UNIVERSITY OF GRONINGEN
FACULTY OF ECONOMICS AND BUSINESS
OLIVIER LECLERC
Global Volatility Accounting
Table of Content
Abstract ... 0
Introduction ... 1
1. Accounting Framework ... 4
1.1. Volatility Level Accounting ... 4
1.2. Volatility Change Accounting ... 6
2. Quantitative Analysis ... 7
2.1. The Source Data ... 7
2.2. Descriptive Analysis ... 8
3. An Account for Global Volatility ... 10
3.1. Volatility Level Accounting and Its Change Over Time ... 10
3.2. Volatility Change Accounting ... 12
4. An Account of Volatility Within Trading Zones... 14
4.1. Motivation ... 14
4.2. Conventionally defined trading zones ... 14
4.3. Alternative definition of trading zones... 17
5. Counterfactuals ... 18
6. Conclusion ... 19
References ... 21
0
Abstract
This paper develops a unifying accounting framework that tracks the level and evolution of global volatility according to a variety of sources, ranging from the ‘Centre’ and the ‘Periphery’, their constituent economies along with the underlying sectors, and their comovements. Our contribution brings together some important and yet independent strands of economic literature—the ‘Great Moderation (GM)’, the macro volatility and development, structural transformation, and international business cycle synchronization. Using a sample of 30
economies at various stages of development over the 1971-2007 period, our results emphasize that the GM is a global phenomenon largely due to the ‘Centre’, while the ‘Periphery’ has experienced an increase in the volatility. There seems to be a an ‘offshoring’ of volatility from the ‘Centre’ to the ‘Periphery’ as the latter gradually proceeds with the process of structural transformation of its constituent economies. This result suggests that, as the ‘Periphery’ gained the status of engine of global economic growth, it also turned into a cushioning of the
fluctuations of the ‘Centre.’ Our analysis shows that, on a global scale, the structural
transformation affecting all countries is akin to a zero-sum phenomenon – a result corroborated by a counterfactual exercise where this process is nullified.
Key words: Volatility accounting, great moderation, developing countries, developed countries,
1
Introduction
During the two decades preceding the Great Recession the global economy underwent a profound mutation driven by lower barriers to trade and investment, technical innovations, as well as the desire of firm for more profitable opportunities. This reshaping of the economy was characterised by a sustained decline in the relative size of the manufacturing sector in the developed nations, parallel to a surge of production of consumer and industrial goods in emerging countries. Good manufacturing was further unbundled, away from the developed economies where their production stages were traditionally concentrated, and into various newly-industrialised nations. This fragmentation of vertically integrated production structures entailed a rise of intra-industry trade between nations, thereby augmenting business cycle synchronisation.
This period also displayed a reversal in the growth trend of the global economy. From the early 1990s, the emergence of developing nations became the engine of world growth and thus attracted the attention of the literature. The link between development level and growth has been well documented and widely understood (see Rodrik 2011 for example); but alongside this phenomenon a more recent line of research has unfolded, investigating the association between level of development and macroeconomic volatility (Lucas 1988). The negative correlation between development and volatility has first been explained by Acemoglu and Zilibotti (1997) in a theory of market incompleteness and financial diversification. Koren and Tenreyro (2013) later extended this theory showing that the lack of technological diversification explained the higher volatility of countries in their early development stages. Koren and Tenreyro (2007) provide the first comprehensive empirical account of the factors driving volatility, highlighting in particular the relative importance of idiosyncratic and macroeconomic shocks and the way in which economies react to them.
2
therein). This period of lower volatility has been analysed by Gali and Gambetti (2009), Stiroh (2010) and Moro (2013), who emphasised the role of the structural change as a possible source for the moderation. In contrast, the literature analysing the macroeconomic volatility in
developing countries focused on a wide range of determinants but did not suitably consider the effect of structural changes.
These concomitant developments of the literature on macroeconomic volatility raise two sets of questions.
First, what is the relationship between the evolution of macroeconomic volatility in developed countries and that of developing countries? As these development groups move up the ladder of structural transformation, the simultaneous evolution of their macroeconomic volatility could amount to a zero-sum game. Bergin et al. (2009) and Coronado (2011) show that
offshoring manufacturing activities from the U.S. to the Mexican maquiladora contributed to an increase in the volatility of the Mexican industry, in effect, volatility was offshored. A similar phenomenon has been documented to take place in the EU by Levasseur (2010), whereby less developed member states experience volatility increases. These findings show the existence of an important interaction between developed and developing countries, largely overlooked by the literature. The intuition suggests that, by offshoring industrial activity, developed countries hedge against volatility while developing countries bear the volatility increase resulting from the relocation of activity. In order to assess this hypothesis, we design a unified framework of global volatility accounting, inclusive of developed and developing countries.
Second, how important is the synchronisation of business cycles between different regions of the globe, and what are the channels leading to the transmission of shocks? These notions relate to the link between trade openness and volatility. For instance, the combination of high
3
sectors in which they specialise are more impactful. But this widely held view has been called into question by a more recent stand of literature. Koren and Tenreyro (2007) explain that idiosyncratic sectoral shocks are as important as macroeconomic shocks in explaining the volatility of developing countries. In addition, Caselli et al. (2014) show that openness to international trade leads to diversification of supply and demand, which in turns reduces volatility.
This paper brings together the various strands of literature outlined above by offering a unifying volatility accounting framework. The framework takes a global perspective fragmented along various dimensions. First, we distinguish two blocs of countries: the Centre (developed countries) and the Periphery (developing countries). Our analysis reports how much each of these blocs and their comovements contribute to global volatility during two periods: the great moderation (1985-2007) and its earlier setting (1971-1984). The contribution of each bloc is then further broken down into the global volatility arising from ‘within’ countries and that arising ‘between’ countries (comovements of sectors across countries of the bloc). The ‘within’
component comprise an intra-sectoral subcomponent as well as an inter-sectoral subcomponent (comovements between the constituent sectors of each country). In doing so, our frameworks captures various types of shocks driving global volatility. For instance, shocks that are
idiosyncratic to sectors are captured by the intra-sectoral subcomponent. Macro shocks can arise from sector-specific shocks that make their way through economies, or from blocs themselves thereby affecting all the constituent sectors. We track macro-shocks with the comovement components in each bloc (inter-sectoral comovements, and comovements between countries). Global shocks are captured by the comovements between each bloc. We then continue our analysis of global volatility by distinguishing two sources of change at all levels of our framework: the effect of the structural transformation and the effect of changes in intrinsic volatility. Finally, we propose a different perspective by considering a global economy partition into blocs according to trading zones instead of development level.
