International Macroeconomics and Finance Sessions 1-2 Nicolas Coeurdacier - nicolas.coeurdacier@sciences-po.fr Master EPP - Fall 2013
Practical Information email : nicolas.coeurdacier@sciences-po.fr References: reading list for articles. Obstfeld and Rogoff "Foundations of International Macroeconomics". Compulsory readings marked with an *onthewebsite. Course website: Link Master EPP on my webpage http://econ.sciences-po.fr/staff/nicolas-coeurdacier
Practical Information Grading Mid-Term Homework: 30% of final grade. Given at lecture 6. To be handed back at lecture 8 Final Exam: 70% of final grade. One (or two) non-technical question(s) related to the course. One (or two) problem set(s).
Objectives of the course current policy and academic issues of globalization provide main academic background and basic technical tools for understanding academic and policy papers in international macro. Both theoretical and empirical contributions key introduction to recent academic work in international macroeconomics and finance. Prerequisite for International Macro II course with P. Martin & P-O Gourinchas. Focus mostly on real aspects. Nominal aspects tackled in International Macro II. prerequisites: master level course in macroeconomics
Road-map for the course Lectures 1-2: Financial integration, globalization and world income Lectures 3-6: International capital flows and current account dynamics Lectures 7-9: International Real Business Cycles Lectures 10-12: International Portfolios, Risk Sharing and Asset Prices
Financial integration, globalization and world income Stylized facts on financial globalization: past and present Gains from international financial integration: theory and empirics
Stylized facts on financial globalization: past and present Financial globalization 6= Trade globalization Measures of trade openness : What are the restrictions (tariffs and regulations) to free trade? (Exports + Imports)/GDP Measures of financial globalization: financial transactions extent of the openness in cross-border
Measures of financial globalization De Jure and de Facto financial openness measures De Jure: What are the restrictions to international capital movements based on the information from the IMF s Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER); example: In October 09, Brazil decided to tax capital inflows to discourage shortterm hot money from flowing in. De Facto: how much international trade in financial assets?
Which financial assets? Characteristics of financial assets: mean to transfer some purchasing power over periods (and states) - Portfolio investment: equity or debt - Foreign direct investment: 10% ownership - Other investments: loans, trade credit - Derivatives (futures, options) - Reserves (central banks)
Financial openness (De Jure) Chinn Ito index based on IMF information on restrictions to capital movements 2,5 0,6 2 0,4 1,5 0,2 0 1-0,2 0,5-0,4 0-0,6-0,5-0,8 2008 2006 2004 2002 2000 1998 1996 1994 1992 1990 1988 1986 1984 1982 1980 1978 1976 1974 1972 1970 Developed Count ries (l eft -axis) Emerging Countries (excluding Central and Eastern Europe) (right-axis) Note: Index between 2.5 and 2.5. 2.5=Closed capital market; 2.5=Fully opened Source : Chinn and Ito, 2008
The world map of financial openness (De Jure) : index based on IMF information on restrictions to capital movements Source: Chinn and Ito, 2008
De facto measure of financial globlization: flows and stocks Flows:thevalueofassetstradedforagivenyear: Stocks: the value of assets held in a given year: = 1 + = 2 + 1 + =... Stocks are the cumulative flows Several measures of financial globalization: Stocks: IFI (International Financial Integration) measure = (Domestic assets held by foreigners + Foreign assets held by domestic agents)/gdp Issueofvaluation(thevalueofassetscanchangeovertime,seelater)
Several measures of financial globalization: Flows: inflows/gdp and outflows/gdp Inflows: capital inflows/gdp: net purchases of domestic assets by foreign investors (for example, a loan by a foreign bank to a domestic firm). Inflows can be negative if a foreign resident sells a domestic asset to a domestic resident Outflows: net purchases of foreign assets by domestic investors (for example, a domestic household buying a bond issued by a foreign government)
International financial openness, 1970 2004 (Domestic assets held by foreigners + Foreign assets held by domestic agents)/ GDP source Lane and Milesi-Ferreti (2007) Strong increase in international assets held in both groups More so in industrialized countries (x7!) than in emerging and dev. countries (x3)
Flows are more volatile than stocks: in the 2008 crisis, collapse of international flows Financial Globalisation
The big retrenchment during the crisis: the end of financial globalization? capital inflows: net purchases of domestic assets by foreign investors 25% 20% Capital Inflows (ratio of world GDP) 15% 10% 5% 0% 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008-5% Advanced Latin America Middle East Emerging A sia Central and Eastern Europe Africa World Source: Milesi-Ferretti and Tille
Two forms of globalization «Real»: trade flows Financial: financial flows Compare the two forms of globalization: = Ratio of financial openness (financial assets) to real openess (goods) =(Domestic assets held by foreigners + Foreign assets held by domestic agents)/(exports + Imports)
The rise and collapse (and rise?) of trade openness: 1995-2009 Source: Amiti and Weinstein, Industrialized countries that accounted for 66% of world GDP and 68% of world exports in 2008.
