This paper describes economic developments in Chile during 1990–95. In part, as a result of the tight monetary policy introduced in early 1992, real GDP growth declined from 11 percent in 1992 to 6 percent in 1993 and further to about 4 percent in 1994. The monetary stance was gradually eased starting in October 1994, and output growth rebounded to 6.6 percent in the first quarter of 1995 compared with the same period of the previous year. Nevertheless, the 12-month rate of inflation fell to 7.9 percent in July 1995.

Abstract

This paper describes economic developments in Chile during 1990–95. In part, as a result of the tight monetary policy introduced in early 1992, real GDP growth declined from 11 percent in 1992 to 6 percent in 1993 and further to about 4 percent in 1994. The monetary stance was gradually eased starting in October 1994, and output growth rebounded to 6.6 percent in the first quarter of 1995 compared with the same period of the previous year. Nevertheless, the 12-month rate of inflation fell to 7.9 percent in July 1995.

II. Indexation and Monetary Policy

This chapter reviews the implementation of monetary policy in Chile in the last ten years. It first gives some background into the evolution of monetary policy up to the implementation of the current system in 1985. It then describes the indexation of the financial system and reviews the current framework for monetary policy, its functioning, and changes that have been recently introduced in 1995. The chapter ends with a simplified analysis of the determinants of inflation in Chile.

1. Background

The end of the fixed exchange rate period in 1982 coincided with a major financial crisis, which turned out to have important consequences for the future conduct of macroeconomic policy. The Central Bank played a key role in rescuing the financial system, including the issue of debt to buy the nonperforming portfolio of commercial banks. 4/ In the event, Central Bank debt relative to GDP rose from 2.5 percent in 1982 to 22.1 percent in 1983 and peaked to 42.0 percent in 1986. 1/ As a result of this episode, the Central Bank of Chile performs a dual role: (i) it carries out monetary policy by placing short-term instruments in the open market; and (ii) it determines the composition and the maturity structure of the public debt (which is practically all Central Bank debt)--a function typically carried out by the Treasury.

During the period 1982-85 the authorities pursued an aggressive policy of nominal devaluation to reverse the real appreciation that had taken place in 1975-82. During 1982-85 the policy of the Central Bank was basically dictated by the need to rescue the financial system and act as a lender of last resort.

In February 1985 the authorities implemented a new monetary policy framework which, subject to some minor modifications, would remain about unchanged until early 1995. Since the authorities wished to maintain a stable real exchange rate as a way to promote export-led growth, the use of the exchange rate for stabilization purposes was not possible. The choice was then reduced to using either some monetary aggregate or an interest rate as the main policy instrument. Given the instability shown by narrow monetary aggregates in Chile (see Rojas, 1993), 2/ it was concluded that setting an interest rate would be the most effective way of implementing monetary policy. The peculiarity of Chile’s financial system, where most financial instruments of over 90 days are denominated in UFs, 3/ implied that the authorities would have to control an indexed interest rate.

2. Indexation mechanisms

An understanding of the indexation mechanisms in Chile is essential for a discussion of monetary policy and inflation. The Chilean economy exhibits widespread indexation, affecting goods and services, labor, foreign exchange and financial markets. The UF is used extensively in the markets for durable goods (real state, automobiles, electronic appliances, etc.) and for services (education, health, and insurance). Although mandatory indexation in the labor market was abandoned in 1982, collective bargaining continues to be based on past inflation plus some measure of past productivity increase. Also, the reference exchange rate for the Chilean peso is adjusted daily by the Central Bank according to the difference between past domestic and estimated foreign inflation. Most financial contracts are expressed in terms of UFs. Interest on indexed contracts is charged with respect to balances denominated in UF on the basis of an annual rate compounded monthly (the premium over UF). Principal is converted back into Chilean pesos on maturity at the corresponding value of the UF.

Since the UF is linked to inflation, the UF interest rate is viewed as a real interest rate. The UF begins to be adjusted on a daily basis on the tenth day of month t, so that by the ninth day of month t+1 it has increased in value by as much as the CPI did in month t-1. The imperfection of this backward indexation rule is reflected in the fact that UF and ex-post real interest rates are not equivalent, as they would be under perfect indexation, reflecting the difference between the ex-post growth rates of the UF and the CPI. 1/ This difference is approximately equal to inflation in the month before a loan is contracted minus inflation in the last month of the contract, and it can be large, particularly when annual inflation is rising or declining, or when monthly inflation fluctuates significantly.

