Euro Area and U.S. External Adjustment: The Role of Commodity Prices and Emerging Market Shocks

Transcription

1 Euro Area and U.S. External Adjustment: The Role of Commodity Prices and Emerging Market Shocks Massimo Giovannini (European Commission, Joint Research Centre) Robert Kollmann (ECARES, Université Libre de Bruxelles and CEPR) (*) Marco Ratto (European Commission, Joint Research Centre) Werner Roeger (European Commission, DG ECFIN) Lukas Vogel (European Commission, DG ECFIN) The views expressed in this presentation are those of the author and not necessarily those of the European Commission

2 1. Introduction Since the early 2000s, the world economy has been undergoing substantial changes. Growth acceleration in emerging economies EA and US experienced a boom-bust cycle. Wide exchange rates fluctuations. Substantial variation in commodity prices Major trade balance adjustments. Before 2009: US trade balance deteriorated markedly reaching about -6% of GDP in , EA trade balance fluctuated around zero in the pre-2009 period. After 2009: EA & US trade balances both rose noticeably.

3 Questions: How these shocks account for movements in the trade balance of these three regions? What can we say about the future evolution of trade balances? Methodology used for answering these questions: The paper uses a fairly standard three-region model of the world economy augmented by a commodity/raw materials sector. The model is estimated (with Bayesian Methods), over the period 1999q1-2017q2 for the EA, US and an aggregate of rest of the world countries.

4 This paper: No mono-causal explanation for the dynamics of global imbalances. Boom- bust cycle view: Can explain widening trade deficit in the US followed by a contracting deficit. But Fails to account for the relatively stable EA trade balance before the crisis (though it declined from 2003). RoW growth divergence: Could account for a rising trade balance in EA and US in recent years. But Fails to explain the US trade balance before the great recession. Savings glut (Bernanke 2005): Possible explanation for the widening pre-crisis US trade deficit But Difficult to reconcile with the fall of the dollar (see Borio and Disyatat (2011)). Commodity price fluctuations: Rise and fall of commodity prices are broadly consistent with trade balance movements. But Could be largely endogenous.

5 Link to the literature: Quantitative analyses of recent oil and commodity fluctuations mostly rely on VAR models ( Kilian (2009), Kilian et al. (2009), Peersman and Van Robays (2009), Caldara et al. (2017). For stylized structural models of the role of energy for international adjustment, see, e.g., Sachs (1981), McKibbin and Sachs (1991), Backus and Crucini (2000) and Gars and Olovsson (2018). The paper here is closest to Forni, Gerali, Notarpietro and Pisani (2015), who estimate a twocountry DSGE model of the EA and the non-ea RoW, using data for Our model differs from that work: 3-region model (EA, US, RoW): analyze differences between EA and US external adjustment to recent global disturbances. Sample period that includes the post-2014 commodity price collapse. We consider a broader bundle of raw materials

6 Major trade balance adjustments. Industrial supplies net exports important for total net exports

7 Large Fluctuations of commodity prices Note: Strong co-movement of industrial supply prices

8 RoW Growth acceleration

9 EA Leveraging and deleveraging

10 Model description EA and US blocks Constrained and unconstrained households, firms and a government. EA and US households provide labor services to firms. Monopolistically competitive firms in the EA and the US using domestic labor and capital. A final good sector in the EA and the US combines domestic and imported intermediates and produces a homogeneous final good that is used for domestic consumption, capital accumulation and exports.

11 Production Production of total output in the EA/US is a four-stage process. 1) Monopolistically competitive firms j produce differentiated intermediate goods using capital and labor. y N FN cu K FC j j j j j 1 j t t ( t t ) ( t t ) t, 2) Differentiated intermediate products are bundled by perfectly competitive firms into domestic value added. 1 j Yt { ( yt) dj} 0 ( 1)/ /( 1) t t t t 3) Perfectly competitive firms combine domestic value added and imported industrial supplies (commodities) to obtain a domestic intermediate good. d d d d d d d d D ((1 s ) ( Y) ( s ) ( IS ) ) is 1/ ( /( 1) is 1/ ( 1)/ /( 1) t t t t t

12 Nominal intermediate good prices and wages are sticky. Domestic and foreign goods are imperfect substitutes. Nearly perfect international capital mobility across countries (up to a risk premium which depends on the net foreign asset position of the country), plus a stochastic exchange risk premium. Exchange rates among all three regions are flexible. Monetary policy is conducted using a Taylor rule. Public expenditure (G, IG, TR) responds to the government balance. Government in EA and US levy distortive taxes and issue debt.