Trade and financial integration, 1970 2004 source Lane and Milesi-Ferreti (2007) Domestic assets held by foreigners + Foreign assets held by domestic agents Exports + Imports
Comparison of industrialized countries share in goods trade and financial trade
The first financial globalization World capital markets very integrated at the end of the 19th century: Share of British wealth invested overseas: 17% in 1870 and 33% in 1913 (larger than any country today). Similar in France, Germany Capital outflows from UK (purchase of foreign assets): mostly to the «New World» with natural resources: Canada + Australia (28%), US (15%), Latin America (24%) source Taylor and Williamson (1994) What form? harbors) Portfolio investment (equity and bonds to invest in railroads,
Capital mobility: Obstfeld and Taylor, 2002 (a narrative based measure)
The first and second globalization: the financial side
Causes and consequences of the 19th century financial globalization Causes: - Transportation and communication (telegraph): information! - High investment/growth in New World Consequence: - Fosters catch-up of the New World - European capital chased European labor: both migrated to New World
A Remark: Capital Flows and the Lucas Puzzle Neoclassical growth model predicts: -Capital flows to capital scarce countries (higher marginal productivity of capital) - Capital flows towards fast growing countries Lucas Puzzle: Capital does not flow towards poorer countries. Less of a puzzle during first globalization wave than now (see Part 2: Assessing long-run international efficiency )
Foreign capital used to flow to poor and rich countries, but now flows mostly to rich countries (The Lucas Puzzle)
The case for Financial Globalisation? Washington Consensus: collection of loosely articulated ideas in the beginning of the 1990s aimed at modernizing, reforming, deregulating and opening economies Mostly came from Latin American governments (IMF, WB, US Treasury came later) Consequence: many emerging markets opened up their capital markets in the 90s (while most developed markets were already opened, thus since the 80s). Important to note restrictions on capital mobility are still more stringent for developing countries (see previous graphs). Less of a consensus now. Why did governments promote financial integration so actively?
Expected gains from financial integration 1) Intertemporal gains: consumption smoothing in response to shocks or in response to capital scarcity - small (Gourinchas and Jeanne (2006)) 2) Intratemporal gains = international risk-sharing - still quite a debate on their magnitude (large if look at asset prices/small if look at real consumption) 3) Growth effects: risk-taking and specialization - scarce empirical evidence 4) The benefits in terms of domestic allocative efficiency: superior foreign technology (FDI), market discipline on domestic policies, social infrastructure, etc... [The] main potential positive role of international capital markets is to discipline policymakers who might be tempted to exploit a captive domestic capital market (Obstfeld 1998).
Expected gains from financial integration - Intertemporal gains from financial integration Gourinchas and Jeanne (2006) - Intratemporal gains = international risk sharing Basic two country (static) extension of Lucas (1978) -Risk diversification and risk-taking Saint Paul (1993) and empirical evidence
Gains from financial integration? Empirics - Cross country regressions using IMF-based measures look at the impact of financial integration on growth. - Results range from no effect (Rodrik (1998)), to (small) significant effects (Quinn(1997,2008), Edwards (2001), Bekaert et al (2005)...). - Stock market liberalization increases equity prices (Henry (2003)) and to a lesser extent investment and growth (Henry (2003), Bekaert et al (2005)) - Not obvious how to translate a given increase in growth in terms of welfare: how permanent is the effect on growth? Does it change output levels in steady state? What share goes to foreigners?