A set of complementary regulations were introduced in Chile’s indexation mechanism to prevent full indexation of the money supply and the financial system. Indexed time deposits are not allowed for maturities below 90 days. In the case of loans, indexed loans are legal for maturities above 30 days. By end-1994 nonindexed deposits amounted to 30 percent of total liabilities to the private sector of the financial system; nonindexed loans accounted for 33 percent of total loans to the private sector.

A system with only real targets may tend to perpetuate inflation and show a higher degree of inflation variability, thus having adverse effects on investment and output. In addition, sectors outside the scope of indexation are not protected from variable inflation. In recent years measures have been taken to reduce the scope of indexation to facilitate disinflation. In particular, the exchange rate rule has been “loosened” by successive revaluations of the reference rate, a widening of the band, and the pegging of the Chilean peso to a basket of currencies, and the Government has introduced projected inflation (in contrast with past inflation) as a key element in determining public sector wage increases.

3. Monetary policy

Until very recently the key instrument of monetary policy in Chile was the interest rate on 90-day indexed Central Bank promissory notes (PRBC). This interest rate is a key indicator around which the market-determined interest rates on indexed loans and deposits in the financial system evolve. The Central Bank places the PRBCs on demand through banks, finance companies, pension funds administrators (AFPs), and insurance companies.

The second instrument used by the Central Bank, a 30-day bill known as PDBC, was sold once a week. The yield on PDBCs is expressed in nominal monthly terms and is set equal to the real yield on PRBC minus 40 basis points. Since the UF is adjusted on the basis of the previous month’s inflation, the rate on PDBCs allows economic agents to estimate the inflation projected by the Central Bank for the current month. This in turn serves as a reference for commercial banks to adjust their nominal interest rates on 30 day deposits. The interest rates on PRBCs and PDBCs establish a floor to lending rates, as they provide an alternative investment opportunity for banks’ reserves.

As noted above, UF interest rates are commonly interpreted as real interest rates, and a widely accepted interpretation of the transmission mechanism in Chile is that, by influencing UF rates via the targeting of the 90-day PRBC rate, the monetary authority aims at closing the gap between the growth rates of potential output and expenditure. As this gap narrows, inflation is expected to decline, albeit with some delay reflecting lags in the transmission mechanism. Given these lags, large and sudden adjustments to the UF rate may reduce inflation at the expense of a substantial slowdown in economic activity; hence, policymakers prefer to operate with small and gradual adjustments, involving some waiting time to allow for the full influence of the change in rates to flow through the system.

The inflation rate, together with the UF interest rate, determine the nominal interest rate and, given real income, real money demand. Money supply (an endogenous variable in this policy framework) would adjust so as to satisfy real money demand. Since monetary aggregates are allowed to adjust, the growth of liquidity has been highly variable over time (Chart I.2) and sensitive to changes in monthly inflation.

In addition to conducting open market operations with its own paper, the Central Bank of Chile also uses other traditional instruments of monetary policy, including rediscounts, operations in the foreign exchange market, and reserve requirements. Rediscount operations take the form of a line of credit for liquidity at an interest rate expressed in nominal terms and with amounts available in three tranches at successively higher interest rates. The use of this line of credit is limited by institution to 60 percent of the reserve requirement of the preceding period. The Central Bank also can influence the money supply by intervening in the foreign exchange market when the interbank rate is within the exchange band. Although reserve requirements on local currency deposits were reduced in January 1994, 1/ the Central Bank has not used them to influence liquidity since the early 1980s. 2/ Reserve requirements on foreign currency deposits have been set at 30 percent to discourage capital inflows. In mid-1993 the Central Bank introduced repurchase operations with its own paper as an additional form of influencing liquidity for short periods of time.

Several episodes can be used to illustrate the workings of the monetary policy framework. The strong growth performance of 1985-90 was accompanied by a reduction in the 12-month inflation rate from 26.5 percent in 1985 to 12.7 percent in 1988. Inflation reached a minimum of 12 percent in the year ended in September 1988, but then accelerated in the last quarter of that year and throughout 1989, boosted by a strong surge in aggregate demand. By the last quarter of 1989 inflation had reached 31 percent. Against this background, monetary policy was tightened further in January 1990, when the PRBC interest rate was raised by about 2 percentage points to 8.7 percent; commercial bank lending rates reached a peak of 16.5 percent in March. Inflation proved resilient and only began to fall in November 1990 and to reach 18.7 percent during the 12 months ending December 1991. At the same time, real GDP growth decelerated to 3 percent in 1990. In late 1990 the Central Bank began to reduce gradually the PRBC UF interest rate which fell back to 4.7 percent at end-1991.

