ISSN 1988-088X

Department of Foundations of Economic Analysis II

University of the Basque Country
Avda. Lehendakari Aguirre 83
48015 Bilbao (SPAIN)
http://www.dfaeii.ehu.es

DFAE-II WP Series
2010-12

Alessandro Maravalle

The role of the terms of trade in the trade

channel of transmission of oil price shocks
 The role of the terms of trade in the trade
channel of propagation of oil price shocks

Alessandro Maravalle∗

Abstract
This paper highlights the role of the terms of trade in the trade chan-
nel of propagation of oil price shocks both empirically and theoretically.
Empirically, I show that oil price shocks have a large, persistent and sta-
tistically significant impact on the US terms of trade. Theoretically, I add
oil in the model by Corsetti and Pesenti (2005) and analyse under what
conditions the terms of trade plays a relevant role in the international
transmission of oil price shocks. With nominal price rigidities and full ex-
change rate pass-through positive oil price shocks depreciate the currency
of the oil importing country. The subsequent negative wealth effect adds
to the recessive effect of the supply channel and may strongly reduce the
consumption in the oil importing country economy. Without exchange
rate pass-through oil shocks transmit to the economy only through the
supply channel. The model suggests that a change in the exchange rate
pass-through might contribute to explain the evidence of a weaker impact
of oil price shocks on the macroeconomic activity in recent times.

Keywords: oil price shocks; macroeconomic interdependence; exchange

rate pass-through; propagation; transmission.

JEL classifications: F31, F41, Q43.

∗ University of the Basque Country, Economics Department, Avenida Lehendakari Aguirre,

83, E-48015 Bilbao, Spain. Email: [email protected]. Alessandro Maravalle ac-

knowledges financial support from the Departamento de Educación Universidades e Investi-
gación del Gobierno Vasco (IT-313-07).

1
1 Introduction
The oil literature has a hard time to match the evidence of a negative correla-
tion between the macroeconomic performance of oil-importing countries and oil
price shocks with empirically sound theoretical mechanisms of propagation of oil
price shocks. The empirical literature has provided quantitative measures of the
impact of oil price shocks on macroeconomic variables such as the growth rate
of GDP and the inflation rate (see, i.e., Hamilton 1983, Leduc and Sill 2004),
employment (see, i.e., Lee and Ni 2002, Davis 1987), the stock market (Kilian
and Park 2009), the terms of trade (Backus and Crucini 2000) and external
balances (see, i.e., Kilian et al 2009). The theoretical literature, in an attempt
to explain this evidence has investigated both demand and supply mechanisms
of propagation of oil shocks to the economic activity.1

Recent contributions of the literature have stressed that the demand chan-
nels, which operate through a reduction of the demand for goods and services
other than oil, are the key to explain the impact of oil price shocks to macroeco-
nomic performance (Nordhaus 2007, Hamilton 2008, Kilian 2008). This paper
follows this direction and focuses on the trade channel of transmission of oil
price shocks, which emphasizes the role of changes in the relative prices and
quantities of imports and exports in the propagation of oil price shocks to the
economy. In particular, as the dollar share of oil imports in total imports is
far larger and volatile than the dollar share of oil expenditures in total output,
data point out the potential relevance of the terms of trade in the mechanism
of adjustment to oil price shocks.2

Firstly, I provide empirical evidence that oil price shocks cause large varia-
tions in the terms of trade of the US, the most important oil-importing country.
To this purpose, I follow the methodology suggested by Kilian and Vigfusson
(2009) that allows me to obtain consistent estimates of the response of the terms
of trade to oil price shocks independently of the presence of asymmetric effects
of oil price increases and decreases (see, i.e., Mork 1989, Hooker 2002, Kilian
and Vigfusson 2009). My main result is that a positive oil price shock has a
large, persistent and statistically significant impact on the US terms of trade. I
also report limited evidence of asymmetric effects of oil price shocks on the US
terms of trade by mean of an impulse-response based test of symmetry. Finally,
I show that the response of the terms of trade to oil price shocks increases more
than proportionally with the size of the shock.
1 The literature on the supply and demand mechanisms of propagation of oil shocks to the

economy is far too vast to be surveyed here. Some of the seminal papers are Hamilton 1988
and 2008, Bernanke 1983, Davis 1987, Davis and Haltwanger 2001, Bresnahan and Ramey
1993, Bernanke et al 1996, Hamilton and Herrera 2004, Kim and Loungani 1992, Rotemberg
and Woodford 1996, Atkenson and Kehoe 1999 and Finn 2000.
2 Complementary to the trade channel there is another international channel of adjustment

to oil price shocks: the financial or valuation effect. The valuation channel works through
changes in the total asset returns differentials. On the role of the valuation effect in the
external adjustment of economies see Ghironi et al. (2007), Lane and Milesi Ferretti (2007),
Gourinchas and Rey (2007) and Deveraux and Sutherland (2008).

2
 Secondly, I analyse the theoretical conditions that permit positive oil price
shocks to produce, through the trade channel, a large negative impact on the
economy of oil importing countries. To this purpose, I add oil in the production
to the two-country general equilibrium model by Corsetti and Pesenti (2005)
and use it to analyse the international transmission of oil price shocks under
different price regimes and financial autarky. With respect to early (see, i.e.,
Fried and Schulze 1975, Dohner 1981 and Bruno and Sachs 1985) and recent
(see, i.e., Bodenstein et al 2007, Backus and Crucini 2000) theoretical analysis of
the trade channel of adjustment to oil shocks my contributions are the following.
The model highlights the role of the degree of exchange rate pass-through in the
propagation mechanism of oil price shocks. A positive oil price shock produces
a trade imbalance that, with nominal price rigidities and full exchange rate
pass-through, forces the currency of the oil importing country to depreciate to
restore the balance. As a consequence of the depreciation, the oil importing
country suffers a negative wealth effect and the oil exporting country a positive
wealth effect, so that the terms of trade redistributes asymmetrically the cost of
adjustments to positive oil price shocks in favor of the oil exporting country. This
transmission mechanism makes it possible for a positive oil price shock to cause
to the oil importing country an economic cost that, in terms of consumption loss,
is larger than the share of oil costs in total output (supply channel). However,
with nominal price rigidities but no exchange rate pass-through, the adjustment
mechanism to an oil price shock changes. As export prices grow insensitive to
the exchange rate the strenght of the trade channel gets weaker, up to the point
in which oil shocks are transmitted to the economy only through the supply
channel. A final contribution of the model is to show that the elasticity of most
economic variables to oil price shocks is not constant but increases with the size
of the shock.

The model suggests that a change in the exchange rate pass-through might
contribute to explain the evidence of a weaker impact of oil price shocks on the
economic activity in recent times (Hooker 1996 and 2002, Hamilton 1996, Balke
et al. 2002, and Blanchard and Galì 2007). In fact, as the degree of exchange
rate pass-through also depends on the proportion of firms that serve foreign
markets by producing locally (i.e. by foreign direct investments) rather than
exporting, the firms’ choice of how serving foreign markets may play a role in
shaping the strength of the trade channel. From these results it follows that
countries more dependent on oil-imports will not necessarily suffer the most in
the aftermath of a surge in the price of oil (see, i.e., Bohi 1999).

Section 2 presents evidence and the empirical analysis. Section 3 describes

the theoretical model and presents results. Section 4 concludes.

2 Empirical Evidence
The top panel of Figure 1 shows the dynamic of the dollar share of US oil
imports in total imports during 1973M1-2010M5 (source OECD). On average,

3
oil imports have represented 16% of all US imports in the period, though this
figure has not been constant over time: it averaged 25% in the 70s, bottomed
around 10% in the mid 80s and through all the 90s, and then rose again to
15% in the 2000s. The bottom panel of figure 1 reports the dynamic of the
real price of oil over the same period. The price of oil is the US crude oil
imported acquisition cost by refiners in dollar per barrel (source EIA) and the
price deflator is the US CPI (source IFS). In both panels the grey-shaded areas
highlight episodes of large and fast variations in the price of oil, mostly related
to political events affecting the oil supply.

Two facts emerge from figure 1: oil imports both are a relevant share of
US total imports and follow a dynamic similar to that of the real price of oil.
Consequently, it i sreasonable to assume that large variations in the price of oil,
by changin the US import price index, also affect its terms of trade, that is the
ratio between the US export price index and the US import price index.

Figure 1 - Oil imports (% of total imports) and the real price of oil in the US.
The terms of trade is the key variable of the trade channel of propagation of
oil shocks to the economy. A change in the terms of trade affects the level and
the composition of both the aggregate demand and the trade balance, though
its final effect on them depends ultimately on factors such as the elasticity of
substitution between imports and exports and the degree to which the interna-
tional financial markets are integrated. The elasticity of substitution between

4
imports and exports dictates by how much a country, after a worsening in its
terms of trade, will switch from imported to domestically produced goods. The
degree to which the international financial markets are integrated affects the
response of the trade balance to a change in the terms of trade, as the trade
balance is unaffacted by movements in the terms of trade in case of complete
markets, while the reaction is maximised in case of financial autarky

Thus, I evaluate if a trade channel is at work after an oil price shock by

focusing on the relationship between the real price of oil and the terms of trade.3

2.1 Data and Methodology

The price of oil is the US crude oil imported acquisition cost by refiners (dollar
per barrel, source EIA). The price index is the CPI all items (index, base year
2005, source IFS). The terms of trade is constructed as the ratio between the
US export price index over the US import price index (index number, base year
2005, source IFS). Data are monthly and cover the period 1973:1-2010:5.

In analysing the impact of oil shocks on the terms of trade I assume that
the real price of oil is predetermined to the terms of trade.4 This assumption
amounts to imply no contemporaneous impact of terms of trade innovations on
the real price of oil, and is regarded as plausible as long as the data frequency
are either monthly or quarterly. The use of oil price is sometimes criticised on
the ground that this measure does not distinguish among different sources of
oil shocks, such as oil supply shocks, specific oil-demand shocks and aggregate
demand shock in the industrial commodity market (Kilian 2009). In spite of
this shortcoming, its use in the present analysis has two justifications. First,
it is always possible to use the price of oil as long as the interest is on the
average oil price innovation and not on specific episodes. Second, as the focus
of the analysis is on the terms of trade, it is reasonable to assume that oil price
shocks both supply and demand driven would produce a similar effect on the
terms of trade (in this respect, in Kilian et al 2009 the external adjustment to oil
shocks results qualitatively and quantitatively similar to both oil demand and oil
supply shocks). In case the source of the oil price shock is the aggregate demand
in the industrial commodity market, both the numerator and the denominator
of the terms of trade would be affected, so that in theory the terms of trade
might go either direction. However, as the role of commodities in US exports is
quantitatively limited with respect to the role of oil in US imports, it is plausible
to assume that even in presence of a shock to the aggregate demand in the
industrial commodities it would be the dynamic of the price of oil to determine
the direction of the change in the terms of trade. I also follow Nordhaus (2007)
3 Foran empirical analysis on the external adjustment to oil shocks see Kilian et al. (2009).
4 The empirical literature has largely debated over both the nature and the identification of
oil shocks (see, i.e., Lee et al 1995, Hamilton 1983 and 2003, Barsky and Kilian 2004, Kilian
and Vigfusson 2009 and Kilian 2009).

5
and scale the price of oil for its economic importance, which in this context is
represented by the share of oil imports in total imports.

In the econometric analysis I allow for increases and decreases in the price
of oil to have different impacts on the terms of trade. Empirical evidence of
asymmetric effects of oil price shocks on economic variables dates back to Mork
(1989) and Olsen and Mysen (1994), and led to rejecting a linear relationship
between oil prices and real activity, arguing that only positive oil price innova-
tions would affect the macroeconomic performance (Mork 1989, Lee et al. 1996,
Hamilton 1996). However, Kilian and Vigfusson (2009) challenge this result on
the ground that it is based on standard censored VAR regression models that
produce biased estimations (see also Rigoboni and Stoker 2007). As an alterna-
tive, they propose a different methodology to both control for the asymmetry
of oil price shocks and to test it. More specifically, they propose to estimate
a bivariate nonlinear structural model and test for symmetry by mean of an
impulse-response based test. The structural nonlinear bivariate model is the
following:

xt = b10 + B11 (L)xt−1 + B12 (L)yt−1 + ε1,t (1)

yt = b20 + B21 (L)xt + B22 (L)x+
t + B23 (L)yt−1 + ε2,t ,

where xt is the logarithm of the first difference of the real price of oil, x+t
captures only the positive values of xt , yt is the logarithm of the terms of
trade and the error terms, ε1,t and ε2,t are mutually uncorrelated error terms,
and (B11 (L), B12 (L)) and (B21 (L), B22 (L), B23 (L)) are polynomial matrices of
coefficients of the first and second equation.
This model allows for a consistent estimation of the coefficients and the
impulse responses regardless of whether the true data generation process of yt is
symmetric or asymmetric. Moreover, an equation-by-equation OLS estimation
applies as the errors are mutually uncorrelated by constructions. To compute
the impulse-response based test I follow Kilian and compute generalised impulse
responses, as nonlinear impulse
responsesrequire to take into account both the
variability of initial values for yt , xt , x+
t and the variability of future shocks
{ε1,t , ε2,t } (see also Koop et al 1996 and Potter 2000).