- Modern empirical growth literature emphasizes conditional convergence (Mankiw Romer and Weil (1992), Barro, Mankiw and Sala-i-Martin (1995)). - Early papers stressed factor accumulation, but recent literature emphasizes total factor productivity (TFP) or social infrastructure (Hall and Jones (1999), Parente and Prescott (2001)) as main drivers of cross-country income differences. - Large welfare gains difficult to reconcile with portfolio home-bias
Empirical evidence of financial liberalization on growth - Evidence based on event study in Henry (2003) for a sample of emerging markets - Tests predictions of standard neoclassical model: (i) financial integration boosts growth and investment. (ii) reduces the cost of capital (or increases asset prices).
Empirical evidence of financial liberalization on growth Bekaert et al. (2003) investigates the opening of stock markets to foreign investors in a sample of 95 emerging markets. Pick up equity market liberalization dates (6= capital account liberalization where effects are found to be smaller/less robust) Find roughly 1% increase in real GDP growth after stock market liberalization. Mostly through capital accumulation but also TFP growth. -Temporary effect? - Is the date exogenous? Is it financial integration of stock markets or just financial development? - Upper bound of the effect?
Financial integration and real GDP growth Source: Bekaert et al. (2003)
Classic Growth Regression and the Impact of Liberalization Sample I II III IV Constant -0.2281-0.2374-0.1493-0.2018 Std. error 0.0179 0.0214 0.0286 0.0658 Log(GDP) -0.0094-0.0088-0.0115-0.0158 Std. error 0.0007 0.0007 0.0008 0.0011 Govt/GDP -0.0039-0.0178-0.0187-0.0301 Std. error 0.0087 0.0098 0.0105 0.0165 Enrollment 0.0305 0.0112 0.0243 0.0566 Std. error 0.0077 0.0097 0.0116 0.0171 Population Growth -0.5594-0.5731-0.8159-1.1013 Std. error 0.0621 0.0691 0.0835 0.1151 Log(Life Expectancy) 0.0755 0.0781 0.0627 0.0838 Std. error 0.0049 0.0056 0.0076 0.0167 Official Liberalization Indicator 0.0095 0.0083 0.0113 0.0130 Std. error 0.0016 0.0017 0.0020 0.0036 Source: Bekaert et al. (2003)
Financial integration boosts investment Impact F inanci al Liberalization on GDP components 0,02 0,015 0,01 0,005 0-0,005-0,01-0,015-0,02-0,025 Investment/GDP Consumption/GDP Government Spendings/GDP Net Exports/GDP Source: Bekaert et al. (2003)
The elusive gains for international financial integration Gourinchas and Jeanne (2006) proposes a new piece of empirical evidence based on calibration of a standard neoclassical growth model Main findings: first class of benefits is small why? - Countries needs to be very capital-scarce or abundant to experience large gains from financial integration. - The convergence-gap accounts for little of the world income inequality. Most inequality explained by long-run cross country differences in productivity or social infrastructure (Hall and Jones (1999)). - Important implications for the research agenda on capital account liberalization.
Gourinchas and Jeanne (2006) - Look at the implications for capital account liberalization. Focus on welfare benefits in response to capital scarcity. - Calibrate variants of the standard (Ramsey-Koopman-Cass) model and compare transition paths towards steady state under two scenarios: - financial autarky; - perfect financial integration with the rest of the world (small open economy).
Dynamic of consumption: autarky
With financial integration: If this is a poor country and ROW in SS: Euler equation becomes: = +1 ( ) 1 so consumption grows at when country is financially integrated. Capital stock jumps to its steady state level. Why?
