Original Sin The Pain, the Mystery And the Road to Redemption

1. Original Sin The Pain, the Mystery And the Road to Redemption Barry Eichengreen, Ricardo Hausmann & Ugo Panizza UC Berkeley, Harvard, and IDB

2. Motivation • Most countries cannot borrow abroad in their own currencies, a problem we referred to three years ago (Eichengreen and Hausmann, “Exchange Rates and Financial Fragility,” Kansas City Fed, ed., New Challenges for Monetary Policy, 1999) as “original sin” • If a country has a net foreign debt, this creates an aggregate currency mismatch • This mismatch is associated with output and capital flow volatility • Monetary policies in such countries are rigid, while their central banks are unable to act as LLRs. • They are vulnerable to crises, of the self-fulfilling variety and otherwise.

3. Implications for Reform of the International Financial Architecture • The conventional prescription of floating is problematic for countries with original sin • Inflation targeting is more difficult for such countries • Stabilizing capital flows is more difficult • Strengthening financial systems is more difficult • Avoiding financial crises is more difficult • All this raises the question of whether the “architecture agenda” can succeed without addressing original sin

4. Key Question from this Point of View • Does the inability to borrow internationally in domestic currency reflect problems with country policies and institutions or systematic problems? • We argue that the problem is too pervasive (and too weakly correlated with country characteristics) to be entirely explicable on the first set of grounds. • This leads us to a systemic explanation for the problem, which in turn leads us to an international proposal for its solution • (We will only touch on causes here; Ugo Panizza will have more to say about them in a later session.)

5. Outline of the Paper • Original sin: measurement • Incidence • Consequences • Causes • Learning from the outliers • The Road to redemption

6. Measurement • Measuring original sin is not straightforward. • In principle, we want to measure external liabilities in own currency as a share of total external liabilities • We use data gathered by the BIS on the currency denomination of bonds and money market instruments • We also consider BIS data on cross-border bank lending, although the data are less complete (both in country coverage and currency breakdown)

7. 1993 - 1998 Total Debt Total Debt Instruments Total debt Share of Share of Instruments Issued by residents in instrument issued in own groups’ Issued by own currency groups’ currency currency currency residents Major financial 939.1 34% 493.6 64% 1868.4 68.1% 52.6% 199.0% centers Euroland 855.9 31% 198.4 26% 647.5 23.6% 23.2% 75.7% Other 390.1 14% 68.6 9% 128.2 4.7% 17.6% 32.9% Developed Countries Developing 269.0 10% 6.3 1 % 16.8 0.6% 2.3% 6.3% Countries International 289.7 11% 0.0 0% 0.0 0.0% 0.0% 0.0% Organizations ECU 0.0 0% 0.0 0% 82.8 3.0% 0.0% 0.0% Total 2743.7 100% 766.8 100% 2743.7 100.0% 27.9% 100.0% 1999 - 2001 Major financial centers Table 1: International bonded debt by country groups and currencies 2597.7 45% 1773.6 61% 3913.8 67.8% 6 8.3% 150.7% Euroland 1885.6 33% 1071.5 37% 1722.2 29.8% 56.8% 91.3% Other 477.6 8% 45.9 2% 89.9 1.6% 9.6% 18.8% Developed Countries Developing 434.0 8% 11.6 0% 47.4 0.8% 2.7% 10.9% Countries International 378.4 7% 0.0 0% 0.0 0.0% 0.0% 0.0 % Organizations ECU 0.0 0% 0.0 0% 0.0 0.0% 0.0% 0.0% Total 5773.3 100% 2902.5 100% 5773.3 100.0% 50.3% 100.0%

8. Total Debt issued by residents (93-98) Developing (10%) Other Developed (14%) Major Financial Centers (34 %) Euroland (31%) Total Debt issued in own currency (93-98) Developing (>1%) Other Developed (9%) Major Financial Centers (64 %) Euroland (26%)

9. Total Debt issued by residents (99-01) International Organizations (7%) Other Developed (8%) Developing (8%) Euroland (33%) Major Financial Centers (45 %) Total Debt issued in own currency (99-01) Other Developed (<2%) Major Financial Centers (61 %) Euroland (37%)