Again in 1993 monetary policy was used to slow the growth of domestic demand to a rate consistent with potential GDP growth and reduce the external current account deficit. As the economy showed signs of overheating, with real GDP growing at 11 percent in 1992, the Central Bank increased in three steps the interest rate on its 90-day indexed paper from 4.7 percent at end-1991 to 6.5 at end-1992 and then left it unchanged at that level until October 1994. Inflation proved to be resilient and remained at over 12 percent until August 1994 when, in the context of favorable terms of trade, it started to decline rapidly. It took almost 24 months of high interest rates and some appreciation of the exchange rate for inflation to come down. 1/

4. Recent changes in monetary policy

Since May 29, 1995, the Central Bank of Chile no longer sets the interest rate on its indexed 90-day PRBCs. This rate is now determined by market forces in preannounced auctions of fixed quantities, and the interest rate on one-day operations between the Central Bank and commercial banks has become the main instrument of monetary policy.

The Central Bank put forward three reasons to change its policy. First, to increase the role of market forces in determining interest rates in Chile and to make changes in the stance of monetary policy less disruptive. Second, to avoid distortions in the term-structure of interest rates stemming from the fact that the Central Bank determined the interest rate on an instrument in the intermediate maturity, while both ends of the yield curve were market-determined. In the previous situation, short-term interest rates tended to increase with an expected decrease in 90-day rates, as holders of pesos would drain liquidity by buying 90-day paper to realize the capital gain implicit in the expected rate reduction. Third, to facilitate Chile’s integration in international financial markets by increasing the volatility of interest rates as a way to discourage capital inflows, and avoid sharp movements in the exchange rate.

To make the new policy framework operative, commercial banks were authorized to hold interest-bearing overnight deposits with the Central Bank. The interest rate was set by the Central Bank at a nominal level equivalent to UF plus 5.2 percent. 1/ It also has lowered the rates charged on its existing one-day line of credit. 2/ Although the Central Bank announces these rates in nominal terms once every month at the time of the publication of the UF, in fact they continue to be determined as a premium over the UF.

The Central Bank still can influence interest rates by changing the amounts of paper offered in the auctions or by using repurchase agreements more actively. At the outset, the scheduled amounts of PRBCs to be placed were originally set at a level below the amounts falling due so as to provide more liquidity to the market. The Central Bank also continues with placements of its 30-day nominal paper (PDBC). The underlying indexed interest rate on PDBCs continues to be 40 basis points below the rate on PRBCs, so that the Central Bank effectively continues issuing an estimate of monthly inflation when setting the nominal rate for the PDBCs. 3/

In summary, the measures adopted in late May 1995 have not changed the basic framework for monetary policy in Chile although they have simplified its implementation. Monetary policy continues to target interest rates on indexed assets, although the transmission mechanism becomes more indirect and there is more room for market forces in determining the time structure of interest rates. Moreover, this partial change of regime increases the volatility of interest rates.

5. Annex: an interpretation

Based on the above discussion, it may be useful to think of inflation in Chile as being determined in the following way: 1/

πt=πt-1+(1-α)(µt-πt-1)+γ[φ(rt)-y¯],0<α<1,γ>0,(1)

which indicates that today’s inflation (πt) is determined by three main elements: (i) last period’s inflation (πt-1); (ii) the difference between the current rate of monetary growth (µ) and last period’s inflation; and (iii) excess aggregate demand, defined as the difference between aggregate demand, φ(r) (which is inversely related to the real interest rate, r) and the full-employment level of output, y.

If policymakers set the rate of monetary growth at a constant level, µ, then in the long run the inflation rate will converge to µ. By providing a long-run value for the inflation rate, the rate of monetary growth acts as a nominal anchor. The speed of disinflation will depend on α, which captures the degree of inflation inertia. The higher α, longer it takes for inflation to converge to its long-run equilibrium. In the extreme case of full indexation (α = 1), today’s inflation equals yesterday’s inflation adjusted by the current excess aggregate demand, and the economy loses its nominal anchor. In contrast, if there is no inertia in the system (α = 0), inflation would immediately reach its long-run value, µ (assuming that excess aggregate demand is zero). Given the high degree of inflation inertia in Chile, it is expected that, even if excess aggregate demand is zero, inflation would converge slowly to its long-run value. Measures to eliminate backward-looking indexation practices would thus help to accelerate the reduction of inflation to international levels.