Dynamic of consumption: autarky versus integration
How much gain? Calculate equivalent variation defined as the % (permanent) increase in consumption that brings domestic welfare under autarky up to its level under integration. Calibration:
Large gains require very large capital deficit (or surpluses): need 1 3 for 2%
This despite (temporary) gains in growth (consistent with literature: after equity market liberalization, GDP growth increases by 1% over the next 5 years) Basic intuition: the distorsion from financial autarky is transitory in nature (the distorsion disappears anyway) - either speed of convergence is fast (Ramsey) and not much gain from going fast to SS - or the problem is not convergence but the level of the SS level of capital, consumption...
Limits and potential extensions to the neoclassical model Deterministic view: risk affects the steady state level of capital stock (precautionary savings). Hence, financial integration by providing risk sharing opportunities modifies the steady state. Can go either way in terms of welfare (see Coeurdacier, Rey and Winant (2011) for a concept of risky steady-state and Coeurdacier, Rey and Winant (2013) for the welfare gains of integration with production & risk). Absence of financial frictions: capital scarcity can be due to credit constraints. If financial integration alleviates credit constraints, can generate permanent welfare gains. Possibility of non-convexities (poverty traps).
Financial integration and international risk-sharing: an introduction -Other gains from financial integration: allow countries to diversify risks and stabilize consumption. - Empirical evidence: hard to reconcile with the data though. 1) some (but scarce) evidence financial integration reduces volatility of consumption (small impact and financial integration increases the probability of crisis) 2) risk-sharing predicts high consumption correlation across countries. But smaller in the data than the correlation of output. "The quantity puzzle". (muchmoreonthisinlectures7-9).
A basic model of international risk-sharing Set-up Two symmetric countries ( ) and( ). One good (numeraire) 2periods: =0et =1 2 states of the world in each country ( = { }) with equal probabilities = 1 2 ; = probability that the weather is the same in both countries (rain or sun); 0 1.
A basic model of international risk-sharing Set-up Representative agent in country ( = ) owns a project at =0. At =1project delivers =(1+ ) if = { } and =(1 ) if = { }. At =1, uncertainty is realized and agents consume according to a standard CRRA utility: ( ) = 1 1
Autarky Agents consume the proceeds of their own project. = (1 + ) if = { } and (1 ) if = { } In country = ( )=1; ( )= 1 h ( 1) 2 +( 1) 2i = 2 2 ( ) = (( 1) ³ 1 ) = 2 (1 ) 2 = 2 2 ( 1 2 ) Cross-country consumption correlation = correlation of output
Financial integration Agent in country ( = ) can sale shares of her project at price (number of shares normalized to unity; share price identical across countries). =number of shares bought in the other country (=number of shares sold to the other country) Consumption at =1in country 6= in state { ; } for a given portfolio ³ ; = +(1 ) Portfolio maximization [can you guess the result?]: max 0[ 1 1 ]
[Technical steps towards the solution] 0 [ 1 1 ]= 2 µ (1+ ) 1 1 (1 ) 1 + 1 + 1 2 µ ( (1+ )+(1 )(1 )) 1 1 ( (1 )+(1 )(1+ )) 1 + 1 0 [ 1 1 ]= 2 µ (1+ ) 1 1 (1 ) 1 + 1 + 1 2 µ (1 +2 ) 1 1 (1+ 2 ) 1 + 1
Financial integration Portfolio maximization Equivalent to (for 6= 1;why?): +2 ) (1 + 2 ) max[(1 + 1 1 2 (1 +2 ) =2 (1 + 2 ) 1 1 ] = 1 2 (if 6= 0;why?) Portfolio which allows perfect pooling of country risk. Portfolio fully diversified internationally (cf. Lucas (1982))
Financial integration In country = ( )=1; ( )= 2 h ( 1) 2 +( 1) 2i = 2 As long as 1, volatility of consumption falls with respect ot autarky: welfare gains from risk sharing due to risk diversification (gains are higher when is low). Note that consumption is not constant (there is still world aggregate risk unless =0). = in all states Perfect correlation of consumption across countries. In particular higher than output correlation.
