13
Discussion of Lewis&De Schryder and Kireyev&Leonidov by Julia Wörz Workshop TASK FORCE ON GLOBAL TRADE Paris, June 20, 2016 Foreign Research Division www.oenb.at
Discussion of Lewis&De Schryder and Kireyev&Leonidov by Julia Wörz Workshop TASK FORCE ON GLOBAL TRADE Paris, June 20, 2016 Foreign Research Division www.oenb.at
www.oenb.at [email protected] - 2 -
John Lewis & Selien De Schryder (BoE): “Export dynamics since the Great Trade Collapse: a cross-country analysis”
Highly needed paper: mixes simple ingredients carefully to construct appropriate
proxies for explaining the great trade collapse
Findings
Elements
Possible extensions
Unclear points
www.oenb.at [email protected] - 3 -
Lewis & De Schryder: Findings
• Great trade collapse (GTC) is a temporary phenomenon, does not indicate
structural shift to lower trade elasticity
• „Unexplained“ part of GTC (magnitude) can be explained by unobserved
common factors among countries, such as trade protectionism, credit
constraints, uncertainty
www.oenb.at [email protected] - 4 -
Lewis & De Schryder: Elements
• Proxies for main factors determining export performance : foreign demand,
structural shifts in global demand, price competitiveness split into unit labor costs
and nominal effective exchange rate
• Panel ECM, PMG estimator with CCE (common external effects)
www.oenb.at [email protected] - 5 -
Lewis & De Schryder: Possible extensions
• Estimation stops in 2011, thus insights about the GTC, but not about current trade weakness
• Coefficients estimated over period 1984-2008 without time variation, hence this average may not (yet) spot a structural change that occurred much prior to 2009
Chart 1.4 TFGT Report: Global imports and GDP growth
Source: IMF WEO. Notes: Imports of goods and services. Global GDP is aggregate at market exchange rates. The last observation refers to 2015.
Chart 1.3 TFGT Report: Ratio of global import growth to
global GDP growth
www.oenb.at [email protected] - 6 -
Lewis & De Schryder: Possible extensions
• Extend to EMEs: these countries may experience more structural change
compared to AEs not trivial given tedious construction of proxies
• Shed light into unobserved common correlated effects this would be a second
paper
• GVC trade but this would be an entirely different paper
www.oenb.at [email protected] - 7 -
Lewis & De Schryder: Unclear points
• Data sources not always indicated
• Trade-weighted world output / GDP growth (Y*) – is this the foreign demand
potential?
• Conditional forecasts: o Unwind previous shocks – is this the adjustment parameter?
o Relative performance – measured relative to what: country specific export potential (=forecast value) or sample average
o “Best performers” – best in terms of highest export growth? But not best in terms of meeting potential (i.e. minimal forecast error)
www.oenb.at [email protected] - 8 -
Alexei Kireyev (IMF) and Andrei Leonidov (Lebedev Institute) : “Can Network Effects and Counter-Shock Policies Weaken International Trade”
Novel and detailed analysis of shock propagation highlighting network effects
Findings
Elements
Possible extensions
Unclear points
www.oenb.at [email protected] - 9 -
Kireyev & Leonidov: Findings
• Direct and indirect effects of a shock differ depending on where the shock
originates and who are the trading partners
• More developed and more open countries propagate shocks more strongly, less
developed and especially commodity exporters often block shocks
• Regardless of the country of shock origin, some countries are always more
strongly affected (i.e. Hong Kong, Kyrgyz Republic)
www.oenb.at [email protected] - 10 -
Kireyev & Leonidov: Elements
• Import demand functions: needed to assess the reaction of a country to a foreign
shock (pass-through coefficients), very detailed estimations (different
specifications, different lags, nominal and real term, for each country)
• Matrix of export-import-links / weights / spatial correlation
• Susceptibility to shocks is determined by exposure to origin country AND trade
elasticity
• Network algorithm
www.oenb.at [email protected] - 11 -
Kireyev & Leonidov: Possible extensions
• Add overview: which origin countries are responsible for the largest
repercussions across global trading system in case of a shock
• Assess the extent of substitution / trade diversion that follows in reaction to a
shock in one specific country
• Account for the crucial role of final demand vs demand for intermediates /
position and participation in global production networks as before, this would
be an entirely different paper
www.oenb.at [email protected] - 12 -
Kireyev & Leonidov: Unclear points
• Discussion of BoP-balances misleading, as focus is on trade only as shock
transmission channel
• Similar unclear if cointegration framework is necessary and used
• “export” vs “import” shock: does this refer to direct vs indirect effects?
