrowing from the market, and avoid borrowing from the central bank. Preferably there should be constraints on Government borrowing from the central bank set out in legislation. In Uganda we have such an agreement in principle to this effect with the Ministry of Finance, but in practice it has not always been fully complied with.
Policy objectives
The monetary policy anchor in an ITL framework is a publicly announced inflation target. Consequently inflation must be the primary objective of monetary policy if the anchor is to play a meaningful role in monetary policy and to ensure that monetary policy has credibility with the public. Our primary objective in Uganda is to hold annual core inflation to an average of 5 percent over the medium term, which we interpret as being over a period of 1-2 years. Having inflation as the primary objective of monetary policy means that controlling inflation must take precedence over any other objectives, but it does not rule out the central bank having secondary objectives for its monetary policy, provided that the secondary objectives are subordinate to the primary objectives. The secondary objective of the BOU's monetary policy is to align real output as close as possible with the estimated potential output of the economy, although there are practical difficulties with estimating potential output.
Why does the BOU target core inflation, a measure of inflation derived from a basket which excludes food crops, fuel, energy and utilities, instead of headline inflation which includes all of the items in the consumer basket? Of course, the consumer cares more about headline inflation than core inflation. The reason why we target core inflation is that we have potentially better control over core inflation than headline inflation. The goods and services whose prices are excluded from core inflation are generally more volatile and more subject to supply price shocks than are other prices in the consumer basket. Because the prices of food crops and fuel are mainly determined by developments on the supply side of the market, such as the abundance of the harvest, they are largely outside of the control of monetary policy. This was articulated very succinctly by the former Deputy Governor of the South African Reserve Bank who wrote that: "... higher interest rates will not help to make the maize crop grow higher".
The forward looking nature of monetary policy
Why does the BOU frame its policy target for inflation in terms of the medium term? Consumer prices, even when excluding volatile commodity prices, are subject to shocks from both the demand and supply side of the economy; for example, movements in the nominal exchange rate have an impact on the domestic prices of imported goods. These shocks can operate over a much shorter horizon than that of the transmission mechanism of monetary policy. The transmission mechanism of monetary policy normally extends over a period of one to two years. What monetary policy can feasibly achieve is to control average inflation over a longer term time horizon, which is why we have defined our inflation target as a medium term target.
The medium term time horizon over which the transmission mechanism of monetary policy operates mean that monetary policy decisions must be made on the basis of a forecast of future developments in the economy. Current or past developments, such as the latest available inflation outturn, are only relevant to the monetary policy decision to the extent that they provide useful information about future developments, which may not be very relevant at all. This entails a qualitative shift in the way in which policy makers think about the economy and make decisions, to embrace a forward looking approach.
Inflation forecasts spanning the time horizon of the monetary policy transmission mechanism are the most important inputs for monetary policy decisions in an ITL framework. Inflation forecasts need to include both forecasts derived from quantitative models and qualitative judgements about how the factors which influence inflation, such as fiscal policy and the exchange rate, will evolve over the medium term. Policymakers also need to think in terms of probabilities and risks; for example, what is the probability that aggregate demand growth will accelerate over the forecast horizon? and given this probability, do we need to take action to forestall a rise in inflationary pressures?
The role of the exchange rate
Policy towards the exchange rate is one of the most difficult and challenging aspects of implementing an ITL monetary policy framework in a low income open economy. Both the BOU's primary and secondary policy monetary policy objectives are domestic objectives: inflation and output stabilization respectively. The policy interest rate is only used to influence these two objectives. But that does not mean that policy makers can be indifferent to the level and rate of change of the exchange rate.
The exchange rate is of crucial importance for low income economies for at least three reasons. First, these economies generally exhibit a strong and rapid pass through of exchange rate movements to domestic prices. Volatility in the nominal exchange rate imparts volatility to domestic inflation. In mid 2011, a very sharp fall in the nominal exchange rate against the dollar, by 18 percent in three months, made a large contribution to the acceleration of inflation in the third quarter of that year; this is not an experience we wish to repeat. Secondly, exchange rate volatility is very disruptive for firms which have to transact in foreign exchange markets. Thirdly, an overvalued real exchange rate is damaging for the competitiveness of traded goods industries and discourages private investment in these industries, which in turn retards long term economic growth and development, as a large volume of research, including research by IMF staff, has demonstrated.
The approach that we have adopted at the BOU is to dampen short term volatility in the nominal exchange rate through sterilized intervention in the interbank foreign exchange market; i.e. sales or purchases of foreign exchange accompanied by secondary market operations to keep the stance of monetary policy (the interest rate) unchanged. Our experience shows that sterilized interventions can be effective in influencing the exchange rate. For example, when the exchange rate was depreciating very rapidly in the last few days of February following announcements by donors that they would suspend aid, the BOU intervened and was able to stem the depreciation. The effectiveness of sterilized intervention most probably arises from portfolio balance effects in the context of imperfect capital mobility.
We also attempt to lean against prolonged overvaluation of the real exchange rate, also using sterilized intervention. However, there are constraints on the extent to which this is practically feasible. Sterilization to stem appreciation involves the issuance of domestic securities to mop up the liquidity created by foreign exchange purchases, which requires that the central bank must have a sufficient quantum of securities at its disposal for this purpose. It is also expensive for the central bank because of the large interest rate differential between domestic currency securities and the assets in which the central bank holds its foreign currency reserves. In Uganda this differential is currently about 11 percentage points.
The speed of transition to ITL
I would like to make a brief comment about our experience with regard to the speed of transition from a monetary targeting framework to an ITL framework. When we first began considering adopting an ITL framework we had envisaged a prolonged transition period not least in order to provide time for strengthening technical capacity. In late 2009 we began the transition the process of transition by adopting what we called a "flexible monetary targeting regime", which involved the introduction of more regular interventions in the money markets to try and dampen volatility in interbank interest rates, although without targeting a particular interest rate, and a monthly review of the monetary targets with the possibility that they might be revised to take account of changes in the macroeconomy. However the transitional monetary policy regime proved somewhat unsatisfactory in that it did not provide a clear anchor for monetary policy. The flexible aspect of the monetary targeting framework arose because the BOU had lost confidence in the efficacy of applying rigid monetary targets, but the framework provided no other monetary policy anchor to replace the monetary targets. As a result, the BOU accelerated the transition and adopted an ITL framework in July 2011, with the inflation forecast as the anchor for monetary policy. Furthermore, our technical capacities have been strengthened by the experience gained from implementing the ITL framework; which has involved substantial "learning by doing".
Communications policy
Good communications are integral to an effective ITL framework. Therefore the BOU completely has revamped its communications policy, the main goal of which is to explain to the public and the markets the reasons for our policy interest rate decision and to make public our forecast for core inflation. As communications policy will be the subject of one of this afternoon's sessions, I will not go into the details except to note that we have been pleasantly surprised at how much coverage the BOU's monetary policy now receives from the media, both local and international, and the close attention which many of the informed market participants, such as the international banks, pay to our monetary policy statements.