International Journal of Application on Economics and Business (IJAEB)

Volume 1, Issue 3, 2023. ISSN: 2987-1972

INFLATION VALUE FORECASTING POST COVID-19

IN DENPASAR USING ARIMA
Ni Putu Ayu Mirah Mariati1*, Luh Pande Eka Setiawati1, Ni Luh Putu Sandrya Dewi1
1
Universitas Mahasaraswati Denpasar, Bali, 80233, Indonesia
*
Email: [email protected]

Submitted: 01-03-2023, Revised: 17-05-2023, Accepted: 03-07-2023

ABSTRACT
Forecasting is used to predict something that will happen in the future so that appropriate actions can be taken.
ARIMA is a time series forecasting method that was developed where the observation data in a time series data
interact. Inflation instability in Denpasar City in the future will make it difficult for the central bank and the
government to determine policy. The Covid-19 pandemic has an impact on the value of inflation in Denpasar City.
The purpose of this study is to estimate inflation in Denpasar City after Covid using the best ARIMA model.
Inflation data was taken from BPS Denpasar City from January 2020 to August 2022. ARIMA analysis was carried
out according to the Box-Jenkins procedure, namely searching the data, estimating parameters and significance
tests, and determining the best ARIMA model. The results of the analysis show that the best ARIMA model is
ARIMA (0,1,1). The results of this study indicate that monthly inflation in Denpasar City is likely to continue to
increase. Based on these results, it is hoped that appropriate policies will be made to reduce inflation.

Keywords: forecasting, inflation, post covid-19, ARIMA

1. INTRODUCTION

Corona virus disease 2019 (Covid-19) throughout the world. One of the impacts of Covid-19
in 2020 is the economic impact. Based on the Official Statistical Gazette published by the
Central Bureau of Statistics on Inflation Developments in March 2020, Indonesia's inflation
rate was 0.10%, especially in Denpasar City the inflation rate was 0.11% [1]. An increase in
the overall price level is called inflation [2]. Inflation is defined as a condition in which the
price of goods increases [3]. Inflation in Denpasar City, which tends to weaken, is caused by
several things, including increasing unemployment due to layoffs (Termination of
Employment) and the work system that has changed to Work From Home (WFH). Research
by Yanuarti et.al stated that the increase in positive cases of Covid had an effect on inflation in
Indonesia [4]. A country wants low and stable inflation. High inflation will certainly cause
people's real income to decrease, the price of goods to increase, and the value of the currency
to decrease. According to Keynesian, if the economy has good job opportunities, an increase
in government spending, an increase in private consumption and an increase in private
investment, it will generate a large demand as well, causing inflation [2].

If we only look at the decline in inflation during the pandemic, this is certainly an encouraging
thing. However, this low inflation of course also has an impact on other things, a decrease in
production, consumption, and investment which of course will end in a decline in economic
growth [5]. This is what happened during the COVID-19 pandemic. The Social Restriction
Policy, lock down, and WFH caused public consumption to decline, production value also
decreased. Inflation research during the Covid-19 pandemic has been carried out by many
researchers such as [6-8]. However, this research focuses on inflation during the Pandemic,
while this research will focus on the inflation value after the Covid 19 Pandemic. This is then
the importance of forecasting or predicting inflation. Forecasting of inflation is important in
economic policy makers. Inflation value predictions have also been carried out in various

https://doi.org/10.24912/ijaeb.v1.i3.1165-1169 1165
International Journal of Application on Economics and Business (IJAEB)
Volume 1, Issue 3, 2023. ISSN: 2987-1972

studies such as (Almosova et al., 2019), (Nyoni, 2019), (Astutik, et al., 2018) with various
forecasting models [9-11]. The focus of the government after the Covid-19 pandemic is of
course how to stabilize the inflation rate. This research is expected to be used as a consideration
for making economic policies, especially in Denpasar City for the future.

2. RESEARCH METHOD

The type of research used in this research is quantitative. The data used is monthly inflation
data taken from the Central Statistics Agency in 2020 to 2022. Data analysis uses the ARIMA
(Autoregressive Moving Average) model with the Box-Jenkins method. Box and Jenkins
introduced the ARIMA model in 1970 which is also known as the Box-Jenkins method which
consists of a series of activities to identify, estimate, and diagnose the ARIMA model with time
series data [12]. This model is well known in the world of finance and economics for
forecasting. The ARIMA model demonstrates an efficient ability to make short-term forecasts.
In the ARIMA model, the prediction value is based on the previous time value along with the
error [13].

3. RESULTS AND DISCUSSIONS

The ARIMA method consists of four stages, the first is model identification, the second is
model estimation, the third is a diagnostic test, and the last is forecasting.

Model Identification Process

Figure 1 Time Series Plot and Autocorrelation Function

In Figure 1, it can be seen that there is no autocorrelation, seen from lag 1 to 8, all values below
0.5 indicate close to 0, and from bar lag 1 to lag 8 nothing exceeds the red boundary line. This
indicates that the data is stationary. After doing the next step is to estimate the parameters
which will be explained below.

Model Estimation and Parameter Significance Test

After identifying the data, the next step is to estimate the parameters. The output results obtained
are as follows.

https://doi.org/10.24912/ijaeb.v1.i3.1165-1169 1166
International Journal of Application on Economics and Business (IJAEB)
Volume 1, Issue 3, 2023. ISSN: 2987-1972

Table 1 ARIMA Model

Model AR (1) MA (1)

pvalue = 0.028
ARIMA (1,1,0)
Significant
ARIMA (0,1,1) pvalue = 0.000 Significant
pvalue = 0.347
ARIMA (1,1,1) pvalue = 0.000 Significant
Not significant

Table 1 explains that from all the models tested, it is found that the models with all significant
parameters are ARIMA (1,1,0) and ARIMA (0,1,1) models so that both models are included in
the best possible model.