10. Things to Note • In 1993-8, Big 4 countries (financial centers: US, UK, Japan, Switzerland) issued only 34% of global debt, but fully 68% of global debt was denominated in their currencies. • (By comparison, countries soon to become members of Euroland issued 31% of global debt but 24% was denominated in their currencies. For them, original sin was a problem, but only a small one.) • Other developed countries issued 14% but had only 5% denominated in their own currencies. Evidently, original sin is a problem even for developed countries aside from financial centers. • Developing countries issued 10% of global debt but had only 1% in their own currencies. • Main difference after 1998 is growing importance of euro as a currency of denomination (relative to the legacy currencies).

11. 1995 - 1998 Total Bank Debt Total debt in Share in Major of residents major five Five Currencies (BIL USD) currencies Major Financial Centers 3,141 44.9% 2,448 44.02% 77.9% Euroland 1,637 23.4% 1,479 26.60% 90.3% Other Developed Countries 263 3.8% 167 3.00 % 63.5% Offshore 502 7.2% 434 7.80% 86.4% Developing Countries 1,305 18.7% 995 17.89% 76.2% International Organizations 23 0.3% 17 0.31% 71.4% Unallocated 127 1.8% 22 0.40% 17.7% Total 6,998 100.0% 5,561 100.00% 79.5% 1999 - 2001 Total Bank Debt Total debt in Share in Major by residents major five Five Currencies (BIL USD) currencies Major Financial Centers 3,691 47.3% 3,146 49.59% 85.2% Euroland 2,263 29.0% 2,080 32.79% 91.9% Other Developed Countries 356 4.6% 223 3.52% 62.8% Offshore 458 5.9% 381 6.01% 83.1% Developing Countries 887 11.4% 673 10.61% 75.8% Table 2: Cross-border bank claims International Organizations 18 0.2% 17 0.27% 93.7% Unallocated 134 1.7% 19 0.30% 14.5% Total 7,808 100.0% 6,344 100.00% 81.3%

12. Things to Note • In 1993-8, Big 4 countries (financial centers: US, UK, Japan, Switzerland) issued only 34% of global debt, but fully 68% of global debt was denominated in their currencies. • (By comparison, countries soon to become members of Euroland issued 31% of global debt but 24% was denominated in their currencies. For them, original sin was a problem, but only a small one.) • Other developed countries issued 14% but had only 5% denominated in their own currencies. Evidently, original sin is a problem even for developed countries aside from financial centers. • Developing countries issued 10% of global debt but had only 1% in their own currencies. • Main difference after 1998 is growing importance of euro as a currency of denomination (relative to the legacy currencies).

13. Measurement Issues • In constructing a measure of original sin, we would like to take into account both bank and bond market data. • We would like to acknowledge the fact that in some cases countries may be able to hedge their currency exposures • To this end, we measure original sin in several alternative ways. • It turns out that are key results emerge almost regardless of which of the following indices we use.

14. A First Measure Considering Only Bond Data (the higher the value, the greater the sin) Securities issued by country i in currency i = - OSIN 1 1 i Securities issued by country i

15. A Second Measure (including loans as well as securities) + Securities Loans issued by country i in major currencies = INDEXA i + Securities Loans issued by country i

16. A Third Measure (which accounts for the fact that debt in currency i issued by other countries creates an opportunity for country i to hedge) Securities in currency i = - INDEXB 1 i Securities issued by country i

17. A Fourth Measure (which eliminates negative values, where there is more debt in currency i than country i has in total, since countries cannot hedge more debt than they issue) æ ö Securities in currency i ç ÷ = - OSIN 3 max 1 , 0 ç ÷ i Securities issued by country i è ø

18. Note that we consider yet additional measures in the paper. • But in Table 1, please focus on the columns for OSIN3, this last measure. • Note that we distinguish the periods before and after the advent of the euro.

