Carmen M. Reinhart, Harvard University (partly based on ongoing work with Vincent Reinhart,

Standish Mellon Asset Management and

Christoph Trebesch, Kiel Institute,)

San Francisco Federal Reserve Asia Economic Policy Conference

San Francisco, November 16-17, 2017

Capital Flow Cycles and the

Case of the Missing Defaults

Reinhart

Roadmap

Data and approach

Capital flow cycles—a long view

Financial conditions in the global financial

centers

The current cycle and “missing defaults”

Reinhart

What we do

Construct a comprehensive data base on cross border flows

back to 1815

Date turning points in the global capital flow cycle, in the

Burns and Mitchell, Harding and Pagan approach. Features of

the cycles are documented.

Role of external factors financial and through commodity

markets—an expanded version of Calvo, Leiderman, and

Reinhart (1993). Unlike the recent literature, we do not include

the VIX or focus on gross flows (post WWI).

Study the interaction among these global cycles and their

connection to sovereign default and global capital mobility.

Explore time variation in those interactions.

Reinhart

What we do (continued)

Construct (this is not complete) a more comprehensive

(beyond short and long rates) profile of financial center

conditions to incorporate the impacts of regulatory changes

and debt management practices and consider secondary

financial centers (Japanese banks in Kaminsky and Reinhart,

2000 and 2001 an UK and Euro area banks, Cerutti, Claessens.

and Ratnovski, 2017).

Highlight the particulars of individual episodes.

Compare the current capital flow cycle to its historical

counterparts.

Reinhart

Is there a global financial cycle?

There is a global cycle in capital flows

When observed over a long period and many countries, with

identifiable peaks and troughs--we count 14 such cycles from 1815.

The “mega-cycles in capital flows” are connected to global capital

mobility.

There is a global cycle in real commodity prices

Despite considerable variation across individual commodities.

Commodity price cycles occur more often (22 cycles) but large ones

are rarer.

There is a cycle in the stance of policy at the global financial

centers

That are evident in the peaks and troughs of real short-term interest

rates of dominant economies. But the policies also involve

manipulating the central bank balance sheet, managing sovereign debt,

and determining regulatory standards

But there may be multiple centers with uneven regional influence, and

they vary over time

Reinhart

Summary of main findings

Reinhart

Is there a global financial cycle?

These cycles combine to influence finance around the world

The impact of financial center interest rates on capital

flows depends on the extent of capital market integration at

the time (both globally and at the level of the capital

importer).

Comovement: It is suggestive that two factors explain

about ½ the variation in “global” flows (our priors did not

suggest a single global factor) Reducing these global

factors to a single indicator of the “global financial cycle”

and a single “push” factor consistently influencing finance

is expecting too much.

Reinhart

Other highlights of our findings

International capital flow cycles have displayed similar patterns over the past 200

years. The magnitudes of the booms are comparable to the busts while booms last

longer than the often abrupt busts (9 years versus 5 years)

The cross-country “incidence” of capital inflows (and reserve accumulation)

suggests the cycle has become more global or inclusive

Real interest rates were 4-to-10 times as volatile pre-WWII as in the more modern

era and cycles were distinctly differ. Nominal rates were far more stable and

“cyclical” in the 19th century.

Capital flow and commodity flow booms often overlap with lower interest rates in

the financial center. These “triple bonanzas” are often followed by “triple busts” and

rising defaults. All of the six major spikes in new defaults (1800-2016) occurred

after a global capital inflow bonanza ended.

The connection of sovereign defaults with commodity cycles is not as systematic as

with capital flows.

All but 2 (1890s and the present one) of the double busts in commodities and capital

flows overlapped with a spike in “global” interest rates.

Reinhart

The Global Capital Flow Database, 1815-2016

1918 - 2016: net flows using current account (CA) and

reserves (gold and FX)

Construct capital account from BOP identity:

CA + KA + ∆RA≡ 0

- Interwar: UN / League of Nations data for 34 countries

- Post-WW2: own constructed series for 61 capital-importing

countries and 7 capital exporters (some series back to 1800s)

- Eurozone post-1999: incorporate Target2 as reserves to

capture within-EZ capital flows

1815-1914: gross flows based on bond issuance

- 1869-1914: UK capital exports to 25 countries Stone (1999)

- 1815-1868: sovereign bond issuance in London, 38 countries,

own data, multiple sources

Reinhart

Other data and ongoing work

o Also constructed the capital account balance for the larger

sample of 145 countries, 1980-2016

o Non-oil primary commodity prices, 1790-2016

o Nominal/real short-/long-term interest rates in financial centers

o Chronologies of regulation/debt management in fin. centers

o New and existing sovereign defaults

o Indices of capital mobility

In progress:

o Integrating gross flows (US, 1920-1930, bank loans 1970-

1980, post-1990 data e.g. Forbes and Warnock)

o Financial conditions in “secondary” financial centers –

expanding role of China

Reinhart

Capital flow cycles: Magnitudes of flows, 1815-2016

Reinhart

Capital flow cycles: Incidence of cross border flows (How “global” is “global”? Capital mobility matters)