2.2 Results
I estimate the model in (1) with 2 lags, as suggested by both the Akaike and
Hannan-Quinn information criteria. The impulse response analysis shows that
oil price shocks have a large, persistent and statistically significant impact on
the US terms of trade. After a one-standard deviation positive oil price shock
(6.25%) the terms of trade starts depreciating, reaching the through after three
months (top panel of Figure), and then slowly goes back to zero. The response of
the terms of trade remains statistically significant (95% confidence interval) up
to two and half a year after the shock. The cumulated depreciation of the terms
of trade is around 14% 30 months after the shock (bottom panel of Figure 2).

6
The large effect of oil price shocks on the terms of trade suggests that the terms
of trade, and so the trade channel, plays an important role in the transmission
mechanism of oil shocks.

To test the presence of asymmetry in the response of the economy to posi-

tive and negative oil price shocks I follow Kiliand and Vigfusson (2009) and use
an impulse-response based symmetry test rather than a standard slope-based
symmetry test. The joint null hypothesis of the impulse-response based symme-
try test is that the response of the terms of trade to positive and negative oil
price shocks is symmetric up to the pre-specified horizon. Table 1 reports the
pre-specified horizon (row 1) and the corresponding p-value of the test statistics
under the null hypothesis of symmetry (row 2). The hypothesis of symmetry
is rejected at the 5% level at the first two horizons, and at the 10% level at
the third horizon (similar results apply when considering a size of the shock of
two standard deviations). The test provides a weak support to the hypothesis
that oil price shocks might have asymmetric effect on the economy through the
terms of trade.

Figure 2 - Impulse (top panel) and cumulated (bottom panel) impulse

response of the Terms of trade to a 1-standard deviation positive oil price
shock. The shaded areas are the 95% confidence interval.

7
Horizon 0 1 2 3 4 5 6 7 8 9 10 11
p-values 0,017 0,025 0,057 0,114 0,194 0,293 0,406 0,522 0,634 0,732 0,814 0,877
Based on1000 simulations. P-values are based on a chi-square with H+1 d.f.

Table 1 - Impulse-Response based Symmetry test

Finally, I investigate whether the response of the terms of trade depends on

the size of the oil price shock. However, by rescaling the real price of oil by the
oil import share, a greater weight is given to those changes in the real price of
oil that occur when the price of oil is high. For oil changes of the same size to
be given the same weight in the dataset I have re-estimate model (1) without
scaling the real price of oil by the oil import share. Results are qualitatively and
quantitatively similar, but they add a further piece of evidence on the effect of
oil price shocks on the terms of trade: the response of the terms of trade to an
oil price shock appear to increase with the size of the shock (Figure 3).

Figure 3 - The response of the Terms of trade to a Positive oil price shock by
shock size. All responses have been scaled by the size of the shocks to ensure
comparability.

3 Theoretical Analysis
This section presents a theoretical analysis of the trade channel of adjustment
to oil price shocks and the role that the terms of trade plays in it. To this

8
purpose, I strictly follow Corsetti and Pesenti (2005) and set up a simple two-
country general equilibrium model with imperfect competition in production
and incomplete international financial markets. Oil is introduced in the model
as a factor of production and is traded from the oil-importing country to the oil-
exporting country at a price that is determined exogenously in the currency of
the home importing country.5 The transmission mechanism of oil price shocks
to the economy is then analysed under three different price regimes: flexible
prices, nominal price rigidity with producer currency pricing (PCP) regime and
nominal price rigidities with local currency pricing (LCP) regime.
In the next section I report the main elements of the model and present
results. For technical details on how to derive the general equilibrium condi-
tions of the model refer to appendix A. The flexible price equilibrium is derived
analytically in appendix B. The equilibrium with nominal price rigidities and
Producer Currency Pricing regime is derived in appendix C. The equilibrium
with nominal price rigidities and Local Currency Pricing is derived analytically
in appendix D.

3.1 The Model

The world consists of two countries of equal size, Home and Foreign, which are
identical in any respect apart from three aspects: total factor productivity, mon-
etary policy and oil endowment. In particular, oil is a Foreign specific resource,
so that Home is the oil importing country and Foreign is the oil exporting coun-
try. Each country is inhabited by a continuum of households and firms: in the
Home country households are indexed by i ∈ [0, 1] and firms are indexed by h
∈ [0, 1]; in the Foreign country households are indexed by i∗ ∈ [0, 1] and firms
are indexed by f ∈ [0, 1]. In what follows I focus on the Home economy since
the Foreign economy can be described symmetrically.

Households Households cannot move across countries, share identical pref-

erences, own domestic firms, derive utility from a final consumption good and
disutility from supplying labor services to firms in exchange for wage income.
The utility of household i at a given period t is given by:

Ut (i) = ln Ct (i) − kLt (i).

The parameter k > 0 determines the disutility of supplying labor and Ct ,
the final consumption good, is not traded internationally and is produced in a
perfectly competitive final good sector by a Cobb Douglas aggregator technology
that combines a Home basket of goods (CH,t ) and a Foreign basket of goods
(CF,t (i)) :
1 1
Ct = CH,t
2
CF,t
2
.
5 This last assumption aims to mimic the fact that while the price of oil is set in US dollar,

the US is an oil-importing country.

9
CH,t (i) and CF,t (i) are CES baskets of, respectively, Home (Ct (h, i)) and Foreign
(Ct (f, i)) differentiated varieties:

1  θ−1
θ 1  θ−1
θ
 
θ−1 θ−1
CH,t (i) =  Ct (h, i) θ dh ; CF,t (i) =  Ct (f, i) θ df  ,
0 0

the parameter θ > 1 is the elasticity of substitution across varieties, is equal

across countries and is higher than the elasticity of substitution between CH
and CF that is 1 as Ct is a Cobb Douglas aggregator.
Solving for the expenditure minimization problem I obtain both the price
level (Pt ) and the households’ demand for a specific variety h (Ct (h, i)) and a
specific variety f (Ct (f, i)):
1 12 1
Pt = 2
PH,t PF,t
2
 −θ  −1
1 Pt (h) PH,t
Ct (h, i) = Ct
2 PH,t Pt
 −θ  −1
1 Pt (f ) PF,t
Ct (f, i) = Ct
2 PF,t Pt
1
PH,t CH,t = PF,t CF,t = Pt Ct ,
2
where PH,t and PF,t are the price indices of, respectively, CH,t and CF,t .
At any period t households face the following budget constraint:

Pt Ct (i) + B t (i) + εt Bt∗ (i) ≤ W t Lt (i) + Πt (i) + B t−1 (i)(1 + it−1 ) (2)
+εt Bt−1∗
(i)(1 + i∗t−1 ).

where Pt Ct (i) is nominal spending on the final consumption good, Wt is the

nominal wage per unit of labor, Πt (i) is household’s share of profit of firms in
the intermediate goods sector. Households can purchase two riskless one-period
bonds at any period t: Bt (i), which is denominated in home currency and yields
a nominal interest it at t + 1, and Bt∗ (i), which is denominated in foreign-
currency and yields a nominal interest rate i∗t at t + 1. The nominal exchange
rate, εt , expresses how many units of Home currency are exchanged per one unit
of Foreign currency.

Households’ decisions on consumption and labor supply solve the following

utility maximization problem:
∞

Max Et β t ln(Ct (i)) − κLt (i),

Ct (i),Lt (i)
t=0

subject to the stream of one-period budget constraints defined in (2) .

10
Firms and pricing regimes The intermediate goods market is characterised
by monopolistic competition: each firm supplies a single variety h that is an
imperfect substitute to all other varieties. Technology is represented by a Cobb
Douglas production function that is common across firms:
1−α
Yt (h) = Zt Lα
t (h)Et (h).

Zt > 0 is the total factor productivity and is common across Home firms.
α ∈ (0, 1) is the share of labor income and is common across firms and countries.
Lt (h) and Et (h) are, respectively, the amount of labor and oil that are employed
in the production of variety h at the period t. The aggregate demand of a variety
h at time t is:

 −θ  −1  −θ  −1
∗
1 Pt (h) PH,t 1 Pt∗ (h) PH,t
Yt (h) = C+ ∗ C ∗. (3)
2 PH,t Pt 2 PH,t Pt∗

Any firm sets two prices for its variety h, one in the Home market and the
other in the Foreign market, to maximize its profit function Πt (h) :

1
M ax Πt (h) = (Pt (h) − MCt (h)) Ct (h, i)di + (εt Pt∗ (h) (4)
Pt (h),εφ ∗
t Pt (h)
0
1
−MCt (h)) Ct∗ (h, i∗ )di∗ ,
0

subject to the aggregate demand for its good (3). The marginal cost of h is:
1−α
(kµ)α Pe,t
M Ct (h) = MCt = Zαα (1−α)1−α , which is identical across varieties h.

The intermediate goods prices are computed under three different scenarios,
each one characterised by either flexible prices or nominal price rigidities, and
by the specific pricing regime that the intermediate goods firms adopt in the
two countries. To define what are the possible pricing regimes that a firm
can adopt let us consider that in general the Home firm, but the same line of
reasoning holds for the Foreign firm, maximizes its profit in the Foreign market
by setting the price P̃t (h) = Pt∗ (h)εφt , where the parameter φ ∈ [0, 1] is the
degree of exchange rate pass-through. If φ = 1 we are under the producer
currency pricing (PCP) regime: the price the firm sets in the Foreign market
is P̃t (h) = Pt∗ (h)εt and is denominated in Home currency, so that any change
in the nominal exchange rate fully transmits to P̃t (h). If φ = 0 we are under
the local currency pricing (LCP) regime: the price the firm sets in the Foreign
market is P̃t (h) = Pt∗ (h) and is denominated in Foreign currency, so that no
change in the nominal exchange rate affects P̃t (h). An interesting way to see

11
at the different degree of exchange rate pass-through involved by the two price
regimes is to interpret it in terms of the geographical distribution of the firm’s
production sites. Let us consider again the case of an Home firm that produces
h and let us assume that it always sets the price in the producer currency, but
now it can choose how to distribute its production between the two countries.
It follows that the influence of the exchange rate regime on the price set in
the Foreign market will depend on how the total production of h is distributed
between Home and Foreign. Full exchange pass-through would correspond to
the case in which h is entirely produced at Home, while the case of no exchange
pass-through would occur when the production of h has been divided across the
two countries as to serve both markets as a local firm.

Under a flexible price regime the choice of the invoice currency for the price
of exports does not have any effect on the propagation mechanism of oil price
shocks, that is both the PCP and the LCP regime lead to the same result.
However, when nominal rigidities are introduced, which one of the two regimes
is on has a large effect on the way oil price shocks transmit to the economy.

Monetary Policy The monetary policy stance is defined by µt = Pt Ct . This

specification of the monetary policy allows the monetary authority to influence
the aggregate nominal spending without specifying what is the exact instrument
that it uses to do it.

The price of oil The nominal price of oil is exogenous 6 , is denominated in

the Home currency and always respects the law of one price:
∗ Pe,t
Pe,t = .
εt

Financial allocations in equilibrium A symmetric equilibrium is com-

puted under the assumption of financial autarky, so that the agents cannot lend
or borrow internationally.

3.1.1 The flexible price regime

The distinction between an oil importing and oil exporting country is the third
source of cross-country asymmetry in the model equilibrium other than differ-
ences across countries in the total factor productivity and the monetary policy
stance. This additional asymmetry, in equilibrium, is captured by the parameter
ϕF LP : θ 
θ−1 + (1 − α)
ϕF LP = θ
> 1. (5)
θ−1 − (1 − α)
6 Though an oil market is not explicitly modelled, the assumption of an exogenus price of

oil amounts to considering a perfectly elastic oil supply (any quantity of oil demanded can
be supplied at the given price) and zero marginal costs. The price of oil is then determined
entirely by a specific tax that is set up by the government of the Foreign country, and whose
revenues are equally redistributed across households as a lump sum transfer.

