Resources Policy 82 (2023) 103573

Contents lists available at ScienceDirect

Resources Policy
journal homepage: www.elsevier.com/locate/resourpol

Oil price shocks and inflation: A cross-national examination in the

ASEAN5+3 countries
David Y. Aharon a, *, Mukhriz Izraf Azman Aziz b, Ido Kallir a
a
Department of Business Administration, Ono Academic College, Israel
b
School of Economics, Finance and Banking, Universiti Utara Malaysia Sintok, Kedah Darul Aman, Malaysia

A R T I C L E I N F O A B S T R A C T

Keywords: We test interaction between the oil price shocks and inflation in the ASEAN5+3 countries utilizing 35 years of
ASEAN5+3 monthly data beginning in 1987–2022. We show that when the COVID-19 pandemic is factored into our sample,
COVID-19 oil-specific demand shocks and aggregate demand shocks had a significant impact on inflation in these countries.
SVAR
These findings hint that the COVID-19 pandemic is likely the fundamental cause of the inflationary impact of
Oil shocks
Inflation
these shocks. The impact of rising inflation sparked by shocks emanating from oil-specific demand and aggregate
demand is evident in Malaysia, Singapore, Thailand, the Philippines, and Japan. We discover evidence that
inflation responds asymmetrically to oil price shocks, depending on whether the shocks are positive or negative.
Our empirical findings have significant policy implications for policymakers as they provide a reasonable
explanation for the ASEAN5+3 countries’ inflationary responses to various oil price shocks.

1. Introduction of Ukraine, which led to restrictions on the export of Russian oil, directly
impacting global oil supply.
Inflation is a key macroeconomic indicator whose fluctuation has a Though the pivotal role of oil and its derivatives in the global
major impact on the global economic system (Aharon and Qadan, 2022; economy is well-known, in the past three years it has become even more
Mensi et al., 2020; Tang et al., 2021; Wei, 2019). In June 2022, the evident. Policymakers nowadays, face a double sword challenge, as in
World Bank reported that annual inflation in advanced economies had addition to the peak of oil prices, heightened inflation levels are wit­
risen to 6.95% from 1.95%, while inflation in emerging and developing nessed around the globe. From this viewpoint, we are motivated in
economies had risen to 9.37% from 4.23% (World Bank, 2022a). High exploring the oil-inflation dynamics. Specifically, the motivation stems
inflation rates indicate the growth in demand as the economy recovered from the following several aspects. First, the examination of the oil-
from COVID-19 and energy prices rebounded (Kilian and Zhou, 2022). inflation relationship, in general, is essential for comprehending the
However, concerns have emerged among policymakers that high infla­ underlying forces behind inflation and its interaction with oil prices for
tion rates are expected to be persistent rather than temporary. making informed decisions related to individuals and businesses plans or
Among the many forces underlying inflation, shocks in oil prices the country’s economic policy. In this sense, our paper suggests an ex­
stand out as particularly important, as most of these shocks are man­ amination of a timely issue, particularly after the recovery of global
ifested as increased costs to consumers. Oil price shocks can lead to price economies from COVID-19 pandemic, that pertains not only to the
increases for commodities and services that rely on oil, such as re­ ASEAN5+3 nations but may be also applicable to other countries.
fineries, chemical products, and transportation. Also, it may restrict the Second, oil is a key component in the economic activity, and its cost
supply of certain products by increasing their production costs. Two can change inflation levels. Central banks regularly employ monetary
recent major events, combined with a longstanding process, created a policy to control inflation, but the impact of their decisions can be
“perfect storm” for oil prices. First, as economies recovered from the challenged by oil price fluctuations. By understanding the oil-inflation
COVID-19 pandemic, demand for oil surged. However, the supply of oil relationship, central banks can better evaluate the impact of oil price
was still limited due to disruptions in supply chains that had collapsed changes on inflation and alter their monetary policy accordingly.
during the health crisis. The second event relates to the Russian invasion Third, the analysis is crucial for countries that are heavily reliant on

* Corresponding author. Zahal 104 street, Kiryat Ono, 55000, Israel.

E-mail addresses: [email protected] (D.Y. Aharon), [email protected] (M.I. Azman Aziz), [email protected] (I. Kallir).

https://doi.org/10.1016/j.resourpol.2023.103573
Received 19 December 2022; Received in revised form 10 March 2023; Accepted 8 April 2023
Available online 17 April 2023
0301-4207/© 2023 Elsevier Ltd. All rights reserved.
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

oil imports, or alternatively, their national income and economic ac­ Armed with the data and figures on how fluctuations in oil prices,
tivity is heavily dependent on exporting oil, as the COVID-19 pandemic particularly, positive spikes in prices, affect inflation, compared with
added a new dimension to the complex interplay between oil prices and negative shocks in prices, policymakers at both the firm and country
inflation. Different countries are affected differently by oil prices based levels, could have better informed decisions pertained to future business
on their level of dependence on tourism and their interest rates, which planning, in their attempts of controlling inflation, as well as main­
leads to various outcomes for foreign investment in venture capital. taining economic growth and its stability.
Thus, nowadays, and perhaps even more than before, realizing the be­
tween oil – inflation relationship is essential for policymakers who 2. Literature review
attempt to balance inflation and maintain economic stability.
The objectives of this study are threefold. First, we are aimed at 2.1. Theoretical underpinnings
exploring the relationship between oil and inflation and the underlying
drivers of inflation. Specifically, our first objective is to adopt the two- There is a long stream of studies dealing with oil price shocks and
stage methodology, as described by Kilian (2009), and assess how inflation and is widely acknowledged that oil fluctuations affect infla­
different oil price shocks might affect inflation. Towards this end, we tion, at least to some extent (e.g., Choi et al., 2018; Chen et al., 2020;
employ the Cholesky decomposition to extract three unique oil shocks Edelstein and Kilian, 2009; Hamilton, 1988; Hooker, 1996; Mork, 1989;
from an estimated structural VAR model. Then, using the cumulative Wen et al., 2021). Nevertheless, the oil-inflation relationship is far from
impulse response functions we calculate the impact of structural shocks being static and has evolved over time due to various factors such as oil
to the price of crude oil on inflation for the ASEAN5+3 countries. crises, financial turmoil, and changes in institutional and technological
Our second objective is to differentiate our examinations under two structures (Elsayed et al., 2021).
definitions of samples: before COVID-19, (January 1987 to February Standard macroeconomic theory predicts that a sudden change in the
2020), and including COVID-19 (January 1987 to April 2022). By doing price of oil will have a negative supply shock on an oil-importing
so, we seek for any difference in oil shocks impact on inflation, in an country, creating ripple effects on real output and inflation. A surge in
attempt to reveal which type of oil shocks is mainly the underlying force oil prices is expected to raise production costs, hence decreasing the
behind inflation levels. economy’s aggregate supply (see Rotemberg and Woodford, 1996).
Lastly, our third objective aims to test the effect of positive versus Sudden spikes in oil prices may also have an impact by forcing a costly
negative oil shocks, and whether they share similar impact in terms of reallocation of resources across industries (Hamilton, 1988). With an
their significance and magnitude. Towards this end, we use Atems unanticipated increase in oil prices, resources may move from
et al.’s (2015) modification of Kilian’s (2009) method to provide energy-intensive sectors like automobiles and chemicals to less
empirical results regarding the issue of asymmetric responses. energy-intensive ones. This costly reallocation process exacerbates the
Our research questions derived from the aims of this study are as effect of an unanticipated change in oil prices beyond the share of GDP
follows: 1) which type of oil shock primarily drives the levels of infla­ allocated to energy expenditures.
tion; 2) whether oil price shocks impacted inflation differently before Oil price shocks can also have an effect on demand-side channels.
and during the COVID-19 pandemic, and 3) whether the effects of pos­ First, energy and, most notably, fuel costs tend to go up when there is a
itive and negative shocks on inflation are symmetrical. Towards this sudden increase in crude oil price. When fuel prices go up, households
end, we use the SVAR method which enables the recognition of struc­ have less disposable income owing to the increased cost of trans­
tural shocks and offers insights into the impact of policy actions. In this portation. Consequently, any unanticipated increase in crude oil prices
sense, the application of SVAR in studying the oil-inflation relationship can lead to decreased spending as they lower purchasing power and
has the potential of yielding novel insights for economic policy makers disposable incomes (Baumeister and Kilian, 2016; Baumeister et al.,
in the context of inflation, which nowadays is of their major concern in 2018). Oil price impact will be greater if energy demand is inelastic, but
their attempts to balance inflation and preserve economic stability. it will be constrained by the proportion of discretionary income that
We contribute to the debate of oil-inflation discussion by focusing on goes towards energy. Alternatively, rising oil prices may have a
ASEAN5+3 nations, a group of countries receiving much less attention moderately deflationary effect if they cause consumers and investors to
in the literature and are gradually becoming an integral part of the cut down on spending and investment in response to declining real in­
global economy.1 Our findings show that supply shocks do not have a come (Edelstein and Kilian, 2007, 2009). As a result, high oil prices may
significant impact on inflation, whereas demand shocks and oil-specific lower inflation by reducing domestic demand.
demand shocks consistently have a significant effect on inflation in most Hamilton (1988) argued that an unanticipated increase in oil prices
ASEAN5+3 economies. It is worth noting that the inflation effects may also trigger a rise in involuntary unemployment by forcing a cost­
resulting from oil-specific demand shocks are more substantial than lier reallocation of labor and capital between sectors. He contended that
those arising from demand shocks. Before the COVID-19 pandemic, none if people are laid off due to rising oil prices, impacted employees will
of the three oil shocks had any influence on inflation, according to the wait until their current industry improves before seeking employment
analysis of the two sub-sample periods. However, when we include data elsewhere. As a result, if there are bottlenecks in resource allocation,
from COVID-19 pandemic, we discovered that demand shocks and oil- increased oil costs could exacerbate recessions and reduce growth.
specific demand shocks have a significant positive impact on inflation Baumeister and Kilian (2016) posit that oil price shocks have the most
for seven of the ASEAN5+3 countries (excluding Indonesia). Our results significant impact on the economy through changes in disposable in­
also confirm the asymmetric effects of oil shocks by revealing that come, which in turn affect consumer spending. According to Baumeister
positive and negative oil shocks lead to higher inflation for most et al. (2018), the two channels—disposable income and terms of
ASEAN5+3 nations when we include data from the pandemic. trade—are identical. This implies that changes in oil prices have a
This paper may bear policy implications for monetary authorities. symmetric effect on the macroeconomic activity.