19. OSIN1 OSIN1 OSIN2 OSIN2 OSIN3 OSIN3 Group 1993 - 1998 1999 - 2001 1993 - 1998 1999 - 2001 1993 - 1998 1999 - 2001 0.58 0 .34 0.37 0.07 0.08 Financial centers 0.53 * 0.86 0.55 0.72 0.53 0.09 Euroland 0.52 0.90 0.80 0.82 0.78 0.72 Other Developed 0.94 0.98 0.95 0.98 0.96 0.87 Offshore 0.97 1.00 0.98 0.99 0.96 0.93 Developing 0.99 1.00 1.00 1.00 0.98 1.00 LAC 1.00 Middle East & 1.00 0.97 0.99 0.95 0.90 Africa 0.9 9 1.00 0.95 0.99 0.99 0.94 Asia & Pacific 0.99 0.99 0.97 0.98 0.91 0.84 Eastern Europe 1.00 Table 3: Measures of original sin by country groupings

20. Graph1: Measures of original sin by country groupings 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Financial Centers Euroland Other Developed Developing OSIN1 OSIN2 OSIN3

21. Graph2: Measures of original sin by regions (developing countries) 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 LAC Middle East & Africa Asia & Pacific Eastern Europe OSIN1 OSIN2 OSIN3

22. Things to Note • Lowest levels of original sin for the financial centers, followed by the Euroland countries • Note that the Euroland countries experience a major decline in original sin following the introduction of the euro • Other industrial countries have higher values. • Developing countries have still higher values. • Among developers, OS is lowest for accession economies, highest in Latin America.

23. Table 4: Countries with OSIN3 <0.8, excluding financial centers Non Euroland Euroland Country 1993 - 98 1999 - 01 Country 1993 - 98 19 99 - 01 Czech Republic 0.0 0.00 Italy 0.00 0.00 Poland 0.82 0.00 France 0.23 0.12 New Zealand 0.63 0.05 Portugal 0.42 0.24 South Africa 0.44 0.10 Belgium 0.76 0.39 Hong Kong 0.72 0.29 Spain 0.59 0.42 Taiwan 1.00 0.54 Netherlands 0.64 0.47 Singapore 0. 96 0.70 Ireland 0.94 0.59 Australia 0.55 0.70 Greece 0.93 0.60 Denmark 0.80 0.71 Finland 0.96 0.62 Canada 0.55 0.76 Austria 0.90 0.68

24. Graph 3: OSIN3 in Euroland 1 0.96 0.94 0.93 0.9 0.9 0.8 0.76 0.68 0.7 0.64 0.62 0.6 0.59 0.59 0.6 0.47 0.5 0.42 0.42 0.39 0.4 0.3 0.24 0.23 0.2 0.12 0.1 0 0 0 Italy France Portugal Belgium Spain Netherlands Ireland Greece Finland Austria 93-98 99-01

25. The Non-Euroland outliers are not all floaters LYS 3 Fixers 3 Intermediate 2 Floaters 1 Australia Canada Poland Czech Rep. South Africa Singapore Denmark Hong Kong New Zealand Average 1993 - 2001

26. Things to Note • The Euroland countries of course feature prominently • Accession economies and overseas regions of 19th century European settlement are also numerous. • Both fixed-rate Hong Kong and floating-rate Singapore and Taiwan appear on this list (raising questions about whether a particular exchange rate regime helps for eliminating original sin)

27. Original Sin is also Persistent • Table 5 in the paper classifies countries as issuing (gold) indexed debt or domestic currency debt in the mid-19th century. (Some cases are mixed.) • It shows a very strong and significant correlation with countries suffering from original sin (issuing foreign currency denominated or indexed debt) today.

28. Original sin is highly persistent OSIN3 and Flandreau-Sussman classification circa 1850

29. The Pain of Original Sin (The consequences)

30. Recall the Problem • If an original sin country has a net foreign debt and it is in foreign currency, then there is an aggregate currency mismatch • If no one else issues debt in its currency, then foreign exposures can then be passed around by residents like a hot potato, but foreign exposure cannot be hedged in the aggregate. • This can have important implications for exchange rate policy, capital flows, and output volatility, among other things. We explore these implications in turn, starting with the exchange rate.