Wouldn’t they necessarily have to be aggregated?
• How are the spillovers computed? There is no VAR, hence no IRFs
• What exactly is W (wij): share of j (=shock origin) in i’s exports (=potential
amplifier, absorber, blocker; = trade weight) * βi (=import elasticity to exports)
www.oenb.at [email protected] - 13 -
Kireyev & Leonidov: Unclear points
• Definition of shock diffusion:
• βI > 1 … amplifying country ok
• 0 < βI /DetectBlends true /DetectCurves 0 /DoThumbnails false /DownsampleColorImages true /DownsampleGrayImages true /DownsampleMonoImages true /EmbedAllFonts true /EmbedJobOptions true /EmbedOpenType false /EmitDSCWarnings false /EncodeColorImages true /EncodeGrayImages true /EncodeMonoImages true /EndPage -1 /GrayACSImageDict > /GrayImageAutoFilterStrategy /JPEG /GrayImageDepth -1 /GrayImageDict > /GrayImageDownsampleThreshold 1 /GrayImageDownsampleType /Bicubic /GrayImageFilter /DCTEncode /GrayImageMinDownsampleDepth 2 /GrayImageMinResolution 200 /GrayImageMinResolutionPolicy /OK /GrayImageResolution 300 /ImageMemory 1048576 /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /LockDistillerParams false /MaxSubsetPct 100 /MonoImageDepth -1 /MonoImageDict > /MonoImageDownsampleThreshold 1 /MonoImageDownsampleType /Bicubic /MonoImageFilter /CCITTFaxEncode /MonoImageMinResolution 800 /MonoImageMinResolutionPolicy /OK /MonoImageResolution 1200 /Namespace [ (Adobe) (Common) (1.0) ] /NeverEmbed [ true ] /OPM 1 /Optimize true /OtherNamespaces [ > > /FormElements true /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MarksOffset 14.17323 /MarksWeight 0.25000 /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /NA /PageMarksFile /RomanDefault /PreserveEditing false /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /LeaveUntagged /UseDocumentBleed true >> ] /PDFX1aCheck false /PDFX3Check false /PDFXBleedBoxToTrimBoxOffset [ 14.17323 14.17323 14.17323 14.17323 ] /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXOutputCondition () /PDFXOutputConditionIdentifier () /PDFXOutputIntentProfile (ISO Coated v2 300% \050ECI\051) /PDFXRegistryName () /PDFXSetBleedBoxToMediaBox false /PDFXTrapped /False /PDFXTrimBoxToMediaBoxOffset [ 0 0 0 0 ] /ParseDSCComments true /ParseDSCCommentsForDocInfo true /ParseICCProfilesInComments true /PassThroughJPEGImages true /PreserveCopyPage true /PreserveDICMYKValues true /PreserveEPSInfo true /PreserveFlatness false /PreserveHalftoneInfo false /PreserveOPIComments false /PreserveOverprintSettings true /StartPage 1 /SubsetFonts true /TransferFunctionInfo /Apply /UCRandBGInfo /Preserve /UsePrologue false /sRGBProfile (sRGB IEC61966-2.1)>> setdistillerparams> setpagedevice