2.2. Symmetric oil price-inflation relationship

1
The ASEAN group of countries has a nominal GDP of more than US$2.8
trillion, putting it behind only the United States, China, Japan, India, and The first study to investigate how oil prices affected the overall US
Germany as the sixth largest economy around the globe and placing it in the top economy and consumer pricing was Hamilton’s (1983), concluding that
tier of fastest-growing regions (Kisswani, 2021). ASEAN5+3 consists of the oil price shock of the post-war era contributed to the US economic
Indonesia, Malaysia, the Philippines, Thailand, Vietnam, China, Japan, and the downturn. Hooker (1999) argued that oil prices have mostly indirect
Republic of Korea. effects on the macroeconomy through the transmission of inflation and

2
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

Fig. 1. Decomposed shocks to the price of oil.

interest rates. Cologni and Manera (2008) examined the impact of oil services, lowering inflation. Working on three ASEAN countries from
prices on inflation for the G7, finding that oil prices had a positive effect, 1978 to 2018, Husaini and Lean (2021) found that a rise in oil price
excluding Japan and the United Kingdom. Sek et al. (2015) found that positively affects inflation in Malaysia, Indonesia, and Thailand.
the price of oil had a direct effect on inflation for nations with low Conversely, a decline in oil price only leads to a reduction in inflation in
oil-dependency but an indirect effect for countries with high Thailand.
oil-dependency. Nasir et al. (2019) examined for the Gulf Cooperation There are, however, counterarguments that downplay the possibility
Council countries and found that the oil price has positive and signifi­ of a significant inflationary effect from rising crude oil prices.2 Leblanc
cant effects on inflation. and Chinn (2004) analyzed how changes in oil prices impacted inflation
For studies involving ASEAN countries, Razmi et al. (2016) analyzed in the G5 countries. The findings indicated a relatively weak impact on
the impact of oil price on inflation in four ASEAN countries during two inflation in the early 2000s. Similarly, Sari and Soytas (2006) showed
different time periods: before and after the 2008 global financial crisis that in Turkey, oil price shocks had a mild effect on the inflation. While
(GFC). The findings indicate that before the GFC, positive oil price fluctuations in oil prices are a primary engine of variations in inflation,
shocks resulted in a significant increase in inflation for all countries, Alvarez et al. (2011) demonstrated that their effect on Spain’s inflation
except Indonesia. Basnet and Upadhyaya (2015) investigated the impact rate is limited. Using a wavelet coherency analysis, Tiwari et al. (2019)
of oil price shocks on inflation in ASEAN-5 using SVAR model found
positive but temporary impacts of oil price on inflation for all ASEAN-5
countries. They claim that once a new equilibrium is reached, increased 2
see for example, the works of Kilian, 2009), as well as Kilian and Lewis
oil prices are gradually absorbed into the costs of other goods and (2011).

3
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

discovered that the impact of crude oil prices on inflation diminishes separating positive from negative shocks, our study examines whether
over time. In addition, panel models demonstrate that higher oil prices the distinct shocks have an asymmetric impact on inflation. Fourth, we
are linked to higher CPI inflation, but to a diminishing extent over time evaluate a distinct sample of countries with substantial intra-regional
(Cuñado et al., 2015; Choi et al., 2018). Gómez-loscos et al. (2012) trade, which includes major oil consumers and producers at varying
employed Qu and Perron’s (2007) methodology to study the impact of stages of economic development. The group of ASEAN5+3 is currently
oil price shocks on G7 macroeconomy. Their findings suggest that oil regarded as the most influential and effective trading bloc in the
price shocks have become less influential, with reduced effects on output developing world (Tan and Diaz, 2020). The ASEAN5+3 economies
and inflation by the late 1990s. have also gained prominence in the oil market and global economy.
According to Guesmi et al. (2017), the economies of the ASEAN5+3
2.3. Asymmetric oil price-inflation relationship countries have rebounded more quickly than those of other developing
market countries since the financial crisis of 2007-8.
Although studies into the correlation between oil price shocks and
inflation have been well common, their results are mixed, and may be
probably due to assuming symmetry between two variables despite the 2.5. Data
prevalence of deviation (Li and Guo, 2022). Another possibility as
mentioned by Hamilton (2003), stems in part from using an incorrect In this study, we focus on the ASEAN5+3 countries as a represen­
functional form and recommends using asymmetric approach to tative for the ASEAN countries. The entire group of ASEAN countries is
modelling oil price shocks. comprised of 10 ASEAN members and 3 East Asian countries (China,
The notion of asymmetry in price transmission suggests that an in­ Japan, and South Korea). Of the 10 ASEAN members, we use the original
crease in oil prices appears to create a quicker reaction in terms of in­ five members of the ASEAN association (Indonesia, Malaysia, The
flationary rates and outputs compared with a decrease in prices. Based Philippines, Thailand, and Vietnam). These countries provide 89% of the
on the innovative multiple threshold nonlinear ARDL model, Pal and total GDP of the 10 ASEAN members in 2019 (Husaini and Lean, 2022).
Mitra (2019) showed that changes in oil prices have an asymmetric Additionally, they account for 79.58% of regional energy consumption
short-term effect on U.S. inflation, suggesting that the magnitude of the (Tan and Uprasen, 2021), making them extremely vulnerable to fluc­
effect is stronger during oil price increases than declines in oil prices. tuations in energy prices (Kisswani, 2017).4
Hammoudeh and Reboredo (2018) also observed irregular and asym­ To summarize, our sample includes the original 5 members of ASEAN
metrical nexus between the movement of oil prices and inflation in the countries (Indonesia, Malaysia, The Philippines, Thailand, and Vietnam)
U.S. Particularly, they find that oil prices exceeding 67 dollars per barrel plus the 3 East Asian countries (China, Japan, and South Korea). For the
had a larger impact on inflation expectations in the intermediate term sake of clarity, we will call the countries we are analyzing ASEAN5+3.
than in the longer term. Likewise, Çatık and Önder (2013) affirmed the Our study utilized monthly data about inflation, oil price, world oil
asymmetric relationship for Turkey. Using a two-regime threshold VAR supplies, and the Index of Global Real Economic Activity (IGREA) from
model, they discovered that inflation and output are significantly January 1987 to April 2022. For inflation, we used the annual inflation
impacted by oil price changes only when the changes surpass a certain rate for each country from the IMF database.5 We computed real oil
threshold level. Choi et al. (2018) confirmed the asymmetrical impact of prices by deflating the cost of procuring crude oil in the United States by
oil price shocks in 72 countries, developing and developed. They showed the U.S. consumer price index (CPI). Information about world oil sup­
that positive shocks, rather than negative ones, have a greater effect. plies came from the U.S. Energy Information Administration.6 We
On the contrary, Li and Guo (2022) demonstrated a significant downloaded the IGREA from the Federal Reserve Bank of Dallas,7 which
asymmetric relationship between oil prices and inflation in China, with produces a business-cycle index based on a panel of worldwide, dollar-
the effect on inflation being more significant when oil prices decline denominated, bulk dry cargo shipping rates. This index is widely used to
than when they increase, particularly in the short term. In the same vein, capture the quantity of shipping in global manufacturing equity mar­
Lopez-Villavicencio and Pourroy (2019) found that in kets. The dataset is an improved version of what Kilian (2009) and Kilian
inflation-targeting nations,3 oil price transmission is greater through and Park (2009) used, after being amended in accordance with the
declines in oil prices than increases in them. Chou and Lin (2013), discussion in Kilian (2019).
Farzanegan and Markwardt (2009), and Ghosh and Kanjilal (2014) also
noted the asymmetry between oil prices and inflation in Taiwan, Iran, 3. Method
and India. Overall, these papers demonstrate that oil price shocks, both
positive and negative, have a sizable effect on inflation, the exact degree 3.1. SVAR advantages
of which varies depending on the severity of the shock.
There are several benefits that motivated us to utilize the SVAR
2.4. The literature gap and contribution model in this study. Structural VAR is a type of unrestricted VAR model
that can be used to make projections about a system of equations. One
Our research adds to the existing studies in at least four ways. First, key characteristic of SVAR models is the ability to use contemporaneous
our sample includes a unique period in which there were different un­ variables as regressors, especially when dealing with long time-series
derlying forces affecting oil prices. Second, despite the existence of
several similar studies (see Basnet and Upadhyaya, 2015; Cuñado et al.,
4
2015; Razmi et al., 2016; Husaini and Lean, 2021), we present a first Thailand and the Philippines rely heavily on foreign oil even though the
attempt to determine how decomposed oil price shocks affected infla­ ASEAN region has almost 40% of the Asia Pacific region’s total oil and gas
reserves (Kisswani, 2021). Indonesia, on the other hand, is a major contributor
tion in the ASEAN5+3 before and during the COVID-19 outbreak. By
to global supplies of liquid natural gas (Basnet and Upadhyaya, 2015). Simi­
distinguishing between the two sample periods, one ending before the
larly, the oil and gas business in Malaysia is regarded as the engine of the
COVID-19 outbreak, and the other including it, we can consider the country’s economy, contributing roughly 15% to its overall GDP (Ajmal et al.,
impact of the 2020–2022 events on oil prices and their effects. Third, by 2022). Given the above data, the analysis for the ASEAN-5 is a good approxi­
mation for the entire ASEAN 10. The remaining countries are either very small
or very poor, even in relative terms.
3 5
Countries where the central bank has publicly committed to a certain https://www.imf.org/external/datamapper/PCPIPCH@WEO/OEMDC
6
inflation rate or range, regularly updates the public on that target, and is held https://www.eia.gov/
7
accountable through institutional mechanisms. https://www.dallasfed.org/research/igrea

