International Finance

Transcription

1 International Finance 7 e édition Christophe Boucher christophe.boucher@u-paris10.fr 1

2 Session 2 7 e édition Six major puzzles in international macroeconomics 2

3 Roadmap 1. Feldstein-Horioka 2. Home bias in trade 3. Home bias in equity portfolios 4. Consumption correlations 5. Pricing puzzle(s) 6. Exchange rate disconnect puzzles(s) 3

4 Introduction National economies product markets are less integrated than macroeconomists typically assume (a feature making international economics a distinct field). Trade costs (and psychological biases) can explain some parts of these puzzles 4

5 Roadmap 1. Feldstein-Horioka 2. Home bias in trade 3. Home bias in equity portfolios 4. Consumption correlations 5. Pricing puzzle(s) 6. Exchange rate disconnect puzzles(s) 5

6 The Feldstein-Horioka puzzle The Original Questions Asked by Feldstein and Horioka From the Martin Feldstein and Charles Horioka's 1980 paper: How internationally mobile is the world's supply of capital? Does capital flow among industrial countries to equalize the yield to investors? Alternatively, does the saving that originates in a country remain to be invested there? 6

7 The Feldstein-Horioka puzzle Imagine a world with perfect capital mobility and a small country in such a world How much people from that small country save is influenced by investment opportunities around the world At the same time, the size of investment in that country is determined by its particular characteristics, and the investment is financed by the worldwide pool of capital Therefore, there should be no relation between domestic saving and domestic investment! Let's see the data 7

8 The Feldstein-Horioka puzzle Feldstein-Horioka Regressions 8

9 The Feldstein-Horioka puzzle THE PUZZLE: Long-period averages of national saving rates are highly correlated with similar averages of domestic investment rates 9

10 The Feldstein-Horioka puzzle THE PUZZLE: Long-period averages of national saving rates are highly correlated with similar averages of domestic investment rates 10

11 The Feldstein-Horioka puzzle For OECD data from 1960-mid 1970s, cross section regression of average investment on saving yielded slope near 1. In more recent OECD data, coefficient is still 0.6, too high to explain easily in a world of capital mobility. If this means capital is immobile, one is puzzled, since nominal returns are very closely arbitraged. One point: cost of international trade can drive a big wedge between countries real interest rates, even under costless international trade in claims to income streams. 11

12 The Feldstein-Horioka puzzle One-Good, Two-Period Model 12

13 The Feldstein-Horioka puzzle 13

14 The Feldstein-Horioka puzzle Two good model 14

15 The Feldstein-Horioka puzzle 15

16 The Feldstein-Horioka puzzle 16

17 The Feldstein-Horioka puzzle 17

18 The Feldstein-Horioka puzzle 18

19 The Feldstein-Horioka puzzle 19

20 The Feldstein-Horioka puzzle 20

21 The Feldstein-Horioka puzzle 21

22 The Feldstein-Horioka puzzle 22

23 The Feldstein-Horioka puzzle 23

24 The Feldstein-Horioka puzzle 24

25 The Feldstein-Horioka puzzle 25

26 The Feldstein-Horioka puzzle 26

27 The Feldstein-Horioka puzzle 27

28 The Feldstein-Horioka puzzle 28

29 The Feldstein-Horioka puzzle 29

30 The Feldstein-Horioka puzzle 30

31 The Feldstein-Horioka puzzle A time-varying puzzle? S-I rates cross-correlations 1 0,8 0,6 0,4 0,2 0-0,2 OCDE non euro zone euro OCDE non euro (lissage) zone euro (lissage)

32 The Feldstein-Horioka puzzle A time-varying puzzle? S-I rates cross-correlations 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0-0,1 Asie OCDE Asia (lissage) OCDE (lissage)

33 Roadmap 1. Feldstein-Horioka 2. Home bias in trade 3. Home bias in equity portfolios 4. Consumption correlations 5. Pricing puzzle(s) 6. Exchange rate disconnect puzzles(s) 33

34 Home bias in trade Export Share of GDP in

35 Home bias in trade Import Share of GDP in

36 Home bias in trade Self-Absorption in

37 Home bias in trade J. McCallum (AER, 1995) showed that in 1998 U.S.-Canada trade data, inter-province trade is 20 times international trade, holding other determinants of trade fixed. Subsequent estimates have whittled this bias down to 6 to 12. OECD data also confirm trade bias, which is fundamental to transfer problem, effects of fiscal policy, etc., and also figures in discussions of low factor content of trade (Trefler). 37

38 Home bias in trade Gravity equations 38

39 Home bias in trade Gravity equations 39

40 Home bias in trade McCallum's Gravity for the U.S. and Canada 40

41 Home bias in trade McCallum's Gravity Results 41

42 Home bias in trade How Wide is the Border Between the US and Canada? 42

43 Home bias in trade Average price volatility 43

44 Home bias in trade Average price volatility 44

45 Home bias in trade Notations of regressions 45

46 Home bias in trade Regression results 46

47 Home bias in trade How wide is the border? 47

48 Home bias in trade Interpreting the results Physical distance plays a significant role in explaining the failure in the law of one price between locations But the physical distance alone does not explain the variability in prices of similar goods if the two locations are in different countries 48