12
 which is an increasing function of both the degree of monopolistic power in
the Home intermediate goods sector (θ) and its share of oil costs (1 − α):
The presence of ϕF LP causes the oil importing currency to depreciate in
equilibrium with respect to the case of no oil in the economy (as in Corsetti and
Pesenti where ϕF LP = 1):
µt
εt = ϕF LP . (6)
µ∗t
The parameter ϕF LP approaches 1 as both (1 − α) goes to 0 (no oil is
required to produce) and as θ approaches 1, as a larger monopolistic power
implies both a lower equilibrium production, which requires fewer oil imports,
F LP
and larger
 ∗ revenues from exports.
 ∗ϕ also drives a wedge in the consumption
C L
ratio C and the labor ratio L , so that in equilibrium Foreign households
work less and consume more than Home households:
Ct∗ Lt
= ϕF LP =⇒ Ct∗ > Ct ; = ϕF LP =⇒ L∗t < Lt .
Ct L∗t
To describe the impact of oil price shocks on a variable of interest Yt I
compute its elasticity to the price of oil. For this purpose, I define by χYt ,Pe,t the
elasticity of the variable Yt to the price of oil, that is the percent change in Yt
when the price of oil Pe,t increases by one percent. The impact of oil price
shocks on the world economy with flexible prices is then summed up by the
following elasticities:

χεt ,Pe,t = 0 (7)

χCt ,Pe,t = χ∗Ct ,Pe,t = −(1 − α) (8)

χPt ,Pe,t = χ∗Pt ,Pe,t = (1 − α) (9)
χLt ,Pe,t = χ∗Lt ,Pe,t = 0 (10)

For example, with flexible prices a positive oil price shock does not affect
either the nominal exchange rate (7) or the level of employment in the two
countries (10) 7 , while, in both countries, the price level increases (9) and con-
sumption decreases (8) by the same amount. 8
Figure 4 shows the effect of a positive oil price shock on the Aggregate Supply
(AS) and the Aggregate Demand (AD) schedules of the Home economy in the
space (C, L) :
7 The increase in the demand of labor that is due to the substitution of labor for oil cancels

out the decrease in the demand of labor that is caused by a lower aggregate demand.
8 It is worth noting that in case of shocks to the real price of oil rather than to the nominal



price of oil the size of χC,P e , χC ∗ ,P e , χP,P e and χP ∗ ,P e increases from |1 − α| to
1−α
α

.This
distinction may matter when trying to estimate the effect of oil price shocks to the economy
in empirical works, though the difference between both measures is negligible if the oil share
is small.

13
 AS : Ct = τ F LP Lt
µ
AD : Ct = t .
Pt

C
AS 0 : C = τ 0 L

AS: C = τ 1L

µ
AD 0 : C =
C0 P0

C1 AD 1 : C = µ 1
P
L
0 1
L =L

Figure 4 - The effect of a positive oil price shock on the economy when prices
are flexible.

The index τ F LP in the AS schedule captures both the supply channel (the
condition of intratemporal marginal substitution between labor and energy) and
the trade channel of transmission of oil price shocks. However, with identical
technology across countries and no change in the nominal exchange rate, oil
price shocks do not change the terms of trade, so that only the supply channel
is at work in the transmission of the oil price shock. A positive oil price shock
lowers τ F LP and rotates the AS schedule downwards, from AS0 to AS1 , because
of the substitution of labor for oil, and also shifts the AD schedule downward,
from AD0 to AD1 , as the price level P increases. In the new equilibrium the
level of employment L does not change, consumption decreases from C 0 to C 1
and the price level increases from P 0 to P 1 . The effect of oil price shocks to the
Foreign economy is perfectly symmetric to that of the Home economy.

14
3.1.2 Nominal Price Rigidities
Nominal price rigidities are introduced by assuming that prices are predeter-
mined one period ahead.9 With nominal rigidities oil price shocks have asym-
metric effects across countries, and the distinction between pricing regimes (PCP
or LCP) changes the propagation mechanism of oil shocks to the economy.

The PCP regime With nominal rigidities and PCP regime the exchange rate
in equilibrium is an increasing function of the price of oil (??), the exchange
rate elasticity to the price of oil is positive (??) and is an increasing function
of the price of oil (??) . It follows that if a given positive oil price shock forces
the oil importing currency to depreciate by δ, an oil shock twice as large will
force a depreciation in the oil importing currency larger than 2δ. This result is
in accordance with the empirical results shown in Figure 3 where the response
of the terms of trade to oil price shocks is increasing in the size of the shock.

µt P CP ∂ϕP CP
εt = µ∗ ϕt , ϕP
t
CP
>1 t
∂Pe >0 (11)
t

χεt ,Pe,t > 0 (12)


∂ χεt ,Pe,t
∂(Pe,t ) >0 (13)

The effects of oil price shocks on the other variables in the economy are
reported below in terms of elasticities to the price of oil and expressed as a
function of χε,PE , which is a measure of the strength of the trade channel itself:

1 1
χCt ,Pe,t = − χε,P e < 0 , χPt ,Pe,t = χεt ,Pe,t > 0 (14)
2 2
1 1
χCt∗ ,Pe,t = χε,P e > 0 , χPt∗ ,Pe,t = − χεt ,Pe,t < 0 (15)
2 2
ϕP
t
CP
(Pe,t )
χLt ,Pe,t =
P CP χ >0 (16)
ϕt (Pe,t ) − 1 εt ,Pe,t
χL∗t ,Pe,t ≷ 0. (17)

Oil price shocks have asymmetric effects to oil importing and oil exporting
countries as the change in the terms of trade redistributes asymmetrically across
countries the cost of adjustment to the shocks. After a positive oil price shock
the currency of the oil importing country depreciates to gain in competitiveness
and make up for the higher price of oil imports. The worsening in the terms
of trade of the oil importing country provokes a negative wealth effect to which
9 Quadratic costs of price adjustment à la Rotemberg or Calvo pricing would introduce a

more realistic dynamic of the economy but would not allow for an analytical solution.

15
households react by both consuming less (11) and, even if real wages have been
driven down by a higher price level, supplying more labor services (11). House-
holds in the oil exporting country, instead, take advantage of the change in the
terms of trade, which makes imports cheaper and reduce the price level, and
increase their consumption.
The final effect on the production/employment is uncertain and depends on
the size of the oil price shock. By the supply channel, the increase in the price
of oil causes a replacement of labor for oil in the production function. However,
as the price of oil is exogenously fixed in Home currency, the input price of oil
for Foreign depends also on χε,P e . It follows that the larger the shock (and so
the larger χε,P e ) the less important the supply channel in the Foreign coun-
try. Eventually, after a very large positive oil price shock the supply channel
might decreases employment in the Foreign country. Thus, the supply channel
increases foreign employment for small oil price shocks (elasticity) and decreases
it for large shock (elasticity). By the trade cahnnel, the foreign currency appre-
ciation both induces a worldwide replacement of oil importing for oil exporting
intermediates, which contracts the demand of labor in the oil exporting coun-
try, and also withers the increase in the price of oil as measured in oil exporting
currency, as the price of oil is denominated in home currency.
This is the terms of trade effect of oil shocks on the oil exporting country
employment. The larger the size of the oil price shock, the bigger the probability
that the latter effect will prevail and so that employment in the oil exporting
country will shrink after a positive oil price shocks.
Figure 5 shows the effect of oil shocks to the oil importing economy using
the AS and AD schedules in the space (C, L) :

AS : Ct = τ P
t
CP
Lt
µt
AD : Ct = .
Pt

As in the case of flexible price, the terms τ Pt

CP
in the AS schedule captures
both the supply and the trade channel of propagation of oil shocks. A positive
oil shock rotates inwards the AS schedule, from AS0 to AS1 , while the increase
in the price level, from P0 to P1 , shifts downwards the AD schedule, from AD0
to AD1. As a result consumption falls from C0 to C1 and the employment in-
creases from L0 to L1 .

16
 C
AS : C = τ0 L

AS1: C = τ1 L

µ
AD0: C =
C0 P0

C1 µ
AD1: C =
P1

L
L0 L1

Figure 5 - The effect of a positive oil price shock on the oil importing economy
with nominal rigidities and PCP pricing.

C*
AS 1: C * = τ1* L*

AS0: C*= τ0 l*

AS2: C * = τ2* L*

C1* µ*
AD1: C * =
P1*
C0* µ*
AD0: C * = *
P0
L*
L1* L0* L2*

Figure 6 - The effect of a positive oil price shock on the oil exporting economy
with nominal rigidities and PCP pricing.

17
 Figure 6 shows the impact of a positive oil price shock to the oil exporting
country with the aid of the AS and AD schedules. The AD schedule always shifts
upwards, from AD0 to AD1 , as the price level always decreases from P0 to P1 .
The AS schedule, instead, may rotate either upwards or downwards, depending
on which one between the supply and the terms of trade effect prevails. For
small oil price shocks the supply effect prevails and the AS curves will rotate
downwards, from AS0 to AS2 , causing employment in the oil exporting country
to increase. For oil price shocks large enough the terms of trade effect prevails
and the AS schedule rotates outwards, from A0 to A1 , shrinking employment in
the oil exporting country.

The LCP regime With a LCP regime the strength of the trade channel of
adjustment to oil shocks gets reduced. The nominal exchange rate still reacts
to oil price shocks (11) but the elasticity of the nominal exchange rate to oil
price shocks is now bounded from above (11).

µt LCP LCP ∂ϕLCP

t
εt = ϕ , ϕ (P e,t ) > 1, >0 (18)
µ∗t t t
∂Pe,t
0 < χεt ,Pe,t < (1 − α) (19)
 
∂ χεt ,Pe,t
0 < . (20)
∂(Pe,t )

Under a LCP regime both economies are partially insulated from oil price shocks
as in both countries both the price level and consumption do not change after
an oil price shock (11). Oil price shocks only change the level of employment
(11) as the ratio of labor to oil in the production function reacts to the change
in the relative prices of the factors of production. However, the supply effect is
always less severe in the oil exporting country (11) as the change in the price of
oil denominated in the oil exporting currency is moderated by the movement in
the nominal exchange rate.

χCt ,Pe,t = χCt∗ ,Pe,t = χPt ,Pe,t = χPt∗ ,Pe,t = 0 (21)

χLt ,Pe,t = (1 − α) (22)
 
χL∗t ,Pe,t = (1 − α) 1 − χεt ,Pe,t < (1 − α) (23)

18
 C
AS 0 : C= τ0 L

A S 1 : C= τ 1 L

µ
AD 0 : C=
C0 P0

L
L0 L1

Figure 7 - The effect of a positive oil price shock on the oil importing economy
with nominal rigidities and LCP pricing.

AS : Ct = ϕLCP
t Lt
µt
AD : Ct = .
Pt
Figure 7 shows the effect of a positive oil price shock on the Home economy
with the help of the AS and AD schedules. After a positive oil price shock the
AD schedule stays put at AD0 . The AS schedule, instead, rotates inwards as
labor replaces oil in the production function. The only change in the economy
of the positive oil price shock is the increase in the level of employment L. The
effect of a positive oil price shock on the oil exporting country is almost identical,
the only difference being a reduced increase in the employment level L∗ .

4 Discussion and Extensions

The model discloses two aspects of the trade channel of propagation of oil shocks,
specifically the role of the exchange rate pass-through and the non linear rela-
tionship between the size of oil shocks and the response of the economy. These
contributions help to better understand when oil price shocks may have a large
recessive impact on oil importing country, and suggest that a change in the ex-
change rate pass-through might contribute to explain the evidence of a weaker
impact of oil price shocks on the economic activity in recent times.

19
 The clarity of these messages comes from the closed form solution of the
model, which is obtained thorugh many simplifying assumptions. Thus, the
model represents a valuable starting point for an analysis of the relationship
between oil and macreoconomic performance which aims to fully reconcile data
with theory. In particular, the model should be extended in three directions:
endogenise the price of oil by introducing a world oil market, a more complex
structure of the international financial market, and the introduction of capital.