4
fluctuations for most countries of the globe, although the increasing synchronicity between the countries of the Periphery stands in sharp contrast with the global trend. The study of the interaction between the Centre and the Periphery reveals a striking result: the structural
transformation, known for its volatility reduction effect in developed countries, appears to have a negligible impact in a global perspective. This is due to its opposite effect on developing
countries. These results confirm the need to consider volatility on a global scale, as the
importance of the interplay between the economies of the globe increases with the acceleration of globalisation.
This paper is structured as follows: section 1 presents the global accounting framework, section 2 describes the data used, section 3 reports the results of our analysis in the Centre-Periphery perspective, section 4 reports the results with the perspective according to trading zones, section 5 shows a complementary counterfactual analysis for both perspectives, and section 6 concludes.
1. Accounting Framework
This section lays out the volatility accounting framework meant to quantify global volatility, to track its evolution, and identify its underlying sources of change. Our simple but articulated framework allows us to quantify the level of global volatility at a given point in time and according to a partition of the world in terms of blocs, economies, and sectors. This approach breakes down global volatility into a number of components reflecting certain types of economic shocks. The first contribution of our framework is to quantify the level global volatility
attributable to each of these components, and to report their evolution over time. In addition, we distinguish between two sources of change: the structural transformation and the intrinsic volatility change.
1.1. Volatility Level Accounting
Global real GDP level in time 𝑡, denoted 𝑌𝑡, is originating from two ‘blocs’: The ‘Centre’ 𝑌𝑡𝐶 and
the ‘Periphery’ 𝑌𝑡𝑃. The annual growth rate of global real GDP can then be expressed as the
5
Δℓ𝑛𝑌𝑡= ∑ 𝜋̅𝑏 𝑏𝑡Δℓ𝑛𝑌𝑏𝑡, (1)
where 𝜋̅𝑏𝑡 is a two-period average share of the nominal GDP of the bloc 𝑏 (= 𝐶, 𝑃) in the global nominal GDP. Consider 𝑦𝑗𝑏𝑡 the level of GDP of the country 𝑗 located in bloc 𝑏, its annual growth rate can then be expressed as the weighted sum of GDP growth of its constituent economies:
Δℓ𝑛𝑌𝑏𝑡 = ∑ 𝜔𝑗 ̅𝑏𝑗𝑡Δℓ𝑛𝑦𝑏𝑗𝑡 , (2) where 𝜔̅𝑗𝑏𝑡 is a two-period average share of the nominal GDP of the economy 𝑗 in the nominal
GDP of bloc 𝑏.
Assuming that each is economy 𝑗 is having its economic activity structured along three main sectors, primary (𝑝), industry (𝑖), and services (𝑠), then the advance of its GDP results from the weighted sum of value added of its constituent sectors, with 𝜐̅𝑗𝑏𝑡𝑘 being a two-period average share of nominal value added of the sector 𝑘 = 𝑝, 𝑖, 𝑠 in the whole economy’s 𝑗 nominal GDP located in bloc 𝑏:
Δℓ𝑛𝑦𝑏𝑗𝑡 = ∑ 𝜐𝑘 𝑏𝑗𝑡𝑘 ∆ℓ𝑛 𝑦𝑏𝑗𝑡𝑘 . (3)
The combination of (1)-(3) yields gives the following decomposition of global nominal GDP growth along blocs, their constituent countries and their related economic activities:
Δℓ𝑛𝑌𝑡= ∑ ∑ ∑ 𝛼̅𝑏 𝑗 𝑘 𝑏𝑗𝑡𝑘 ∆ℓ𝑛 𝑦𝑏𝑗𝑡𝑘 , (4)
with 𝛼̅𝑏𝑗𝑡𝑘 = 𝜋̅𝑏𝑡𝜔̅𝑏𝑗𝑡𝜐̅𝑏𝑗𝑡𝑘 the share of nominal value added of sector 𝑘 in the total real value added of the global economy.
Using (4), the level of volatility of the global nominal GDP growth (global volatility from now on) can then be expressed as:
6
Equation (5) allows for the partitioning of the level of global volatility into a number of
components and subcomponents involving a complex interplay across various blocs, economies and their constituent sectors:
- The volatility that arises from each bloc (𝑏 = 𝑏′).
As each bloc comprises multiple countries, we can distinguish between the volatility arising ‘within’ and ‘between’ them.
o Within economies (𝑗 = 𝑗′, 𝑏 = 𝑏′).
This component reports the global volatility that is specific to each economy of the bloc. This global volatility contribution can be further broken in two.
Intra-sectoral (𝑘 = 𝑘′, 𝑗 = 𝑗′, 𝑏 = 𝑏′).
The volatility specific to each sector. Inter-sectoral (𝑘 ≠ 𝑘′, 𝑗 = 𝑗′, 𝑏 = 𝑏′).
The comovement of sectors within each of the countries of the bloc. o Between economies (𝑗 ≠ 𝑗′ , 𝑏 = 𝑏′).
This component captures the comovements between the sectors of different countries in the bloc.
- The volatility that arises from comovements between blocs (𝑏 ≠ 𝑏′).
This partitioning of global volatility can be extended to a perspective including more than just two blocs, such as the analysis of global volatility according to trading zones that we perform.
