The Portfolio Channel of Capital Flows: A Small Open Economy - PowerPoint PPT Presentation

Porfolio Shocks The Portfolio Channel of Capital Flows: A Small Open Economy Approach Carlos Montoro (BCRP) & Marco Ortiz (Universidad del Pac´ ıfico) Presented by: Marco Ortiz 2020 First Conference on Financial Stability and Sustainability ma.ortizs@up.edu.pe Marco Ortiz January 2020 1/29

Porfolio Shocks Contents Motivation The Model Results Conclusions Marco Ortiz January 2020 2/29

Porfolio Shocks Motivation ◮ Exchange rate determination has been one of the major puzzles in International Macroeconomics. ◮ Meese & Rogoff (1983): Asset price models fail to explain variations in exchange rates. ◮ Meese (1990): The proportion of exchange rate fluctuations that current economic models can predict is essentially zero. ◮ Microestucture Approach: Evans & Lyons (2002), Payne (2003) find a strong positive between order flow and returns in the FX market. ◮ Breedon & Vitale (2010) show that the order flow impacts in the returns of the exchange rate through portfolio and and information channels. Marco Ortiz January 2020 3/29

Porfolio Shocks Motivation (2) ◮ Microstrucure in th FX Market ◮ Lyons (1997), Evans & Lyons (2004), Bacchetta & van Wincoop (2006), Hau & Rey (2006), Breedon & Vitale (2010) present models on information and liquidity. ◮ FX Intervention ◮ Vitale (2011): FXI model with a signalling a a portfolio channel. ◮ Vitale (2003): FXI and monetary policy model. ◮ None of these models is full-scale SOE NK-DSGE model. Marco Ortiz January 2020 4/29

Porfolio Shocks Motivation (3) Figure 1: FX Intervention - Selected Countries (2002 - 2016) Source: Domanski, Kohlscheen & Moreno (2016) Marco Ortiz January 2020 5/29

Porfolio Shocks Motivation (4) ◮ The workhorse SOE NK-DSGE used by central banks for policy analysis regularly assumes that the exchange rate is determined by a UIP condition and that the foreign position for domestic currency bonds is zero. ◮ We observe in the data that the position of non-resident investors in domestic currency bonds shows a strong correlation with the exchange rate. Marco Ortiz January 2020 6/29

Porfolio Shocks Motivaci´ on (5) Figure: Non-Resident Investor Holdings in Fixed Return Market and Nominal Exchange Rate (PEN/USD) Source: BCRP Marco Ortiz January 2020 7/29

Porfolio Shocks Motivation (6) Questions that we need to address ◮ How does portfolio changes affect the economy? ◮ How does FXI operates? ◮ What are the channels? ◮ What is the optimal monetary and FXI policy design? ◮ Do results change when we move from partial equilibrium to a fully scale SOE NK-DSGE model? Marco Ortiz January 2020 8/29

Porfolio Shocks What has it been done? FXI in DSGE Models: ◮ Montoro & Ortiz (2013), Benes et al. (2015), Vargas et al. (2013), Escud´ e (2012), Blanchard et al. (2015); Cavallino (2015); Engle (2011); Adler, Lama & Medina (2016); Fanelli & Straub (2016); Gabaix & Maggiori (2015), Chang (2018); Itskhoski & Muhkin (mimeo). Marco Ortiz January 2020 9/29

Porfolio Shocks What do we do? 1) Extend the standard NK SOE DSGE model to incude: ◮ A market of FX dealers. ◮ An explicit role for FX volatility. ◮ An explicit role for Non-Resident portfolio holdings. ◮ An interaction between monetary and FXI policies. Marco Ortiz January 2020 10/29

Porfolio Shocks What do we find? (ongoing) Porfolio shocks... ◮ impact FX markets and are transmitted to the rest of the economy. ◮ though the portfolio channel has an additional stabilization role through the current account. FX intervention... ◮ Is a useful instrument for the stabilization of the economy. ◮ The stabilization role goes beyond offsetting the portfolio shocks. ◮ There are important interactions between monetary and FXI policies. Marco Ortiz January 2020 11/29

Porfolio Shocks The Model ◮ NK-SOE DSGE with an FX market composed by risk averse dealers. ◮ Each dealer d pays the domestic interest rate i t for the household assets for one period (myopic) and invest in foreign and domestic assets. Dealers maximize: − E t e − γ Ω ι max t +1 ̟ ι,d, ∗ t where E t is the rational expectations operator, γ is the absolute risk aversion coefficient. Ω ι t +1 is the total return in domestic currency of the portfolio, given by: � � t +1 = (1 + i t ) ̟ d,ι t ) S t +1 ̟ ι, ∗ ,d A ι,S Ω ι + (1 + i ∗ − (1 + i t ) t t t � � = (1 + i t ) ̟ ι,d t ) S t +1 ̟ ι, ∗ ,d ̟ ι,d + S t ̟ ι, ∗ ,d + (1 + i ∗ − (1 + i t ) t t t t t − i t + s t +1 − s t ) ̟ ι, ∗ ,d ≈ ( i ∗ t Marco Ortiz January 2020 12/29