49 Home bias in trade The border is wide 49

50 Home bias in trade A simple model: Utility is 50

51 Home bias in trade 51

52 Home bias in trade 52

53 Home bias in trade 53

54 Home bias in trade 54

55 Home bias in trade 55

56 Home bias in trade The nonlinear relationship between trade costs and home bias in trade 56

57 Roadmap 1. Feldstein-Horioka 2. Home bias in trade 3. Home bias in equity portfolios 4. Consumption correlations 5. Pricing puzzle(s) 6. Exchange rate disconnect puzzles(s) 57

58 Home bias in equity portfolios THE PUZZLE: Despite the growth of international capital markets, stock-market investors maintain a puzzling preference for home assets The first paper to highlight that puzzle was by French and Poterba Investor diversification and international equity markets (AER, 1991): At the end of the 1980s Americans held roughly 94% of their equity in the U.S. stock market The corresponding number for Japanese was 98% U.S. equity holders hold only about 11.7% (O-R, 2000) of equity abroad 58

59 Home bias in equity portfolios 59

60 Home bias in equity portfolios The home equity bias is smaller for smaller countries and has shown some tendency to decline over time Still, according to standard economics models of optimal international portfolio diversification, equity investors are not diversified internationally nearly as much as they should be 60

61 Home bias in equity portfolios Coval and Moskowitz (1999) showed that home bias is not limited to international portfolios, but that the preference for investing close to home also applies to portfolios of domestic stocks. Bias also in bond portfol (Fidora, Fratzscher and Thimann, 2006) 61

62 Home bias in equity portfolios Attempts to resolve the puzzle Hypothesis that investors have superior access to information about local firms or economic conditions (information immobility instead of capital immobility): information asymmetries and frictions (e.g. Coval and Moskowitz (1999, 2001), Bae, Stulz and Tan (2005)). The role of policies and of the quality of domestic institutions, such as corporate governance (e.g. Dahlquist, Pinkowitz, Stulz and Williamsom, 2002) Behavioural explanations such as patriotism (Morse and Shrive (2004)) or investors who maximise expected wealth relative to a group of peers (Gomez, Priestley and Zapatero (2002)). 62

63 Roadmap 1. Feldstein-Horioka 2. Home bias in trade 3. Home bias in equity portfolios 4. Consumption correlations 5. Pricing puzzle(s) 6. Exchange rate disconnect puzzles(s) 63

64 Consumption correlations Imagine a world of financially integrated economies, where consumers from different countries can insure their consumption by investing in other countries If there are no obstacles to investments, consumers will be investing in a country where they expect bigger returns On average, then, expected returns from investment in different countries should be similar So, consumers from different countries would get similar returns That means that (i) consumption levels, or (ii) changes in consumption levels would be correlated? 64

65 Consumption correlations Correlations in per Capita Private Consumption Growth,

66 Consumption correlations Consumption and Output: Correlations between Domestic and World Growth Rates,

67 Consumption correlations The International Consumption Correlation Puzzle PUZZLE 1: Consumption correlations are low PUZZLE 2: Consumption correlations are lower than output correlations 67

68 Consumption correlations Unlike other puzzles, the various consumption correlation puzzles tend to be very model specific. Much less fundamental than other puzzles Simplest model with PPP and homogeneous tastes predicts equal consumption growth worldwide: But this fails miserably. Basically a corollary of the home bias in risky asset allocation, and Feldstein Horioka puzzles 68

69 Consumption correlations the Backus-Kehoe-Kydland (JPE 1992) puzzle that per capita consumptions are less correlated internationally than outputs; see table 5. Why this is a puzzle is not obvious. Consumptions are more correlated that levels of consumable output, Y I - G, which would appear to be the more relevant comparison 69

70 Roadmap 1. Feldstein-Horioka 2. Home bias in trade 3. Home bias in equity portfolios 4. Consumption correlations 5. Pricing puzzle(s) 6. Exchange rate disconnect puzzles(s) 70

71 Pricing puzzle(s) The Law of One Price Abstract away from all and any frictions that might exists in markets Then identical goods in different locations must sell for the same price when prices are expressed in a common currency 71

72 Pricing puzzle(s) The Real Exchange Rate 72

73 Pricing puzzle(s) Absolute Purchasing Power Parity 73

74 Pricing puzzle(s) Problems with PPP

75 Pricing puzzle(s) Problems with PPP

76 Pricing puzzle(s) Relative PPP 76

77 Pricing puzzle(s) Alternative Formulation of Relative PPP 77

78 Pricing puzzle(s) PPP theory is an old idea Some versions of it have been traced to the Salamanca school in 16th century Spain, and to the writings of Gerrard de Malynes appearing in 1601 in England The 19th century classical economists Ricardo, Mill, and Marshall talk about PPP theory 78