31. Original sin and exchange rate flexibility • We use the Levy-Yeyati and Sturzenegger classification, M2/Reserves, and exchange rate variability/reserve variability as 3 measures of the dependent variable. • Note that there are good reasons to think that these things will be affected by original sin. • There are of course the theoretical reasons mentioned earlier. But there are also empirical reasons. Consider:

32. Levels of Reserves Reserves/M2 0.879 Singapore 0.637 Peru 0.485 Chile 0.446 Poland 0.409 Colombia 0.362 Greece 0.335 Indonesia 0.314 Czech. Republic 0.303 Guatemala 0.302 Mexico 0.287 Norway 0.262 Israel 0.259 Paraguay 0.253 Brazil 0.245 Jamaica 0.243 Korea 0.242 Philippines 0.230 Thailand 0.137 Sweden 0.133 India 0.105 Germany 0.094 Switzerland 0.087 Dominican Rep. 0.064 South Africa 0.062 Australia 0.058 Canada 0.057 New Zealand 0.048 Japan 0.021 United Kingdom 0.013 United States 0 0.2 0.4 0.6 0.8 1

33. VEN .341941 KAZ PER MDA EST CHL HKG BHR COL HUN NOR MLT LVA SVN MYS OMN BRB BHS ARG BGR DNK CHE CZE ISL CRI e( RESM2 | X) POL MUS CYP ISR KOR ROM THA TUR LBN BRA LKA TTO SVK MEX JPN JAM URY NZD AUS JOR GTM CAN RUS SUR SLV USA TUN PHL UKR NIC IDN MAR ZWE PNG BOL DOM ZAF IND PAK -.410813 -.878934 .307706 e( SIN33_A | X ) Original sin and reserves over M2

34. Intervention using reserves Relative Volatilities: Exchange Rate and Reserves 30.454 Japan .380 United States 17.946 United Kingdom 12.682 New Zealand 6.913 Australia 6.617 Thailand 3.369 Canada 2.920 Brazil 2.842 Germany 2.466 South Africa 2.324 Philippines 2.272 Switzerland 2.154 Indonesia 1.583 Dominican Rep. 1.345 Korea 1.260 Czech 1.215 India 0.976 Sweden 0.926 Colombia 0.838 Mexico 0.760 Israel 0.690 Singapore 0.615 Paraguay 0.506 Peru 0.424 Chile 0.417 Guatemala 0.416 Poland 0.390 Greece 0.362 Norway 0.272 Jamaica 0 5 10 15 20 25 30 35 std(dev)/std(res/m2)

35. 1.61765 JPN BHS BRB BHR IND ZAF PAK AUS RUS CHE IDN PER e( RVER | X) KOR COL TUR ARG ROM THA BOL MEX URY GTM ZWE CHL UKR VEN CAN MAR DNK LBN ECU USA NZD MDA ISL SLV POL LKA PHL KAZ CYP NIC HUN CRI TUN BGR NOR TTO LVA DOM PNG ISR MUS SVN OMN SVK SGP EST JOR CZE HKG JAM MLT MYS -1.08857 -.878092 .309288 e( SIN33_A | X ) Original Sin and relative volatility of exchange and reserves

36. Multivariate Analysis • We control for GDP per capita, openness, and foreign debt/GNP, and focus on the coefficient of OSIN3. • We run weighted LS (weighted by share of securities in total debt) and also ordered probit for the LYS measure. • OS is negatively correlated with exchange rate flexibility using all 3 measures. (Time series and cross section evidence, and instrument variables regressions, support that this is not reserve causality.) • The effect is important: eliminating OS is enough to move countries by a full category in the LYS classification.

37. Table 6: Original sin and exchange rate flexibility (1) (2) (3) LYS RESM2 RVER OSIN3 0.984 0.248 - 0.801 (2.98)*** (3.74)*** (2.02)** LGDP_PC 0.268 - 0.053 0.026 (3.61)*** (1.85)* (0.61) OPEN 0.178 - 0.014 1.017 (1.85)* (0.41) (2.88)*** SHARE2 58.719 - 35.858 - 569.562 (0.46) (0.66) (2.36)** Constant - 1.389 0.531 0.104 (1.79)* (1.73)* (0.17) Observations 75 65 65 R - squared 0.22 0.37 0.62

38. Output and Capital Flow Volatility • Output volatility is measured as the SD of GDP growth. • Capital flow volatility is the SD of capital flow/domestic credit ratio • We control for level of development, openness, foreign debt, and terms of trade volatility. • Again, we run weighted least squares • Original sin account for more than ¼ of the difference in volatility between developed and developing countries.