Porfolio Shocks The Model (2) ◮ Each individual dealer d demand for foreign currency is given by: t − i t + E t s t +1 − s t = i ∗ ̟ ι,d ∗ t γσ 2 where σ 2 = var t (∆ s t +1 ) is the unconditional variance of the exchange rate. We assume this variance as a constant, determined by the volatility of shocks and the FXI policy. ◮ Aggregating across dealers we obtain a modified UIP condition: t + γσ 2 ( B d, ∗ E t s t +1 − s t = i t − i ∗ ) t Marco Ortiz January 2020 13/29

Porfolio Shocks The Model (3) ◮ Non-resident investors (or carry traders) will have an exogenous demand for domestic bonds. t + S t B ∗ ,c B c = 0 t where: � ρ bc, ∗ exp( ε b c, ∗ B c, ∗ B c, ∗ � = ) t t − 1 t ◮ The non resident investors will requiere to increase the supply of foreign currency instruments to increase their domestic currency portfolio. Marco Ortiz January 2020 14/29

Porfolio Shocks Central Bank ◮ We introduce a Central Bank, which intervenes in the FX market (in addition to its inflation stabilization role.) The balance sheet of the Central Bank is given by: S t B cb, ∗ + B cb t = M s t + NW cb t ◮ where B cb, ∗ represents the NIR and B cb are the bonds issued by the t t central bank. M s t is the money supply and NW cb represents the Central t Bank’s net worth. Its flow constraint is given by: t +1 + S t +1 B cb, ∗ t +(1+ i cb, ∗ ) S t +1 B cb, ∗ B cb t +1 − M s t +1 + P t Γ cb t = (1+ i cb t ) B cb − M s t t t where Γ cb are the transfers to families. t ◮ We abatract from central bank’s net worth and money supply and assume: t + S t B cb, ∗ = 0 B cb ◮ Thus, when the centrla bank sells foreign instruments, it will do so against domestic currency ones, sterilizing all FXI. Marco Ortiz January 2020 15/29

Porfolio Shocks Bond Market Equilibrium and the Current Account ◮ Domestic bonds market equilibrium: B d t + B cb t + B c t = 0 . ◮ Holdings of foreign currency bonds by dealers (families) are determiend by the current account (log-linearized): b t − β − 1 b t − 1 b cb t − β − 1 b cb t − β − 1 b t − 1 � � � � + φ b ∗ � � φ b + φ b cb rer t + b ∗ + . . . t − 1 � � rer t + b ∗ ,cb − β − 1 b t − 1 . . . + φ b ∗ ,cb = t + y t − φ C c t + φ b + φ b cb ( i t − 1 − π t ) + φ b ∗ + φ b ∗ ,cb = t def � � i ∗ t − 1 + rer − π ∗ t t β β Marco Ortiz January 2020 16/29

Porfolio Shocks FX Intervention ◮ “Discretion” benchmark: = ε cb, 0 B ∗ cb t t ◮ Nominal exchange rate rule: = − φ ∆ s ∆ s t + ε cb, 1 B ∗ cb t t ◮ Reaction to Portfolio shocks: = − φ rer rer t + ε cb, 2 B ∗ cb t t Marco Ortiz January 2020 17/29

Porfolio Shocks Other equations of interest ◮ Aggregate demand ( y t ) y t = φ C ( c t ) + φ X ( x t ) − φ M ( m t ) + g t (1) ◮ Real exchange rate ( rer t ) rer t = rer t − 1 + ∆ s t + π ∗ t − π t (2) ◮ Total CPI ( π t ): π t = ψπ H t + (1 − ψ ) π M + µ t (3) t Marco Ortiz January 2020 18/29

Porfolio Shocks Other equations of interest (2) ◮ Domestic goods Phillips Curve ( π H t ): π H mc t − t H + βE t π H � � t = κ H (4) t +1 t ◮ Imported Goods Phillips Curve ( π M t ): π M = κ M mc M + βE t π M (5) t t t +1 ◮ Exported Goods Phillips Curve ( π X t ) π X t = κ X mc X t + βE t π X (6) t +1 ◮ Taylor Rule ( i t ) ı t = ϕ π ( π t ) + ϕ y ( y t ) + ε int ˆ (7) t Marco Ortiz January 2020 19/29

Porfolio Shocks Baseline Calibration Marco Ortiz January 2020 20/29

Porfolio Shocks Results (1) - Equilibrium (a) No Intervenci´ (b) Intervenci´ on ( ϕ ∆ s = 10 ) on Marco Ortiz January 2020 21/29

Porfolio Shocks Results (2) - Rules vs. “Discrection”) FX Intervention FX Intervention Pct. Dev. Steady State Pct. Dev. Steady State 2 1.5 RER Rule RER Rule 1 Disc. Disc. 1 0.5 0 0 -1 -0.5 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Quarters after shock Quarters after shock Real Exchange Rate Real Exchange Rate Pct. Dev. Steady State Pct. Dev. Steady State 0.05 0.2 0 0.1 0 -0.05 -0.1 -0.1 -0.2 -0.15 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Quarters after shock Quarters after shock Dep.Rate Dep.Rate Pct. Dev. Steady State Pct. Dev. Steady State 0.4 0.2 0.2 0.1 0 0 -0.1 -0.2 -0.2 -0.4 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 Quarters after shock Quarters after shock (c) Int. Rule 1 (d) Int. Rule 2 Marco Ortiz January 2020 22/29