4
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

data (e.g., monthly). In contrast to the unrestricted VAR model, the occur due to shocks in either aggregate demand or oil-specific demand.
SVAR model imposes a few significant constraints that determine the A supply shock occurs when geopolitical turbulence, such as conflicts or
behavior of specific variables. We may use a particular economic theory shifts in OPEC members’ output restrictions, have an impact on the
and add restrictions to the SVAR model, which allow us to study the supply curve. Given the substantial costs associated with short-term
contemporaneous shocks of one variable to another. In other words, one changes in production (i.e., one month), oil producers are hesitant to
can assess the impact of a distinct shock by setting the off-diagonal parts promptly adjust output in response to shifts in demand.
of the residual covariance matrix to zero. In our oil price-inflation Second, there is a lag of at least a month before we see the effects of a
relationship analysis, such restrictions are advantageous when exam­ change in aggregate demand due to shocks to oil supplies. An unantic­
ining how each oil shock transmit its shocks to the system. To illustrate, ipated surge in economic activity follows a disturbance in demand
using SVAR, both the aggregate demand and oil-specific demand shocks caused by global economic movements and activities. A rise in global
can be adjusted to have no effect on the supply shock in the short run. consumption would raise the demand for oil, driving up oil prices and
However, the model can be relaxed so that aggregate demand shock is increasing supply. Yet, not all oil price surges are attributable to a
impacted in the short run by supply shock but not oil-specific shock. persistent expansion of economic activity, as they may result from
Thirdly, we can impose condition to allow oil-specific shock to respond concerns about the availability of oil in the future.
immediately to supply shock and aggregate demand shock. Third, changes in global economic activity and oil supplies are likely
to have an immediate effect on crude oil prices that is called oil-specific
3.2. How SVAR model adds value to the domain knowledge? demand shocks.8 It should be noted that, while oil supplies and prices
are likely to increase, these shocks are unlikely to be associated with a
The SVAR model was first developed by Sims (1980) as a response to positive impact on global economic activity. Additional details on these
his concerns about the employment of highly complex econometric factors can be found in Kilian (2009, pp. 1059–1060).
models for policy studies and forecasting work, which he claimed were Fig. 1 depicts the decomposed structural shocks to the price of oil
being identified using extraordinary (unjustified) exclusionary derived from Eq. (3). Following the 2008 subprime crisis, we see a
constraints. substantial inverse shock to the demand for oil in the late 2000s.
The findings of SVAR models are easy to analyze, unlike those of Hamilton (2009) argued that oil prices had already risen above their
typical large-scale macroeconometric models, which are hidden within a fundamental value by the time they reached $140 per barrel in June
complex framework (the black box). According to Sims (1980), SVAR 2008. The outbreak of COVID-19 is largely responsible for the significant
models offer a more methodical way to impose restrictions, which may negative supply and oil-specific demand shocks in early 2020. The
allow the researcher to discover empirical regularities otherwise missed pandemic had major repercussions on oil supplies because of dramatic
by the methodologies that are commonly used. In contrast, users of declines in output and demand (Sharif et al., 2020).
large-scale macroeconometric models can easily alter their results by
making subjective ex post decisions. 3.4. Time varying granger causality
Moreover, the relative popularity of SVAR models is due to the lack
of a general agreement for the correct specification for a simultaneous A major drawback of the VAR method is that without adjustment, it
equation models (such as those produced by the Cowles Commission). ignores nonlinearities, drifts, and breaks in parameter values (Stock and
As such, the SVAR approach by Kilian (2009) that we adopt for this Watson, 2001). This problem is especially true for oil markets, which
paper can be found in many related papers (e.g., Alsalman and Karaki, have exhibited a great deal of volatility and substantial upward drifts
2019; Atems et al., 2015; Baek and Yoon, 2022; Basher et al., 2016; Hu over the last decade (Mensi et al., 2014), resulting in structural breaks in
et al., 2018; Jiménez-Rodríguez, 2022; Lee et al., 2022; Li et al., 2022; time series properties.
Qin, 2020; Wang et al., 2014). To account for the limitation in the VAR model and conduct
robustness checks, we examine the time-varying causality between
3.3. Structural VAR for decomposed oil price shocks inflation and decomposed oil price shocks for the ASEAN5+3 countries
utilizing the causality approach developed by Shi et al. (2020). Given
We define the structural VAR (SVAR) in Eq. (1) below: that causality tests rely directly on time periods, different time series
samples may lead to different results. Therefore, the dynamic Granger
A0 yt = A(L)yt− 1 + εt , (1)
causality method across time can capture the dynamic behavior of the
where yt denotes (i) the world’s oil supply, (ii) the IGREA, and (iii) the oil markets and determine whether oil price shocks Granger-cause
real oil price; εt is a vector of structural innovations with an economic inflation over time.
interpretation that are serially and mutually uncorrelated. Exclusionary Shi et al. (2020) utilized three dynamic causality algorithms: forward
restrictions are imposed on A−0 1 in et = A−0 1 εt , where et is a vector of recursive causality, rolling causality, and recursive evolving causality.
errors in a VAR: They used these algorithms to identify phases of instability. The meth­
odology relies on a VAR model to obtain Wald test statistics, which are
yt = A−0 1 A(L)yt− 1 + A−0 1 εt . (2) then calculated exhaustively via recursive calculations. Equation (4)
illustrates our method.
We consider three structural shocks. The oil supply shock, εst , de­
notes shocks to global crude oil supplies, εyt denotes global demand ∑
m ∑
m
y1t = ∅(1)
0 + ∅(1)
1k y1t− k + ∅(1)
2k y2t− k + ε1t , (4)
shocks for all industrial goods due to real economic activity globally k=1 k=1
(henceforth, aggregate demand shock), and εot indicates shocks to oil-
market-specific demands, which we refer to as oil-specific demand
shocks. In Eq. (2), the identification of A−0 1 is obtained by enforcing the
following exclusion restrictions: 8
The oil-specific demand shock is referred to as the “other demand shock” or
⎛ ⎞
⎛ ⎞ ⎡ ⎤ oil supply shock the “speculative demand shock” by Baumeister and Peersman (2013) and Kilian
est a11 0 0 ⎜ ε1t ⎟ and Murphy (2014), respectively. The idea here is to consider them indepen­
et = ⎝ eyt ⎠ = ⎣ a21 a22 0 ⎦⎜ ⎝ ε2t
agregate demand shock ⎟,
⎠ (3) dently of the prevailing economic demand. More accurately, these shocks result
eot a31 a32 a33 ε oll− specific demand shock
from shifts in economic sentiment, meaning, agents’ forward-looking behav­
3t
iour, which leads to a stronger demand for oil. Furthermore, the sign constraints
Eq. (3) has three implications. First, the oil supply shock does not used by most studies are practically identical.

