International Credit Flows, and Pecuniary Externalities. Princeton Initiative Princeton University. Brunnermeier & Sannikov

Transcription

1 International Credit Flows and Pecuniary Externalities Markus K. Brunnermeier & Princeton University International Credit Flows, Yuliy Sannikov Princeton Initiative 2017 Princeton, NJ, Sept. 9 th, 2017

2 Motivation Old Washington consensus in decline Free trade: flow of goods/services intratemporal Free finance: flow of capital intertemporal When does full capital account liberalization reduce (capital controls/macropru regulation improve) welfare? 1. Liquidity mismatch can lead to sudden stop runs Technological illiquidity: Market illiquidity: Funding illiquidity: irreversibility (adjustment costs) redeployability/specificity not in this paper short-term debt, hot money Type of capital flow matters: FDI, portfolio flows (equity), long-term debt 2. Terms of trade hedge (Cole-Obstfeld) can be undermined when Industry s output is not easily substitutable. Consumers cannot easily find substitutes No strong competitors in other countries Natural resources: oil, copper for Chile, Hard drives in Thailand, Bananas in Ecuador 2

3 Asset side Motivation Old Washington consensus in decline Free trade: flow of goods/services intratemporal Free finance: flow of capital intertemporal When does full capital account liberalization reduce (capital controls/macropru regulation improve) welfare? 1. Sudden stop including runs due to liquidity mismatch Technological illiquidity: Market illiquidity: irreversibility (adjustment costs) redeployability/specificity not this paper Funding illiquidity: short-term debt, hot money Type of capital flow matters: FDI, portfolio flows (equity), long-term debt 2. Terms of trade hedge (Cole-Obstfeld) can be undermined when Industry s output is not easily substitutable. Consumers cannot easily find substitutes No strong competitors in other countries Natural resources: oil, copper for Chile, Hard drives in Thailand, Bananas in Ecuador 3

4 Liability Asset side Motivation Old Washington consensus in decline Free trade: flow of goods/services intratemporal Free finance: flow of capital intertemporal When does full capital account liberalization reduce (capital controls/macropru regulation improve) welfare? 1. Sudden stop including runs due to liquidity mismatch Technological illiquidity: Market illiquidity: Funding illiquidity: irreversibility (adjustment costs) redeployability/specificity not this paper short-term debt, hot money Type of capital flow matters: FDI, portfolio flows (equity), long-term debt 2. Terms of trade hedge (Cole-Obstfeld) can be undermined when Industry s output is not easily substitutable. Consumers cannot easily find substitutes No strong competitors in other countries Natural resources: oil, copper for Chile, Hard drives in Thailand, Bananas in Ecuador 4

5 Motivation Old Washington consensus in decline Free trade: flow of goods/services intratemporal Free finance: flow of capital intertemporal When does full capital account liberalization reduce (capital controls/macropru regulation improve) welfare? 1. Sudden stop including runs due to liquidity mismatch Technological illiquidity: Market illiquidity: Funding illiquidity: irreversibility (adjustment costs) redeployability/specificity not in this paper short-term debt, hot money Type of capital flow matters: FDI, portfolio flows (equity), long-term debt 2. Terms of trade hedge (Cole-Obstfeld) can be undermined when Industry s output is not easily substitutable. Consumers cannot easily find substitutes No strong competitors in other countries Natural resources: oil, copper for Chile, Hard drives in Thailand, Bananas in Ecuador 5

6 Role of (international) financial markets Better allocation of physical capital/resources Better allocation of risk (sharing) Complete markets (1) and (2) can be controlled separately Pecuniary externalities have 2 nd order w-effects Frictions/incomplete markets (1) and (2) are interlinked Pecuniary externalities have welfare effects First Best Second Best 8

7 Literature Macro, Money with financial frictions BGG, Kiyotaki & Moore 1997/2008, Gertler & Kiyotaki, Mendoza, Bianchi, Brunnermeier & Sannikov 2012/14, He & Krishnamurthy 2013, Basak & Cuoco 1998 includes volatility dynamics + uncertainty how long crisis lasts terms of trade hedge Cole & Obstfeld 1991, Martin 2010 Constrained inefficiency, pecuniary/firesale externalities Incomplete markets: Stiglitz 1982, Newsbury & Stiglitz 1984, Geanakoplos & Polemarchakis 1986, Coeurdacier, Rey & Winant 2013, He & Kondor 2013 Debt collateral constraint (that depends on price) Stiglitz & Greenwald, Lorenzoni 2005, Bianchi 2011, Bianchi & Mendozza 2012, Jeanne & Korinek 2012, Stein 2012, Davilia & Korinek 2017, International Complete markets: Heathcote & Perri 2013, Backus, Kehoe, Kydland 1994 ToT manipulation: Keynes 1929, Costinout, Lorezoni & Werning 2014 Sticky prices: Farhi & Werning 2013, Schmitt-Grohe & Uribe 2012/3, Empirical: Obstfeld & Taylor 2004, Calvo