79 Pricing puzzle(s) The term Purchasing Power Parity was coined by the Swedish economist Gustav Cassel in his 1918th paper Abnormal Deviations in International Exchanges Cassel claimed that the (relative) PPP is supported empirically, and used it as a theory for determination of nominal exchange rates What was happening around 1918 so that Cassel thought about issues related to nominal exchange rates? 79

80 Pricing puzzle(s) Relative PPP: a caveat 80

81 Pricing puzzle(s) Empirical evidence 81

82 Pricing puzzle(s) Why PPP Fails in the Short Run 82

83 Pricing puzzle(s) Why PPP is an Important Concept 83

84 Pricing puzzle(s) The PPP Puzzle 84

85 Pricing puzzle(s) The Speed of Convergence of PPP 85

86 Roadmap 1. Feldstein-Horioka 2. Home bias in trade 3. Home bias in equity portfolios 4. Consumption correlations 5. Pricing puzzle(s) 6. Exchange rate disconnect puzzles(s) 86

87 Exchange rate disconnect puzzles(s) TWO PUZZLES: The Forward Premium Puzzle Remarkably weak short-term feedback links between the nominal exchange rate and the rest of the economy See illustrations on countries with no big inflation episodes during with monthly data 87

88 Exchange rate disconnect puzzles(s) XR volatility and CPI volatility 88

89 Exchange rate disconnect puzzles(s) XR volatility and interest rate volatility 89

90 Exchange rate disconnect puzzles(s) XR volatility and Money aggregate volatility 90

91 Exchange rate disconnect puzzles(s) XR volatility and output volatility 91

92 Exchange rate disconnect puzzles(s) Arbitrage and Interest Rates An important question for investors is in which currency they should hold their liquid cash balances: Hold money in USD, exchange to Euros in the end of the period Put money in Euro right away The key issue here is the exchange rate risk. There are two possible choices to address this risk: Riskless arbitrage: use a forward contract Risky arbitrage: wait and use a spot contract 92

93 Exchange rate disconnect puzzles(s) Riskless Arbitrage: Covered Interest Parity 93

94 The forward premium puzzle Covered Interest parity 94

95 The forward premium puzzle Uncovered Interest Parity 95

96 The forward premium puzzle FOREX Equilbrium 96

97 The forward premium puzzle Interest rate differentials 97

98 The forward premium puzzle Forward Premium Puzzle 98

99 The forward premium puzzle Forward Premium Puzzle 99

100 The forward premium puzzle Safe and Risky Assest 100

101 The forward premium puzzle Risk Aversion and Deviations from UIP 101

102 The forward premium puzzle Partial Equilibrium Risk Premia 102

103 The forward premium puzzle Peso Problems 103

104 The forward premium puzzle Peso Problems 104

105 The forward premium puzzle Peso Problems 105

106 The forward premium puzzle FRU Again 106

107 Exchange rate disconnect puzzles(s) Evidence on Uncovered Interest Parity 107

108 Exchange rate disconnect puzzles(s) The Forward Premium Puzzle In data, high interest currencies tend to appreciate 108

109 Exchange rate disconnect puzzles(s) THE PUZZLE: higher-interest rate currencies, i.e., those that come with a premium, tend to appreciate If CIP holds, that means that the forward exchange rate does not predict the future spot exchange rate In other words, UIP does not hold The puzzle was first formulated by Eugene Fama (JME, 1984) 109

110 Exchange rate disconnect puzzles(s) Models of Exchange rate Determination 110

111 Exchange rate disconnect puzzles(s) Overview of Models 111

112 Exchange rate disconnect puzzles(s) The flexible price approach 112

113 Exchange rate disconnect puzzles(s) Solution to the Flexible Price Model 113

114 Exchange rate disconnect puzzles(s) Sticky price approach: Dornbusch «overshooting» model 114

115 Exchange rate disconnect puzzles(s) Obstfeld and Rogoff's Microfounded Model of the Exchange Rate 115

116 Exchange rate disconnect puzzles(s) Bacchetta and Van Wincoop's Scapegoat Model of X Rates 116

117 Exchange rate disconnect puzzles(s) Details on the Scapegoat Model 117

118 Exchange rate disconnect puzzles(s) Main result 118

119 Exchange rate disconnect puzzles(s) Forecasting the exchange rate 119

120 Exchange rate disconnect puzzles(s) Meese and Rogo (1983, JIE) 120

121 Exchange rate disconnect puzzles(s) Mean Squared Errors 121

122 Exchange rate disconnect puzzles(s) Evidence on Long Horizon Prediction 122

123 Exchange rate disconnect puzzles(s) Engel and West (2005, JPE) and Mark, Engel and West (2007, NBER Macro Annual) 123

124 See you next week. 124