The introduction of an international oil market would permit the model to

distinguish the effects of oil price shocks according to the source of the shock.
A more complex modelling of international financial markets would matter not
only because it affects the strength of the trade channel but also because of the
so-called evaluation effect (though it might play a minor role for oil importing
country under the assumption that the assets weightings in a country’s Port-
folio reflected the relative economic importance of the country). Finally, the
introduction of capital would allow for a more realistic and interesting dynamic
of the adjustment process of the economy to oil shocks.

20
References
[1] Atkenson, A.J. and Kehoe, P.J. (1999), "Models of Energy Use: Putty-
Putty versus Putty-Clay", American Economic Review 89, 1028-1043.
[2] Backus, D.K. and M.J. Crucini (2000), "Oil Prices and the terms of Trade",
Journal of International Economics 50, 185-213.
[3] Balke, N.S., Brown, S.P.A. and M.K. Yucel (2002), "Oil Price Shocks and
the U.S. Economy: Where Does the Asymmetry Originate?," The Energy
Journal 23(3), 27-52.
[4] Barsky, R.B. and L. Kilian (2004), "Oil and the Macroeconomy Since the
1970," Journal of Economic Perspective 18, 115-134.
[5] Bernanke, B.S. (1983), "Irreversibility, Uncertainty, and Cyclical Invest-
ment," Quarterly Journal of Economics 98, 85-106.
[6] Bernanke, Ben S., M. Gertler, and M. Watson (1997), “Systematic Mone-
tary Policy and the Effects of Oil Price Shocks,” BPEA 1, 91—142.
[7] Blanchard, O.J., and J. Galì (2007), " The Macroeconomic Effects of Oil
Price Shocks: Why are the 1970s so different from the 2000s?", NBER
Working Papers 13368.
[8] Bodenstein, M., C.J. Erceg, and L. Guerrieri (2007), "Oil Shocks and Ex-
ternal Adjustments," mimeo, Federal Reserve Board.
[9] Bohi, D.R. (1991), "Energy Price Shocks and Macroeconomic perfor-
mance", Resource and Energy 13(2),145-162.
[10] Bresnahan, T. and V. Ramey (1993), "Segments Shifts and Capacity Uti-
lization in the U.S. Automobile Industry, " American Economy Review
Papers and Proceedings 83, 213-218.
[11] Bruno, M.R., and J. Sachs (1985), "Economics of Worldwide Stagflation",
Harvard University Press.
[12] Corsetti, G. and P. Pesenti (2005) "The Simple Geometry of Transmission
and Stabilization in Closed and Open Economies", NBER Working Paper
11341.
[13] Davis, S.J. (1987), "Allocative Disturbances and Specific Capital in Real
Business Cycle Theories," American Economic Review Papers and Pro-
ceedings 77, 326-332.
[14] Davis, S.J. and J. Haltiwanger (2001), "Sectoral Job Creation and Distruc-
tion Responses to Oil Price Changes," Journal of Monetary Economics 48,
465-512.

21
[15] Devereux, M.B. and A. Sutherland (2010), "Valuation effects and the dy-
namics of net external assets," Journal of International Economics 80(1),
129-143.
[16] Dohner, R.S. (1981), "Energy Prices, Economic Activity and Inflation: sur-
vey of issues and results " in K.A. Mork (Ed.): Energy Prices, Inflation and
Economic Activity. Cambridge, MA: Ballinger.
[17] Finn,M.G. (2000), "Perfect Competition and the Effects of Energy Price
Increases on Economic Activity", Journal of Money Credit and Banking
32, 400-416.
[18] Fogli, A. and F. Perri. (2006), "The Great Moderation and the US External
Imbalance," Monetary and Economic Studies 24(S1), 209-225.
[19] Ghironi, G, J. Lee and A. Rebucci (2007), "The Valuation Channel of
External Adjustment," NBER Working Papers 12937.
[20] Gourinchas, P.O. and H. Rey (2007), "International Financial Adjust-
ment," Journal of Political Economy 115(4), 665-703.
[21] Hamilton, J.D. (1983), "Oil and the Macroeconomy since World War II,"
Journal of Political Economy 91, 228-248.
[22] Hamilton, J.D. (1988), " A Neoclassical Model of Unemployment and the
Business cycle", Journal of Political Economy 96, 593-618.
[23] Hamilton, J. (1996), “This Is What Happened to the Oil-Macroeconomy
Relationship”, Journal of Monetary Economics 38 (2), 215-220.
[24] Hamilton, J.D. (2003), "What is an Oil Shock," Journal of Econometrics
113, 363-398.
[25] Hamilton, J.D. (2008), " Oil and the Macroeconomy", in S.Durlauf and L.
Blume (eds), The New Palgrave Dictionary of Economics, 2nd ed., Palgrave
MacMillan Ltd.
[26] Hamilton, J. D. and A. M. Herrera (2004), “Oil Shocks and Aggregate
Macroeconomic Behavior: The Role of Monetary Policy: A Comment,”
Journal of Money, Credit, and Banking 36 (2), 265—86.
[27] Hooker, M. (1996), "What Happened to the oil price-macroeconomy rela-
tionship?", Journal of Monetary Economics 38, 195-213.
[28] Hooker, M.A. (2002), “Are Oil Shocks Inflationary? Asymmetric and Non-
linear Specifications Versus Changes in Regime,” Journal of Money, Credit,
and Banking34, 540-561.
[29] Kilian, L. (2008), "The economic effects of Energy Price Shocks", Journal
of Economic Literature 46(4),871-909.

22
[30] Kilian, L., A. Rebucci and N. Spatafora (2009), "Oil Shocks and External
Balances", Journal of International Economics 77, 181-194.
[31] Kilian, L. and R. Vigfusson (2009), "Are the Responses of the US Econ-
omy Asymmetric in Energy Price Increases and Decreases," International
Finance Discussion Papers.
[32] Kilian, L. (2009), "Not All Oil Price Shocks Are Alike: Disentagling De-
mand and Supply Shocks in the Crude Oil Market," American Economic
Review 99(3), 1053-1069.
[33] Kilian, L. and C. Park (2009), " The Impact of Oil Price Shocks on the
U.S. Stock Market", International Economic Review 50(4), 1267-1287.
[34] Koop, G. and M.H. Pesaran and S.M. Potter (1996), "Impulse Response
Analysis in Nonlinear Multivariate Models", Journal of Econometrics 74,
119-147.
[35] Lane,P.R: and G. M. Milesi-Ferretti, (2005) "A Global Perspective on Ex-
ternal Positions," NBER Working Papers 11589.
[36] Leduc, S. and K. Sill (2004), “A Quantitative Analysis of Oil Price Shocks,
Systematic Monetary Policy and Downturns,” Journal of Monetary Eco-
nomics 51, 781-808.
[37] Lee, K. and S. Ni (2002), "On the Dynamics Effects of Oil Price Shocks,"
Journal of Monetary Economics 49, 823-852.
[38] Lee, K., S. Ni and R.A. Ratti (1995), “Oil Shocks and the Macroeconomy:
The Role of Price Variability,” Energy Journal 16, 39-56.
[39] Loungani, P., and M.K. Yucel (2000), "How Robust is the Oil-
unemployment nexus?", mimeo.
[40] Mork, K.A. (1989), “Oil and the Macroeconomy. When Prices Go Up and
Down: An Extension of Hamilton’s Results,” Journal of Political Economy
97, 740-744.
[41] Mendoza, E., Quadrini, V. and Rios-Rull J.V. (2007), "Financial Inte-
gration, Financial Development and Global Imbalances", NBER Working
Papers 12909.
[42] Nordhaus, W. (2007), "Who’s Afraid of a Big Bad Oil Shock," Brooking
Papers on Economic Activity 2007-2, 219-238.
[43] Oystein, O., and H. T. Misen (1994), "Macroeconomic Responses to Oil
Price Increases and Decreases in seven OECD Countries", Energy Jour-
nal15 (4), 19-35.
[44] Rigobon, R and T.M. Stoker(2009) "Bias From Censored Regressors," Jour-
nal of Business & Economic Statistics 27(3), 340-353.

23
APPENDIX A - General equilibrium conditions
Households The world consists of two countries of equal size, Home and
Foreign, which are identical in any respect apart from total factor productivity,
monetary policy and oil endowment. Oil is a Foreign specific resource, so that
Home is the oil importing economy and Foreign is the oil exporting economy.
Each country is inhabited by a continuum of households and firms: in the Home
country households are indexed by i ∈ [0, 1] and firms are indexed by h ∈ [0, 1];
in the Foreign country households are indexed by i∗ ∈ [0, 1] and firms are indexed
by f ∈ [0, 1]. I follow the convention of labelling with an asterisk the variables
of the foreign country and, when reporting results in a table, reporting results
for the Home country in the first column and results for the Foreign country in
the second column.
In each country households share preferences, own domestic firms, derive utility
from a final consumption good (Ct or Ct∗ ) and disutility from supplying labor
services (Lt or L∗t ) to firms in exchange for wage income (Wt or Wt∗ ). The
utility of home household i at a given period t is given by:

Ut (i) = ln Ct (i) − kLt (i).

The parameter k > 0 determines the disutility of supplying labor and Ct is
produced in a perfectly competitive final good sector by a Cobb Douglas aggre-
gator technology that combines a Home (CH,t ) and a Foreign basket of goods
(CF,t ) with equal weights:
1 1
Ct = CH,t
2
CF,t
2
.
The final consumption good cannot be traded and is entirely consumed. CH,t (i) and CF,t (i) are
CES baskets of, respectively, Home (Ct (h, i)) and Foreign (Ct (f, i)) differenti-
ated varieties:

1  θ−1
θ 1  θ−1
θ
 
θ−1 θ−1
CH,t (i) =  Ct (h, i) θ dh ; CF,t (i) =  Ct (f, i) θ df  ,
0 0

the parameter θ > 1 is the elasticity of substitution across varieties, is equal

across countries and is higher than the elasticity of substitution between CH
and CF that is 1.
Symmetrically, for the foreign household i∗ we have:

∗  12
∗  12
Ut∗ (i∗ ) = ln Ct∗ (i∗ ) − k∗ L∗t (i∗ ), Ct∗ = CH,t CF,t
1  θ−1
θ 1  θ−1
θ
 
θ−1 θ−1
∗
CH,t (i∗ ) =  Ct∗ (h, i∗ ) θ dh ∗
, CF,t (i∗ ) =  Ct∗ (f, i∗ ) θ df  .
0 0

24

∗ ∗

Given the vector of prices PH,t , PH,t , PF,t , PF,t , household’s demands for
∗
(CH,t (i), CF,t (i)) in the Home country and household’s demands for CH,t (i∗ ), CF,t
∗
(i∗ )
in the Foreign country, are obtained as solutions to the following cost minimiza-
tion problems:

M in PH,t CH,t (i) + PF,t CF,t (i) Min P ∗ CH,t

∗ (i∗ ),C ∗ (i∗ ) H,t
∗
(i∗ ) + PF,t
∗ ∗
CF,t (i∗ )
CH (i),CF (i) CH F
1 1 1 1
s.t. Ct (i) = CH,t (i) 2 CF,t (i) 2 s.t. Ct∗ (i∗ ) = CH,t
∗
(i∗ ) 2 CF,t
∗
(i∗ ) 2 .