8 Model setup - symmetric Preferences E න 0 e ρt c 1 γ t 1 γ dt Same preference discount rate r saving out of constraint Two output goods y a and y b - imperfect substitutes y t = 1 2 y s 1 a t s + 1 s 1 s/(s 1) 2 y b s t (Comparative) advantages: Good a Good b Country A തak t ak t Country B ak t തak t 12

9 Two country/sector model World capital shares: ψ t Aa + ψ t Ab + ψ t Ba + ψ t Bb = 1 World supply of (output) goods: Y t a = ψ t Aa a + ψ t Ba a K t Y t b = ψ t Bb a + ψ t Ab a K t Price of output goods a and b in terms of price of y P t a = 1 2 Y t Y t a 1/s and Pt b = 1 2 Y t Y t b 1/s Terms of trade P t a /P t b 13

10 Two country/sector model Capital evolution for dk t = Φ ι t δ k t dt + σ A k t dz t A in country A dk t = Φ ι t δ k t dt + σ B k t dz t B in country B Φ concavity technological illiquidity Single type of capital Investment in composite good Shocks are Two dimensional Affect global capital stock dz t A + dz t B Redistributive (initial shock + amplification) affects wealth share, η t Example: Apple vs. Samsung lawsuit 14

11 Market structures Trade Finance Markets Complete Markets Full integration/first Best Open credit account (equity home bias) Closed credit account Output y a, y b Physical capital K Debt Equity X X X X X X X X X Add taxes/capital controls intratemporal intertemporal 15

12 Returns on physical capital dk t /k t = Φ ι t Postulate dq t /q t δ dt + σ A dz t A = μ t q dt + σ t qa dz t A + σ t qb dz t B Returns from holding physical capital Ito product rule: d X t Y t = dx t Y t + X t dy t + σ X σ Y dt dr t Aa = ap t a ι t q t + μ t q + Φ ι t δ + σ A σ t qa dt + + σ A + σ t qa dz t A + σ t qb dz t B dr t Ab = ap t b ι t q t + μ t q + Φ ι t δ + σ A σ t qa dt + + σ A + σ t qa dz t A + σ t qb dz t B 16

13 market structure specific For any market structure The 3 step solution procedure 1. Derive equilibrium conditions Optimality and asset pricing conditions (from postulated processes) Consumption with log-utility: c t = rn t (no precautionary savings) Asset pricing (from above) with log-utility: Sharpe Ratio of asset = volatility of net worth Internal investment rate ι t : qφ ι t 1 = 0 Market clearing conditions 2. Derive evolution of state variable η t = N t q t K t 3. Express in terms of ODE All μ postolated and σ postulated are expressed in terms of q (η), q (η), 17

14 Market structures Trade Finance Markets Complete Markets Full integration/first Best Open credit account (equity home bias) Closed credit account Output y a, y b Physical capital K Debt Equity X X X X X X X X X Add taxes/capital controls intratemporal intertemporal 22

15 Market structures 1. Complete markets First best 2. Incomplete markets (equity home bias) Levered short-term debt financing Sudden stops: (varying technological illiquidity) Amplification Runs due to sunspots 3. Closed capital account: capital controls (no equity, no debt) 4. Welfare analysis 23

16 1. Complete markets: First Best 1. Perfect specialization Full specialization ψ Aa t = ψ Bb t = 1/2 Investment rate equalization ι A B t = ι t Output equalization y t a = y t b Y t = a K t 2 2. Perfect risk sharing Consumption (intensity) shares where λ A and λ B are Pareto weights dz t A +dz t B 2 dz t (standard Brownian) ζ t A = λ A ζ t, ζ t B = λ B ζ t Global capital evolution dk t = Φ(ι t ) δ K t dt + σ K 2 t dz t A +dz B t x 2 dz t 24

17 1. Complete markets: First Best prices Risk-free rate: r F = ρ + Φ ι δ σ2 2 Price of capital: q = a 2 ι t r F t + σ2 [Φ ι δ] 2 Gordon Growth Formula d r g 25

18 1. Complete markets: First Best Remarks Perfect capital allocation + perfect risk sharing Prices are constant and independent of shocks Economy shrinks/expands with (multiplicative) shocks Elasticity of substitution, s, has no impact on prices 27

19 Market structures 1. Complete markets First best 2. Incomplete markets (equity home bias) Levered (short-term) debt financing Sudden stops: (varying technological illiquidity, irreversibility) Amplification Runs due to sunspots 3. Closed capital account: capital controls (no equity, no debt) 4. Welfare analysis 33