5
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

Table 1
Descriptive statistics.
Mean Max Min SD Skewness Kurtosis JB PP

Economic activity index 0.021 1.880 − 1.633 0.588 0.728 3.907 51.900*** − 3.08***
Real oil price 3.728 4.851 2.569 0.499 0.262 2.193 16.334*** − 10.73***
Oil supply 0.001 0.045 − 0.146 0.012 − 4.289 52.833 45064.9*** − 21.75***
Inflation Malaysia 2.486 8.515 − 2.890 1.617 0.023 4.783 56.059*** − 14.22***
Inflation Indonesia 8.751 82.416 − 1.165 10.861 4.815 28.612 13195.7*** − 8.71***
Inflation Thailand 2.940 10.576 − 4.426 2.474 0.202 3.056 2.928*** − 14.69***
Inflation Singapore 1.675 7.572 − 1.566 1.794 0.821 3.434 50.831*** − 20.43***
Inflation Philippines 5.871 21.111 − 0.649 4.064 1.403 5.215 225.224*** − 12.99***
Inflation Japan 0.499 3.924 − 2.515 1.214 0.796 3.361 46.954*** − 17.56***
Inflation Korea 3.574 10.134 − 0.426 2.361 0.738 2.906 38.601*** − 15.11***
Inflation China 4.840 28.400 − 2.200 6.546 2.024 6.597 516.801*** − 15.06***

Note: SD, JB, and PP are abbreviations for standard deviation, the Jarque Bera, and the Phillips-Perron tests, respectively. The uppercase asterisk *, ** and ***
represent significance level at 10%, 5% and 1%, respectively.

where y1t and y2t are the oil price shocks and inflation variables, signs of non-normality.
respectively. A Wald test is employed to assess the joint significance of Table 2 reports the results of Perron (1997) structural break unit root
test and the BDS nonlinearity test suggested by Brock et al. (1996). As
∅1k (k = 1, …, m) of the null hypothesis of no Granger causality from y1
(2)
can be seen, the findings from Table 2 demonstrate that all series exhibit
(oil shocks) to y2 (inflation). According to Shi et al. (2018, 2020), a stationarity at the first difference level, which support the findings from
rolling window estimation of the standard Wald statistics outperforms the standard unit root test (PP in Table 1). The BDS test indicate that all
other techniques. They made this determination by comparing several variables have non-linear dependence, thereby confirming the chaotic
statistical measures against a series of Granger causality test statistics, behavior of the time series. This finding supports our use of the SVAR
one for each period under consideration. Shi et al. (2020) determined model to estimate the asymmetric effect of oil shocks on inflation for all
that the best result is obtained by using the recursive evolving window ASEAN5+3 countries.
method. Wf2 (f1 ) denotes the Wald statistics across [f1 , f2 ] with a per­
centage of the sample size equal to fw = f2 − f1 ≥ f0 . The sup Wald sta­
tistics are written as follows: 3.6. The effects of oil shocks on inflation
s { }
Wf (f0 ) = wf2 (f1 ) , (5)
(f1 , f2 ) ∈ ̂0, f2 = f To determine the impact of the three decomposed oil price shocks on
inflation, we computed the impulse response functions for all inflation
where 0 ̂ = {(f1 , f2 ) : 0 < f0 + f1 ≤ f2 ≤ 1, and 0 ≤ f1 ≤ 1 − f0 }. In this rates for the ASEAN5+3 economies. Based on Eq. (8), we plot the cu­
mulative impulse response functions with one standard deviation shock
setting, ̂f e and ̂f f are used to signify the initial estimated chronological
for up to 12 months ahead, along with their 95% confidence intervals:
observations with statistics that either surpass or fall short of the critical
values for the beginning and end of the causal relationship, respectively. ∑
12 ∑
12 ∑
12
CPIt = β0 + β1i εs,t− i + β2i εy,t− i + β3i εo,t− i + et , (8)
inf { }
(6)
i=0 i=0 i=0
̂f e = f : SWf (f0 ) > tcv ,
f ε[f0 , 1]
where CPIt indicates the change (%) in the inflation series and εst , εyt , and
and εot, denote the structural oil supply shocks (henceforth, SS), aggregate
demand shocks (henceforth, DD), and oil-specific demand shocks
inf { }
̂f f = f : SWf (f0 ) > tcv , (7) (henceforth, OIL), respectively. Using the cumulative impulse functions
̂
f ε[ f e , 1] from the SVAR model, we can estimate the reactions of inflation to the
various crude oil market shocks using a dynamic OLS technique. To
where tcv denotes the critical values of the SWf statistics. Since the
ensure robustness of our estimates, we follow Atems et al. (2015) by
recursive evolving window is the best approach of the three time-
implementing 1000 replications for each cumulative impulse functions
varying causality algorithms, we use it to investigate the causality be­
for symmetric and asymmetric analyses.
tween oil price shocks and inflation for the ASEAN5+3.
Fig. 2 depicts the responses of inflation to the three types of oil
shocks derived from Eq. (8), for up to 12 months ahead. Panel A of Fig. 2
3.5. Descriptive statistics illustrates the responses before the COVID-19 outbreak (January
1987–February 2020), and Panel B indicates the responses in the full
Kilian (2009) identified three oil price shocks from the structural sample including the COVID-19 outbreak (January 1987–April 2022). In
VAR model based on the log return of world oil supplies, the detrended general, all structural oil shocks have an insignificant impact on infla­
IGREA, and the log-level of real oil prices. In accordance with Basher tion in the ASEAN5+3 countries before the COVID-19 outbreak. How­
et al. (2016), we use the same constructs of the variables. For inflation, ever, when we include the COVID-19 data, these shocks have a
we convert all series into first differences to ensure the stationarity of the significant impact on inflation in seven ASEAN5+3 countries, origi­
data. nating specifically from OIL.
A descriptive analysis is summarized in Table 1. As can be seen, the Fig. 2 demonstrates that for Malaysia, SS prompt inflation to rise
mean inflation level is highest for Indonesia at 8.7%, followed by the marginally before and during the pandemic. However, the impact is not
Philippines at 5.8%. Japan has the lowest mean inflation level of 0.5%. statistically significant. OIL and DD are also statistically insignificant
All the series have skewness values that are not zero and kurtosis values before COVID-19. Nevertheless, when we include data during the
that are typically more than 3, providing indication of asymmetric and COVID-19 period, DD have a positively significant impact on inflation.
leptokurtic distributions. Based on Jarque-Bera test, the variables show One reasonable explanation for this relationship is that after the massive