13 The RoW block Representative household with CES preferences for domestic and imported goods. New Keynesian Phillips curve. Taylor rule for monetary policy. Production of manufactures + industrial supplies Manufactures: Production technology that uses labor as the sole factor input. Industrial supplies: A competitive sector supplies energy and non-energy materials, to domestic and foreign firms. The RoW commodity supply is increasing function of the real commodity prices. Commodity supply is subject to exogenous shocks.

14 Table 1. Prior and posterior distributions of key estimated model parameters Posterior distributions EA US Prior distributions Mode Std Mode Std Distrib. Mean Std (1) (2) (3) (4) (5) (6) (7) (8) Preferences Consumption habit persistence B Risk aversion G Inverse labor supply elasticity G Import price elasticity G Steady state consumption share of Ricardian households B Nominal and real frictions Price adjustment cost G Nominal wage adj. cost G 5 2 Real wage rigidity B Monetary policy Interest rate persistence B Response to inflation B Response to GDP B Fiscal policy Taxes persistence B Taxes response to deficit B Taxes response to debt B Commodities Commodities demand elasticity B

15 Commodities demand elasticity B Autocorrelations of forcing variables Permanent TFP growth B Subjective discount factor B Investment risk premium B Trade share B Commodities demand AR(1) N Commodities demand AR(2) N Standard deviations (%) of innovations to forcing variables Monetary policy G Gov. transfers G Permanent TFP level G Permanent TFP growth G Subjective discount factor G Investment risk premium G Domestic price mark-up G Trade share G Commodities demand G Commodities supply G RoW Commodities Inverse supply elasticity B Autocorr. Oil supply shock B Autocorr. Materials supply shock B Notes: Cols. (1) lists model parameters. Cols. (2)-(3) and Cols. (4)-(5) show the mode and the standard deviation (Std) of the posterior distributions of EA parameters and of US parameters, respectively. Cols. (6) (labelled Distrib. ) indicates the prior distribution function (B: Beta distribution; G: Gamma distribution; N: Normal distribution). Identical priors are assumed for EA and US parameters.

16 Fig. 7a Dynamic effects of a positive shock (1 standard deviation) to trend growth rate of RoW TFP Persistent growth shocks have only small TB effects, because demand (expected income growth) and supply move together. Sign of spillover effect uncertain: import demand (+) man. prices (+), commodity prices (-)

17 Fig. 7b. Dynamic effects of a negative demand shock in RoW (1 standard deviation) Strong effect of TB: RoW Demand (-) + RoW Devaluation Partly offset by IS price movements Spillover to EA larger (smaller responsiveness of domestic demand in EA).

18 Fig. 7c. Dynamic effects of positive shock to RoW commodity supply (1 standard deviation) Permanent decline of commodity prices. Rise in EA and US GDP (+RoW GDP) but lower RoW income. EA and US TB rises temporarily until aggregate demand has adjusted to permanently higher level.

19 Fig. 8a. Historical shock decomposition: Industrial supplies price, in Euro (yoy growth) Commodity prices are partly driven by demand and supply But remain subject to large supply and demand shocks (esp. after 2012)

20 Fig. 8b. Historical shock decomposition: RoW GDP growth (yoy) Persistent productivity growth shocks, interrupted in Demand shocks less important Pro-cyclical demand shocks from EA and US

21 Fig. 8c. Historical shock decomposition: EA GDP growth (yoy) Demand shocks are important for (boom-bust) cycle Some pro-cyclical movements from RoW savings Some counter-cyclical movements from bond premia, commodity (D+S)

22 Fig. 8d. Historical shock decomposition: US GDP growth (yoy) Demand shocks are important for (boom-bust) cycle Less pro-cyclical movements from RoW savings (see IRF) Some counter-cyclical movements from bond premia, commodity (D+S)

23 Fig. 8e. Historical shock decomposition: EA trade balance/gdp ratio Pre 2009: Savings RoW+Bond premium+domestic demand+commodity(s+d) After 2009: Domestic demand (09+11)+commodity(S+D)+bond premium+saving RoW

24 Fig. 8f. Historical shock decomposition: US trade balance/gdp ratio Pre 2009: Savings RoW+domestic demand+commodity(d) After 2009: Domestic demand (09)+commodity(D)+saving RoW

25 Conclusions Both the boom bust cycle in EA and US plus higher savings in emerging economies have contributed to fluctuations in the trade balance. Recent changes of the trade balance in EA and US are strongly influenced by large commodity price shocks. What does this imply for the likely evolution of the EA trade balance? The period 2014 until 2016 is characterized by an unusual decline of industrial supply prices which has increased the trade balance. Even if industrial supply prices would remain permanently at a lower level this should not have a permanent positive impact on the trade balance. The still sizeable contribution from domestic demand is likely to decline as positive savings shocks vanish. However rising aggregate demand in the RoW keeps the trade balance up.

26

27 Reserve Slides

28 Source: ECB

29

30

31