The first order conditions of each cost minimization problem give the demands
for the baskets of Home and Foreign goods, and the lagrangian multiplier associ-
ated to the technology constraint gives the price index of the final consumption
good, which is then interpreted as the minimum expenditure required to con-
sume one unit of the final good. The following table reports Home (column 1)
and Foreign (column 2) demands for the baskets of Home (A.2a, A.2b) and For-
eign goods (A.1a, A.1b), the prices of the final consumption good (A.3a, A.3b)
and the optimal composition of nominal spending (A.4a, A.4b)

  12  ∗  12
PH,t ∗ PH,t
CF,t (i) = Ct (i) PF,t (A.1a) CF,t (i∗ ) = Ct∗ (i∗ ) ∗
PF,t (A.1b)
  12  P ∗  12
PF,t ∗
CH,t (i) = Ct (i) PH,t (A.2a) CH,t (i∗ ) = Ct∗ (i∗ ) P F,t
∗ (A.2b)
H,t
1 1
 1
 1
Pt = 2PF,t PH,t
2 2
(A.3a) Pt∗ = 2 PF,t∗ 2 ∗
PH,t 2
(A.3b)
Pt Ct = 2PH,t CH,t = 2PF,t CF,t (A.4a) Pt∗ Ct∗ = 2PH,t
∗ ∗
CH,t ∗
= 2PF,t ∗
CF,t . (A.4b)

Home and Foreign households’ demands for varieties h and f (Ct (h, i), Ct (f, i) and Ct∗ (h, i∗ ), Ct∗ (f, i∗ ))
solve the following cost minimization problems:

   
M in ph,t Ct (h, i) + pf,t Cf,t (i) Min
∗ (i∗ ),C ∗ (i)
p∗t (h)Ct∗ (h, i∗ ) + p∗t (f )Ct (f, i∗ )
Ch,t (i),Cf,t (i) Ch,t f,t
1  θ−1
θ 1  θ−1
θ
 
θ−1 θ−1
s.t. CH,t (i) =  Ct (h, i) θ dh ∗
s.t. CH,t (i∗ ) =  Ct∗ (h, i∗ ) θ dh
0 0
1  θ−1
θ 1  θ−1
θ
 
θ−1 θ−1
s.t CF,t (i) =  Ct (f, i) θ df  ∗
s.t CF,t (i∗ ) =  Ct (f, i∗ ) θ df  .
0 0

From the first order conditions we obtain :

25
  −θ  −θ  −1
1 Pt (h) 1 Pt (h) PH,t
Ct (h, i) = 2 PH,t CH,t (i) = 2 PH,t Pt Ct (i) (A.5a)
 −θ  −θ  −1
1 Pt (f) 1 Pt (f ) PF,t
Ct (f, i) = 2 PF,t CF,t (i) = 2 PF,t Pt Ct (i) (A.6a)
 −θ  −θ  ∗ −1
1 Pt∗ (h) 1 Pt∗ (h) PH,t
Ct∗ (h, i∗ ) = 2 ∗
PH,t
∗
CH,t (i∗ ) = 2 ∗
PH,t Pt∗ Ct∗ (i∗ ) (A.5b)
 −θ  −θ  P ∗ −1
1 Pt∗ (f ) 1 Pt∗ (f )
Ct (f, i∗ ) = 2 ∗
PF,t
∗
CF,t (i∗ ) = 2 ∗
PF,t
F,t
Pt∗ Ct∗ (i∗ ). (A.6b)

At any period t the home household i face the following one-period budget
constraint:
Pt Ct (i) + Bt (i) + εt Bt∗ (i) ≤ Wt Lt (i) + Πt (i) + Bt−1 (i)(1 + it−1 ) (A.7a)
∗
+εt Bt−1 (i)(1 + i∗t−1 ),

where Pt Ct (i) is nominal spending on the final consumption good, Wt is the

nominal wage per unit of labor, Πt (i) is household’s share of profit of firms in
the intermediate goods sector. Households can purchase two riskless one-period
bonds at any period t: Bt (i), which is denominated in home currency and yields
a nominal interest it at t + 1, and Bt∗ (i), which is denominated in foreign-
currency and yields a nominal interest rate i∗t at t + 1. The nominal exchange
rate, εt , expresses how many units of Home currency are exchanged per one unit
of Foreign currency. Similarly, the one-period budget constraint for the foreign
household i∗ is:

1 Bt−1 ∗
Pt∗ Ct∗ (i∗ ) + ∗
εt Bt (i ) + Bt∗ (i∗ ) ≤ Wt∗ L∗t (i∗ ) + Π∗t (i∗ ) + εt (i )(1 + it−1 ) (A.7b)
∗
+Bt−1 (i∗ )(1 + i∗t−1 ).

In both countries households’ decision on consumption, labor supply and finan-

cial accumulation maximize the present discounted value of utility subject to
the flow of one-period budget constraints:

∞

∞

M ax Et β t ln(Ct (i)) − κLt (i) M ax Et β t ln(Ct∗ (i∗ )) − κ∗ L∗t (i∗ )

Ct (i),Lt (i),Bt (i),Bt∗ (i) Ct∗ (i∗ ),L∗ ∗ ∗
t (i ),Bt (i ),Bt (i )
∗ ∗
t=0 t=0

s.t. (A.7a) s.t. (A.7b)

From the first order conditions we obtain:

1 1
λt = Pt Ct(i) (A.8a) λ∗t = Pt∗ Ct∗(i∗ ) (A.8b)
 

1 1 C∗ P ∗
1+i∗ = Et β CCt+1
t Pt
Pt+1 εt+1 (A.9a) 1+i∗ = Et β C ∗t P ∗t = Et Q∗t,t+1 (i∗ ) (A.9b)
t t t+1 t+1
Wt = −κPt Ct (i)  (A.10a) Wt∗ = −κ∗ Pt∗ Ct∗ (i∗ )  (A.10b)
∗ ∗
1 Ct Pt
1+it = Et β Ct+1 Pt+1 = Et (Qt,t+1 ) (A.11a) 1
1+it = Et β CC∗t PP∗t εt+1 1
. (A.11b)
t+1 t+1

26
From (A.10a), it follows that Ct (i) = Ct as κ, Pt and Wt are common to all home
households. Similarly, by (A.10b) Ct∗ (i) = Ct∗ . Qt,t+1 (A.11a) is the nominal
discount rate of home households, and Q∗t,t+1 (A.9b) is the nominal discount
rate of foreign households.

Intermediate goods sector In each country the intermediate goods sector

is characterised by monopolistic competition: any firm supplies a single variety,
h in the Home country and f in the Foreign country, that is an imperfect
substitute to all other varieties. Technology is represented by a Cobb Douglas
production function and is common across all firms in the same country:

∗ 1−α
Yt (h) = Zt Et1−α (h)Lα ∗ ∗
t (h) (A.12a) Yt (f) = Zt (Et ) (f ) (L∗t )α (f ). (A.12b)
Zt > 0 (Zt∗ > 0) is the total factor productivity and is common across Home
(Foreign) firms. α ∈ (0, 1) is the share of labor income and is common across
firms and countries. Lt (h) and Et (h) (L∗t (f) and Et∗ (f )) are, respectively, the
amount of labor and oil that are employed in the production of variety h (f) at
the period t. The aggregate demand for a given variety, h or f, at time t is
given by aggregating Home and Foreign households’ demands for that variety
(A.1a, A.1b, A.2a, A.2b):

   −θ  −1  −θ  P ∗ −1

1 Pt (h) PH,t 1 Pt∗ (h)
Yt (h) = Ct (h, i)di + Ct∗ (h, i∗ )di∗ = 2 PH,t Pt Ct + 2 ∗
PH,t P∗
H,t
Ct∗ (A.12c)
   −θ  −1  −θ  P ∗t −1
1 Pt (f ) PF,t 1 Pt∗ (f )
Yt∗ (f) = Ct (f, i)di + Ct∗ (f, i∗ )di∗ = 2 PF,t Pt Ct + 2 ∗
PF,t Pt∗
F,t
Ct∗ (A.12d)
 
Ct ≡ Ct (i)di, Ct∗ ≡ Ct∗ (i∗ )di∗ .
In

each∗ country intermediate goods firms take the prices of inputs as given
∗
Wt , Wt , Pe , Pe,t and demand the quantity of labor services (Lt (h) or L∗t (f))
and oil (Et (h) or Et∗ (f )) that solve the following cost minimization problem
subject to the technology constraint:

∗
Min Pe,t Et (h) + Wt Lt (h) M in Pe,t Et∗ (f ) + Wt∗ L∗t (f)
E(h),L(h) Et∗ (f),L∗
t (f )

s.t.Yt (h) = Zt E 1−α (h)Lα (h) s.t.Yt∗ (f ) = Zt∗ (Et∗ )1−α (f) (L∗t )α (f ).
From the first order conditions we obtain the optimal ratio of factor inputs that
is adopted by the Home (A.13a) and Foreign (A.13b) intermediate goods firm,
while the lagrangian multiplier associated to the technology constraint provides
the marginal cost of producing an additional unit of variety h (A.13a) or f
(A.13b):

Wt α Et (h) Wt∗ ∗
α Et (f )
= 1−α Lt (h) (A.13a) ∗ = 1−α L∗ (A.13b)
Pe,t PE,t t (f)
1−α 1−α
PE,t Wtα ( ∗
PE,t ) (Wt∗ )α
M Ct (h) = MCt = Zt αα (1−α)(1−α)
(A.13c) MCt∗ (f) = M Ct∗ = Zt∗ αα (1−α)(1−α)
. (A.13d)

27
Any firm has to set two prices for the variety it produces, one in the Home
market and the other in the Foreign market. These two prices are set by the
firm to maximize its own profit function, which is Πt (h) for home firms and
Π∗t (f ) for foreign firms. The profit functions Πt (h) and Π∗t (f )are defined as
follows:

1 1
Πt (h) = Pt (h) Ct (h, i)di + εt Pt∗ (h) Ct∗ (h, i∗ )di∗ − Wt Lt (h) − Pe,t Et (h)
0 0 (A.14a)
1 1
= (Pt (h) − MCt (h)) Ct (h, i)di + (εt Pt∗ (h) − M Ct (h)) Ct∗ (h, i∗ )di∗
0 0
1 1
Pt (f )
Π∗t (f ) = εt Ct (f, i)di + (Pt∗ (f) − M Ct (f)) Ct∗ (f, i∗ )di∗ − Wt∗ L∗t (f) − Pe,t
∗
Et∗ (f )
0 0 , (A.14b)
1 1
(Pt (f )−MCt∗ (f ))
= εt Ct (f, i)di + (Pt∗ (f) − M Ct∗ (f )) Ct∗ (f, i∗ )di∗
0 0

where the second part of the right hand side of A.(14a) follows from applying
(A.12a), (A.12c) and (A.13a), and the second part of the right hand side of
(A.14b) results from applying (A.12b), (A.12d) and (A.13b).

Monetary Policy In each country the monetary policy stance determines the
aggregate nominal spending by setting a parameter, µt in the Home country
and µ∗t in the Foreign country. This specification allows the monetary authority
to influence the aggregate nominal spending without having to specify the exact
instrument that monetary policy uses for it.
µt = Pt Ct (A.15a) µ∗t = Pt∗ Ct∗ . (A.15b)

The price of oil The nominal price of oil is exogenous 10 and is denominated
in the oil importing country currency. Moreover, for the price of oil it always
holds the law of one price:
∗ Pe,t
Pe,t = .
εt

Symmetric Equilibrium Conditions In a symmetric equilibrium the fol-

lowing conditions hold: Lt (h) = LH,t , Et (h) = Et , L∗t (f ) = L∗F,t , Et∗ (f) =
Et∗ , Pt (f ) = PF,t , Pt (h) = PH,t , Pt∗ (f ) =, PF,t
∗
, Pt∗ (h) = PH,t
∗
, Ct (f, i) =
CF,t , Ct (h, i) = CH,t , Ct (f, i ) = CF,t , Ct (h, i ) = CH,t , Bt = Bt (i), Bt∗ =
∗ ∗ ∗ ∗ ∗ ∗

Bt∗ (i∗ ).
1 0 Though an oil market is not explicitly modelled, the assumption of an exogenus price of

oil amounts to considering a perfectly elastic oil supply (any quantity of oil demanded can
be supplied at the given price) and zero marginal costs. The price of oil is then determined
entirely by a specific tax that is set up by the government of the Foreign country, and whose
revenus are redistributed across households as a lump sum transfer.

28
Clearing Market Conditions Clearing market conditions for the labor mar-
kets, the intermediate goods sector and the final good sector require:

   
Lt (h)dh = Lt (i)di = Lt (A.16a) L∗t (f)df =
L∗t (i∗ )di∗ = L∗t (A.16b)
 1 1  ∗ ∗
∗  12
∗  12
Ct (i)di = Ct = CH,t CF,t 2 2
(A.16c) Ct (i )di = Ct∗ = CH,t CF,t (A.16d)
    
Ct (h, i)di + C(h, i∗ )di∗ dh = Zt Et1−α (h)Lα
t (h)dh ⇒

∗
CH,t + CH,t = Zt Et1−α Lα
t (A.17a)
    
Ct∗ (f, i∗ )di∗ + C(f, i)di df = Zt∗ (Et∗ (f))1−α (L∗t (f ))α df ⇒

∗
CF,t + CF,t = Zt∗ (Et∗ )1−α (L∗t )α . (A.17b)

General condition for the equilibrium nominal exchange rate A gen-

eral condition for the equilibrium nominal exchange rate, which is independent
of the pricing regime, is found by aggregating over Home households’ budget
constraint:

   
Pt Ct − Wt Lt − Πt + Bt (i)di + Bt∗ (i)di + (1 + it−1 )εt Bt−1 (i)di + (1 + i∗t−1 )εt Bt∗ (i)di = 0.

