Page 1 of 9
134
Archives of Business Review – Vol. 8, No.7
Publication Date: July 25, 2020
DOI: 10.14738/abr.87.8590.
Melo, M. M., & Gomes, J. W. F. (2020). The Role of Monetary Policy using the Taylor ́s Rule under High Level Indebtedness: The
Brazilian Case. Archives of Business Research, 8(7). 134-142.
The Role of Monetary Policy using the Taylor ́s Rule under High
Level Indebtedness: The Brazilian Case
Marcelo Miranda de Melo
MSc in Construction Management (UMIST-U.K.), PhD in Economics
(CAEN-UFC), Research Professor at Federal University of Ceará (CAEN- UFC), Economics and Finance Professor at UFC, Ceará, Brazil
Jose Weligton Felix Gomes
PhD in Economics (CAEN-UFC), Economics and Finance
Professor at Federal University of Ceará, Research
Professor at Federal University of Ceará
(CAEN-UFC),Ceará, Brazil
ABSTRACT
This research investigated if the Central Bank of Brazil follows the
Taylor ́s rule under high indebtedness. Research data collected
covered last 25 years from Dec/1995 until Feb/2020 from the Central
Bank of Brazil and methodology applied was vector auto-regression
(VECM) analysis with five macroeconomic variables as follows:
Government net Debt to GDP (DEBT/GDP), Gross Domestic Product
(GDP), Exchange Rate (EXRATE), Consumer Price Index (IPCA) and
Interest Rate (SELIC). Government ́s indebtedness in relation to
Brazilian GDP doubled from January/2014 up to Mar/2020, what make
this research an outstanding issue. Conclusively the high-level debt
environment interferes in the Central Bank of Brazil policy. Therefore,
the Taylor ́s rule is not being followed as it is expected for inflation rate
targeting control. With respect economic activity monitoring the
Central Bank of Brazil still applies the Taylor ́s rule, increasing the
interest rate in overheated economic activity periods.
Keywords: Taylor ́s Rule, VECM, Indebtedness, DEBT/GDP.
INTRODUCTION
Over the last decades monetary policy has changed frequently. Most macroeconomic frameworks
used by major central banks all over the world still apply the Phillips Curve, Taylor ́s Rule and the
Quantitative Theory of Money (QTM) in their macroeconomic forecasting models. However, since
the American Crisis in 2008, most macroeconomic concepts are under investigation. Nowadays,
there is a room of facts that the stable relationship between money and prices and the exogenous
money offer is no longer consistent. The Fiscal Theory of Price Level (FTPL) and the Modern
Monetary Theory (MMT) try to explain, at least partially, most achieved results of empirical
research.
Page 2 of 9
Archives of Business Research (ABR) Vol.8, Issue 7, July-2020
135
The Brazilian case is particularly interesting due to the fact that the Central Bank of Brazil used to
apply for long period of time real interest rate policy to face persistent inflation. In recent times,
especially since early 2014, the DEBT/GDP ratio is continuously increasing. The purpose of this
research is to investigate if this increase in government ́s liabilities may affect the Central Bank of
Brazil monetary policy conduction applying the Taylor ́s Rule.
As monetary policy tool, the Taylor ́s Rule establishes that nominal interest rate should be greater
than inflation in order to ensure real interest rates when inflation increases. On the other side,
interest rate should fall when product would be under its potential level and increase if product
level is above. By feedback process applying real interest rate government ́s indebtedness tends to
increase affecting inflation subsequently. The main question is: Does the Central Bank of Brazil still
follow the Taylor ́s Rule under high level indebtedness?
Government ́s indebtedness in relation to Brazilian GDP doubled from January/2014 up to
Mar/2020, what make this research an outstanding issue. Unforeseen events like the corona virus
pandemic tend to get this situation even worse. Government’s expenses increased dramatically in
medical supplies, extra income for low income population, increases in budget availability for
states and counties, etc. Population isolation was the main measure to face this pandemic according
to Health World Organization (HWO). For that reason, Brazilian government expects relevant drop
in GDP and a considerable increase in net government ́s debt in post-pandemic environment.
Research data collected covered last 25 years from Dec/1995 until Feb/2020 from the Central
Bank of Brazil and methodology applied was vector auto-regression (VECM) analysis with five
macroeconomic variables as follows: net government ́s debt/gdp (DEBT/GDP), gross domestic
product (GDP), exchange rate (EXRATE), consumer price index (IPCA) and interest rate (SELIC).
This methodology is especially appropriate to consider bi-directional effects among these
macroeconomic variables. Granger causality methodology was also applied to identify short-run
effects among these variables as well as the impulse response function to reveal the long run ones.
