Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.
The trade and demand nexus:
Do global value chains matter?
Julia Wörz
Foreign Research Division
Oesterreichische Nationalb...
Rubric
www.oenb.at oenb.info@oenb.at2
Overview
1
2
3
Literature
Empirical set-up
Motivation
4
5 Concluding remarks
Results
Rubric
www.oenb.at oenb.info@oenb.at3
1
2
3
Literature
Empirical set-up
Motivation
4
5 Concluding remarks
Results
Overview
Rubric
www.oenb.at oenb.info@oenb.at4
Puzzle 1: Trade grew consistently faster than GDP.
Puzzle 2: The trade-GDP ratio dec...
Rubric
www.oenb.at oenb.info@oenb.at5
Puzzle 1: Trade grew consistently faster than GDP.
Puzzle 2: The trade-GDP ratio dec...
Rubric
www.oenb.at oenb.info@oenb.at6
• Standard trade models failed to explain changes in trade-GDP
growth ratio
• Reason...
Rubric
www.oenb.at oenb.info@oenb.at9
1
2
3
Literature
Empirical set-up
Motivation
4
5 Concluding remarks
Results
Overview
Rubric
www.oenb.at oenb.info@oenb.at10
• Yi (2001):
 Falling trade barriers too weak and too early to explain high trade
...
Rubric
www.oenb.at oenb.info@oenb.at11
1
2
3
Literature
Empirical set-up
Motivation
4
5 Concluding remarks
Results
Overview
Rubric
www.oenb.at oenb.info@oenb.at
Import demand function …
… augmented by an index of GVC participation:
 Most G20 and...
Rubric
www.oenb.at oenb.info@oenb.at13
Data sources
IMF DOTS
Value of imports between each country and its
trading partner...
Rubric
www.oenb.at oenb.info@oenb.at
1st GVC-integration measure: VSI
14
Vertical Specialisation Index (VSI)
2010=100
Sour...
Rubric
www.oenb.at oenb.info@oenb.at
2nd GVC-integration measure: GVC-participation
15
GVC participation index
2010=100
So...
Rubric
www.oenb.at oenb.info@oenb.at16
1
2
3
Literature
Empirical set-up
Motivation
4
5 Concluding remarks
Results
Overview
Rubric
www.oenb.at oenb.info@oenb.at
Results
17
Regression results advanced and emerging economies, 1980-2012
Pooled mean ...
Rubric
www.oenb.at oenb.info@oenb.at
Results
18
Regression results advanced and emerging economies, 1995-2012
Pooled mean ...
Rubric
www.oenb.at oenb.info@oenb.at
Results
19
1995-2011: alternative GVC-indicator (and alternative estimation methods)
...
Rubric
www.oenb.at oenb.info@oenb.at
Results
20
Recursive estimates, including TFE*GVC-part interaction term
Fixed effects...
Rubric
www.oenb.at oenb.info@oenb.at
Results
21
Recursive estimates, including TFE*VSI interaction
Advanced economies
lnye...
Rubric
www.oenb.at oenb.info@oenb.at22
1
2
3
Literature
Empirical set-up
Motivation
4
5 Concluding remarks
Results
Overview
Rubric
www.oenb.at oenb.info@oenb.at
Conclusions and way forward
23
• Countries with stronger involvement in GVCs tend to ...
Rubric
www.oenb.at oenb.info@oenb.at
Thank you !
Upcoming SlideShare
Loading in …5
×

Julia Wörz. The trade and demand nexus: Do global value chains matter?

705 views

Published on

Open seminar, Eesti Pank 03.09.2015

Published in: Economy & Finance
  • Be the first to comment

  • Be the first to like this

Julia Wörz. The trade and demand nexus: Do global value chains matter?

