Forecasting Framework

(Source: Monetary Policy Statement 2009)

The Bank’s inflation forecasting framework has two complementary elements, namely, the Near-Term Forecasting (NTF) model and the Core Model for Medium-Term Forecasting that, in addition to expert analysis and judgement, are used to support monetary policy formulation.  These are based on assumptions and research on the policy and price transmission process.  Figure A shows, in broad terms, the transmission process. The NTF projects inflation for four quarters ahead and incorporates the influence of South Africa’s inflation, rand/Pula exchange rate, inflation persistence (represented by past inflation – a quarter earlier) and known/projected adjustments to administered prices (e.g., fuel prices and utility tariffs), government levies and consumption taxes. Prospective developments with respect to these factors, therefore, determine the forecast path for inflation over the next four quarters.

The Core Model is a structural representation that captures key relationships (the transmission process) in the Botswana economy.  The model allows for consistent projections of up to 12 quarters ahead for key macroeconomic variables such as GDP, inflation, interest rates and exchange rates.  In this formulation, the real variables are expressed in terms of the deviation from their trend values.  An important feature of the model is the built-in policy response, in terms of: (a) adjustment to the policy interest rate (Bank Rate) to bring inflation to the target range, and (b) adjustment of the nominal effective exchange rate (through adjustment to the rate of crawl) in line with the inflation objective and underlying real trends.

The influences on the key variables are specified as follows:

• The output gap is determined by its past level, real monetary conditions index, South Africa’s output gap and any aggregate demand shocks (e.g., increase in incomes, taxes, etc.);
• Inflation is explained by its past level, inflation expectations, import prices, output gap and shocks to aggregate supply;
• Real marginal cost influences are in the form of imported inflation and domestic demand pressures arising from the output gap;
• Exchange rate changes are consistent with the crawling band arrangement; and
• Adjustment of the policy rate (Bank Rate) to stabilise inflation around the objective and maintain trend output growth.

The output gap estimates the extent to which actual GDP is either above or below the long-run trend that is consistent with price stability.  Excessively high output compared to trend is inflationary, while much lower output compared to trend is disinflationary.  Real monetary conditions refer to the combination of the level of real interest rates and real exchange rates that define the extent to which monetary policy is restrictive or accommodative of output expansion in terms of financing and demand.

The process for generating price increases and inflation is based on three key assumptions:

(a) Monopolistic competition: a market structure that has many producers/sellers of similar, but slightly differentiated products (e.g., Nike versus Adidas; Mercedes versus BMW; Sony versus LG). Each producer/seller can set its price without affecting the market.

(b) Imperfect substitution between factor inputs: faced with an increase in demand, suppliers would need to increase production and supply. However, while adjusting they cannot, in the short term, maintain the long term optimal relative proportions of the factors of production. For example, land and capital cannot be readily adjusted, but labour can be more quickly adjusted. Hence, workers are required to work more hours and/or there is an increase in the number of workers for the same amount of land and capital to meet an increase in demand. Therefore, initially labour productivity tends to decline and marginal costs of additional production increase. Hence, in addition to increasing supply, firms respond to an increase in demand by raising prices.

(c) The magnitude of the price increase would, however, reflect both backward-looking (inflation persistence) and forward-looking inflation expectations.

Demand, output gap and real monetary conditions: All things equal, an increase in the Bank Rate results in an increase in other market interest rates and adjustment of the exchange rate, to the extent it is flexible.  These developments change the real monetary conditions and, consequently, the desire for borrowing, which affects aggregate demand.  The response of firms to the lower demand determines the margin of the output gap and the rate of increase of prices.  

Foreign price increases: Other things remaining the same, local vendors would tend to pass on to consumers price increases of foreign-sourced goods and services.

Exchange rate developments: A change in the exchange rate will affect the import price of foreign sourced goods and services.  The exchange rate, on the other hand, can be affected by capital flows (cross-border financial transactions) that respond to the variation in interest rate differentials. This channel is, however, weak in Botswana given that the exchange rate is not fully flexible and, therefore, does not respond to changes in interest rates.  Nevertheless, a change in the Pula exchange rate from any source will be transmitted through this channel. 

Other price shocks: For Botswana, these would include an increase in administered prices (mostly the cost of utilities and government services and others, such as fuel prices and transport fares).  To the extent that these are artificially held below market levels for a long time, subsequent adjustments tend to be much larger than the general price trends and the inflation objective. 

Expectations: Among other considerations, suppliers of goods and services and workers base their decisions regarding price increases and wage adjustments on what they expect inflation to be. Expectations can be both backward and forward looking. Those with backward looking expectations see inflation persisting at past levels and will be slow to respond to changes in policy actions that affect inflation.  On the other hand, setting the inflation objective and entrenching the credibility of the policy framework can have a significant influence in encouraging forward looking expectations, which take account of the current and prospective conditions in determining expectations of future inflation.

While there are significant foreign influences and other occasional shocks to domestic inflation (as shown in the bottom part of Figure A), the monetary policy framework is premised on an understanding that the rate of domestic price increases reflects, in the main, local demand conditions and the policy environment. Monetary policy, therefore, has an impact on expectations and other second-round effects.  First, the extent to which local vendors pass on foreign price increases to consumers will depend on local conditions with respect to competitiveness and demand for their goods and services.  In turn, demand is influenced by the monetary policy stance and its effect on real monetary conditions.  Second, a well articulated and credible policy influences expectations as a source of inflation; thus, generally, price increases and demand for wage adjustments would be related to the monetary policy stance and the extent to which the public believes the inflation objective to be achievable on a sustained basis. Third, a widely accepted price stability objective can influence the rate of increase in administered prices to be consistent with the inflation objective. In the circumstances, there is scope for policy coordination and a measured approach to such cost increases.