The remainder of the paper is organized as follows: beyond this introduction, present relevant
literature review in 2; the time series co-integration, VAR and Granger Causality methodologies are
explained in 3; empirical results are presented in 4; especially data collected, stationary and co- integration results are presented. In 5 conclusions are addressed and finally, in 6 all references are
listed.
LITERATURE REVIEW
This research studied how the government must balance the provision of sufficient liquidity
against the risk of adverse expectations regarding future debt prices when private liquidity has
dried up [1]. The socially optimal balance is captured in a Taylor-like rule that sets a target for real
public debt and manages expectations by overreacting to deviations from the target value.
Another relevant academic wok estimated the natural rate and the Taylor Rule for the Brazilian
economy from 2003 to 2015. The natural rate in a small open economy is equal to the international
real rate of interest, adjusted for the premium due to country risk and exchange rate risk. This
framework allows decomposing the interest rate into components to understand why the Brazilian
interbank market rate is so high when compared to other countries. This natural rate is not
constant, and we use it to estimate the Taylor rule. They tested the hypotheses that the Brazilian
Page 3 of 9
URL: http://dx.doi.org/10.14738/abr.86.8590 136
Melo, M. M., & Gomes, J. W. F. (2020). The Role of Monetary Policy using the Taylor ́s Rule under High Level Indebtedness: The Brazilian Case.
Archives of Business Research, 8(7). 134-142.
Central Bank has changed its response to inflation, output and exchange rate during President
Dilma Roussef ’s first mandate. They cannot reject this hypothesis with respect to inflation and
output gaps. During the research period the Brazilian Central Bank became much more lenient with
inflation target rate than in other periods. Government’s priority seemed to be economic growth
[2].
A model-based fiscal Taylor rule was presented and a toolkit to assess the fiscal stance, defined as
the change in the structural primary balance. A simple fiscal Taylor rule prescribes the fiscal stance
as a function of past government debt, past output gap and the past structural primary balance.
The appropriate reaction to shocks depends on the debt level as the model features rising marginal
costs of debt [3].
The relationship between the Central Bank’s reaction, also known as Taylor Rule, and the Brazilian
public debt was analyzed. Results showed that when the Central Bank increases the interest rate,
it manages to decrease inflation and the GDP growth. However, these impacts are smoothed by the
increase of the debt/GDP. Debt/GDP path affected about 11% of inflation rate. On the other hand,
debt/GDP affected about 7% of interest rate [4].
Inflation stabilization and employment maximization through a Taylor Rule for fiscal policy, similar
to John Taylor’s foundational examination of the behavior of the Federal Reserve were examined.
This paper compared the historical data with the rule. When the predictions of the Deficit Rule are
compared to historical data from 1965, they found that fiscal policy aligns with what the Deficit
Rule predicts with two exceptions: the stagflation of the 1970s and the current increases in budget
deficits. This paper contributes a simplified intermediate target rule for MMT [5]. While this rule
does not provide a “how” for attaining maximized unemployment and inflation rates, it does offer
a controllable intermediate target for the federal government in the same way a central bank can
target short term interest rates.
This research examined the implications for optimal inflation of changes in the level and maturity
of government debt under the assumption where fiscal and monetary policies co-ordinate, and in
the case of an independent central bank following a Taylor rule. Under co-ordination, inflation
persistence and volatility depend on the sign, size and maturity of debt. Higher debt leads to higher
inflation and longer maturity leads to more persistent inflation although inflation plays a minor
role in achieving fiscal sustainability. Under an independent monetary authority, inflation is higher,
more volatile and more persistent and plays a significant role in achieving fiscal solvency [6].
METHODOLOGY
Time series modeling is based on two main approaches: structural and non-structural models. In
the structural approach, time series modeling makes plenty use of economic theory in order to
obtain the relationships among different variables. An underlying point of this approach is that the
theory is not always able to establish an arrangement of the dynamics of behavior among these
variables, since it is necessary to take into account the estimates and inference the endogenous
relationships that exist among the explanatory variables of the model. In this sense, vector
autoregression model (VAR) and vector error correction model (VECM) were developed to
establish the relationships among economic variables in non-structural models, that is, they do not
start from concepts already established in theory.
Page 4 of 9
Archives of Business Research (ABR) Vol.8, Issue 7, July-2020
137
The vector autoregression (VAR) is commonly used for forecasting systems of interrelated time
series and for analyzing the dynamic impact of random disturbances on the system of variables.
The reduced form VAR approach sidesteps the need for structural modeling by treating every
endogenous variable in the system as a function of p-lagged values of all of the endogenous
variables in the system. A vector error correction (VEC) model is a restricted VAR designed for use
with nonstationary series that are known to be cointegrated. The VEC has cointegration relations
built into the specification so that it restricts the long-run behavior of the endogenous variables to
converge to their cointegrating relationships while allowing for short-run adjustment dynamics.