Reinhart

0

10

20

30

40

50

60

70

80

18

15

18

21

18

27

18

33

18

39

18

45

18

51

18

57

18

63

18

69

18

75

18

81

18

87

18

93

18

99

19

05

19

11

19

17

19

23

19

29

19

35

19

41

19

47

19

53

19

59

19

65

19

71

19

77

19

83

19

89

19

95

20

01

20

07

20

13

Share of countries issuing

at least one bond

or syndicated loan

(gross flows > 0)

solid line

Share of countries with

net capital inflows,

dashed line

World Wars

are shaded

percent

Co-movement of capital flows across countries Factor Analysis and Principal Components, 1870-2016

Reinhart

Not a WOW…. but two factors explain about

½ of the variation in “global” flows – now and then

Percent

ExplainedTotal

Percent

ExplainedTotal

First Factor 26% 26% 32% 32%

Second 24% 50% 18% 51%

Third 17% 67% 13% 64%

 1870 - 1914 1950 - 2016

The interaction of cycles

in capital flows and…

1) interest rates in financial center

and global financial conditions

2) commodity prices (non-oil)

3) sovereign default

Early 19th century: capital flows & long-term rates Debt management matters: Debt conversions in the UK helped

foster the search for yield in the periphery

Reinhart

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

0

5

10

15

20

25

30

35Share of countries issuing

at least one bond

(bars, right axis)

Yield on a UK consol

(dashed line)

Coupon decline in four successful

debt conversions from 5 to 3%

(solid line, left axis)

Interest rates and capital flows: 1870-present

The post-war era of capital

controls stands out for the

absence of boom-bust in

international capital flows.

Reinhart

Reinhart

Share of Developing or Emerging Market Countries with Reserve

Accumulation Greater than 15% over 3 Years (47 countries),

and Real US Federal Funds Rate: Peaks and Troughs, 1960-2016

Real rate, Percent Share of countries

10

20

30

40

50

60

70

80

90

100

-6

-4

-2

0

2

4

6

8

10

1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015

Reseves

solid line

right scale

Interest rates

dashed line

left scale

“Global” interest rates and capital flows

1815-2016 – the role of capital moblity

Time Period 1815 - 1869 1870 - 1914 1918 - 1975 1976 - 2016

Capital Mobility Low High Low High

post-1820

defaults

Gold standard,

financial global.

Wars and capital

controls

Rising to a

new peak

Real interest rate 0.216 -1.317** 0.582* -0.936***

 in financial center (0.163) (0.494) (0.305) (0.327)

Observations 53 45 50 41

R2 0.023 0.098 0.114 0.120

Notes: The dependent variable is the value of global capital flows as percent of GDP. The explanatory variable is the

interest rate in the financial center (UK until 1918, US thereafter, see Data Appendix) Robust s.e. in parentheses. *, **, and

***, indicate significance at the 10%, 5% and 1% - level, respectively.

The relationship between capital flows and sovereign

default goes in both directions. Furthermore, defaults can

dampen the role of international interest rates

Reinhart

Commodity cycles: 1790-2016 (real, non-oil prices)

Reinhart

-60

-40

-20

0

20

40

60

80

1795 1815 1835 1855 1875 1895 1915 1935 1955 1975 1995 2015

Per

cen

t ch

an

ge

Trough-to-peak (boom) and peak-to-trough

(bust) changes in real commodity prices

(percent)

Commodity Booms:

Trough-to-peak increase ≥ 15%

Commodity Busts:

Peak-to-trough declines ≥ 15%

Reinhart

Example: Interwar – Capital Flows & Defaults

-4

-2

0

2

4

6

8

10

12

0

5

10

15

20

25

30

35

40

45

50

1922 1924 1926 1928 1930 1932 1934 1936 1938

Per

cent

of

GD

P

Per

cen

t

Net capital inflows

(26 debtor countries)

as a % of UK GDP

(bars, right scale)

Share of

sovereigns

entering

a new default,

(5-year sum)

(line, left scale)

Dangerous liasons: commodity and capital flow

“double busts” and sovereign defaults

Reinhart 21

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0

5

10

15

20

25

30

35

40

45

50

1815 1830 1845 1860 1875 1890 1905 1920 1935 1950 1965 1980 1995 2010

Per

cent

of

countr

ies

ente

ring a

new

def

ault

Share of

sovereigns

entering a

new default

3-year sum

solid line

“Double Busts”

joint declines in

capital flows and

commodity prices

(>1 year)

red shading

The current cycle

in historical perspective

Reinhart

Reinhart

Share of New Defaults before and after capital flow

surges, 1815-2016

Despite the magnitude of the recent boom and

reversal, the increase in defaults is modest.