  1. 1. The trade and demand nexus: Do global value chains matter? Julia Wörz Foreign Research Division Oesterreichische Nationalbank Joint work with Alexander Al-Haschimi, Frauke Skudelny and Elena Vaccarino (ECB) Eesti Pank Seminar, 3 September 2015 www.oenb.at WORK IN PROGRESS…
  2. 2. Rubric www.oenb.at oenb.info@oenb.at2 Overview 1 2 3 Literature Empirical set-up Motivation 4 5 Concluding remarks Results
  3. 3. Rubric www.oenb.at oenb.info@oenb.at3 1 2 3 Literature Empirical set-up Motivation 4 5 Concluding remarks Results Overview
  4. 4. Rubric www.oenb.at oenb.info@oenb.at4 Puzzle 1: Trade grew consistently faster than GDP. Puzzle 2: The trade-GDP ratio declined. Ratio of global import growth to GDP growth (in PPP) Source: ECB staff calculations. Motivation
  5. 5. Rubric www.oenb.at oenb.info@oenb.at5 Puzzle 1: Trade grew consistently faster than GDP. Puzzle 2: The trade-GDP ratio declined since the mid-90s. Ratio of global import growth to GDP growth (in PPP) (in USD) Source: ECB staff calculations. Motivation
  6. 6. Rubric www.oenb.at oenb.info@oenb.at6 • Standard trade models failed to explain changes in trade-GDP growth ratio • Reasons for these changes are unknown, cyclical or structural? • This paper analyses the role of GVCs in this context (= possible structural reason). • We include indicators for the participation in GVCs in a standard import demand equation. • Alternatively, one could analyse differences between gross and value added trade. Motivation
  7. 7. Rubric www.oenb.at oenb.info@oenb.at9 1 2 3 Literature Empirical set-up Motivation 4 5 Concluding remarks Results Overview
  8. 8. Rubric www.oenb.at oenb.info@oenb.at10 • Yi (2001):  Falling trade barriers too weak and too early to explain high trade growth, but vertical specialisation can explain the puzzle • Eaton, Kortum & Romalis (2011):  During the crisis, spending shifted away from durable goods  composition effect hypothesis • Alessandria, Kaboski & Midrigan (2010); Altomonte, di Mauro, Ottaviano, Rungi & Vicard (2012); Bems, Johnson & Yi (2012):  Disproportionally large inventories in GVC trade, higher sensitivity of trade to foreign income shocks, bullwhip effect  supply chain effect hypothesis • Constantinescu, Mattoo & Ruta (2015):  decline in trade-GDP ratio started long before the crisis, thus reflecting longer term structural reasons • Ollivaud & Schwellnus (2015):  No decline in the ratio with correct GDP measurement and treatment of intra-EU flows, thus weak global demand is responsible Literature
  9. 9. Rubric www.oenb.at oenb.info@oenb.at11 1 2 3 Literature Empirical set-up Motivation 4 5 Concluding remarks Results Overview
  10. 10. Rubric www.oenb.at oenb.info@oenb.at Import demand function … … augmented by an index of GVC participation:  Most G20 and EU countries  Estimations for 1980-2012 and 1995-2012  2 alternative GVC integration measures:  Vertical specialisation index (only backward integration)  GVC participation (based on decomposition of gross exports) Empirical set-up 12 𝒍𝒏 𝑴𝒊𝒋𝒕 = 𝜶𝒊𝒋 + 𝜶 𝟏 𝒍𝒏 𝑻𝑭𝑬𝒊𝒕 + 𝜶 𝟐 𝒍𝒏 𝑷𝒋𝒕 𝑷𝒊𝒕 + 𝜶 𝟑 𝒍𝒏 𝑬𝑹𝒊𝒋𝒕 + 𝜺𝒊𝒋𝒕 𝒍𝒏 𝑴𝒊𝒋𝒕 = 𝒊𝒋 +  𝟏 𝒍𝒏 𝑻𝑭𝑬𝒊𝒕 +  𝟐 𝒍𝒏 𝑷𝒋𝒕 𝑷𝒊𝒕 +  𝟑 𝒍𝒏 𝑬𝑹𝒊𝒋𝒕 + 𝟒 𝒍𝒏 𝑮𝑽𝑪_𝒑𝒂𝒓𝒕𝒊𝒕 + 𝜺𝒊𝒋𝒕