The cointegration term is known as the error correction term since the deviation from long-run
equilibrium is corrected gradually through a series of partial short-run adjustments.
[7] combined cointegration and error correction models to establish the trace error correction
model. As long as there is a cointegration relationship between variables, the error correction
model can be derived from the autoregressive distributed lag model. And each equation in the VAR
model is an autoregressive distributed lag model; therefore, it can be considered that the VEC
model is a VAR model with cointegration constraints. Because there is a cointegration relationship
in the VEC model, when there is a large range of short-term dynamic fluctuation, VEC expressions
can restrict long-term behavior of the endogenous variables and be convergent to their
cointegration relation [8].
Following Zou (2018), let !R = (!%R, !ÄR , ... , !FR )Í be as Î-dimensional stochastic time series, t =
1, 2, ..., å and !R~Ì(1), each !ER~Ì(1), ⁄ = 1, 2, ... , Î is affected by exogenous time series of d- dimension DR = (D%R, DÄR, ..., DÓR)′; then the VAR model can be written by:
!R = À%!R&% + ÀÄ!R&Ä + ⋯ + Ày!R&y + ÃDR + ÒR,t = 1, 2, ..., å
If !R is not affected by exogenous time series of d-dimension DR = (D%R, DÄR, ... , DÓR)′, then the VAR
model of formula (1) can be written as follows:
!R = À%!R&% + ÀÄ!R&Ä + ⋯ + Ày!R&y + ÒR,t = 1, 2, ... , å
With cointegration transformation of formula (2), we can get that
Δ!R = ∏!R&% + ÙΓE
y&%
E]%
Δ!R&E + ÒR
Where
∏ = ÙAE
y
E]%
− Ì,
ΓE = − Ù AZ
y
Z]E*%
(4)
If !R has cointegration relationship, then ∏!R&%~Ì(0) and formula (3) can be written as follows:
Page 5 of 9
URL: http://dx.doi.org/10.14738/abr.86.8590 138
Melo, M. M., & Gomes, J. W. F. (2020). The Role of Monetary Policy using the Taylor ́s Rule under High Level Indebtedness: The Brazilian Case.
Archives of Business Research, 8(7). 134-142.
Δ!R = AB′!R&% + ÙΓE
y&%
E]%
Δ!R&E + ÒR
where BÍ
!R&% = QÕˆR&% is the error correction term, which reflects long-term equilibrium
relationships between variables, and the above formula can be written as follows:
Δ!R = AQÕˆR&% + ÙΓE
y&%
E]%
Δ!R&E + ÒR
The equation (6) is the vector error correction model (VECM), in which each equation is an error
correction model.
In order to check stationary conditions unit root tests were applied. The following time series were
used in this research: net government ́s debt/gdp (DEBT/GDP), gross domestic product (GDP),
exchange rate (EXRATE), consumer price index (IPCA) and interest rate (SELIC). The
differentiation process is applied if a time series has unit root. Therefore, ADF and Phillips-Perron
Tests can be used for the same purpose. Suppose êR e X ̄ integrated time series of order 1 (I(1)),
stationary in first difference, the residues of equation are also Ì(1), what is like say that those time
series are not stationary in level.
The Johansen Test is the most effective econometric test to detect time series co-integration [9].
The approach proposed by [10] performs trace and eigenvalue statistics for co-integration
detection. In addition, we used the Granger causality test for determining if one time series is
valuable in forecasting another one. A time series X ̄ is accepted to Granger-cause Y ̄ if it can be
shown, usually through a series of t-tests and F-tests on lagged values of . The number of lags to be
included is usually chosen using an information criterion such as the Akaike, Schwarz or Hannan
Quinn information criterions.[9].
EMPIRICAL RESULTS
The stationary condition was tested in all-time series to make possible for the application of co- integration test, Johansen Test. The co-integration methodology is only applicable using time series
not stationary in level. The ADF and Phillips-Perron tests were applied (Table 1).
Table 1: Stationary Test for Macroeconomic Time Series.
Time Series Specification ADF ADF
critical values
Phillips
Perron
Phillips Perron
critical values Significance
DEBT/GDP Level* -1.50 -3.99 +0.51 -3.99 1%
GDP Level* -0.22 -3.99 -0.91 -3.99 1%
EXRATE Level* -1.58 -3.99 -1.52 -3.99 1%
IPCA Level** -1.91 -2.57 -4.58 -2.57 1%
SELIC Level*** -3.05 -3.45 -3.25 -3.45 1%
Source: Own research. Legend: *drift and linear tendency, **none, ***intercept.