6
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

Table 2 government has been promoting ethanol for more than a decade to
Unit root and BDS nonlinearity test. reduce imports of petroleum (Silalertruksa and Gheewala, 2009).
Series PP unit root break test BDS nonlinear test Singapore is one of the top global export refining centers. The oil
industry has led to advancements in the chemical industry and oil and
test statistics test statistics
gas equipment manufacturing. The surge in crude oil prices resulted in
Economic activity index − 5.41263** 0.176178*** an increase in the prices of distillates and related products. It can be
Real oil price 13.6553*** 0.000013**
assumed that these two processes moderated, if not balanced, each
−
Oil supply − 21.1319*** 0.012147***
Inflation Malaysia − 15.95714*** 0.027474*** other.
Inflation Indonesia − 7.44361*** 0.056829*** After two oil crises in the 1970s, Japan tried to diversify its energy
Inflation Thailand − 15.1718*** 0.030622*** resources to increase its security. Despite the plan to close all nuclear
Inflation Singapore 9.943267*** 0.022456***
−
reactors, as of 2022, about 10.4% of the country’s electricity still comes
Inflation Philippines − 13.64118*** 0.041952***
Inflation Japan − 18.09604*** − 0.004964* from nuclear energy. Japan is a world leader in electricity production
Inflation Korea − 14.41976*** 0.017707*** from renewable sources. If in 2001, 57% of Japan’s energy consumption
Inflation China − 9.145806*** 0.03454*** was from oil, in 2021 it dropped to only 37%.10 Nevertheless, even in
Table 2 reports the test statistics results of Perron (1997) structural break unit Japan, as in Thailand, local demand did not recover as quickly as in
root test and the BDS nonlinearity test suggested by Brock et al. (1996). The other countries after the disappearance of COVID-19. A striking example
uppercase asterisk *, ** and *** represent significance level at 10%, 5% and 1%, is tourism to Japan, which remained very limited even at the end of
respectively. 2022.
Finally, Fig. 2 illustrates that SS do not exhibit a corresponding sig­
drop in global oil prices in April 2020, oil prices rebounded and nificance in terms of their immediate effects on inflation (except for
increased by more than 300% between January 2021 and April 2022.9 Indonesia). This result is evident in both sample periods before and
Given that oil prices and Malaysian inflation tend to be positively including the outbreak of COVID-19. In line with Kilian (2009), our
correlated (Majuca, 2020), rising oil prices have an immediate impact findings indicate that shocks to OIL have a greater impact on output than
on transportation costs. The latter, which makes up 14.6% of the do shocks to SS. In contrast, the latter have no impact on inflation.
Malaysian CPI basket, has a solid effect on Malaysia’s overall inflation
rate.
3.7. Specifications of asymmetric effects
In the case of Indonesia, the impulse response shows that OIL and DD
are inflationary before COVID-19, after which the response is insignifi­
After establishing that inflation reacts distinctively to oil price shocks
cant during the pandemic (Fig. 2). This result might be due to the
in two sample periods, one before and the other including COVID-19, we
collapse in domestic demand for oil in Indonesia, mainly due to the large
proceed to examine the direction of these reactions. Accordingly, we
number of daily cases of COVID-19 infections in mid-2021. At that point,
estimate the following function based on Atems et al. (2015).
Indonesia was the epicenter of the coronavirus in Asia, surpassing India
in terms of the largest number of new cases daily per million people. ∑
12 ∑
12 ∑
12 ∑
12

Fig. 2 also reveals that the responses of inflation to OIL in Thailand, CPIt = β0 + β1i ε+
s,t− i + β2i ε−s,t− i + β3i ε+
y,t− i + β4i ε−y,t− i

Singapore, and Japan during the spread of COVID-19 are identical. (9)
i=0 i=0 i=0 i=0

Specifically, OIL typically have significant inflationary effects, and these ∑

12
∑
12
+ β5i +
ε + β6i −o,t− i
ε + et
effects tend to last for a full year after the initial shock. For Korea and
o,t− i
i=0 i=0
China, we show that DD are insignificant with their impact before the
COVID-19 outbreak or when we include data during the pandemic and where ε+
st , εyt , and εot denote positive values from the shocks and
+ +

the Russian invasion of Ukraine. Our results for Korea are consistent 0 otherwise, and ε−st , ε−yt , and ε−ot denote negative values from the shocks
with Cuñado et al. (2015), who attribute their findings to the heavily and 0 otherwise.
export-dependent industry structures in Korea. Without a discernible Fig. 3 depicts the impulse response functions estimated from Eq. (9)
increase in global demand, an increase in the price of oil can quickly with a single standard deviation shock. The figure supports this form of
cause a slump because it drives up the price of some goods. According to asymmetry in several ways. First, there is little evidence that shocks to
Chen et al. (2020), the pass-through effects of oil price shocks on Chinese the world’s oil supply had any (asymmetric) effects on inflation over any
inflation have weakened since the global financial crisis. One explana­ time horizon, either before or during the recent pandemic. However,
tion is the adjustment the Chinese have made in their economic structure positive SS during COVID-19 reduced inflation significantly in
and the steps they have taken to promote domestic demand following Singapore, Japan, and China between the fourth and sixth month after
the global financial crisis. These changes have focused on increasing the shock. This positive supply shock-induced downward inflationary
domestic demand with the aim of reducing the likelihood of internal and pressure is primarily due to COVID-19’s initial outbreak, which forced
external shocks, such as rising oil prices. China to halt its global trade. In reaction, the rest of the world limited or
The similarity in our results makes it tempting to look for similarities suspended air travel to and from China (Aziz et al., 2022), causing a 435
between the economies of the three countries. Although Thailand is a kb/d contraction in the demand for oil in 10 years.11 Apart from the
fast-growing country and an importer of oil, it is also exceptional in abrupt drop in the demand for oil, the glut in oil supplies was exacer­
several regards. First, it has a very large tourism sector, a sector that was bated during the Russia-Saudi oil price war in March 2020. The price of a
deeply and negatively affected by COVID-19 until and including the first barrel of West Texas crude plunged to zero U.S. dollars in mid-April
quarter of 2022. Second, the Thai industrial sector is mainly light in­ 2020, mainly because the world was running out of storage facilities
dustry, which consumes relatively little energy. The population of and suppliers had to pay to get rid of it.12
Thailand constitutes 0.93% of the global population, but accounts for Second, when we include the COVID-19 pandemic, the negative DD
only 0.69% of the world’s energy consumption. Additionally, the

10
https://www.enecho.meti.go.jp/en/category/special/article/detail_171.
9
Lockdowns across the globe significantly reduced the demand for energy html.
11
products. The Organization of the Petroleum Exporting Countries and its https://www.iea.org/reports/oil-market-reportfebruary-2020
12
partners, commonly known as OPEC+, were unable to come to an agreement on https://www.ig.com/en/news-and-trade-ideas/what-do-negative-oil-price
production cutbacks, which made the slump worse (Maojun and Ghani, 2020). s-mean–200507

7
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

Fig. 2. Response of inflation to OSS, ODD and OIL, respectively.

8
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

Fig. 3. Response of inflation to asymmetric shocks (cont’d).

9
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

Fig. 3. (continued).

10
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

Fig. 3. (continued).

11
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

Fig. 3. (continued).