39. Table 7: Original sin and volatility (1) (2) VOL_GROWTH VOL_FLOW OSIN3 0.011 7.103 (1.96)* (3.58)*** LGDP_PC - 0.012 - 3.214 (2.14)** (2.56)** OPEN - 0.001 - 4.181 (0.12) (1.20) VOL_TOT - 0.000 0.223 (0.86) (1.08) SHARE2 - 1 4.287 147.265 (1.72)* (0.04) Constant 0.135 32.825 (2.25)** (2.39)** Observations 77 33 R - squared 0.40 0.64

40. Original Sin & Credit Ratings • Countries with original sin will have more volatile debt service ratios and deficits, because their service obligations will be more exchange-rate and interest-rate sensitive. • Greater fiscal volatility means lower credit ratings, other things equal. • This can explain, inter alia, why the correlation of credit ratings with debt and deficit variables is so low (as shown on the next slide).

41. The Weak Relationship Between Debt/GDP and Credit Ratings NOR JPN GBR AUT DEU USA 19 SWE DNK CAN BEL AUS ESP FIN ITA PRT CYP ISL SVN rating foreign currency CZE ISR EST CHN GRC LVA HUN TUN POL TTO PAN IND MEX CRI ARG MAR DOM BRA JOR PRY TUR PAK 5 -.291965 1.13803 net_debt/gdp

42. Multivariate Analysis • We control for standard determinants of credit rates, public debt/GDP, public debt/tax revenues, and level of development. • Estimates are by weighted least squares and double censored Tobit. • Results, on next slide, suggest that eliminating original sin increases country credit rating by three notches.

43. Table 8: Original sin and credit ratings (1) (2) (3) (4) RATING1 RATING1 RATING1 RATING1 DE_GDP2 - 1.553 - 1.815 (1.91)* (2.19)** DE_RE2 - 0.599 - 0.665 (1.40) (1.52) LGDP_PC 3.189 3.051 2.884 2.76 4 (8.54)*** (7.59)*** (6.47)*** (5.68)*** OSIN3 - 3.429 - 3.324 - 4.883 - 4.435 (3.85)*** (3.49)*** (3.49)*** (3.11)*** Constant - 12.369 - 11.059 - 8.751 - 7.889 (3.16)*** (2.60)** (1.89)* (1.57) Observations 56 49 51 44 R - squared 0.82 0.81 0.81 0.80

44. Causes of original sin Immaculate conception and the road to redemption

45. Miner’s canary or cause? Is OS one more symptom of poor institutions, or is it a different kind of phenomenon?

46. Recent Theories • Underdevelopment of institutions and policies in general • Inadequate monetary credibility • Fiscal profligacy • Weak contract enforcement • Presence or absence of trade sanctions • Political economy stories • We will leave our discussion of these theories and evidence to Ugo tomorrow, but here we need to make one important point:

47. Inadequacy of Conventional Explanations • None of a variety of measures of levels of development, monetary credibility, fiscal profligacy, strength of contract enforcement, trade dependence, or political economy have much traction. • Same is true for a variety of samples, specifications and econometric treatments. • Only robust correlates are country size and financial-center status (which is, in a sense, what we are trying to explain).

48. Why Might a Few Important Currencies So Dominate the Global Portfolio? • Additional currencies add decreasing diversification benefits but constant transaction costs. • International transaction costs and heterogeneity • Network externalities may give a small number of vehicle currencies special attraction. • Countries seeking to add their currencies to the global portfolio thus face an uphill battle. • And each that succeeds makes life tougher for the others.

49. Bottom Line • Original sin is not merely a problem of country policies (one need not deny the relevance of these, of course). • It is also a problem with the operation of the international system (given transactions costs, a world of heterogeneous countries, and network effects that lock in the status quo). • Redemption therefore requires international action to overcome the inertia in the system.

50. Lessons from outliers • Countries that have recently escaped original sin seem to have done so through non-nationals issuing debt in domestic currency • IFIs have played a major role in this process • Borrowers swap their obligations with residents