1.2. Volatility Change Accounting
From the level volatility accounting, we now move to the sources underlying its changes over time. The changes to the level of global volatility can be traced to the structural transformation and the changes in the intrinsic volatility of sectors in the global economy. The structural transformation is modelled as the change in the share of nominal value added of each sector in the total nominal value added of the global economy (∆𝛼̅𝑏𝑗𝑡𝑘 , ∆𝛼̅𝑏𝑘′𝑗′𝑡
′
). In contrast, the intrinsic volatility change is the change in the correlation between sectors as well as the change in the variance of specific sectors (∆𝐶𝑜𝑣 (∆ℓ𝑛 𝑦𝑏𝑘′𝑗′𝑡
′
7 ∆𝑉𝑎𝑟(Δℓ𝑛𝑌𝑡) = ∆ ∑ ∑ ∑ ∑ ∑ ∑ 𝛼̅𝑏𝑘′𝑗′𝑡 ′ 𝛼̅𝑏𝑗𝑡𝑘 𝐶𝑜𝑣 (∆ℓ𝑛 𝑦𝑏𝑘′𝑗′𝑡 ′ , ∆ℓ𝑛 𝑦𝑏𝑗𝑡𝑘 ) 𝑘′ 𝑘 𝑗′ 𝑗 𝑏′ 𝑏 = ∑ ∑ ∑ ∑ ∑ ∑ ∆𝛼̅𝑏𝑘′𝑗′𝑡 ′ [𝛼̅𝑏𝑗𝑡𝑘 𝐶𝑜𝑣(∆ℓ𝑛 𝑦𝑏𝑘′𝑗′𝑡 ′ , ∆ℓ𝑛 𝑦𝑏𝑗𝑡𝑘 )] ̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ 𝑘′ 𝑘 𝑗′ 𝑗 𝑏′ 𝑏 (a) + ∑ ∑ ∑ ∑ ∑ ∑ ∆𝛼̅𝑏𝑗𝑡𝑘 [𝛼̅𝑏𝑘′𝑗′𝑡 ′ 𝐶𝑜𝑣(∆ℓ𝑛 𝑦𝑏𝑘′𝑗′𝑡 ′ , ∆ℓ𝑛 𝑦𝑏𝑗𝑡𝑘 )] ̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ 𝑘′ 𝑘 𝑗′ 𝑗 𝑏′ 𝑏 (b) + ∑ ∑ ∑ ∑ ∑ ∑ ∆𝐶𝑜𝑣 (∆ℓ𝑛 𝑦𝑏𝑘′𝑗′𝑡 ′ , ∆ℓ𝑛 𝑦𝑏𝑗𝑡𝑘 ) [𝛼̅𝑏𝑘′𝑗′𝑡 ′ 𝛼̅𝑏𝑗𝑡𝑘 ] ̅̅̅̅̅̅̅̅̅̅̅̅̅̅ 𝑘′ 𝑘 𝑗′ 𝑗 𝑏′ 𝑏 (c)
where (a) and (b) constitute the change in global volatility caused by the structural
transformation, and (c) is the change in global volatility caused by changes in the intrinsic volatility of sectors.
The change in sectoral intrinsic volatility outlined above comprises both changes in the variance of specific sectors as well as changes in the covariance between different sectors. This allows us to decompose this source of global volatility change (c) into a number of components according to bloc, economies and their constituent sectors, similarly to the decomposition applicable to (5) explained above.
2. Quantitative Analysis
2.1. The Source Data
The main data source for our study is the GGDC 10-Sector Database. For 46 countries at
different stages of economic development, this panel data tracks employment, as well as current and constant price series of value added at national currencies (2005 prices), over various time spans starting from 1950. For each country, the data are structured according to the International Standard Industrial Classification (Revision 3.1), with a total economy broken down into 10 main sectors: (1) agriculture, (2) mining, (3) manufacturing, (4) utilities, (5) construction, (6) trade services, (7) transport services, (8) business services, (9) government services, and (10) personal services.
8
We discard a number of countries from the available sample, according to time spans covered and international price convertibility. Malaysia, for example, displays data series beginning in 1975, which is too short to reliably analyse the evolution of volatility between the pre-1984 and the post-1984 periods. Similarly, we exclude countries like Taiwan due to the lack of PPPs at the sectoral level. In contrast, we enrich our sample with the addition of two countries: Canada (using World-KLEMS 2012) and unified Germany (using EU-KLEMS 2009). We aggregate the 10 sectors available into 3 broad sectors, the construction of which is explained in Table A of the appendix. Such restructuring limits the analysis of structural transformation, as information is lost by merging manufacturing with utilities and market services with non-market services. However, this delineation is commonly used in the literature (see Duarte and Restuccia 2010 for example) and it increases the fluidity of results as well as their interpretation. Finally, the
international price convertibility is not possible for all countries in the source data, as a large number of them lack sector specific PPPs. Therefore we complete this information available for Sub-Saharan countries by using the PPPs constructed by Inklaar and Timmer (2012). This enables to convert the local currency value added data at constant 2005 prices to international prices using 2005 PPPs.
[Insert Table A Here]
These adjustments yield a data set covering a large variety of countries displaying different stages of economic development over the 1970-2007 period: 12 developed countries (Centre) and 18 developing countries1 (Periphery).
2.2. Descriptive Analysis
One of the most important dynamics characterising our data is the structural transformation taking place in developed and developing countries. This process, illustrated in Figure 1, is well documented in the literature, but our global perspective establishes some new facts. The
importance of the Centre is confirmed by our results, despite a decline over the two periods. A
1 Although South Korea remained a developing country until 1997, we choose to classify it with the
9
notable feature is that the industry sector remains largely stable over time in the Centre, while it is normally declining for those countries. This is mainly due to our restructuring of the sectoral data, aggregating construction with manufacturing, and to the inclusion of countries like Japan and Germany where the manufacturing sector has not experienced the sharp relative decline characteristic of the U.S.. Countries in the Periphery show a surge in their share of world output in the industry and services sectors. However, their share of primary sector production remains steady, likely due to the slow structural transformation in Africa and Latin America compared to Asian countries. In fact, the structural transformation in the Periphery is mostly driven by
countries like China and India (see Tables 1). In terms of global production share, the evolution of China (from 1% to 3.6% for the industry and from 1% to 3.9% for services) and India (from 1.4% to 2.4% for the industry and from 1.4% to 2.5% for services) are particularly striking.
[Insert Figure 1 Here]
Tables 1 report the level of volatility and its change over time for the different countries, along with the sectors constituent of the Centre (Table 1a) and the Periphery (Table 1b). We observe a very large difference between the ranges of magnitudes of volatility across economies in the Periphery compared to that of the Centre. For example, during the 1971-1984 period, volatility in the Periphery ranges from 4.93 for Tanzania to 128.4 for Nigeria (with a median of 14.6) while volatility in the Centre ranges from 0.7 for France to 5.47 for Japan (with a median of 4.1). The difference between the magnitudes of the median of the two blocs reveals a much higher
volatility level in the Periphery. As we inspect volatility at the sectoral level, the order magnitude of volatility in the Periphery appears even more striking. For instance, the primary sector in Senegal displays a level of volatility of 334 while Botswana and Nigeria display high levels of volatility in all sectors. These results are explained by the low level of diversification
characterising developing countries: Koren and Tenreyro (2007) show that exposure to sectoral shocks accounts for most of the volatility difference between developed and developing
countries.
10
There is a broad consensus on the stylised facts of the great moderation. However, most of the literature on this topic limited its scope to the Centre countries and little attention has been granted to countries in the Periphery. Our results in Table 1a confirm the well-established view on the volatility decrease in developed countries around 1984: all countries apart from France show a decline in volatility and the median for the Centre falls from 4.1 to 2.2. Noticeably, the countries in the Periphery show a decline in their median volatility from 14.6 to 6.1. Although this fall in volatility is larger than that of the developed countries, the developing countries still report a range of magnitude of volatility much larger than that of developed countries. Overall the evolution of volatility in the Centre and the Periphery suggest that the great moderation was a global phenomenon for our sample of countries, although countries in the Periphery remain at substantially higher volatility levels than the Centre countries.