Welfare implications Income smoothing across states provide positive welfare gains. Financial integration complete financial markets. How big are these gains? Small if one looks at consumption and standard CRRA preferences (unless very large coefficient of risk aversion). Similar to Lucas calculations about the gains from removing business cycles. But asset prices extremely volatile: welfare gains implied by asset prices usually larger. Still hard to reconcile both views (see Lewis (2000)).
Implications of financial integration and discussion of the evidence 1) Risk diversification lowers consumption volatility. Scarce evidence and small effects but yes (see Bekaert et al. (2006)). Issue: also seems to increase the probability of crisis. 2) Raises consumption correlation (and in particular above output correlation). Consumption far to be perfectly correlated across countries in the data. Points the lack of risk sharing but at least countries which exhibit higher degree of integration have higher correlation of consumption (see Imbs (2006)). 3) Risk diversification implies full international portfolio diversification. Not true in the data although international portfolio diversification has been increasing over the last two decades. Investors still holds a disproportionate share of local assets: Home bias in equities (French and Poterba (1991)). More evidence in lectures 10-12.
The limits of financial integration - Consumption correlation across countries quite low and lower than output for most countries ("Quantity Puzzle"; see lectures 4-6) - Home bias in equity puzzle (see also lectures 10-12) Investors tend to hold a disproportionate share of their local assets. Goes against the view of a large decrease in barriers to international investments; unless it is actually optimal to hold undiversified portfolios. Note: Useful measure of Home Bias: =1 Share of Foreign of Equity Holdings Share of Foreign Stocks in World Market Capitalisation.Why?
1 0.9 Japan 0.8 Oceania North America 0.7 Europe 0.6 0.5 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Home Bias (HB) across time for selected regions
Financial markets, risk diversification and growth - International financial markets provide risk diversification (risk sharing). - Saint Paul (1993) and Obstfeld (1994) go one step further: risk diversification allows more risk taking towards more productive activities (specialization). (see also Acemoglu and Fabrizio Zilibotti (1997)) - Financial integration improves the risk/return of savers = risk taking/specialization = growth effect of integration beyond standard neoclassical and risk-sharing effects
Risk-taking and financial integration adapted from Saint Paul (1993) Set-up One country ( ) and 2 technologies of production (1 and 2) 2 states of the world ( =1or 2) with equal probabilities 2periods: =0et =1 Entrepreneurs have one unit of capital at =0 Production technologies are risky with constant return to scale
- technology 1 pays at =1in state =1(proba = 1 2 ) and gives per unit of capital invested at =0(and gives 0 is state =2)) - technology 2 pays at =1in state =2(proba = 1 2 ) and gives per unit of capital invested at =0(and gives 0 is state =1)) We assume ( is more efficient in technology 1) At =0, entrepreneurs invest their inital dotation in each technology (technological choice) (resp. (1 )) = capital invested in technology 1 (resp. 2).
At =1, uncertainty is realized and entrepreneurs consume according to a standard CRRA utility: ( ) = 1 1 1) Autarky Technological choice max 0 [ 1 1 ]=max(1 2 ( ) 1 1 + 1 ( (1 )) 2 1 1 ) Optimal technological choice satisfies 1 =( )1 1 = ( )1 1 1+( )1 1
Entrepreneurs invest in both production to diversify risk. More in 1 if 1 (the opposite if 1). Very risk averse entrepreneurs have a strong desire to smooth consumption across states invest more in the low return technology Consumption allocation across states 1 = ; 2 = (1 ) = + (1 ) 2 Specialization choice: trade off between volatility of consumption and average level of consumption
2) Financial integration Second country ( ) completely symmetric in all respects except: - technology 1 pays at =1in state =1(proba = 1 2 ) and gives per unit of capital invested at =0(and gives 0 is state =2)) - technology 2 pays at =1in state =2(proba = 1 2 ) and gives per unit of capital invested at =0(and gives 0 is state =1)) is more efficient in technology 2 (gains from specialization)
Entrepreneurs can buy and sell claims on future output (portfolio choice made at =0) in international financial markets. Buying shares of firms in country gives right to a share of country production at =1(number of shares normalized to unity). Idem for country. Before trading claims, entrepreneurs hold all the shares of the country. Same technological choice as earlier. Once technological choice has been made, entrepreneurs trade shares of their firms with foreigners. We denote by the number of shares of country held by entrepreneurs in country (and (1 ) the number of shares held in country by entrepreneurs in country ).