12
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

affect inflation in the majority of the ASEAN5+3 countries. In contrast, impact of OIL during the pandemic.
positive DD have no significant impact on inflation. Specifically, the
impact on inflation due to negative DD in Malaysia, Thailand, the 4. Discussion
Philippines, Japan, and Korea begins to rise in the third or fourth month
and remains positive permanently. This result remains despite a The oil-inflation relationship is a central economic concept. In the
decrease in global output due to reintroduced lockdowns in China and seminal work of Darby (1982) it is demonstrated how Lucas-Barro’s
the slowdown in OECD countries.13 Thus, negative DD and the Russian theoretical model behaved under the influence of the rise in oil prices
invasion of Ukraine have kept the pressure on inflation high in many during the 1970s on the developed countries’ economies at that time.
countries, including the ASEAN5+3. When oil prices are nearly the exclusive indicator of energy prices, any
Third, when we add the COVID-19 data to the sample period, both supply model that uses energy as one of the production parameters will
positive and negative OIL cause inflation to go up in all of the cause a movement of the supply curve if oil prices increase. The resulting
ASEAN5+3 countries, except Indonesia. For Indonesia, positive OIL is new equilibrium will have higher prices, that is, inflation.
negatively significant to inflation, which infers that an increase in OIL is The impact of oil price shocks on macroeconomic events has been
translated to a lower price level. Result for negative OIL on the contrary extensively studied and several theoretical models have been developed
is not statistically significant. Our result in Indonesia is closely aligned (e.g., Rasche and Tatom, 1981; Hamilton, 1988). Various reasons exist
with Razmi et al. (2016). Razmi et al. (2016) attribute this reducing as to why an oil shock can impact macroeconomic variables, with some
impact of oil price on Indonesia’s inflation to the country’s status as an necessitating a non-linear approach to interpreting the relationship. For
oil exporter. In contrast, our findings for the rest of the ASEAN5+3 example, an oil shock may decrease aggregate demand by redistributing
countries for OIL echo the outcomes of Cuñado et al. (2015). OIL (also income between net oil-importing and exporting countries, or it may
known as speculative demand shocks) is inextricably linked to oil lead firms to consume less energy, which in turn reduces aggregate
market uncertainty (Cuñado et al., 2015). The recent Russia-Ukraine supply by decreasing capital and labor productivity. Such a decrease in
conflict that has elevated geopolitical risks in the financial and com­ factor productivity could lead to reduced wages and a decrease in po­
modities markets may have exacerbated uncertainty in the oil markets tential output if workers choose to withdraw their labor voluntarily,
(Umar et al., 2022). This uncertainty has resulted in spikes in oil prices thereby exacerbating the direct effects of lower productivity.
and hence, inflation. In fact, a negative OIL stemming from the elevated Research has shown that oil price shocks have had an impact on both
geopolitical risk of this war may have a limited deflationary influence on output and inflation in European and North American countries, as seen
price levels.14 According to Friedman (1977), if uncertainty about in studies by Hamilton (1983, 2003), Hooker (2002), Huntington
inflation is to blame for falling output, then uncertainty about oil prices (1998), and Kahn and Hampton (1990). However, since the 1980s, there
might also be to blame for both the price levels and decreasing output. have been no significant and sustained changes in oil prices. In this
period, ASEAN countries, which were not industrialized thirty years ago,
3.8. Robustness tests have become integrated into the industrialized world. Despite this body
of research, there are still many questions that remain unresolved, and
To assess for robustness, we run a recursive time-varying Granger our work aims to address some of these questions.
causality test (TVGC) to verify if oil price shocks and inflation have time- Firstly, past empirical evidence on oil price shocks has shown that
varying causality before and including the COVID-19 outbreak. they do not Granger-cause macroeconomic variables beyond 1985,
Although inflation may influence oil price shocks, our focus is the causal following the market fall of that year. This suggests that the relationship
results or the impact of shocks in terms of decomposed oil prices on between oil prices and the macroeconomy may be non-linear (Hamilton,
inflation in the ASEAN5+3. As a result, we do not examine the causality 1996; Lee et al., 1995).
of inflation on oil price shocks. The null hypothesis for each of the Secondly, in the context of our paper, an oil shock may have varying
decomposed shocks to the price of oil is that they do not Granger-cause impacts on different country clusters (NAFTA, EU, or ASEAN) and
inflation. within each country in these groups. These impacts can be influenced by
The full sample includes 424 months. The rolling window covers 72 factors such as the countries’ sectoral composition, involvement in other
months, which is around 15% of the length of the sample. This pro­ aspects of the oil industry, degree of oil net importation, demand flex­
cedure allows us to better track and monitor the dynamic evolution of ibility, and differential tax structures.
Granger causality for the relationship between oil prices and inflation Consistent with Alvarez et al. (2011), Kilian and Lewis (2011), and
across different periods. We calculate the right-tail Monte Carlo critical Tiwari et al. (2019), we found that oil price shocks had a limited impact
values for the 90th and 95th percentiles based on the bootstrapping on inflation before the recent pandemic. Apart from the considerable
advocated by Shi et al. (2018, 2020) using 1000 replications. The lag variation in crude oil prices,15 our findings corroborate evidence of
orders regarding the corresponding sequences are defined by the Akaike declining inflation in the 50 years preceding the pandemic in emerging
information criteria (AIC). Fig. 4 plots the TVGC from supply shocks, market and developing economies (EMDEs hereafter). The declines are
demand shocks, and oil-specific demand shocks to inflation. Panel A attributed to a combination of factors, including a rise in globalization,
illustrates the causality of supply shocks, Panel B does the same for the the liberalization of financial, labor, and product markets, as well as a
demand shocks, and Panel C presents the results for the oil-specific de­ greater emphasis on price stability by monetary authorities (Ha et al.,
mand shocks to inflation. 2019a). In fact, the 1990s and 2000s saw such a sharp decline in infla­
Overall, we find evidence supporting our main findings from the tion that by the early 2000s, some industrialized economies were seri­
SVAR in Fig. 4 for most ASEAN5+3 countries. Specifically, SS do not ously concerned about deflation. To illustrate, inflation in industrialized
Granger-cause inflation for six countries (except for Singapore and countries dropped from 15.3% to 1.3% between 1974 and 2019, with
Japan) during the full sample period. For DD, there is no evidence of EMDEs recording the largest decline, from 17.5% to 2.6% during the
Granger causality in Malaysia, Singapore, Korea, and China during same period (World Bank, 2022b). The World Bank (2022b) also re­
COVID-19. However, there is evidence that OIL Granger-cause inflation ported that from 2012 to 2019, advanced economies with an
in Malaysia, Thailand, Singapore, Japan, Korea, and China. This result
generally conforms with the previous findings showing the significant
15
Between January 2015 and February 2020, crude oil prices averaged $52
per barrel. In contrast, between January 2021 and February 2022, oil was
13
https://www.iea.org/reports/oil-market-report-september-2022 trading at about $70 a barrel, before spiking to $100 per barrel in March 2022
14
Since March 2022, oil price barrel has soared to more than $100. following Russia’s invasion of Ukraine.

13
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

Fig. 4. Time-varying Granger causality of decomposed oil shocks to inflation for the ASEAN5+3. (Panel A) Oil supply shocks Granger-cause inflation; (Panel B)
Demand shocks Granger-cause inflation (Panel C) Oil-Specific demand shocks Granger-cause inflation.

14
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

Fig. 4. (continued).

inflation-targeting regime saw low levels of inflation, while more than World Bank (2022a) projects that the global inflation rate will remain
half of inflation-targeting EMDEs were able to keep inflation rates within above its average, despite a forecasted decline to about 3% in mid-2023.
target ranges. In addition, Ha et al. (2019b) demonstrated that a 50% increase in crude
In contrast, when we include data for the period during the COVID- oil correlates with a 4.4% increase in inflation after two years. In short,
19 pandemic, we show that both OIL and DD tend to increase inflation, these inflationary effects originating from OIL are consistent with Bau­
with several of the ASEAN5+3 economies exhibiting statistically sig­ meister and Peersman (2013) and Kilian and Murphy (2014), who
nificant impacts. Our findings from the impulse response are consistent suggested that OIL are “speculative shocks” that react to traders’ per­
with recent global economic developments. While DD cause inflation to ceptions. Often, they are not related to fundamentals, but rather to oil
increase after the second or third month, OIL prompt a positive response market uncertainty (Cuñado et al., 2015).
from the first month, and their effects last indefinitely. In comparison with prior research on ASEAN5+3 economies, our
We confirm the World Bank’s view (2022b) that the current situation results can be summarized as follows. Firstly, apart from Indonesia,
is analogous to the oil price shocks of 1973 and 1979–1980. New positive oil shocks originating from demand and oil-specific raise
pandemic outbreaks, mobility restrictions in China, and Russia’s inva­ inflation for all ASEAN5+3 nations. These findings conform to Basnet
sion of Ukraine are all factors that contribute to higher crude oil prices,16 and Upadhyaya (2015), Razmi et al. (2016), and Husaini and Lean
which in turn are driving up inflation in many countries. In fact, the (2021). Secondly, we find medium and long-term inflationary impacts
from increasing oil prices (from demand and oil-specific) in most
ASEAN5+3 economies, contradicting the short-term inflationary effects
found by Basnet and Upadhyaya (2015) and Razmi et al. (2016).
16
In two years, (April 2020 to April 2022) crude oil prices rose by 350% in Thirdly, we observe significant asymmetry between oil prices and
nominal terms, the most for any comparable two-year period since 1973 (World inflation for all ASEAN5+3 countries, except Indonesia, echoing the
Bank, 2022b).