The aggregate profit Πt is defines as follows:



  Πt =
 Πt (h)dh  
= Pt (h) Ct (h, i)di + εt Pt∗ (h) Ct (h, i∗ )di − Wt Lt (h) − Pe,t Et (h) dh
∗ ∗
= PH,t CH,t + εt PH,t CH,t − Wt Lt − Pe,t Et . (A.18)

Let’s impose financial autarky as the equilibrium financial allocations. To do

this (a) consider that both Bt and Bt∗ are in zero net-supply worldwide,
   
Bt (i)di + Bt (i∗ )di∗ = 0, Bt∗ (i)di + Bt∗ (i∗ )di∗ = 0,

(b) assume the following financial allocations: Bt = 0, εt Bt∗ = 0, ∀t. The

aggregate budget constraint then becomes: Pt Ct − Wt Lt − Πt = 0. Replace in
this last expression (A.18) and (A.4a) to find:
∗ ∗
PH,t CH,t = εt PH,t CH,t + Pe,t Et .

This expression states that the nominal exchange rate moves to offset imbalances
in the trade balance. By replacing in it (A.2a), (A.2b), (A.4a), (A 4b), (A.10a) , (A.13a) ,
(A.15a), (A 15b), and (A.17a) I find a general condition for the exchange rate
that depends only on the intermediate goods equilibrium prices:
   ∗ 
µt µ∗ µt 1 µt
PH,t + P ∗
t
= MCt 1−α εtµ − 1 . (A.19)
H,t t

29
 To find the equilibrium of the model the last conditions that are required are
the equilibrium prices of the intermediated goods. These prices are computed
under three different scenarios, each one characterised either by flexible prices or
nominal price rigidities, and by the specific pricing regime that the intermediate
goods firms adopt in the two countries. To define what are the possible pricing
regimes that a firm can adopt let us consider that in general the Home firm
maximizes its profit in the Foreign market by setting the price P̃t (h) = Pt∗ (h)εφt ,
where the parameter φ ∈ [0, 1] is the degree of exchange rate pass-through. If
φ = 1 we are under the producer currency pricing (PCP) regime: the price the
firm sets in the Foreign market is P̃t (h) = Pt∗ (h)εt and is set in Home currency,
so that any change in the nominal exchange rate fully transmits to P̃t (h). If
φ = 0 we are under the local currency pricing (LCP) regime: the price the firm
sets in the Foreign market is P̃t (h) = Pt∗ (h) and is set in Foreign currency,
so that no change in the nominal exchange rate affects P̃t (h). These two price
regimes can also be interpreted in term of the geographical distribution of the
firm’s production sites. Let us consider again the case of an Home firm that
produces h and let us assume that it always sets the price in the producer
currency, but now it can choose how to distribute its production between the
two countries. It follows that the influence of the exchange rate regime on the
price set in the Foreign market will depend on how the total production of h
is distributed between Home and Foreign. Full exchange pass-through would
correspond to the case in which h is entirely produced at Home, while the case
of no exchange pass-through would occur when the production of h has been
divided across the two countries as to serve both markets as a local firm.
The general condition to compute the Aggregate Supply (AS) of the oil im-
porting country can be derived by replacing (A.2a) , (A.2b) , (A.13a) , (A.15a) in
(A.17a):
  12  P ∗  12
PF,t
Ct PH,t + Ct∗ F,t
∗
PH,t = Lt κµ 1
α MCt .
t
(A.20a)

Similarly, the general condition to compute the AS of the oil exporting coun-
try is obtained by applying (A.1a) , (A.1b) , (A.13b) , (A.15b) in (A.17b):
  12  P ∗  12 ∗
PH,t
Ct PF,t + Ct∗ H,t
∗
PF,t = L∗t κµ
α
t 1
MCt∗ . (A.20b)

30
APPENDIX B - The symmetric equilibrium in a
Flexible Price Regime
In a flexible price regime the choice of the invoice currency for sales abroad does
not affect the equilibrium prices of the intermediate goods, which are obtained
by solving the following profit maximization problems:

Home’s firm profit maximization problem

 −θ  −1  −θ  P ∗ −1
Pt (h) PH,t Pt∗ (h)
Max∗ (Pt (h) − M Ct (h)) 12 PH,t Pt Ct +(ǫt Pt∗ (h) − M Ct (h)) 12 ∗
PH,t
H,t
Pt∗ Ct∗
Pt (h),Pt (h)

Foreign’s firms profit maximization problem

 −θ  ∗ −1    −θ  −1
Pt∗ (f ) PF,t
Ct∗ + Ptǫ(f) 1 Pt (f ) PF,t
Max∗ (Pt∗ (f ) − M Ct (f )) 12 ∗
PF,t Pt∗ t
− MCt (f ) 2 PF,t Pt Ct .
Pt (f ),Pt (f )

From the first order conditions we obtain:

θ ∗ θ MCt
PH,t = θ−1 M Ct (B.1a) PH,t = θ−1 ǫt (B.1b)
θ θ
PF,t = θ−1 ǫt M Ct∗ (B.2a) PF,t = θ−1 MCt∗ .
∗
(B.2b)

Given the intermediate goods prices, we can use the conditions previously de-
rived in appendix A to compute the equilibrium of the two-country economy.
The equilibrium nominal exchange rate is found by replacing (B.1a) and (B.1b)
in (A.19) :
 θ 
+(1−α)
εt = ϕF LP µµ∗t , ϕF P = θ−1
θ
−(1−α)
> 1. (B.3)
t θ−1

The parameter ϕF LP captures the asymmetries in the equilibrium that are

caused by the distinction between an oil importing and oil exporting country.
The presence of ϕF LP causes the currency of the oil importing country to de-
preciate in equilibrium with respect to the case of no oil in the economy (as
in Corsetti and Pesenti). ϕF LP is an increasing function of both the degree of
monopolistic power in the Home intermediate goods sector (θ) and its share of
oil costs (1 − α). It approaches 1 as both (1 − α) goes to 0 (no oil is required
to produce) and θ approaches 1, as a larger monopolistic power implies both
a lower equilibrium production, which requires fewer oil imports, and larger
revenues from exports.
The equilibrium value of consumption is found by applying (A.1a) , (A.2a) , (A.4a),
1 1
(A.15a) , (B.1a) and (B.1b) to Ct = CH,t 2 CF,t 2

α α  1−α 1
(1 − α) PµE,t
θ−1 α
Ct = 2θ κ
t
(Zt Zt∗ ) 2 ϕF LP − 2 . (B.4a)

Similarly, Ct∗ is obtained from applying (A.1b) , (A.2b) , (A.4b) , (A.15b) ,

∗ 12 ∗ 12
and (B.2a) , (B.2b) to Ct∗ = CH,t CF,t :

31

α α  (1−α) 1
(Zt Zt∗ ) 2 ϕF LP (1− 2 ) . (B.4b)
µt α
θ−1
Ct∗ = 2θ κ (1 − α) PE,t

Equilibrium employment in the oil importing country is obtained by replacing

(B.3) in (A.19), and in the oil exporting country by substituting (A.4b), (A.13b),
(A.15b) in (A.17b) :

   
α ϕF LP α 1
Lt = L = κ θ (B.5a) L∗t = L∗ = κ θ . (B.5b)
θ−1 +(1−α) θ−1 +(1−α)

To get Et , the quantity of oil used in equilibrium in the oil importing country,
replace (B.5a) in (A.13a) , and similarly for Et∗ replace (B.5b) in (A.13b):
  
µ∗
Et = PµE,t
t
θ
1−α
−(1−α)
(B.6a) E t
∗
= θ
1−α
+(1−α) P ∗
t
. (B.6b)
θ−1 θ−1 E,t

The equilibrium price level of each country is found replacing the equilib-
rium consumption level, (B.4a) for Ct and (B.4b) for Ct∗ , in the correspond-
ing monetary policy stance equation, (A.15a) for the oil importing country
and (A.15b) for the oil exporting country:
1
∗ − 2 2θ

κ α  PE,t 1−α F LP α
Pt = µα
t (Zt Zt ) θ−1 α 1−α ϕ 2 (B.7a)

 α  −(1−α) 1
(Zt Zt∗ )− 2 ϕF LP −(1− 2 ) . (B.7b)
α
2θ
Pt∗ = θ−1 κ
α (1 − α) PµE,t
t

The ratio of foreign to home consumption in equilibrium is easily computed

from (B.4a) and (B.4b) :
Ct∗
Ct = ϕF LP > 1 ⇒ Ct∗ > Ct .

Similarly, (B.7a) and (B.7b) give the ratio of foreign to home employment in
equilibrium:
L∗ 1
t
Lt = ϕF LP < 1 ⇒ L∗t > Lt .

Finally, the AS in the flexible price regime is obtained by applying (B.1a) , (B.1b) ,
(B.2a) , (B.2b) , (B.4a) and (B.4b) to(A.20a) :

  12  α
−1
κµt 1 M̃Ct∗ MCt∗
AS F LP : Ct = τ F
t
LP
Lt , τ F
t
LP
= α MCt MCt εt2 1 + ϕF LP , M̃ Ct∗ = εt1−α
.
Supply Channel T rade Channel

The index τ F t
LP
in the AS schedule captures both the supply channel (the
condition of intratemporal marginal substitution between labor and energy) and
the trade channel of transmission of oil price shocks. However, with identical
technology across countries and no change in the nominal exchange rate, oil
price shocks do not change the terms of trade, so that only the supply channel
is at work in the transmission of oil price shocks.

32
 The effect of price shocks to the oil importing and oil
exporting economy in a Flexible Price Regime
To understand the effects of oil price shocks to the economy I compute the
elasticity of any variable of interest to the nominal price of oil. In particular,
I define the elasticity of the variable Yt to the nominal price of oil Pe,t as
∂Yt Pe,t
χYt ,P e,t = ∂Pe,t Yt
, that is the percent change in Yt as the price of oil rises
by one percent.

Proposition 1 In a flexible price regime the nominal exchange rate is unaf-

fected by oil price shocks.

Proof. To prove it, it is sufficient to show χεt ,Pe,t = 0, which follows immedi-
ately from (B.3) .

Proposition 2 In a flexible price regime, oil price shocks decrease the aggregate
consumption in both the home importing and the oil exporting country.

Proof. To prove it, it is sufficient to compute the elasticity of Ct and Ct∗ to

Pe,t . From (B.4a) and (B.4b) I obtain χCt ,Pe,t = χCt∗ ,Pe,t = − (1 − α) < 0.

Proposition 3 In any price regime, χCt ,Pe,t = −χPt ,Pe,t and χCt∗ ,Pe,t = −χPt∗ ,Pe,t .
µ ∂Pt
∂Ct Pe,t ∂ ( Pt ) Pe,t −µt ∂P
e,t Pt
Proof. By (A.15a) follows that χCt ,Pe,t = ∂Pe,t Ct = t
∂Pe,t P µt = Pt2 µt Pe,t =
t
∂Pt Pe,t
− ∂P e,t Pt
= −χPt ,Pe,t . Similarly, it is possible to compute χCt∗ ,Pe,t = −χPt∗ ,Pe,t .

By Proposition 3 it follows that with flexible prices oil price shocks increase
the price level in both country as χPt ,Pe,t = χPt∗ ,Pe,t = (1 − α) .