[Insert Table 1b Here]
3. An Account for Global Volatility
3.1. Volatility Level Accounting and Its Change Over Time
Panel A of Table 2 reports the level of global volatility for the 1971-1984 and 1985-2007 periods and its partition in terms of its components. During the 1971-1984 period, the level of global volatility was 1.47, of which 75.4% is attributable to the Centre, 6% for the Periphery, and 20% for the comovements between the two blocs. Focusing on the Centre, about 70%
((0.452+0.320)/1.106)) of the 1.106 level of volatility is attributable to the comovements within this bloc. These comovements consist of two components: the inter-sectoral comovements within economies, amounting to 29% (0.320/1.106) of the volatility of the bloc; and the comovements between economies, amounting to 41% (0.452/1.106) of the volatility of the bloc. The remaining 30% (0.333/1.106) of the Centre volatility arise from each of the sectors constituting the
11
Periphery is intra-sectoral within countries, while 33% ((0.019+0.046)/0.081) is attributable to comovements. These comovements are small because they consist of two conflicting forces: the inter-sectoral comovements within economies, amounting to 57% (0.046/0.081) of the volatility of the bloc; and the comovements between economies, counterbalancing the volatility of the Periphery by 24% (-0.019/0.081). This last result suggests that countries in the Periphery are inversely synchronised over this period.
[Insert Table 2 Here]
We observe an evolution in all components of global volatility for the 1985-2007 period. The most striking change is the increased importance of the Periphery, amounting to 27% of global volatility, while the share of global volatility attributable to the Centre falls to 57%. The Periphery is also marked by a dramatic rise in the importance of the comovements between its economies, now amounting to 64% ((0.043+0.055)/0.154) of the volatility of the bloc. This last result suggests an increase in the synchronicity across countries in the Periphery, a phenomenon which could be caused by the uniformity gains in terms of structural transformation.
The changes between the two periods are tacked in Panel B of Table 2, and the contribution of each component to the change in global volatility are displayed in panel C. These changes reflect some well-established facts concerning the great moderation, but they also emphasise new developments. For instance, the Periphery displays an increase of 90% of its global volatility, which drastically contrasts with the 70% decline in the global volatility attributable to the Centre. In addition, we observe a 68% decline in the comovements between the Centre and the
12
The evolutions depicted for the Centre and the Periphery are the results of internal changes that are strikingly different for each bloc. The 86.9% contribution to global volatility decline
attributable to the Centre is almost equally divided among the between country component (41.8%) and the within country components (45.1%), the latter resulting from equal relative contribution of its comprising components (intra-sectoral and inter-sectoral). The decline in the comovements between countries in the Centre suggests a tendency for the business cycles across economies to become less synchronised. This result is similar to the findings of Heathcote and Perri (2003), for example, who state that from 1960 to 2002 the U.S. business cycle became less synchronized with the cycles in 15 countries of the European Union and Japan. In contrast, the negative contribution (-8.2%) to the global decline in the Periphery is overwhelmingly
attributable to comovements between countries in this bloc. In effect, countries in the Periphery moved from a state of negative synchronicity (-0.019) during 1971-1984 to a state of relatively high positive synchronicity (0.043) during the globalisation period of 1985-2007.
3.2. Volatility Change Accounting
In this section we inspect the forces that are responsible for the evolution of global volatility depicted above. Panel D of Table 2 shows the sources of change, quantifying how much of the change in global volatility is attributable to each of two forces studied: the structural
13
Up to this point we have analysed the sources of volatility change from a global perspective. We now move on to quantify the effect of the structural transformation and that of the intrinsic volatility change by focusing on each bloc and the comovements between them. Table 3 reports the results of the sources of change in global volatility from this focused perspective. As in Table 2, Panel A displays the level of volatility for each period and Panel B shows the change between periods in absolute terms as well as in terms relative to the total change for each block, while the sources of volatility change in the focused perspective are reported in Panel C.
[Insert Table 3 Here]
As previously noted, the intrinsic volatility is the main factor driving the change in the global volatility contribution of the Centre (88%). This source of change is evenly distributed between the within country component (47%) and the between country component (41%). The structural transformation in the Centre also contributed to its decline in global volatility, with an effect amounting to 12% of the total change in the bloc.
In contrast, the structural transformation caused an increase in the level of global volatility attributable to the Periphery. This effect, amounting to 115% of the global volatility increase of the block, is mitigated by the intrinsic volatility effect which amounts to -15%. This suggests that the countries in the Periphery would have experienced a fall in their bloc level of global volatility if they were not subject to the structural transformation. However, the contribution of the
intrinsic volatility results from two conflicting effects: the within country part contributes to -228% while the between country part amounts to 213%. While the former confirms the results displayed in Table 1b, suggesting that the Great Moderation is a global phenomenon, the latter shows a drastic increase in the synchronicity of the countries in the Periphery.
14
From this analysis, it appears that the only two sources contributing to an increase in global volatility are the structural transformation in the Periphery, and the intrinsic volatility of comovements between countries in this bloc. However, the influence of these two forces combined mitigates the global decline in volatility by no less than 31%.
4. An Account of Volatility Within Trading Zones
4.1. Motivation
So far our analysis has been performed by considering a global economy structured along two blocs distinguished by their level of development: the Centre and the Periphery. This classic view, traditionally used in the literature, enables to set the stage. However, the emergence of common currency areas and free trade agreements gave way to a stronger regionalisation of the world. The rise of trade flows inside regions was considered by Kose et al. (2013) as contributing to shift the importance away from global factors and into regional factors on the matter of
business cycle synchronisation. In light of these developments, we study the evolution of global volatility by considering two definitions of trading zones: a conventional definition according to geographic proximity, and a more recent definition emphasising cultural proximity.
4.2. Conventionally defined trading zones
We propose to study the evolution of global volatility in the perspective of conventional trading zones by constructing five regions, inspired from the most important trading agreements for the countries in our sample: the North American Free Trade Agreement (Nafta), the European Union (EU), the Asean Free Trade Area (Afta) and South Asia Free Trade Agreement (Safta) in Asia, and the Mercado Común del Sur (Mercosur) in South America. We disclose the countries of each region in the appendix.