15
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

findings of Husaini and Lean (2021) and Li and Guo (2022). Specifically, curation; Methodology; Formal analysis; Writing – original draft;
our findings on the inflationary effects resulting from negative oil price Writing – review & editing.
shocks (from demand and oil-specific) are consistent with Li and Guo’s Ido Kallir: Investigation; Writing – review & editing.
(2022) findings on China’s economy.
Declaration of competing interest
5. Conclusion
The authors declare that they have no known competing financial
In contribution to the literature, we utilized Kilian’s (2009) meth­
interests or personal relationships that could have appeared to influence
odology to examine how oil price shocks impacted inflation in the
the work reported in this paper.
ASEAN5+3 countries, before and during the COVID-19 pandemic. Using
35 years’ worth of data, the results hint that the outcomes of combined
Data availability
OIL and DD on inflation are significant for most of the ASEAN5+3
countries when we include data from the COVID-19 period. In contrast,
Data will be made available on request.
SS have a relatively mild impact on inflation. The impulse responses for
Malaysia, Singapore, Thailand, and Japan demonstrate that when we
References
include COVID-19 pandemic data, the effect of OIL and DD on inflation
is positive and statistically significant. However, inflation in China, Aharon, D.Y., Qadan, M., 2022. Infection, invasion, and inflation: recent lessons. Finance
Korea, and the Philippines remains relatively stable prior to and Res. Lett. 50, 103307.
throughout the COVID-19 spread, due to government policies designed Ajmal, M., Isha, A.S.N., Nordin, S.M., Rasheed, S., Al-Mekhlafi, A.B.A., Naji, G.M.A.,
2022. Safety management and safety outcomes in oil and gas industry in Malaysia:
to minimize the effect and risk of external shocks. In addition, for the safety compliance as a mediator. Process Saf. Prog. 41, S10–S16.
majority of the ASEAN5+3 countries, SS have little to no impact on Alsalman, Z.N., Karaki, M.B., 2019. Oil prices and personal consumption expenditures:
inflation. In contrast, inflation is significantly influenced by OIL, does the source of the shock matter? Oxf. Bull. Econ. Stat. 81 (2), 250–270.
Alvarez, L.J., Hurtado, S., Sánchez, I., Thomas, C., 2011. The impact of oil price changes
particularly when we include the COVID-19 pandemic period. on Spanish and euro area consumer price inflation. Econ. Modell. 28 (1–2), 422–431.
Our findings underscore those shocks in crude fuel prices, particu­ Atems, B., Kapper, D., Lam, E., 2015. Do exchange rates respond asymmetrically to
larly those resulting from the reduction in supply, can be quick and shocks in the crude oil market? Energy Econ. 49, 227–238.
Aziz, M.I.A., Azhari, A., Mobin, M.A., 2022. Detecting asset price bubbles during the
direct triggers for the acceleration of inflation. The last twenty years
Covid-19 crisis and its implications: evidence from the stock and oil market. Int. J
have shown the industrialized world how dangerous inflation can be, Bank. Finance 17 (2), 91–114.
how fast and seemingly unpredictable its acceleration can be, and the Baek, J., Yoon, J.H., 2022. Do macroeconomic activities respond differently to oil price
price it exacts from everyone. Hundreds of millions of households are shocks? New evidence from Indonesia. Econ. Anal. Pol. 76, 852–862.
Basher, S.A., Haug, A.A., Sadorsky, P., 2016. The impact of oil shocks on exchange rates:
experiencing a rapid increase in their living expenses due to general a Markov-switching approach. Energy Econ. 54, 11–23.
inflation and, in particular, in heating costs due to the increase in energy Basnet, H.C., Upadhyaya, K.P., 2015. Impact of oil price shocks on output, inflation and
prices. They are also dealing with the rapid rise in interest rates, the sole the real exchange rate: evidence from selected ASEAN countries. Appl. Econ. 47
(29), 3078–3091.
remedy that central banks seem to have for dealing with inflation. Baumeister, C., Kilian, L., 2016. Lower oil prices and the US economy: is this time
The special challenge facing the ASEAN5+3 countries stem from different? Brookings Pap. Econ. Activ. 2016 (2), 287–357.
their commitment to returning to significant growth in their GDP (Japan Baumeister, C., Kilian, L., Zhou, X., 2018. Is the discretionary income effect of oil price
shocks a hoax? Energy J. 39 (S2), 117–138.
and South Korea) or to maintaining it over time in the face of substantial Baumeister, C., Peersman, G., 2013. The role of time-varying price elasticities in
demographic changes. The rapid increase in life expectancy (Japan and accounting for volatility changes in the crude oil market. J. Appl. Econom. 28 (7),
Singapore lead the world in this regard) throughout the ASEAN5+3 has 1087–1109.
Brock, W.A., Dechert, W.D., Scheinkman, J.A., LeBaron, B., 1996. A test for
obscured the long-term trend of changes in the nature of the population. independence based on the correlation dimension. Econom. Rev. 15 (3), 197–235.
All of the countries in this group except Malaysia have a fertility rate less Çatık, A.N., Önder, A.Ö., 2013. An asymmetric analysis of the relationship between oil
than 2. In Malaysia, it is 2.1, in South Korea, less than 1, in Japan and prices and output: the case of Turkey. Econ. Modell. 33, 884–892.
Chen, J., Zhu, X., Li, H., 2020. The pass-through effects of oil price shocks on China’s
China, 1.1, and even in Thailand, it is less than 1.5. The weakness of the
inflation: a time-varying analysis. Energy Econ. 86, 104695.
natural growth mechanisms in the ASEAN5+3 countries make them Choi, S., Furceri, D., Loungani, P., Mishra, S., Poplawski-Ribeiro, M., 2018. Oil price and
particularly sensitive to any inflationary shocks. Thus, we provide a inflation dynamics: evidence from advanced and developing economies. J. Int.
practical contribution to the literature by emphasizing the importance of Money Finance 82, 71–96.
Chou, K.W., Lin, P.C., 2013. Oil price shocks and producer prices in Taiwan: an
policymakers paying more attention, and earlier, to any shock to oil application of non-linear error-correction models. J. Chin. Econ. Bus. Stud. 11 (1),
prices and supplies. Such an indicator should prompt them to take im­ 59–72.
mediate measures to moderate the shock if it cannot be avoided, Cologni, A., Manera, M., 2008. Oil prices, inflation and interest rates in a structural
cointegrated VAR model for the G-7 countries. Energy Econ. 30 (3), 856–888.
whether by establishing strategic oil reserves or by more aggressive Cuñado, Juncal, Jo, Soojin, Pérez de Gracia, Fernando, 2015. Macroeconomic impacts of
actions to find substitutes. oil price shocks in Asian economies. Energy Pol. 86 (C), 867–879.
Our study has several limitations which future studies could build Darby, M.R., 1982. The price of oil and world inflation and recession. Am. Econ. Rev. 72
(4), 738–751.
upon. Firstly, the inflation data is gathered monthly, skipping potential Edelstein, P., Kilian, L., 2007. The response of business fixed investment to changes in
fluctuations that might be observed within shorter time settings. Thus, it energy prices: a test of some hypotheses about the transmission of energy price
would be useful to employ derived inflation measures from a more shocks. B E J. Macroecon. 7 (1).
Edelstein, P., Kilian, L., 2009. How sensitive are consumer expenditures to retail energy
frequent interval, such as weekly or daily observations. Secondly, while prices? J. Monetary Econ. 56 (6), 766–779.
we focus on ex-post inflation, future works may test the oil-inflation Elsayed, A.H., Hammoudeh, S., Sousa, R.M., 2021. Inflation synchronization among the
nexus using ex-ante inflation. Lastly, we focus on a specific class of G7and China: the important role of oil inflation. Energy Econ. 100, 105332.
Farzanegan, M.R., Markwardt, G., 2009. The effects of oil price shocks on the Iranian
ASEAN countries, whereas future studies could investigate the connec­
economy. Energy Econ. 31 (1), 134–151.
tion between oil and inflation in different country groups or compare oil- Friedman, J.W., 1977. Oligopoly and the Theory of Games. North Holland, Amsterdam.
importing countries with oil-exporting ones. Ghosh, S., Kanjilal, K., 2014. Oil price shocks on Indian economy: evidence from Toda
Yamamoto and Markov regime-switching VAR. Macroeconomics Finance Emerg.
Market Economies 7 (1), 122–139.
Author contribution statement Gómez-Loscos, A., Gadea, M.D., Montañés, A., 2012. Economic growth, inflation and oil
shocks: are the 1970s coming back? Appl. Econ. 44 (35), 4575–4589.
David Y. Aharon: Conceptualization; Investigation; Resources; Guesmi, K., Kaabia, O., Abid, I., 2017. ASEAN plus three stock markets integration.
J. Quant. Econ. 15 (3), 565–581.
Formal analysis; Writing – original draft; Writing – review & editing. Ha, J., Ivanova, A., Ohnsorge, F., Unsal, D.F., 2019a. Inflation: Concepts, Evolution, and
Mukhriz Izraf Azman Aziz: Conceptualization; Investigation; Data Correlates. World Bank Policy Research Working Paper, 8738.