  11. 11. Rubric www.oenb.at oenb.info@oenb.at13 Data sources IMF DOTS Value of imports between each country and its trading partners IMF WEO + IFS GDP, deflators, prices, exchanges rates WIOD Available across 40 countries over the period 1995-2011 CEPII – CHELEM Database Bilateral trade values, 79 countries, 1967-2012 GVC participation Index à la Koopman et al. (2014) Vertical Specialisation Index (VSI) à la Amador and Cabral (2009) Trade elasticities
  12. 12. Rubric www.oenb.at oenb.info@oenb.at 1st GVC-integration measure: VSI 14 Vertical Specialisation Index (VSI) 2010=100 Source: update of Amador & Cabral, 2009. 0 20 40 60 80 100 120 140 160 180 200 1967 1972 1977 1982 1987 1992 1997 2002 2007 2012 China India Brazil 0 20 40 60 80 100 120 140 160 180 200 1967 1972 1977 1982 1987 1992 1997 2002 2007 2012 United States Germany United Kingdom We observe increasing vertical specialisation, especially by EMEs
  13. 13. Rubric www.oenb.at oenb.info@oenb.at 2nd GVC-integration measure: GVC-participation 15 GVC participation index 2010=100 Source: authors’ calculations based on WIOD. Global Crisis induced a reversal in GVC participation intensity 𝑮𝑽𝑪 𝑷𝒂𝒓𝒕𝒊𝒄𝒊𝒑𝒂𝒕𝒊𝒐𝒏𝒊𝒕 = 𝑽𝑺𝟏𝒊𝒕 + 𝑭𝑽𝒊𝒋𝒕
  14. 14. Rubric www.oenb.at oenb.info@oenb.at16 1 2 3 Literature Empirical set-up Motivation 4 5 Concluding remarks Results Overview
  15. 15. Rubric www.oenb.at oenb.info@oenb.at Results 17 Regression results advanced and emerging economies, 1980-2012 Pooled mean group estimator, dependent variable: import volumes *** p<0.01, ** p<0.05, * p<0.1 Source: authors’ calculations. (1) (2) (3) long-run ln(TFE) 1.51*** 0.69*** 0.79*** ln(rel.M-prices) -0.96*** -0.97*** -0.97*** ln(er) 0.019** 0.022** 0.021** ln(VSI) 0.33*** ln(TFE) x ln(VSI) 0.050*** EC 0.18*** 0.17*** 0.17*** short-run d_ln(TFE) 3.21*** 3.02*** 3.38*** d_ln(rel.M-prices) -0.83*** -0.82*** -0.82*** d_ln(er) -0.031*** -0.027** -0.029** d_ln(VSI) 0.095*** d_[ln(TFE) x ln(VSI)] 0.36 const -0.35*** -0.76*** -0.75*** ADVANCED ECONOMIES 1980-2012 (4) (5) (6) 1.36*** 0.60 0.25 -0.13** -0.13** -0.12** 0.033 0.045 0.048 0.29** 0.081*** 0.13*** 0.14*** 0.14*** 3.85*** 2.85*** 3.83*** -0.50*** -0.48*** -0.49*** -0.14*** -0.14*** -0.15*** 0.30*** 2.37*** 0.084 -0.19 -0.38 EMERGING ECONOMIES 1980-2012
  16. 16. Rubric www.oenb.at oenb.info@oenb.at Results 18 Regression results advanced and emerging economies, 1995-2012 Pooled mean group estimator, dependent variable: import volumes *** p<0.01, ** p<0.05, * p<0.1 Source: authors’ calculations. (7) (8) (9) long-run ln(TFE) 1.55*** 1.00*** 0.96*** ln(rel.M-prices) -0.78*** -0.79*** -0.79*** ln(er) -0.086* -0.044 -0.047 ln(VSI) 0.23*** ln(TFE) x ln(VSI) 0.037*** EC 0.28*** 0.28*** 0.28*** short-run d_ln(TFE) 3.00*** 2.87*** 3.17*** d_ln(rel.M-prices) -0.88*** -0.87*** -0.87*** d_ln(er) -0.071** -0.084*** -0.084*** d_ln(VSI) 0.067*** d_[ln(TFE) x ln(VSI)] -0.14 const -0.42** -0.80*** -0.92*** ADVANCED ECONOMIES 1995-2012 (10) (11) (12) 1.57*** 0.72** 0.43 -0.27 -0.24 -0.27* -0.17 -0.089 -0.12 0.35*** 0.085*** 0.21*** 0.22*** 0.22*** 3.82*** 2.62*** 3.72*** -0.68*** -0.69*** -0.64*** -0.20*** -0.25*** -0.24*** 0.32*** 1.42** 0.058 -0.32 -0.66* EMERGING ECONOMIES 1995-2012