15
between trade zones in this study is overwhelmingly superior to that of the comovements Centre-Periphery reported in the first part of our analysis. In fact, comovements among trading zones constitute 41% (0.606/1.467) of the level of global volatility during 1971-1984. We distinguish a number of pair comovements between specific trading zones, and these reveal high levels for EU-Nafta (0.241) and Nafta-Asia (0.13). This suggests a strong synchronicity for the business cycles between Europe and North America as well as between Europe and Asia, even before the second wave of globalisation.
[Insert Table 4 Here]
For the 1971-1984 period, the sum of the global volatility arising from each trading zone amounts to 59% of the total. This comes with a range of variation for each trading zone, culminating with 43% for Nafta and leaving only 8% for Asia and 4% for Europe. The contribution of the European trading zone seems small considered on its own, but this zone displays some of the highest comovements with other trading zones. The European trading zone differs from its North American counterpart in that most of its level of global volatility arise from comovements between countries: this component contributes to 78% (0.046/0.059) of the European zone volatility, compared to 54% (0.348/0.642) for Nafta.
The period 1985-2007 shows a decline in global volatility from 1.467 to 0.572 which is
characterised by two main facts. The magnitude of comovements between trade zones fell from 41% to 28% (0.160/0.572) of the volatility level of the period, implying an increase in the volatility share attributable to each trading zone. In additional, the share of the Asian trading zone soared to reach 24%, a threefold increase, while the contribution of most other trading zones remained fairly constant. This increase in the Asian trading bloc is overwhelmingly caused by a rise of comovements across countries in this trading zone. These results suggest a shift away from comovements between trade zones, apart for the EU-Asia pair, and towards the
comovements of countries within the Asian bloc.
16
primary source of global volatility decline, with a contribution amounting to 50% of the global change. In particular, the EU-Nafta and Nafta-Asia comovements together account for 31% (respectively 20% and 11%). The remaining 50% of the global volatility decline is attributable to each trading bloc, with Nafta alone accounting for a considerable 44%. This result is due to the large relative size of the US economy in the global economy (40% on average, see Table 1). Panel D reports the sources underlying the change in global volatility. We observe that the two largest contributors to the global change – Nafta and the comovements between blocs – report a volatility evolution mainly driven by intrinsic volatility. The sum of their intrinsic volatility sources of change accounts for 82% of the global change (43% for comovements between trading zones, and 39% for Nafta).
Similarly to the change to a perspective focused on trade blocs operated for the Centre-Periphery analysis, we now delve into each trading zone to analyse their sources of global volatility
change. The results of this study are reported in Table 5. The contribution of the structural transformation to the global volatility change varies considerably across trading zones. Nafta, displaying the largest fall in global volatility (-0.392), is only affected by the structural transformation up to a modest 11%. In contrast the second largest global volatility decline, Europe (-0.043), is explained at 33% by the structural transformation. Duarte and Restuccia (2010) hint at a possible explanation for this phenomenon by showing that the structural transformation in countries like Spain and Italy took place later than for other European countries. As for the Asian trading zone, the structural transformation effect overwhelmed the effect of the intrinsic volatility. This appears in accordance with the recent facts about the progress of China and India with regards to their structural transformation. Finally, the intrinsic volatility effect dominates in Africa and Latin America, which confirms the fact that these blocs experienced a lack of structural transformation (see Timmer et al. 2014 and MacMillan and Rodrik 2014).
17
4.3. Alternative definition of trading zones
The conventional definition of trading zones, based on geographical proximity, is built upon a number of concepts core to bilateral trade such as iceberg costs, comparative advantage, tariffs, and trade agreements. Depending upon the importance of comovements between countries in trading zone compared we can assess the effective interconnectedness of trade regions. If these comovements are low then trade within the region cannot be considered as more important than overall trade. In effect, this would imply that the notion of iceberg costs does not apply. The results of our analysis are mixed: comovements in Europe and Asia amount to 85% and 70% of the change in global volatility in each trade zone respectively (Panel C), but this component is low for the other regions considered. These considerations raise the following question: is our conception of trading zone adequate? It is possible that various notions other than geographical proximity contribute to stronger comovements between regions. Neville et al. (2012) show that trading regions synchronicity can be the result of cultural proximity and institutional similarity among other non-geographical factors.
Using non-geographical factors we sort the 30 countries of our sample into three clusters,
defining an alternate approach to trading zones (Table 6). Cluster 1 consist of countries culturally close to the Commonwealth of Nations. Cluster 2 regroups other industrialised nations. Cluster 3 contains developing countries that are similarly sensitive to sectoral idiosyncratic shocks.
[Insert Table 6 Here]
The results of our analysis of global volatility according to our alternative definition of trading zones are presented in Table 7. The contribution of comovements to the change of global
volatility in each clusters ranges from 11% (0.072/0.307) for cluster 1 to 57% (-0.104/-0.181) for cluster 2, leaving cluster 3 with 35% (0.019/0.055). In comparison with our earlier findings, this alternative definition does not seem to yield higher synchronicity for trading zones. It appears that in our case the notion of geographical distance takes precedence over cultural proximity.
18
5. Counterfactuals
This last section of our results revisits the Centre-Periphery and the trading zone analysis in order to ascertain the impact of the structural transformation on global volatility in these two perspectives. The counterfactuals are produced by analysing global volatility with no structural transformation. In effect, we assume that the snapshot of structural organisation taken during the 1971-1984 period remains unchanged for the subsequent period. This implies that the Centre stands at a stage of structural transformation much more advanced than the Periphery, for which the process is barely taking place. In addition, the two blocs are no longer interacting by
offshoring activity from the Centre to the Periphery. This counterfactual analysis enables to assess whether the deployment of production activities from Centre to the Periphery also triggered a deployment of volatility along the way.
We report our results in Tables 8 and 9, which are structured similarly to Tables 2 and 4, holding sectoral share of global production constant at the 1971-1984 level. Consequently, we are not reporting the sources of global volatility change, as only the intrinsic volatility is allowed to take effect. This yields results that are differing only slightly for the Centre, but the Periphery
displays drastic changes (comparing Tables 2 and 8). The Centre is still affected by the great moderation (65% actual decline in global volatility compared to 70% with structural
transformation) and shows no significant alteration in its constituting within and between components. In contrast, the counterfactuals show a 22% decline in the global volatility on the Periphery, which drastically differs from the global volatility increase this bloc displays under actual circumstances. It appears clearly that without structural changes, the Periphery does not anymore look like the cushion of the Centre’s business cycle fluctuations. Most importantly, the decline in global volatility reported by this counterfactual analysis (-0.930) is almost equal to the decline report in the actual analysis (-0.895); which confirms our earlier finding that the
structural transformation had little effect when considered in a global perspective.