Portfolio choice for a given Duetosymmetryifcountry chooses to invest in technology 1, country chooses to invest in technology 2 Portfolio choice maximization: max max (1 2 0[ 1 1 ] ( +(1 ) (1 )) 1 1 + 1 ( (1 )+(1 ) ) 2 1 1 )
Portfolio choice for a given FOC: ( (1 )) ³ 0 ( 1 ) 0 ( 2 ) =0 1 = 2 Then: +(1 ) (1 ) = (1 )+(1 ) = 1 2 Perfect pooling of risk. Such a portfolio allows optimal risk diversification for any value of
Technological choice Given optimal, expected utility of an entrepreneur is: ( ) = 1 2 ( + (1 ) 2 ) 1 1 + 1 2 ( + (1 ) 2 ) 1 1 = 1 (( ) + ) 1 2 1 1 ( ) with since. Efficient specialization requires =1. Intuition?
Consumption allocation across states Sharing of world output (perfect risk-sharing) 1 = 2 ; 2 = 2 = 2 (i) Consumption is not risky anymore (despite higher specialization in each country) = standard risk-sharing (intratemporal) gains. (ii) Average consumption has increased due to efficient specialization (iii) Volatility of output has increased.
Risk diversification, risk taking and growth: empirical evidence? Very scarce, only indirect evidence but... Kalemli-Ozcan Sorensen and Yosha (2003) Hypothesis tested: Two regions better integrated financially are more specialized as they can share risks better. Sample of non-eu OECD countries and regions within countries for EU, Canada, US.
Two stages: 1) measure the degree of "lack of risk-sharing" (=sensitivity of incomes to local output = 1 ) in a given region (i): log {z } = + 1 log {z } + Risk-sharing is measured by =1 1 2) investigates the impact of ( =1 1 ) on regions specialization (under the assumption that countries devote more ressources to sectors where they are the most productive). Specialization Herfindahl Index at 2-digits ISIC category.
Empirical issues Endogeneity: more specialization of countries can increase the need for financial integration and risk sharing (reverse causality). Use variables related to financial development (such as shareholder rights, the size of the financial sector (relative to GDP), legal systems) as instruments for the amount of risk sharing. Similar findings.
Risk diversification, risk taking and growth: empirical evidence? Thesmar and Thoenig (2012) Investigates the liberalization of Paris-Bourse over the period 1986-1990 on risk-taking. Liberalization made French firms where able to sell shares to diversified international investors Significant increase in the volatility of cash-flows of listed firms (compared to non-listed ones)
Risk diversification, risk taking and growth: empirical evidence? Kalemli-Ozcan, Sorensen. and Volosovych (2010) Similar evidence: Firms with more diversified (international ownership) are more volatile sales growth and operating revenues. Large sample of firms in Europe (AMADEUS) over 1996-2006. 16 countries Important additional finding: result survives when aggregating up at the macro-level (regional level). Not obvious à priori, why? Control as much as possible for reverse causality by controlling for firm FE and also using IV ( exogenous harmonization of financial regulation between EU countries and trust ).
Cross-sectional regressions Firm in region in country and sector log( )= + + log(1 + )+ + = average measure of firm-level volatility = % of foreign ownership (also use dummies when largest shareholder is foreign), country and sector dummies, set of control variables
Panel regressions Firm in region in country, sector and date log( )= + + + + log(1+ )+ + = time varying measure of firm-level volatility = % of foreign ownership, country and sector dummies, set of control variables
Aggregating up at the regional level Aggregate volatility of sales/operating revenues at the regional level Aggregate foreign ownership at the regional level ( 100 regions) = X with =weight of firm in region Compute also median variables in the region. Same results. log( )= + + + log(1 + ) + +