33
 APPENDIX C - The symmetric equilibrium
with Nominal Price rigidities and PCP Regime
Nominal price rigidities are introduced in the intermediate sector as firms set
the prices one period ahead. Under the producer currency pricing (PCP) the
firms fix their prices in the producer currency, and the equilibrium prices solve
the following profit maximization problems:

Home Price setting

  −θ  ∗ −θ 
Pt (h) ∗ εt Pt (h) ∗
M ax Et−1 Qt−1,t (Pt (h) − MCt (h)) PH,t CH,t + (ǫt Pt (h) − M Ct (h)) εt P ∗ CH,t
Pt (h),εt Pt∗ (h) H,t

foreign price setting

  −θ  ∗ −θ 
Pt (f ) (f ) P (f )
Max Et−1 Q∗t−1,t ǫt − M Ct (f ) ǫǫttPPtF,t CF,t + (Pt∗ (f ) − MCt (f )) Pt ∗ ∗
CF,t
Pt (f ) ∗ F,t
ǫt ,Pt (f )

Replacing (A.4a, A.11a, A.15a and A.15b) in the first order conditions of the
Home profit maximization problem and replacing (A.4b, A.9b, A.15a and A.15b)
in the of the Foreign profit maximization problem we obtain:

 µ∗ 
Et−1 t MCt
θ ∗ θ µt
PH,t = θ−1 Et−1 [M Ct ] (C.1a) PH,t = θ−1
 µ∗  (C.1b)
εt Et−1 µt
 t 
µ
Et−1 µ∗t MCt∗
∗ θ ∗ θ
PF,t = θ−1 Et−1 [MCt ] (C.2a) PF,t = θ−1 εt
t

µ
 . (C.2b)
Et−1 µ∗t
t

The equilibrium nominal exchange rate is found by replacing (C.1a) and (C.1b)
in (A.19):
 
CP µt 1+k0 MCt
εt = ϕPt µ∗ (C.3) ϕ P CP
t = 1−k1 MCt > 1,
t  µ∗ 
(1−α) Et−1 µt
θ−1 θ−1
k0 = θ Et−1 [MCt ] > 0, k1 = (1 − α) θ E
 t
µ∗
t
 > 0.
t−1 MCt µ
t

As it makes little economic sense to have a negative nominal exchange rate,

condition (C.3) sets an upper bound to the magnitude of the oil price shock
µ∗
Et−1 [MCt ]Et−1 µt
MCt θ 1 1
as E[MCt ] < θ−1 (1−α)
 µ∗
t = k1 . It is also plausible to assume
Et−1 MCt µt
   1−α
t 
∗ Pe,t kµα ∗
Cov MCt , µµt = Cov tt
, µt
Zt αα (1−α)(1−α) µt
≤ 0, which amounts to set the fol-
t

lowing weak restrictions on the joint distribution of the exogenous variables

(Zt , µ∗t , µt ) : (i) Zt is uncorrelated to both monetary policies and (ii) µ∗t is not
perfectly positively correlated with µt .
 
µ∗
Proposition 4 If Cov M Ct , µt ≤ 0 then k1 ≥ k0 .
t

34
  µ∗   µ∗ 
Et−1 t Et−1 µt
θ 1 1 µt
 µ∗  θ 1 θ 1
Proof. k0 = θ−1 (1−α) Et−1 [MCt ] = t θ−1 (1−α) = θ−1 (1−α) E
 µ∗
 t

t −Cov MC , µt
∗ ≤
Et−1 [MCt ]Et−1 µt t−1 MCt µ t µ
t t
 µ∗ 
Et−1 µt
θ 1
θ−1 (1−α) E
 t
µ∗
t
 = k1 .
t−1 MCt µ
t

To find the equilibrium value of consumption in the home oil importing coun-
try we apply (A.1a) , (A.2a) , (A.3a) , (A.4a), (A.15a) and (C.1a) , (C.1b) to Ct =
1 1
CH,t 2 CF,t 2 :
 
µt
  12
1 θ−1
1 Et−1 µ∗

P CP − 12
Ct = 2 θ (µt µ∗t ) 2 
µt
t ϕt . (C.4a)
Et−1 µ∗
MCt∗ Et−1 [MCt ]
t

Similarly, to get the equilibrium value of consumption in the oil exporting coun-
try we apply (A.1b) , (A.2b) , (A.3b) , (A.4b) , (A.15a) and (C.2a) , (C.2b) to Ct∗ =

∗  12
∗  12
CH,t CF,t :
  µ∗   12
1 θ−1
1 Et−1 t
µt

P CP  12
Ct∗ = 2 θ (µt µ∗t ) 2  µ∗
t
 ϕt . (C.4b)
Et−1 µt MCt Et−1 [MCt∗ ]

Equilibrium employment in the oil importing country is obtained by replacing

(C.3) in (A.19) :
  
µ∗t
 
Et−1 [MCt ]Et−1
 µ t 
 1+ 
µ∗
 
α θ−1 MCt
 Et−1 MCt t
µt 
Lt =    , (C.5a)
2κ θ Et−1 (MCt )  MCt Et−1
µ∗
t 
 µt 
1− θ−1
θ (1−α)

µ∗

Et−1 MCt t
µt

while employment in the oil exporting country is obtained from applying (A.4a), (A.4b),
(A.13b), (A.15b), (C.2a), (C.2b) to (A.17b),
  
µt
MCt∗ Et−1 µ∗
Et−1 [MCt∗ ]
α θ−1 1
L∗t = 2k θ Et−1 [MCt∗ ] ϕP CP
t
µt
 + 1 . (C.5b)
t Et−1 µ∗
MCt∗
t

To get Et , the quantity of oil used in equilibrium in the oil importing country,
replace (C.5a) in (A.13a) , and similarly for Et∗ replace (C.5b) in (A.13b):
  
µ∗
 
Et−1 [MCt ]Et−1 t
 µ t 
 1+ 
µ∗
 
1−α θ−1 MCt µt  Et−1 MCt t
µt 
Et =    (C.6a)
2 θ Et−1 (MCt ) Pet  MCt Et−1
µ∗
t 
 µt 
1− θ−1
θ (1−α)

µ∗

Et−1 MCt t
µt
  
µt
MCt∗ µt Et−1 µ∗
Et−1 [MCt∗ ]
1−α θ−1
Et∗ = 2 θ Et−1 [MCt∗ ] Pe,t
∗
t
µt
 + ϕP
t
CP
. (C.6b)
Et−1 µ∗
MCt∗
t

35
The equilibrium price level of each country is found replacing the equilib-
rium consumption level, (C.4a) for Ct and (C.4b) for Ct∗ , in the correspond-
ing monetary policy stance equation, (A.15a) for the oil importing country
and (A.15b) for the oil exporting country:
   − 12
  12 Et−1
µt

P CP  12
2θ µt µ∗
Pt = θ−1 µ∗

µt
t ϕt (C.7a)
t Et−1 µ∗
MCt∗ Et−1 [MCt ]
t
  µ∗   12
  12 Et−1 [MCt∗ ]Et−1 µt MCt
P CP − 12
µ∗ 2θ
Pt∗ = t
µt θ−1
 µ∗  t
ϕt . (C.7b)
Et−1 µt
t

Finally, with nominal rigidities and under PCP prices the AS of the oil importing
country is obtained by applying (C.1a) , (C.1b) , (C.2a) , (C.2b) , (C.4a) and
(C.4b) to(A.20a) :

 1
−1
κµt 1
AS P CP : Ct = τ P
t
CP
Lt , τP
t
CP
= α MCt εt2 ΥP
P
CP
+ ΥP CP P CP P CP
P ∗ ϕt ΥC
Supply Channel T rade Channel
 
µ∗
1
2
    12
t µ

Et−1
µt  Et−1 µ∗t MCt∗
 
µ∗
 ΥP
P
CP
= 
µ
t εt
Et−1 t MC E
µt t t−1 [MCt ]
∗ Et−1 µ∗t Et−1 [MCt ]
ΥP
C
CP
=   1 , 
t
 µ∗   12 .
µt 2
Et−1 Et−1 t
µ∗ µt

 
µt
t

 ΥP
P∗
CP
=  µ∗
t
 εt
Et−1
µ∗
MCt∗ Et−1 [MCt ] Et−1 µt MCt Et−1 [MCt∗ ]
t

Following a similar procedure, the AS for the oil exporting country is:

 −1
1
κµ∗ 1 1 1 1
AS P CP : Ct∗ = τ ∗P
t
CP
Lt τ ∗P
t
CP
= 2
α MCt∗ εt
t
ϕP CP ΥP CP ΥP CP + 1
t C P ε 2 ΥP
P∗
CP

Computing the effect of oil price shocks to the oil

importing and oil exporting economy

Proposition 5 In presence of nominal price rigidities and under a PCP Regime

a positive (negative) oil price shock depreciates (appreciates) the currency of the
oil importing country

Proof. To prove it, it is sufficient to show that the elasticity of εt to the price
of oil is positive. From (C.3) , χεt ,Pe,t = (1 − α) MCt (1−k1 MCk0t +k
)(1+k0 MCt ) > 0.
1

Proposition 6 In presence of nominal price rigidities and under a PCP Regime

the impact of an oil price shock to the economy increases more than proportion-
ally to the size of the shock.

36
Proof. To prove it, it is sufficient to show that χεt ,Pe,t increases with the size
of the oil price
 shock,
 Pe,t . From proposition 5 is easy to obtain: 
∂χεt ,P e,t 
∂Pe,t = ∂Pe,t (1 − α) (k0 + k1 ) [(1 − k1 MCt ) (1 + k0 MCt )]−2 1 + k0 k1 MCt2 >
∂MCt

0.

Proposition 7 In presence of nominal price rigidities and under a PCP Regime

a positive (negative) oil price shock reduces (increases) the aggregate consump-
tion in the oil importing country.

Proof. To prove it, it is sufficient to show that the elasticity of consumption of

the oil importing country to the price of oil is negative, that is,
    12
µt
−1 Et−1 µ∗
1 θ−1
given Ct = kC εt 2 , kC = 2 µt θ
t 
µt
 > 0, it
Et−1 [MCt (h)]Et−1 µ∗
MCt∗
t
follows that:
∂Ct ∂εt Pe,t εt − 32 ∂εt Pe,t εt ∂εt Pe,t 1
χCt ,Pe,t = ∂εt ∂Pe,t Ct εt = − 12 kC εt ∂Pe,t Ct εt = − 12 Ct ∂P e,t εt Ct
=
1
− 2 χεt ,Pe,t .

Proposition 8 In presence of nominal price rigidities and under a PCP Regime

a positive (negative) oil price shock increases (decreases) the aggregate consump-
tion in the oil exporting country.

Proof. To prove it, it is sufficient to show that the elasticity of consumption to

the price of oil is in the oil exporting country is positive, that is, given
  ∗   12 µ
1 Et−1 µt
1 θ−1 ∗ 1
Ct∗ = kC
∗ 2 ∗
εt , kC = 2 θ µt Et−1 [MCt∗ ] E
 µ∗ t
t
 > 0, it follows that:
t−1 µ MCt
t
∂Ct∗ ∂εt Pe,t εt − 12 ∂εt Pe,t εt
χCt∗ ,Pe,t = ∂εt ∂Pe,t Ct∗ εt = 12 kC εt ∂Pe,t Ct∗ εt = 12 Ct∗ χεt ,Pe,t C1∗ = 12 χεt ,Pe,t >
t
0.

Proposition 9 In presence of nominal price rigidities and under a PCP Regime

after a positive (negative) oil price shock employment in the oil importing coun-
try increases (decreases) more than proportionally to the depreciation (appreci-
ation) of the nominal exchange rate.

Proof. To prove it, it is sufficient to show that the elasticity of the oil im-
porting country employment to the price of oil is positive and larger than the
elasticity of the nominal exchange rate to the price of oil. From (C.5a) it fol-
t ∂εt Pe,t ∂Lt Pe,t ∂εt Pe,t εt ∂Lt Pe,t εt
lows that χLt ,Pe,t = ∂L
∂εt ∂Pe,t Lt = ∂εt Lt ∂Pe,t εt Pe,t = ∂εt Lt Pe,t χε,PE =
∗
1 α µt εt ϕP CP
ϕP CP

2κ 1−α µt Lt χε,Pe,t = t
χεt ,Pe,t > χεt ,Pe,t ., as t
>1
(ϕPt CP −1) (ϕPt CP −1)

Proposition 10 In presence of nominal price rigidities and under a PCP Regime

the larger the size of a positive oil price shock the bigger the probability that
employment in the oil exporting country will shrink.

37
Proof. It is sufficient to show that the elasticity of foreign employment to
the price of oil is positive for oil price shocks of small size but turns neg-
ative for oil shocks of large size. From (C.5b) we can compute χL∗ ,Pe,t =
 
∂L∗ t Pe,t 1 [CF,t −CF,t ]
∗
3  [C ∗ −CF,t ]
∗ = (1 − α) + χ ε,Pe 2 [C ∗ +CF,t ] − − α . As CF,t < 1, it
∂Pe,t Lt
F,t
2 [ F,t
∗ +C
F,t ]
 
[C ∗ −CF,t ]
3 
follows that 12 CF,t − 2 −α < 0,thus χL∗t ,Pe,t < 0 for χεt ,Pe,t >
[ F,t F,t ]
∗ +C

(1−α) < (1 − α) . By proposition 6 χ

εt ,Pe,t increases in the mag-

[C∗ −CF,t ]
( 2 −α)− 12 CF,t
3
[ F,t
∗ +C
F,t ]
nitude of the shocks and so the larger the size of the oil price shock the more
likely that L∗t will decrease.