19
Comparing Tables 4 and 9 sheds light on the quantitative impact of structural transformation on the evolution of global volatility in trading zones. The most important difference appears in Asia, which now shows signs of the Great Moderation along with the other trading zones. This is contrasted by a higher level of global volatility in Nafta and Europe (+0.012 for the EU, and +0.033 for Nafta). This increased level of global volatility in Nafta and Europe amounts to 56% of the decrease in Asia in the absence of structural transformation, which suggests that much of the actual global volatility increase observed in Table 4 for Asia stems from the relocation of activity from Europe and Nafta. Finally, a striking result is that the only component contributing to an increase in global volatility worldwide without structural transformation is the
comovements between countries in Asia (contributing -1.7% of the global decline). Therefore, Asian economies show signs of increased synchronicity even without the relocation of volatile activity from Europe and Nafta.
[Insert Table 9 Here]
6. Conclusion
This paper brings together several literature strands in order to analyse the evolution of volatility on a global scale. In a unified framework of volatility accounting, we inspect the contributions arising from different components of global volatility. These components reflect the interplay of different country blocs, the economies comprising them, and their constituent sectors. We apply this framework to a perspective partitioning the world into blocs according to their development level (the Centre and the Periphery), as well as a perspective distinguishing multiple trading regions. In addition, we identify the contributions to global volatility change attributable to two underlying sources of volatility evolution: the structural transformation and the change in the intrinsic volatility of sectors.
20
resulting in a substantial increase in the contribution to global volatility of this bloc. However, if we consider the interaction between the Centre and the Periphery, we observe that the structural transformation taking place in these blocs is akin to a zero-sum game worldwide: the fall in global volatility caused by this phenomenon in the Centre is matched by a similar increase in the global volatility arising from the Periphery.
21
References
Acemoglu, Daron, and Fabrizio Zilibotti. 1997.’Was Prometheus unbound by chance? Risk, diversification, and growth.’ Journal of Political Economy 105, no. 4: 709-751.
Bergin, Paul R., Robert C. Feenstra, and Gordon H. Hanson.2009. ‘Offshoring and volatility: evidence from Mexico's maquiladora industry.’ The American Economic Review 99, no. 4: 1664-1671.
Caselli, Francesco, Miklos Koren, Milan Lisicky, and Silvana Tenreyro. 2014. "Diversification through trade." January http://personal.lse.ac.uk/tenreyro/volatilitytrade.pdf
Coronado, Roberto A. 2011. Offshoring and Volatility: More Evidence from Mexico’s Maquiladora Industry. Federal Reserve Bank of Dallas, WP No. 1106.
https://www.dallasfed.org/assets/documents/research/papers/2011/wp1106.pdf
Davis, Steven J., and James A. Kahn. 2008. ‘Interpreting the Great Moderation: Changes in the Volatility of Economic Activity at the Macro and Micro Levels.’ Journal of Economic
Perspectives, 22(4): 155-80.
Francis, Neville, Michael T. Owyang, and Özge Savascin. 2012. ‘An endogenously clustered factor approach to international business cycles.’ Federal Reserve Bank of St. Louis Working Paper Series 2012-014 (2012).
Galí, Jordi, and Luca Gambetti. 2009. ‘On the Sources of the Great Moderation.’ American Economic Journal: Macroeconomics, 1(1): 26-57.
Giovanni (di), Julian, and Andrei A. Levchenko. 2010. "Putting the Parts Together: Trade, Vertical Linkages, and Business Cycle Comovement." American Economic Journal: Macroeconomics, 2(2): 95-124.
Heathcote, Jonathan, and Fabrizio Perri. 2003. ‘Why has the US economy become less correlated with the rest of the world?.’ American Economic Review: 63-69.
Imbs, Jean. 2004. "Trade, finance, specialization, and synchronization." Review of Economics and Statistics 86, no. 3: 723-734.
Inklaar, Robert, and Marcel P. Timmer. 2014. ‘The relative price of services. Review of Income and Wealth 60, no. 4: 727-746.
Koren, Miklós, and Silvana Tenreyro. 2007. ‘Volatility and development.’ The Quarterly Journal of Economics: 243-287.
22
Levasseur Sandrine. 2011. Production Under Foreign Ownership and Domesti Volatility: An Empirical Investigation at the Sectoral Level. Document de Travail de l’OFCE, no. 1.
http://www.ofce.sciences-po.fr/pdf/dtravail/WP2011-01.pdf
Lucas, Robert E. 1988. ‘On the mechanics of economic development.’ Journal of Monetary Economics 22, no. 1: 3-42.
McMillan, M. S., Rodrik, D., Verduzco-Gallo, _I. 2014. Globalization, structural change, and productivity growth, with an update on Africa. World Development, 63, 11-32.
Moro, Alessio. 2015. ‘Structural Change, Growth, and Volatility,’ American Economic Journal: Macroeconomics Forthcoming.
Ngouana, Mr Constant Lonkeng. 2013. Structural Transformation and the Volatility of Aggregate Output in OECD Countries. No. 13-43. International Monetary Fund.
http://www.imf.org/external/pubs/ft/wp/2013/wp1343.pdf
Rodrik, Dani. 2011. ‘The future of economic convergence.’ National Bureau of Economic Research, WP No. 17400, September. http://www.nber.org/papers/w17400.