  17. 17. Rubric www.oenb.at oenb.info@oenb.at Results 19 1995-2011: alternative GVC-indicator (and alternative estimation methods) Advanced economies Emerging economies Fixed effects with AR(1) dist. Dynamic panel Fixed effects with AR(1) dist. Dynamic panel Lagged dep. 0.72*** 0.72*** 0.69*** 0.69*** TFE 1.51*** 1.41*** 0.35*** 0.25*** 1.42*** 1.40*** 0.48*** 0.36*** Long-term coef. 1.27 0.90 1.55 1.16 Relative prices -0.02 -0.05* 0.21*** 0.13*** -0.023 -0.06 0.15*** 0.07* Long-term coef. 0.76 0.46 0.47 0.24 ER 0.43*** 0.3*** 0.13*** 0.06** 0.40*** 0.38*** 0.11*** -0.01 Long-term coef. 0.46 0.21 0.37 -0.03 TFE*GVC_part 0.05*** 0.03*** 0.03*** 0.065*** Long-term coef. 0.11 0.21
  18. 18. Rubric www.oenb.at oenb.info@oenb.at Results 20 Recursive estimates, including TFE*GVC-part interaction term Fixed effects with AR, advanced economies recursive lntfe_d 0.00 0.50 1.00 1.50 2.00 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 with interaction term recursive lntfe_d 0.00 0.50 1.00 1.50 2.00 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 lntfe_d_part 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 with interaction term recursive lntfe_d 0.00 0.50 1.00 1.50 2.00 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 lntfe_d_part -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Fixed effects with AR, emerging economies recursive lntfe_d 0.00 0.50 1.00 1.50 2.00 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Results
  19. 19. Rubric www.oenb.at oenb.info@oenb.at Results 21 Recursive estimates, including TFE*VSI interaction Advanced economies lnyer_usd_d 0.00 0.50 1.00 1.50 2.00 2.50 3.00 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 Emerging economies lnyer_usd_d 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 lnyer_usd_d_part 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 lnyer_usd_d_part 0.00 0.05 0.10 0.15 0.20 0.25 0.30 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 lnyer_usd_d 0.00 0.50 1.00 1.50 2.00 2.50 3.00 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 lnyer_usd_d 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012
  20. 20. Rubric www.oenb.at oenb.info@oenb.at22 1 2 3 Literature Empirical set-up Motivation 4 5 Concluding remarks Results Overview
  21. 21. Rubric www.oenb.at oenb.info@oenb.at Conclusions and way forward 23 • Countries with stronger involvement in GVCs tend to import more, beyond the effect of the demand variable • Omitting these variables leads to higher estimates of income elasticities  larger difference for EME and for longer sample  Confirming our hypothesis that GVCs explain part of the high trade to GDP growth ratio • Income elasticity higher and GVC elasticity lower in advanced than emerging countries • Sensitivity of imports to GVC-integration rising for EMEs, but declining for AEs • On the agenda: more (and proper) robustness checks, allow for time- varying effect of GVC-participation, differentiate demand components
  22. 22. Rubric www.oenb.at oenb.info@oenb.at Thank you !

×