The economic intuition behind Proposition 10 comes from considering that oil
price shocks affect oil exporting country employment through a supply and a
trade channel. From (A17b) and (A13b) we obtain:
−(1−α)

L∗t = KL∗ (Pe,t )1−α εt ∗
CF,t + CF,t , KL∗ > 0,
Supply Eff ect T rade channel

where KL∗ is a combination of terms independent of Pe,t . The supply channel

1−α
of oil price shocks on L∗t passes through Pe,t . As the production function is a
Cobb Douglas, the elasticity of substitution of labor for oil is 1, so that after a
positive oil price shock firms replace labor for oil. However, as the price of oil is
∗
exogenously fixed in Home currency, Pe,t depends also on εt , that is the trade
channel contrasts the supply channel. To see let’s compute the elasticity of the
price of oil in foreign currency to Pe,t :

P 

e,t εt
∂ εt Pe,t εt −Pe,t ∂ Pe,t
χPe,t
∗ ,P
e,t
= Pe,t ∗
Pe,t = ε2t
εt = (1 − χεt ,Pe,t ).

It follows that for sufficiently small positive oil price shocks the supply chan-
nel increases foreign employment, while for large shocks it decreases it. How-
ever, the trade channel affects foreign employment also via a wealth and a
substitution

effect on the aggregate demand for foreign intermediate goods
∗
YF,t = CF,t + CF,t . By the wealth effect home households feel poorer, con-
sume less and so lower the demand for foreign intermediate goods (CF,t ), as
opposed to foreign households who feel richer,
∗consume
 more and so increase
the demand for foreign intermediate goods CF,t . By the substitution effect
both Home and Foreign firms in the final good sector switch their demand
for inputs from foreign to home produced intermediate goods. The final effect
of the trade channel on the aggregate demand for foreign intermediate goods
(YF,t ) depends on the specific functional forms of preferences and technology.
Within the model, the trade channel shrinks (YF,t ) for small shocks, while its
impact on (YF,t ) is ambiguous for large shocks. To prove it let’s compute the
elasticity of YF,t to the price of oil. From (A.1b), (C.2a) and (C.1b) we ob-
∗ −1 ∗
tain : CF,t = εt 2 KCF Ct∗ , KCF
∗ ∗
> 0, where KCF is a combination of terms
independent of Pe,t . It follows that:

38
     12
−1  µ∗ 
∗
∂ KCF Ct∗ εt 2   Et−1 µt MCt
P e,t
χCF,t
∗ ,P
e,t
= ∂(Pe,t ) ∗
CF,t = 12 χεt ,Pe,t χεt ,Pe,t − 1 ≶ 0 , KCF
∗
=2  µ∗  t >0
Et−1 µt Et−1 [MCt∗ ]
t

−1
Similarly, from (A.1a), (C.1a) and (C.2b) we obtain : CF,t = εt 2 KCF Ct , KCF >
0, and KCF is a combination of terms independent of Pe,t . It follows that

    12
  µ
Et−1 [MCt ]Et−1 µ∗t
χCF,t ,Pe,t = − 12 χεt ,Pe,t χεt ,Pe,t + 1 < 0, KCF = 2 
µ
t >0
Et−1 µ∗t MCt∗
t

Thus, the trade channel always reduces the oil importing country demand for
oil exporting intermediate goods, while it reduces the oil exporting country
demand for oil exporting intermediate goods only sufficiently small shocks (as
∂χεt ,Pe,t
∂Pe,t > 0 and χCF,t
∗ ,P e,t < 0 for χε ,P
t e,t
> 1). The final effect of oil price
shocks on the aggregate demand for YF is given by:

∗    ∗ 
CF,t CF,t CF,t
χYF,t ,Pe,t = ∗ χCF,t ,Pe,t
CF,t +CF,t + ∗ χC ∗ ,Pe,t
CF,t +CF,t F,t
= 12 χεt ,Pe,t − CF,tC+C
F,t
∗ 1 + χεt ,Pe,t + ∗
CF,t +CF,t χε
F,t

39
 APPENDIX D - The symmetric equilibrium
with Nominal Price rigidities and LCP Regime
Nominal price rigidities are introduced in the intermediate sector as firms
set the prices one period ahead. Under the local currency pricing (LCP) firms
in the intermediate goods sector fix the price in the export market in the export
market currency, and set the price to solve the following profit maximization
problems:

Home Price setting

  −θ  ∗ −θ 
(h) ε P (h)
M ax Et−1 Qt−1,t (Pt (h) − MCt (h)) PPtH,t CH,t + (ǫt Pt∗ (h) − M Ct (h)) εtt Pt ∗ ∗
CH,t
Pt (h),εt Pt∗ (h) H,t

foreign price setting

  −θ  ∗ −θ 
Pt (f ) ǫt Pt (f ) Pt (f )
Max Et−1 Q∗t−1,t ǫt − M Ct (f ) ǫt PF,t CF,t + (Pt
∗
(f ) − MCt (f )) P ∗ C ∗
F,t
Pt (f ) ∗ F,t
ǫt ,Pt (f )

Replacing (A.4a, A.11a, A.15a and A.15b) in the first order conditions of the
Home profit maximization problem and replacing (A.4b, A.9b, A.15a and A.15b)
in the of the Foreign profit maximization problem we obtain:

 µ∗ 
Et−1 µt MC
θ θ
Pt (h) = PH,t = θ−1 Et−1 [MCt ] (D.1a) Pt∗ (h) = ∗
PH,t= θ−1  t
µ∗
 (D.1b)
Et−1 µt εt
 t 
µt
Et−1 µ∗ MCt∗
θ θ
Pt∗ (f) = PF,t
∗
= ∗
θ−1 Et−1 [MCt ] (D.2a) Pt (f) = PF,t = θ−1 t
µ
 . (D.2b)
Et−1 εtµt∗
t

The equilibrium nominal exchange rate is found by replacing (D.1a) and (D.1b)
in (A.19):
µt LCP
εt = µ∗ ϕt (Pe,t ) (D.3)
t

  
µ∗
θ−1 MCt MCt µt
ϕLCP
t (Pe,t ) = 1+ θ (1 − α) Et−1 [MCt ] +ϕ F LP
µ∗ t  >1
Et−1 µt MCt
t

To find the equilibrium value of consumption in the home oil importing country
we apply (A.1a) , (A.2a) , (A.3a) , (A.4a), (A.15a) and (D.1a) , (D.1b) to Ct =
1 1
CH,t 2 CF,t 2 :
    12
µt
Et−1 εt µ∗
1 θ−1 1
Ct = 2 θ µt Et−1 [MCt ] E

µt
t  . (D.4a)
t−1 µ∗
MCt∗
t

Similarly, to get the equilibrium value of consumption in the oil exporting coun-
try we apply (A.1b) , (A.2b) , (A.3b) , (A.4b) , (A.15a) and (D.2a) , (D.2b) to Ct∗ =

∗  12
∗  12
CH,t CF,t :

40
   12
FP
Ct∗ = 1 θ−1 ∗
ϕµ∗ 1
(D.4b)
2 θ µt .
 ∗
Et−1 µt MCt Et−1 [MCt ]
t

Equilibrium employment in the oil importing country is obtained by replacing

(D.3) in (A.19) :
 ∗

µt
α θ−1 1
Lt = 2κ θ M Ct Et−1 [MCt ] + ϕF LP µ
t
t

µ∗  . (D.5a)
Et−1 µt MCt

Employment in the oil exporting country is obtained from applying (A.4a), (A.4b), (A.13b),
(A.15b), (D.2a), (D.2b) to (A.17b),:

  µ
MCt∗ µ∗t ∗
∗ α θ−1 µt MC
Lt = 2κ θ Et−1 εt µ∗ 
µt
t
∗
 + t
Et−1 [MC ∗ ] (D.5b)
t Et−1 µ∗
MCt t
t

To get Et , the quantity of oil used in equilibrium in the oil importing country,
replace (C.5a) in (A.13a) , and similarly to get Et∗ replace (D.5b) in (A.13b):

 
κµt 1−α θ−1 1 µ∗
Et = + ϕF LP µt
µ1∗ (D.6a)
θ MCt

Pe,t 2κ Et−1 [MCt ] t Et−1 t
µ MCt
t 
  MCt∗ µ∗t
µ
1−α kµ θ−1 LCP µt MCt∗
Et∗ = 2κ Pe,t θ ϕt Et−1 εt µ∗

µ
t  + Et−1 [MCt∗ ] (D.6b)
t Et−1 µ∗t MCt∗
t

The equilibrium price level of each country is found replacing the equilib-
rium consumption level, (D.4a) for Ct and (D.4b) for Ct∗ , in the correspond-
ing monetary policy stance equation, (A.15a) for the oil importing country
and (A.15b) for the oil exporting country:
  
µt
1 Et−1 µ∗
MCt∗
2θ
Pt = θ−1 (Et−1 [M Ct ]) 2 t
µt
 (D.7a)
Et−1 εtµ∗
t
  ∗   12
2 θ µ
Pt∗ = ϕLCP θ−1 Et−1 µt MCt Et−1 [MCt∗ ] (D.7b)
t t

Finally, with nominal rigidities and under LCP prices the AS of the oil importing
country is obtained by applying (D.1a) , (D.1b) , (D.2a) , (D.2b) , (D.4a) and (D.4b) to(A.20a) :

κµt 1  LCP −1

AS LCP : Ct = τ LCP
t Lt τ LCP
t = α MCt ΥP + ΥLCP
C ΥLCP
P∗
Supply Channel T rade Channel

 
µt
 12     12
MCt∗ µt
Et−1
µ∗ Et−1 µ∗
MCt∗
  t
µ∗

 ΥLCP
P = 
µt
t ,
µ∗  F LP Et−1 t MC
µt t  Et−1 εtµ∗
Et−1 [MCt ]
ΥLCP = t ϕ  
t
C µt  Et−1 [MCt∗ ]

µt 
µ∗   12
 Et−1 [MCt ] Et−1 εt µ∗  Et−1 [MCt∗ ]Et−1 µt εt
t
ΥLCP
P∗ =
µ∗ t
Et−1 µt MCt
t

41
 Computing the effect of oil price shocks to the oil
importing and oil exporting economy

Proposition 11 In presence of nominal price rigidities and under a LCP Regime

a positive (negative) oil price shock depreciates (appreciates) the currency of the
oil importing country.

Proof. To prove it, it is sufficient to show that the elasticity of the nominal
exchange rate to the price of oil
is greater than
 zero. From D.3,  µ∗ 
MCt 1
Et−1 [MC] ϕF LP +ϕE,t MCt Et−1 [MC]Et−1 µt
χεt ,Pe,t = (1 − α)

MCt


1
 < (1 − α) , ϕE,t = Et−1 [MC]
 µ∗
t >
1+ E +ϕE,t Et−1 MC µt
t−1 [MC] ϕF LP t
0.

Proposition 12 In presence of nominal price rigidities and under a LCP Regime

in both the oil importing and oil exporting country oil price shocks affect neither
the consumption nor the price level

Proof. To prove it, let’s compute the elasticity of Ct and Ct∗ to the price of oil.
From (D.4a) and (D.4b), χCt ,Pe,t = χCt∗ ,Pe,t = 0. By proposition 3 it follows
that χPt ,Pe,t = −χCt ,Pe,t = 0 and χPt∗ ,Pe,t = −χCt∗ ,Pe,t = 0.

Proposition 13 In presence of nominal price rigidities and under a LCP Regime

a positive (negative) oil price shock increases (decreases) employment in the oil
importing country.

Proof. To prove it, let’s compute the elasticity of Lt to the price of oil. From
∂Lt Pe,t t Pe,t
(D.5a) . χLt ,Pe,t = ∂Pe,t Lt
= (1 − α) PLE,t Lt = (1 − α) .

Proposition 14 In presence of nominal price rigidities and under a LCP Regime

after a positive (negative) oil price shock employment in the oil exporting country
increases (decreases), but by less than the increase in the oil importing country.

Proof. To prove it, let’s compute the elasticity

 of Lt to the price of oil. From 
∂L∗ L∗
(D.5b). χL∗t ,Pe,t = ∂Pe,t L∗ = Pe,t (1 − α) 1 − εt ∂Pe,t PLe,t
t Pe,t t Pe,t ∂εt
t
= (1 − α) 1 − χε t ,Pe,t
<
t
(1 − α) = χLt ,Pe,t .

42