20 2. Equilibrium characterization: state variable Equilibrium is a map Histories of shocks {Z s A, Z s B, s t} prices allocation q t, ψ t Aa, ι t A, ι t B, ζ t A, ζ t B wealth distribution η t = N t q t K t 0,1 A s wealth share ψ Aa t + ψ Ab t + ψ Ba t + ψ Bb t = 1 and C A t + C B t = Y t ι t K t ψ Aa Portfolio weights: t, ψ Ab t, 1 ψ t Aa +ψ Ab t η t η t η t Consumption rates: ζ A t = C A t /N t ζ B t = C B t /(q t K t N t ) 35

21 Net worth dynamics Agent I {A, B} Consumption propensity ζ t i = c t i /n t i Portfolio weights ψ t Aa /η t fraction held in capital that will produce output a ψ Ab t /η t b 1 ψ Aa t /η t ψ Ab t /η t fraction held in international debt/bond No equity of derivatives Net worth dynamics dn I t /n I t = ψ t Aa dr Ia η t + ψ Ab t dr Ib t η t + 1 ψ Aa t +ψt Ab t η t Solvency constraint: n t 0 (together form budget constraint) dr t F ζ t I dt No exogenous debt constraint. 36

22 Net worth state variable dynamics N t dn t /N t = r t dt + γ ψ t aa η t Net worth of A risk premium a dt + γ ψ t ba η t dt C t N t dt + dz t terms q t K t Value of total capital d(q t K t )/(q t K t ) = r t dt + γ 4 risk premium terms dt Y t ι t K t q t K t γ + dz t terms η t = N t /q t K t dη t /η t =. Wealth share use Ito s product rule Close model 37

23 2. State variable: 3 regions Wealth share η Three regions Full specialization A produces a Full specialization a a, b B produces a, b b b Symmetric 0 1 1/2 ψ t Aa = η t ψ t Bb = 1 η t ψ t Ba = ψ t Ab = 0 η 39

24 2. Capital share, terms of trade, price of capital Numerical: ρ = 5%, a = 14%, a = 4%, δ = 5%, κ = 2, σ A = σ B = 10% wealth share Three different elasticities of substitution: s = {.5,1, } 40

25 TOT: Supply vs. demand shock Supply versus demand shock TOT improve for A as η t declines for η t ഥη,. 5 can be due to dz A < 0: Negative supply shock World recession dz B > 0: Positive demand shock World boom TOT: Output price but fire-sale of (physical) capital stock k t 41

26 2. Stability, Phoenix Miracle for different s Stationary distribution drift volatility Masspoint at {0,1} Phoenix miracle Three different elasticities of substitution: s = {.5,1, } wealth share Difference to Cole & Obstfeld 1991: persistence of capital, δ < 42

27 Overview 1. Complete markets First best 2. Incomplete markets (equity home bias) Levered short-term debt financing Sudden stops: (varying technological illiquidity) Amplification Runs due to sunspots 3. Closed capital account: capital controls (no equity, no debt) 4. Welfare analysis 44

28 2. Amplification σ t ηa = Aa ψ t (1 η η t ) t 1 [ψ t Aa η t ] q (η t) q η t σ A Leverage effect Loss spiral ψ Aa t /η t 1/{1 ψ Aa q t η η t t } q η t (infinite sum) Technological illiquidity (κ, δ) market illiquidity q η (dis)investment adjustment cost 46

29 2. Amplification leverage σ t ηa = Aa ψ t (1 η η t ) t 1 [ψ t Aa η t ] q (η t) q η t σ A Leverage effect Loss spiral ψ Aa t /η t 1/{1 ψ Aa q t η η t t } q η t (infinite sum) Technological illiquidity (κ, δ) market illiquidity q η (dis)investment adjustment cost 47

30 2. Amplification leverage σ t ηa = Aa ψ t (1 η η t ) t 1 [ψ t Aa η t ] q (η t) q η t σ A Market illiquidity (price impact) Leverage effect ψ t Aa /η t Loss spiral 1/{1 ψ t Aa η t q η t q η t } (infinite sum) chnological illiquidity (κ, δ) market illiquidity q η (dis)investment adjustment cost 48

31 2. Amplification leverage σ t ηa = Aa ψ t (1 η η t ) t 1 [ψ t Aa η t ] q (η t) q η t σ A Market illiquidity (price impact) Leverage effect ψ t Aa /η t Loss spiral 1/{1 ψ t Aa η t q η t q η t } (infinite sum) Technological illiquidity (κ, δ) market illiquidity q η (dis)investment adjustment cost 49

32 2. Technological (κ, δ) market illiquidity q η Quadratic adjustment cost Investment rate of Φ + 1 κ Φ2 generates new capital at rate Φ Φ ι = 1 κ 1 + 2κι 1 Three cases κ = 0 q = 1 κ = 2 κ ι<0 = 100 and κ ι>0 = 2 52