16
D.Y. Aharon et al. Resources Policy 82 (2023) 103573

Ha, J., Kose, M.A., Ohnsorge, F., Yilmazkuday, H., 2019b. Sources of inflation: global and Mensi, W., Beljid, M., Managi, S., 2014. Structural breaks and the time-varying levels of
domestic drivers. In: Inflation in Emerging and Developing Economies: Evolution, weak-form efficiency in crude oil markets: evidence from the Hurst exponent and
Drivers, and Policies, pp. 143–199. Shannon entropy methods. International Economics 140, 89–106.
Hamilton, J.D., 1983. Oil and the macroeconomy since world war II. J. Polit. Econ. 91 Mensi, W., Rehman, M.U., Al-Yahyaee, K.H., 2020. Time-frequency co-movements
(2), 228–248. between oil prices and interest rates: evidence from a wavelet-based approach.
Hamilton, J.D., 1988. A neoclassical model of unemployment and the business cycle. N. Am. J. Econ. Finance 51, 100836.
J. Polit. Econ. 96 (3), 593–617. Mork, K.A., 1989. Oil and the macroeconomy when prices go up and down: an extension
Hamilton, J.D., 1996. This is what happened to the oil price-macroeconomy relationship. of Hamilton’s results. J. Polit. Econ. 97 (3), 740–744.
J. Monetary Econ. 38 (2), 215–220. Nasir, M.A., Al-Emadi, A.A., Shahbaz, M., Hammoudeh, S., 2019. Importance of oil
Hamilton, J.D., 2003. What is an oil shock? J. Econom. 113 (2), 363–398. shocks and the GCC macroeconomy: a structural VAR analysis. Resour. Pol. 61,
Hamilton, J.D., 2009. Causes And Consequences Of the Oil Shock Of 2007-08 (No. 166–179.
W15002). National Bureau of Economic Research. Pal, D., Mitra, S.K., 2019. Asymmetric oil price transmission to the purchasing power of
Hammoudeh, S., Reboredo, J.C., 2018. Oil price dynamics and market-based inflation the US dollar: a multiple threshold NARDL modelling approach. Resour. Pol. 64,
expectations. Energy Econ. 75, 484–491. 101508.
Hooker, M.A., 1996. What happened to the oil price-macroeconomy relationship? Perron, P., 1997. Further evidence on breaking trend functions in macroeconomic
J. Monetary Econ. 38 (2), 195–213. variables. J. Econom. 80 (2), 355–385.
Hooker, M.A., 1999. Oil Macroeconomy Revisited. Available at SSRN 186014. Qin, X., 2020. Oil shocks and financial systemic stress: international evidence. Energy
Hooker, M.A., 2002. Are oil shocks inflationary? Asymmetric and nonlinear Econ. 92, 104945.
specifications versus changes in regime. J. Money Credit Bank. 540–561. Rasche, R.H., Tatom, J.A., 1981. Energy price shocks, aggregate supply and monetary
Hu, C., Liu, X., Pan, B., Chen, B., Xia, X., 2018. Asymmetric impact of oil price shock on policy: the theory and the international evidence. In Carnegie-Rochester Conference.
stock market in China: a combination analysis based on SVAR model and NARDL January Series Pub. Pol. 14, 9–93 (. North-Holland).
model. Emerg. Mark. Finance Trade 54 (8), 1693–1705. Razmi, F., Azali, M., Chin, L., Habibullah, M.S., 2016. The role of monetary transmission
Huntington, H.G., 1998. Crude oil prices and US economic performance: where does the channels in transmitting oil price shocks to prices in ASEAN-4 countries during pre-
asymmetry reside? Energy J. 19 (4). and post-global financial crisis. Energy 101, 581–591.
Husaini, D.H., Lean, H.H., 2021. Asymmetric impact of oil price and exchange rate on Rotemberg, J.J., Woodford, M., 1996. Imperfect competition and the effects of energy
disaggregation price inflation. Resour. Pol. 73, 102175. price increases on economic activity. J. Money Credit Bank. 28 (4), 549.
Husaini, D.H., Lean, H.H., 2022. Digitalization and energy sustainability in ASEAN. Sari, R., Soytas, U., 2006. The relationship between stock returns, crude oil prices,
Resour. Conserv. Recycl. 184, 106377. interest rates, and output: evidence from a developing economy. Empirical
Jiménez-Rodríguez, R., 2022. Oil Shocks and Global Economy. Energy Economics, Economics Lett. 5 (4), 205–220.
106373. Sek, S.K., Teo, X.Q., Wong, Y.N., 2015. A comparative study on the effects of oil price
Kahn, G.A., Hampton, R., 1990. Possible monetary policy responses to the Iraqi oil shock. changes on inflation. Procedia Econ. Finance 26, 630–636.
Federal Reserve Bank of Kansas City Economic Rev. 2, 19–32. Sharif, A., Aloui, C., Yarovaya, L., 2020. COVID-19 pandemic, oil prices, stock market,
Kilian, L., 2009. Not all oil price shocks are alike: disentangling demand and supply geopolitical risk and policy uncertainty nexus in the US economy: fresh evidence
shocks in the crude oil market. Am. Econ. Rev. 99 (3), 1053–1069. from the wavelet-based approach. Int. Rev. Financ. Anal. 70, 101496.
Kilian, L., 2019. Measuring global real economic activity: do recent critiques hold up to Shi, S., Hurn, S., Phillips, P.C., 2020. Causal change detection in possibly integrated
scrutiny? Econ. Lett. 178, 106–110. systems: revisiting the money–income relationship. J. Financ. Econom. 18 (1),
Kilian, L., Lewis, L.T., 2011. Does the Fed respond to oil price shocks? Econ. J. 121 (555), 158–180.
1047–1072. Shi, S., Phillips, P.C., Hurn, S., 2018. Change detection and the causal impact of the yield
Kilian, L., Murphy, D.P., 2014. The role of inventories and speculative trading in the curve. J. Time Anal. 39 (6), 966–987.
global market for crude oil. J. Appl. Econom. 29 (3), 454–478. Silalertruksa, T., Gheewala, S.H., 2009. Environmental sustainability assessment of bio-
Kilian, L., Park, C., 2009. The impact of oil price shocks on the US stock market. Int. ethanol production in Thailand. Energy 34 (11), 1933–1946.
Econ. Rev. 50 (4), 1267–1287. Sims, C.A., 1980. Comparison of interwar and postwar business cycles: monetarism
Kilian, L., Zhou, X., 2022. The impact of rising oil prices on US inflation and inflation reconsidered. Am. Econ. Rev. 70 (2), 250–257.
expectations in 2020–23. Energy Econ. 113, 106228. Stock, J.H., Watson, M.W., 2001. Vector autoregressions. J. Econ. Perspect. 15 (4),
Kisswani, K.M., 2017. Evaluating the GDP–energy consumption nexus for the ASEAN-5 101–115.
countries using nonlinear ARDL model. OPEC Energy Rev. 41 (4), 318–343. Tan, G.L., Diaz, J.F., 2020. Real GDP growth rates of the ASEAN region: evidence of
Kisswani, K.M., 2021. The dynamic links between oil prices and economic growth: recent spillovers and asymmetric volatility effects. Labuan Bull. Int. Bus. Finance (LBIBF)
evidence from nonlinear cointegration analysis for the ASEAN-5 countries. Emerg. 39–61.
Mark. Finance Trade 57 (11), 3153–3166. Tan, Y., Uprasen, U., 2021. Carbon neutrality potential of the ASEAN-5 countries:
LeBlanc, M., Chinn, M.D., 2004. Do high oil prices presage inflation? The evidence from implications from asymmetric effects of income inequality on renewable energy
G-5 countries. UC Santa Cruz Economics Working Paper 561, 04-04. consumption. J. Environ. Manag. 299, 113635.
Lee, K., Ni, S., Ratti, R.A., 1995. Oil shocks and the macroeconomy: the role of price Tang, Y., Xiao, X., Wahab, M.I.M., Ma, F., 2021. The role of oil futures intraday
variability. Energy J. 16 (4). information on predicting US stock market volatility. J. Manag. Sci. Eng. 6 (1),
Lee, C.C., Tang, H., Li, D., 2022. The roles of oil shocks and geopolitical uncertainties on 64–74.
China’s green bond returns. Econ. Anal. Pol. 74, 494–505. Tiwari, A.K., Cunado, J., Hatemi-J, A., Gupta, R., 2019. Oil price-inflation pass-through
Li, Y., Guo, J., 2022. The asymmetric impacts of oil price and shocks on inflation in in the United States over 1871 to 2018: a wavelet coherency analysis. Struct. Change
BRICS: a multiple threshold nonlinear ARDL model. Appl. Econ. 54 (12), 1377–1395. Econ. Dynam. 50, 51–55.
Li, Z., Huang, Z., Failler, P., 2022. Dynamic correlation between crude oil price and Umar, Z., Polat, O., Choi, S.Y., Teplova, T., 2022. The impact of the Russia-Ukraine
investor sentiment in China: heterogeneous and asymmetric effect. Energies 15 (3), conflict on the connectedness of financial markets. Finance Res. Lett., 102976
687. Wang, Y., Wu, C., Yang, L., 2014. Oil price shocks and agricultural commodity prices.
Lopez-Villavicencio, A., Pourroy, M., 2019. Inflation target and (a) symmetries in the oil Energy Econ. 44, 22–35.
price pass-through to inflation. Energy Econ. 80, 860–875. Wei, Y., 2019. Oil price shocks, economic policy uncertainty and China’s trade: a
Majuca, R.P., 2020. Assessing the Impact of Oil Prices on the Malaysian Economy. http quantitative structural analysis. N. Am. J. Econ. Finance 48, 20–31.
s://www.amro-asia.org/wp-content/uploads/2020/04/Impact-of-Oil-Prices-on-Mal Wen, F., Zhang, K., Gong, X., 2021. The effects of oil price shocks on inflation in the G7
aysia_final.pdf. countries. N. Am. J. Econ. Finance 57, 101391.
Maojun, P.A., Ghani, N.A.A., 2020. COVID-19: Impact on Inflation. https://www.bnm. World Bank, 2022a. Commodity Markets Outlook: the Impact of the War in Ukraine on
gov.my/documents/20124/1067369/p3ba1.pdf. Commodity Markets. April 2022.
World Bank, 2022b. Global Economic Prospects. The World Bank. June 2022.

17