Stiroh, Kevin J. 2009. ‘Volatility accounting: A production perspective on increased economic stability.’ Journal of the European Economic Association 7, no. 4: 671-69
23
Table 1a. Volatility of Production and Relative Sectoral Share: Developed Economies
24
Table 1b. Volatility of Production and Relative Sectoral Share: Developing Economies
Country
25
Table 1b. Volatility of Production and Relative Sectoral Share: Developing Economies (Continued)
Country
1971 - 1984 (1) 1985 - 2007 (2) Change between periods Volatility Share Volatility Share Volatility
[((2)/(1)-1)×100] Share [(2)-(1)] Whole economy Kenya 8.1 0.1 2.2 0.1 -72.6 0.0 Primary 9.0 0.0 13.2 0.0 46.9 0.0 Industry 65.0 0.0 5.9 0.0 -90.9 0.0 Services 8.0 0.1 2.2 0.1 -72.7 0.0 Whole economy Malawi 25.9 0.0 3.9 0.0 -85.1 0.0 Primary 38.1 0.0 222.3 0.0 482.9 0.0 Industry 67.1 0.0 14.7 0.0 -78.1 0.0 Services 31.3 0.0 1.8 0.0 -94.2 0.0 Whole economy Mauritius 29.6 0.0 23.0 0.0 -22.3 0.0 Primary 373.7 0.0 93.7 0.0 -74.9 0.0 Industry 96.2 0.0 48.4 0.0 -49.6 0.0 Services 40.0 0.0 27.3 0.0 -31.8 0.0 Whole economy Mexico 11.3 3.6 9.2 3.3 -18.7 -0.4 Primary 19.5 0.4 5.4 0.4 -72.1 0.0 Industry 11.6 1.6 10.8 1.5 -7.1 -0.2 Services 11.9 1.6 11.0 1.4 -8.1 -0.2 Whole economy Nigeria 128.4 0.3 17.6 0.3 -86.3 -0.1 Primary 250.7 0.2 40.0 0.1 -84.0 -0.1 Industry 186.1 0.0 24.0 0.0 -87.1 0.0 Services 95.1 0.1 26.2 0.1 -72.4 0.0 Whole economy Senegal 15.3 0.0 4.0 0.0 -73.9 0.0 Primary 333.9 0.0 95.2 0.0 -71.5 0.0 Industry 34.0 0.0 6.2 0.0 -81.7 0.0 Services 23.4 0.0 4.0 0.0 -82.8 0.0 Whole economy South Africa 10.7 0.1 8.8 0.1 -18.3 0.0 Primary 15.2 0.1 17.3 0.1 13.9 0.0 Industry 125.8 0.0 121.2 0.0 -3.7 0.0 Services 7.3 0.0 8.8 0.0 19.9 0.0 Whole economy Tanzania 4.9 0.7 4.3 0.6 -12.6 -0.1 Primary 9.9 0.0 1.2 0.0 -88.2 0.0 Industry 19.3 0.2 12.5 0.1 -35.4 -0.1 Services 2.7 0.5 3.3 0.4 21.1 -0.1 Whole economy Zambia 13.9 0.0 9.0 0.0 -35.3 0.0 Primary 68.9 0.0 69.2 0.0 0.4 0.0 Industry 41.8 0.0 39.9 0.0 -4.6 0.0 Services 24.8 0.0 7.0 0.0 -71.9 0.0
27
Table 2. Level of Global Volatility: Within and Between Centre-Periphery, Change Over Time and Sources
A-Volatility Level B-Change in
Volatility C-Contribution to Change in Global Volatility (%) D-Sources of Global Volatility Change (Contribution %) 1971-1984 1985-2007 Absolute %
Global economy 1.467 0.572 -0.895 -61.0 Structural
28
Table 3. Contribution of Structural Transformation and Intrinsic Volatility to the Change in the Volatility Within and Between Blocs
29
Table 4. Level of Global Volatility: Within and Between Trading Zones, Change Over Time and Sources
A-Volatility Level B-Change in
Volatility Contribution C-to Change in Global Volatility (%) D- Sources of Global Volatility Change (Contribution %) 1971-1984 1985-2007 Absolute % Global economy 1.467 0.572 -0.895 -61.0 Structural Transformatio n Intrinsic Volatility EU 0.059 0.016 -0.043 -72.9 4.8 2.0 3.0 Within economies 0.022 0.009 -0.013 -59.1 1.5 Intra sectoral 0.013 0.006 -0.007 -53.8 0.8 Inter sectoral 0.009 0.003 -0.006 -66.7 0.7 Between economies 0.037 0.007 -0.030 -81.1 3.4 NAFTA 0.641 0.249 -0.392 -61.2 43.8 5.0 39.0 Within economies 0.577 0.223 -0.354 -61.4 39.6 Intra sectoral 0.293 0.115 -0.178 -60.8 19.9 Inter sectoral 0.284 0.108 -0.176 -62.0 19.7 Between economies 0.064 0.026 -0.038 -59.4 4.2 Asia 0.110 0.137 0.027 24.5 -3.0 -12.0 9.0 Within economies 0.114 0.118 0.004 3.5 -0.4 Intra sectoral 0.060 0.060 0.000 0.0 0.0 Inter sectoral 0.054 0.058 0.004 7.4 -0.4 Between economies -0.004 0.019 0.023 -575.0 -2.6 South America 0.045 0.011 -0.034 -75.6 3.8 1.0 3.0 Within economies 0.037 0.011 -0.026 -70.3 2.9 Intra sectoral 0.019 0.005 -0.014 -73.7 1.6 Inter sectoral 0.018 0.006 -0.012 -66.7 1.3 Between economies 0.008 0.000 -0.008 -100.0 0.9 Africa 0.003 0.000 -0.003 -100.0 0.3 0.0 0.0 Within economies 0.002 0.000 -0.002 -100.0 0.2 Intra sectoral 0.002 0.000 -0.002 -100.0 0.2 Inter sectoral 0.000 0.000 0.000 0.000 0.0 Between economies 0.001 0.000 -0.001 -100.0 0.1
Comovements b/ Trading Zones 0.606 0.160 -0.446 -73.6 49.8 7.0 43.0
EU/NAFTA 0.241 0.059 -0.182 -75.5 20.3 4.0 16.0
EU/Asia -0.021 -0.004 0.017 -81.0 -1.9 2.0 -4.0
NAFTA/Asia 0.130 0.032 -0.098 -75.4 10.9 -5.0 16.0
30
Table 5. Contribution of Structural Transformation and Intrinsic Volatility to the Change in the Volatility Within and Between Trading Zones
A-Volatility Level B-Change in Volatility C-Contribution to Change in Zone Volatility (%)