33 Sudden stops: amplification & runs Sudden stop Adverse fundamental triggers %-decline in debt that exceeds %-decline in net worth; (ψaa η) η > 1 ψaa > ψaa η ψ Aa η η η pro-cyclical leverage hyperbola 54

34 Sudden stops: amplification & runs Sudden stop Adverse fundamental triggers %-decline in debt that exceeds %-decline in net worth; (ψaa η) η > 1 ψaa > ψaa η ψ Aa η η η pro-cyclical leverage Slope of tangent vs. secant hyperbola 55

35 Sudden stops: amplification & runs Sudden stop Adverse fundamental triggers %-decline in debt that exceeds %-decline in net worth; (ψaa η) η > 1 ψaa > ψaa η ψ Aa η η η pro-cyclical leverage σ t Debt A = 1 + ψ t Aa η t / η t η t ψ t Aa η t + ψ t Aa η t (1 η t ) 1 [ψ t Aa η t ] q (η t) q η t σ A An unanticipated sunspot triggers a sudden capital price drop from q to q, accompanied by a drop in η to η. q η = max ηq + ψ Aa q q, 0 58

36 Sudden stops: amplification & runs Sudden stop Adverse fundamental triggers %-decline in debt that exceeds %-decline in net worth; (ψaa η) η > 1 ψaa > ψaa η ψ Aa η η η pro-cyclical leverage σ t Debt A = 1 + ψ t Aa η t / η t η t ψ t Aa η t + ψ t Aa η t (1 η t ) 1 [ψ t Aa η t ] q (η t) q η t σ A An unanticipated sunspot triggers a sudden capital price drop from q to q, accompanied by a drop in η to η. q = max ηq + ψaa q q, 0 hyperbola η 59

37 Sudden stop due to sunspot 60

38 Sudden stop due to sunspot: Zoomed in 61

39 Overview 1. Complete markets First best 2. Incomplete markets (equity home bias) Levered short-term debt financing Sudden stops: (varying technological illiquidity) Amplification Runs due to sunspots 3. Closed capital account: capital controls (no equity, no debt) 4. Welfare analysis 62

40 Market structures Trade Finance Markets Complete Markets Full integration/first Best Open credit account (equity home bias) Closed credit account Output y a, y b Physical capital K Debt Equity X X X X X X X X X Add taxes/capital controls intratemporal intertemporal 64

41 3. Credit account: open vs. closed ρ = 5%, a = 14%, a = 4%, δ = 5%, κ = 2, σ A = σ B = 10%, s = 1 Capital price lower Lower input price Destabilizes balance sheet 65

42 3. Credit account: open vs. closed ρ = 5%, a = 14%, a = 4%, δ = 5%, κ = 2, σ A = σ B = 10%, s = 1 More stability Less growth Phoenix miracle slightly smaller Stability 66

43 Overview 1. Complete markets First best 2. Incomplete markets (equity home bias) 3. Closed capital account: capital controls (no equity, no debt) 4. Welfare analysis Pecuniary externalities Welfare calculations + Pareto improving redistributions 68

44 4. When are credit flows excessive? Constrained inefficiency (in incomplete market setting) due to pecuniary externalities Price of capital: Price of output good: Price taking behavior undermined this hedge fire sale externality if leverage is high terms of trade hedge restrained competition shock automatic hedge Price taking behavior 69

45 4. Welfare comparison r = 5%, a = 14%, a = 4%, δ = 5%, κ = 2, σ A = σ B = 10%, Full specialization 75

46 4. Welfare comparison r = 5%, a = 14%, a = 4%, δ = 5%, κ = 2, σ A = σ B = 10%, Inefficiency at the extremes: Role for redistributive Policy default/bail-out/debt-relief Full specialization Pareto improving Intuition: Other country s output price is high 76

47 4. Welfare comparison r = 5%, a = 14%, a = 4%, δ = 5%, κ = 2, σ A = σ B = 10%, Full specialization 77

48 4. Welfare comparison Any monotone transformation of η would be equally good state variable Normalization: take CDF of η Uniform stationary distribution! 79

49 normative positive Conclusion Sudden stops Amplification of fundamental shock Runs due to sunspots vulnerability region Phoenix miracle Tradeoff between capital allocation & risk sharing Terms of trade hedge When are short-term credit flows excessive? When can capital controls (financial liberalization) be welfare enhancing (reducing)? Pecuniary externality Price of physical capital Price of output goods: fire-sales externality technological illiquidity terms of trade hedge externality Bailout/Restructuring Redistributive policy can be Pareto improving if one country is sufficiently balance sheet impaired Reduces output good price 82