- Home >
- Análisis Económicos >
- Publicaciones >
- Macro Vision

Macro Vision

< Volver*• Well-anchored inflation expectations, the recent and widespread decline in current inflation due to the high level of idle capacity in the economy, and important institutional reforms (such as the recently approved public-spending cap) pave the way for a sustained drop in interest rates. A reduction in the inflation target would reinforce this trend.*

*• In this study we show that a reduction in the inflation target does not necessarily limit the monetary-easing cycle in the short run. Quite the opposite. If monetary policy is perfectly credible, a reduction in the inflation target enables lower inflation and lower nominal interest rates. Real interest rates, however, would be temporarily higher. *

In a recent interview, Brazilian Central Bank Governor Ilan Goldfajn stated that inflation is near a breaking point. In a disinflation process, expectations must be anchored first, after which real (ex-ante) interest rates rise and inflation retreats, creating room for a sustained decline in interest rates.

This happened in Israel, as mentioned by the central bank governor and in our recent article (“Aconteceu em Israel” by Mário Mesquita). A recession in the early 2000s widened the output gap and, consequently, disinflationary pressures. The Bank of Israel responded by maintaining a conservative monetary-policy stance, and thus managed to convince the public of its commitment to price stability. As current inflation retreated, Israeli officials sought to halt inflation readings at permanently lower levels. To that end, they reduced the inflation target from 7%-10% in 1998 to 4% in 1999, and later to 3%-4% in 2000.

A similar process is taking place in Brazil. In recent months, the Brazilian Central Bank concentrated its efforts on anchoring inflation expectations and was successful. Market estimates for inflation in different time periods declined, as shown in the chart below. With well-anchored expectations and plenty of slack in the economy, current inflation has broadly retreated.

These drivers, along with key institutional changes (such as the approval of the spending cap), may enable a change in the country’s inflation levels. But that will only be possible if Brazil maintains its commitment to pursuing ever-lower inflation readings, as Israel did. In other words, if the inflation target is reduced and monetary policy is credible, this will translate into lower inflation in the coming years, without the need for tighter monetary policy.

**Impact of a lower inflation target on inflation forecasts **

In order to quantify the impact of a lower inflation target on inflation forecasts, we estimated two Phillips curves in annual frequency, from 2002 until 2016: one for inflation in market-set prices excluding food and another for food inflation, as described below.

In the equations, in addition to a measure of the gap in the labor market, we included past inflation and inflation expectations three years ahead. We used inflation expectations three years ahead as a proxy for the inflation target that economists perceive as the one actually pursued by the central bank.^{[1]}

If monetary policy is credible, a revision in the inflation target triggers immediate revisions (of the same magnitude) in inflation expectation in longer terms. Hence, although specific shocks may cause inflation to deviate from the target in the short run, economic agents trust that a credible central bank will act to drive inflation to the target set by the National Monetary Council (CMN).

Inflation in market - set prices ex - food^{[2],[3]}

= α_{1}Unemployment gap+ α_{2}Inertia+ α_{3}Commodity prices in BRL

+α_{4}Inflation in regulated prices+ α_{5}Proxy for the inflation target

Where: α_{1}=-0.5, α_{2}=0.25, α_{3}=0.04, α_{4}=0.07 and α_{5}=0.64

Food inflation

= β_{1} Agricultural commodity prices in BRL+β_{2} Proxy for the inflation target

Where: β_{1}=0.15 and β_{2}=0.85

Based on the estimated coefficient and assuming full credibility of the Central Bank, a 100-bp reduction in the inflation target reduces market-set prices by approximately 70 bps, meaning an impact of 50 bps on headline inflation.

**Impact of a lower inflation target on the Selic benchmark rate: **

In order to quantify the impact of a lower inflation target on the nominal interest rate, we used the Taylor Rule, as described below[4]:

Nominal interest rate

= Nominal neutral interest rate+ α.(Inflation expectation - Inflation target)- β.Gap

Nominal neutral interest rate=Real neutral interest rate +Inflation target

A lower inflation target affects the nominal interest rate through two channels:

1. The first one is a direct and mathematical effect. A lower inflation target reduces the nominal neutral interest rate, calculated as the real neutral interest rate plus the inflation target. The impact on the nominal rate has the same magnitude as the reduction in the target. In other words, a 100-bp reduction in the inflation target causes a 100-bp reduction in the Selic rate (determined by the Taylor Rule).

2. The second channel depends on the effect of a lower target on inflation expectations. To calculate that impact, we estimated a regression in which inflation expectations 12 months ahead are a function of a measure of the labor-market gap, inflationary inertia and inflation expectations three years ahead (as a proxy for the inflation target).

Inflation expectation= δ_{1}Inertia+ (1 - δ_{1}) Proxy for inflation target + δ_{2}Gap

Where: δ_{1}= 0.3 and δ_{2}= -1.6

A reduction in the target that economists believe that the Central Bank is pursuing (proxy for the inflation target) reduced expectations 12 months ahead by 70 bps (represented in the equation by the coefficient (1-δ_{1}) = 0.7).

In that case, the deviation in expected inflation from the inflation target widens by 30 bps[5]. Considering the estimated coefficient ��=2, a 30-bp increase in the deviation of inflation expectations from the target increases the Selic rate by 60 bps.

Hence, a 100-bp-lower inflation target may translate into interest rates that are 40 bps lower, calculated as the impact on neutral interest rates (-100 bps) plus the impact on the deviation of expectations from the target (+60 bps).

Although a lower inflation target translates into lower nominal interest rates, real interest rates increase because the drop in nominal interest rates (-40 bps) is milder than the decline in inflation expectations (-70 bps), producing higher real interest rates (+30 bps).

**What if monetary policy is not credible?**

Up to now, we only assumed that monetary policy is perfectly credible, i.e., economic agents believe that the Central Bank is committed to delivering inflation on target and that price adjustments tend to be close to this number. In this case, long-term inflation forecasts converge to the new target immediately and inflation expectations 12 months ahead decline, impacting current inflation as well as nominal interest rates, which recede as well. As shown in the table below, only real interest rates are temporarily higher under the assumption of perfect credibility.

On the extreme opposite end of the spectrum, if the Central Bank has low credibility (or none at the limit), a reduction in the inflation target will not translate into lower inflation forecasts. In that case, expectations remain consistently above the target set by the CMN, pressuring nominal and real interest rates.

**Conclusion:**

Well-anchored expectations, widespread disinflation and significant institutional reforms (such as the recently approved public spending cap) indicate a possible change in Brazil’s inflation level in the next years. The country should seize this opportunity to renew its commitment to the pursuit of lower inflation, i.e., to reduce the inflation target. A lower target does not limit a decline in interest rates. Quite the opposite. If there is perfect credibility, it will translate into lower inflation and nominal interest rates, although, in the short run, real interest rates could increase.

**Felipe Salles
Julia Gottlieb**

< Volver