Best Model Selection and Diagnostic Checking

In the Best Model Selection and Diagnostic Checking, the best model selection will be carried
out based on the ARIMA criteria.

Table 2 Model Selection

Log
Model MSE RMSE MAE AIC
likelihood
ARIMA
0.27 0.43 0.35 120.22 -70.23
(1,1,0)
ARIMA
0.22 0.32 0.23 100.56 -74.76
(0,1,1)

Based on Table 2, the three significant models, one of the best models was selected which has
the smallest MSE, RMSE, MAE, and AIC values, and the largest log likelihood value. From
Table 2 it is known that the model that meets the criteria is ARIMA (0,1,1). Based on the
ARIMA model used, the best model is ARIMA (0,1,1). The model obtained is as follows:
ARIMA (0,1,1): xt = 0,94 + xt −1 − xt −2 +  t − 0,02 t −1 .

Forecasting

From the estimation results, the following results are obtained: it turns out that from the three
models, the ARIMA model (0,1,1) has a significant parameter, White Noise, and has the
smallest MS value. Thus, the correct model for this data is ARIMA (0,1,1). Forecasting results
for the next 10 months can be seen in table 3 below.

Table 3 Forecasting Results

Year Month Forecasting

2022 September 0.609
October 0.631
November 0.652

https://doi.org/10.24912/ijaeb.v1.i3.1165-1169 1167
International Journal of Application on Economics and Business (IJAEB)
Volume 1, Issue 3, 2023. ISSN: 2987-1972

December 0.674
2023 January 0.695
February 0.716
March 0.738
April 0.759
May 0.780
Juni 0.802

4. CONCLUSIONS AND SUGGESTIONS

Based on the results of forecasting using the ARIMA model, the inflation rate in Denpasar in
2022 will continue to increase. If this is allowed to happen, of course there will be a spike in
inflation. Therefore, based on these results, the Denpasar City government must adjust the
policies taken in stabilizing the inflation rate in the midst of efforts to restore economic growth.
It would be better if the government gradually restores economic growth while keeping
inflation down. Of course, this effort cannot be carried out with hasty policies, nor can it be
carried out by only emphasizing on one thing or one field.

REFERENCES

[1] Badan Pusat Statistik, Berita Resmi Statistik: Perkembangan Indeks Harga Konsumen /
Inflasi, 2020.

[2] N. G. Mankiw, Principles of Economics (8th Edition). Cengage Learning, 2018.

[3] Habarier, G., Inflation, Its Cause and Cures, America Enterprise Association, 1960.

[4] D. Yuniarti, D. Rosadi, and Abdurakhman, Inflation of Indonesia during the COVID-19
pandemic, Journal of Physics: Conference Series, 2021, 1821(1). DOI: https://doi.org/
10.1088/1742-6596/1821/1/012039.

[5] A. Wahyuni, Prediksi Nilai Inflasi Post Covid 19 di Indonesia, Indonesian Journal of
Islamic Economics and Finance, 2022, 2(1), pp; 57-65, DOI: https://doi.org/10.37680/
ijief.v2i1.1606 .

[6] W. R. Salam, Inflasi di Tengah Pandemi dalam Perspektif Islam, Jurnal Syntax
Tranformation, 2020, 1(5), pp.1–9, DOI: https://doi.org/10.46799/jst.v1i5.72 .

[7] A. Priyadi, A.A.Susamto, and H. Purwoto, Dampak COVID-19 Terhadap Tingkat Inflasi
Indonesia, 2021, Thesis: Universitas Gadjah Mada, http://etd.repository.ugm.ac.id/
penelitian/detail/203233.

[8] Hartati, Penggunaan Metode Arima dalam Meramal Pergerakan Inflasi, Jurnal
Matematika Sains dan Teknologi, 2017, 18(1), https://doi.org/10.33830/jmst.v18i1.
163.2017 .

[9] A. Almosova, N. Andresen, Nonlinear Inflation Forecasting with Recurrent Neural

Networks., Ecb.Europa.Eu, 2019, https://www.ecb.europa.eu/pub/conferences/shared/
pdf/20190923_inflation_conference/L2_Almosova.pdf .

https://doi.org/10.24912/ijaeb.v1.i3.1165-1169 1168
International Journal of Application on Economics and Business (IJAEB)
Volume 1, Issue 3, 2023. ISSN: 2987-1972

[10] T. Nyoni, Modeling and forecasting inflation in Burundi using ARIMA models, Munich
Personal RePEc Archive, 2019, from: https://core.ac.uk/download/pdf/214011504.pdf .

[11] S.R.P. Astutik, Sukestiyarno, P.Hendikawati, Peramalan Inflasi di Demak Menggunakan

Metode ARIMA Berbantuan Software R dan MINITAB. PRISMA, Prosiding Seminar
Nasional Matematika, 2018, pp745-754, https://journal.unnes.ac.id/sju/index.php/prisma/
article/view/20356 .

[12] G. E. P., Box and J. Jenkins, Time Series Analysis, Forecasting, and Control, 1970.

[13] I. Ghozali, Ekonometrika: Teori, konsep dan aplikasi dengan SPSS 17. Semarang: Badan
Penerbit Universitas Diponegoro, 2009.

https://doi.org/10.24912/ijaeb.v1.i3.1165-1169 1169