As described above, monetary policy in Chile operates by setting an interest rate on an inflation-indexed asset (the premium over the UF). Equation (1) suggests that, by setting this “real” interest rate, policymakers affect excess aggregate demand, φ(r). In principle, the real interest rate could be set so as to eliminate excess aggregate demand; however, eliminating excess aggregate demand is a necessary but not a sufficient condition to control inflation. Indeed, suppose that the real interest rate is set so as to eliminate excess aggregate demand. Then, equation (1) reduces to

πt=πt-1+(1-α)(µt-πt-1)(2)

which indicates that if inflation is currently higher than µ, inflation will fall gradually toward its long-run value µ, with the speed of disinflation being determined by the degree of inflation inertia. Note that the role of µ is critical since it guides inflation to its long-run value. If there were a nominal anchor in the system (either the exchange rate or a monetary aggregate), then µ would simply be the rate of growth of the nominal anchor. In the case of Chile, however, there is no explicit nominal anchor. Hence, even if monetary policy ensures that there are no aggregate demand pressures, µ is not determined by the system and there is no long-run value of the inflation rate for the system to converge to.

Even if there is no explicit nominal anchor, however, the announcement of an inflation target (as the Chilean authorities do in their annual report to Congress) could provide a nominal anchor to the system. In effect, if the authorities announced an inflation target of µ and the public fully believed that the authorities would take the measures needed to achieve that target, then inflation would converge to that inflation target. In particular, an interest rate rule whereby nominal interest rates are raised (lowered) whenever inflation is above (below) the inflation target can provide under certain conditions the same degree of monetary control as a monetary aggregate or the exchange rate. Hence, Chile’s monetary policy framework can be interpreted as being based on an inflation target and interest rate rules. Even if monetary policy responds to other objectives in the short run, the system will remain anchored as long as the inflation objective takes priority in the medium and long run.

References

  • Budnevich, Carlos and Jorge Pérez, “Monetary Policy: The Recent Chilean Experience,” mimeo (Banco Central de Chile, 1994).

  • Mendoza, Enrique, “Fisherian Transmission and Efficient Arbitrage under Partial Financial Indexation,” IMF Staff Papers, Vol. 39 (March 1992).

    • Search Google Scholar
    • Export Citation
  • Mendoza, Enrique and Fernando Fernández, “Monetary Transmission and Financial Indexation: Evidence from the Chilean Economy,” IMF Paper on Policy Analysis and Assessment, PPAA 94/17 (1994).

    • Search Google Scholar
    • Export Citation
  • Rojas, Patricio, “El Dinero como un Objetivo Intermedio de Política Monetaria en Chile: Un Análisis Empirico,” Cuadernos de Economia, Vol. 30 (1993), pp. 139-178.

    • Search Google Scholar
    • Export Citation
  • Végh, Carlos, “Inflation Targets and Interest Rate Rules”, mimeo (International Monetary Fund, 1995).

4/

See Chapter IV for a description of these rescue operations.

2/

Broader aggregates proved more stable but they were weakly related with the target (inflation) and the task of controlling them was increasingly difficult.

3/

The “unidad de fomento” (UF) that was introduced in 1967 is a unit of account linked to the consumer price index with a one-month delay.

1/

For a full description of the operations of indexed contracts in Chile, see Mendoza (1992). For an analysis of the difference between interest rates on UF and real ex-post interest rates, see Mendoza and Fernández (1994).

1/

As of January 6, 1994, time deposits are subject to a 3.6 percent reserve requirement and demand deposits are subject to 9 percent.

1/

The decline in inflation was both in tradeable and nontradeable goods and there was also a deceleration of nominal wage growth.

1/

On Friday, June 23, 1995, the Central Bank lowered the rate on the daily deposits at the Central Bank to 5 percent.

2/

The percent rates on Central Bank rediscounts were reduced from 6.3 to 5.7 for the first tranche; from 6.7 to 6.0 for the second tranche; and from 7.1 to 6.5 for the third tranche. The interest rate in the first auction of 90-day PRBCs (Tuesday, May 30) declined from 6.1 percent to 5.7 percent. On August 10 the rate on PRBCS was 5.6 percent.

3/

Because the rate of the PRBCs is now market determined, the reference real rate for the PDBCs is now the clearing rate of the last auction of PRBCs.

1/

A theoretical development of the arguments below can be found in Végh (1995).