D- Sources of Zone Volatility Change (Contribution %) 1971-1984 1985-2007 Structural Transformation Intrinsic Volatility EU 0.059 0.016 -0.043 33.0 67.0 Within economies 0.022 0.009 -0.013 30.2 13.0 Intra sectoral 0.013 0.006 -0.007 16.3 4.0 Inter sectoral 0.009 0.003 -0.006 14.0 9.0 Between economies 0.037 0.007 -0.030 69.8 54.0 NAFTA 0.641 0.249 -0.392 11.0 88.0 Within economies 0.577 0.223 -0.354 90.3 81.0 Intra sectoral 0.293 0.115 -0.178 45.4 40.0 Inter sectoral 0.284 0.108 -0.176 44.9 41.0 Between economies 0.064 0.026 -0.038 9.7 7.0 ASIA 0.110 0.137 0.027 413.0 -313.0 Within economies 0.114 0.118 0.004 14.8 -356.0 Intra sectoral 0.060 0.060 0.000 0.00 -151.0 Inter sectoral 0.054 0.058 0.004 14.8 -205.0 Between economies -0.004 0.019 0.023 85.2 43.0 Latin America 0.045 0.011 -0.034 19.0 82.0 Within economies 0.037 0.011 -0.026 76.5 58.0 Intra sectoral 0.019 0.005 -0.014 41.2 30.0 Inter sectoral 0.018 0.006 -0.012 35.3 28.0 Between economies 0.008 0.000 -0.008 23.5 24.0 Africa 0.003 0.000 -0.003 30.0 70.0 Within economies 0.002 0.000 -0.002 66.7 49.0 Intra sectoral 0.002 0.000 -0.002 66.7 50.0 Inter sectoral 0.000 0.000 0.000 0.0 -1.0 Between economies 0.001 0.000 -0.001 33.3 21.0
Table 6. Alternate Definition of Trading Zones
Cluster 1 Cluster 2 Cluster 3
Botswana Brazil Argentina
Canada France Chile
Denmark Germany China
Ghana Italy Ethiopia
India Japan Mexico
Indonesia Netherlands Senegal
Malawi South Korea
31
Table 7. Level of Global Volatility: Within and Between Alternate Notions of Trading Zones, Change Over Time and Sources
A-Volatility Level B-Change in Volatility C-Contribution to Change in Global Volatility (%) 1971-1984 1985-2007 Absolute % Global economy 1.467 0.572 -0.895 -61.0 Cluster 1 0.703 0.307 -0.396 -56.3 44.2 Within economies 0.586 0.235 -0.351 -59.9 39.2 Intra sectoral 0.305 0.123 -0.182 -59.7 20.3 Inter sectoral 0.281 0.112 -0.169 -60.1 18.9 Between economies 0.117 0.072 -0.045 -38.5 5.0 Cluster 2 0.26 0.079 -0.181 -69.6 20.2 Within economies 0.121 0.044 -0.077 -63.6 8.6 Intra sectoral 0.061 0.023 -0.038 -62.3 4.2 Inter sectoral 0.06 0.021 -0.039 -65.0 4.4 Between economies 0.139 0.035 -0.104 -74.8 11.6 Cluster 3 0.034 0.090 +0.056 +0.056 +165% Within economies 0.020 0.040 +0.021 +0.021 +105% Intra sectoral 0.025 0.042 +0.017 +0.017 +65% Inter sectoral -0.011 0.008 +0.019 +0.019 -178% Between economies 0.034 0.090 +0.056 +0.056 +165%
Comovements b/ Alternate Trading Zones 0.469 0.097 -0.372 -79.3 41.6
Clusters 1-2 0.388 0.048 -0.340 -87.6 38.0
Clusters 1-3 0.083 0.064 -0.019 -22.9 2.1
32
Table 8. Level of Global Volatility : Within and Between Center-Periphery, Change Over Time and Sources
(Counterfactuals 1: No Structural Transformation over the Two Periods)
33
Table 9. Level of Global Volatility : Within and Between Trading Zones, Change Over Time and Sources
(Counterfactuals 2: No Structural Transformation over the Two Periods)
A-Volatility Level B-Change in
Volatility C-Sources of Change in Global Volatility (%) 1971-1984 1985-2007 Absolute % Global economy 1.467 0.537 -0.930 -63.4 EU 0.059 0.028 -0.031 -52.5 3.3 Within economies 0.022 0.016 -0.006 -27.3 0.6 Intra sectoral 0.013 0.011 -0.002 -15.4 0.2 Inter sectoral 0.009 0.005 -0.004 -44.4 0.4 Between economies 0.037 0.012 -0.025 -67.6 2.7 NAFTA 0.641 0.282 -0.359 -56.0 38.6 Within economies 0.577 0.250 -0.327 -56.7 35.2 Intra sectoral 0.293 0.131 -0.162 -55.3 17.4 Inter sectoral 0.284 0.119 -0.165 -58.1 17.7 Between economies 0.064 0.032 -0.032 -50.0 3.4 Asia 0.110 0.058 -0.052 -47.3 5.6 Within economies 0.114 0.046 -0.068 -59.6 7.3 Intra sectoral 0.060 0.025 -0.035 -58.3 3.8 Inter sectoral 0.054 0.021 -0.033 -61.1 3.5 Between economies -0.004 0.012 0.016 -400.0 -1.7 Latin America 0.045 0.016 -0.029 -64.4 3.1 Within economies 0.037 0.016 -0.021 -56.8 2.3 Intra sectoral 0.019 0.008 -0.011 -57.9 1.2 Inter sectoral 0.018 0.008 -0.01 -55.6 1.1 Between economies 0.008 0.000 -0.008 -100.0 0.9 Africa 0.003 0.000 -0.003 -100.0 0.3 Within economies 0.002 0.000 -0.002 -100.0 0.2 Intra sectoral 0.002 0.000 -0.002 -100.0 0.2 Inter sectoral 0.000 0.000 0.000 0.000 0.0 Between economies 0.001 0.000 -0.001 -100.0 0.1
Comovements between Trading Zones 0.606 0.155 -0.451 -74.4 48.5
EU/NAFTA 0.241 0.086 -0.155 -64.3 16.7
EU/Asia -0.021 0.001 0.022 -104.8 -2.4
NAFTA/Asia 0.130 0.006 -0.124 -95.4 13.3
34
Appendix
Table A—Sectors and Countries in Scope The 10-Sector Database
Our Coverage ISIC Rev.
3.1 code
ASD sector
name ISIC Rev. 3.1 description Broad Sectors Regions Countries
A,B Agriculture Agriculture, Hunting and Forestry, Fishing
Primary
Africa
Botswana, Ethiopia, Ghana, Malawi, Mauritius, Nigeria, Senegal, South Africa, Tanzania, Zambia
C Mining Mining and Quarrying
D Manufacturing Manufacturing
Industry
E Utilities Electricity, Gas and Water supply
F Construction Construction
G,H Trade services
Wholesale and Retail trade; repair of motor vehicles, motorcycles and personal and household goods, Hotels and Restaurants
Services
Asia
China, Honk Kong, India, Indonesia, Japan, South Korea
I Transport
services Transport, Storage and Communications
Europe
Denmark, France, Germany, Italy, Spain, Sweden, The
Netherlands, United Kingdom
J+K Business
services
Financial Intermediation, Renting and Business Activities (excluding owner occupied rents)
L,M,N Government
services
Public Administration and Defence, Education, Health and Social work
Latin
America Argentina, Brazil, Chile
O,P Personal
services
Other Community, Social and Personal service
activities, Activities of Private Households North
America
Canada, Mexico and United States