0

5

10

15

20

25

1824 1834 1843 1852 1865 1873 1890 1897 1914 1929 1938 1979 1991 2011

Per

cen

t

Peak 1-2 years later Size of boom

Share of new defaults at

peak capital inflows (stripe)

Share of defaults

1-2 years post-peak (pale)

Magnitude of capital

inflow boom (dark)

Double and Triple Busts

and the “missing” defaults since 2011

Double bust Capital flow Commodity Interest Rate Share of Countries

episodes Bust Bust Spike (real)? in Default

(in peak year)

1824 - 1828 yes yes yes 43.75

1890 - 1894 yes yes no 18.60

1914 - 1918 yes yes yes 17.65

1929 - 1933 yes yes yes 46.43

1981 - 1986 yes yes yes 42.74

1991 - 1999 yes yes yes 46.34

2011 - 2016 yes yes no 13.82

About 15-20 new defaults “missing” since 2011 (in hist. comparison)

• Lower rates (post inflation stabilization in US)?

• Better macroeconomic management? Mismeasurement?

• China’s emergence as a push factor (both real and in finance)?

Reinhart

Some (not necessarily) mutually exclusive

hypotheses for the missing defaults

Better macroeconomic management, including macroprudential

and more hedging. Frankel and Vegh (2013), McGettigan et al.

(2015) on fiscal and monetary policy. Ostry, Ghosh and Qureshi

(2012 and 2016) and Cerutti, Claessens and Laeven (2016) on

macroprudential and Donders et al (2017) on private

commodity hedging, among others.

Better luck, external factors have remained more favorable than

in past cycles

low interest rates in US and other advanced economies

high growth in China (the newest push factor)?

Mismeasured (under-reported) defaults (Reinhart and Trebesch,

2016, arrears on official creditors. Arrears on Chinese lending

to low income commodity producers?

Reinhart

Vulnerability to Changes in International Interest

Rates, 1970-2016

Reinhart

20

25

30

35

40

45

50

55

197

0

197

1

197

2

197

3

197

4

197

5

197

6

197

7

197

8

197

9

198

0

198

1

198

2

198

3

198

4

198

5

198

6

198

7

198

8

198

9

199

0

199

1

199

2

199

3

199

4

199

5

199

6

199

7

199

8

199

9

200

0

200

1

200

2

200

3

200

4

200

5

200

6

200

7

200

8

200

9

201

0

201

1

201

2

201

3

201

4

201

5

201

6

Perc

ent

Share of variable rate

external debt, 137

developing and EMs

Reinhart

China’s inroads in

world trade are widely

recognized and cuts

across advanced and

emerging economies

and nearly all regions.

Shares of China in BIS

broad effective

exchange rate indexes

based on bilateral trade

shares.

0 10 20 30 40

AlgeriaArgentinaAustralia

AustriaBelgium

BrazilBulgariaCanada

ChileChina

Chinese TaipeiColombia

CroatiaCyprus

Czech RepublicDenmark

EstoniaEuro area

FinlandFrance

GermanyGreece

Hong Kong SARHungary

IcelandIndia

IndonesiaIreland

IsraelItaly

JapanKoreaLatvia

LithuaniaLuxembourg

MalaysiaMalta

MexicoNetherlands

New ZealandNorway

PeruPhilippines

PolandPortugalRomania

RussiaSaudi Arabia

SingaporeSlovakiaSlovenia

South AfricaSpain

SwedenSwitzerland

ThailandTurkey

United Arab EmiratesUnited Kingdom

United StatesVenezuela

2011-13

1993-95

Percent

Shares in the FRB broad exchange rate index for

the US based on bilateral trade shares, percent

Reinhart

57

26 22

1 0

10

20

30

40

50

60

70

2018201320082003199819931988198319781973

Other important trading partners of which China

Trade inroads have their counterpart in China’s

International Lending—especially low income countries: Estimated Share of Total Cross-Border Claims as of 2016:Q1

Reinhart

0 20 40 60 80 100

AlgeriaAngola

BeninBotswana

Burkina FasoBurundi

CameroonCape Verde

ChadCongo

CongoDRCote D'Ivoire

DjboutiEgypt

EthiopiaGabon

GambiaGhana

GuineaKenya

LesothoLiberia

MadagascarMalawi

MaliMauritania

MoroccoMozambique

NamibiaNiger

NigeriaRwanda

SeychellesSierra Leone

So AfricaSudan

TanzaniaTogo

TunisiaUgandaZambia

ZimbabweAverage

Percent of all external claims

0 10 20 30 40 50 60 70 80 90

Bangladesh

Cambodia

China

India

Indonesia

Korea

Laos

Malaysia

Mongolia

Myanmar

Nepal

Pakistan

Philippines

Sri Lanka

Thailand

Vietnam

Average

Percent of all external claims

Africa

Asia

Are some of these official Chinese

loans in arrears?

Currency Composition of Public and Publicly

Guaranteed Debt: All Other Currencies:

Renminbi’s Growing Share?

Reinhart

8

9

10

11

12

13

14

15

16

17

Percent

The Incidence of Default With and Without Official Creditors:

1952-2016

Defaults and arrears to non-China official creditors are much higher

than commonly reported—but still no recent upturn

Reinhart

There is evidence to suggest that the macroeconomic

management of capital inflow surges has been improving in

emerging markets as a whole.

Yet, one has to recall that prior to the Global Financial Crisis

of 2007-2009, a widely accepted view was that the advanced

economies had tamed the business cycle. This was the short-

lived era of the so-called “Great Moderation.”

Perhaps, the change is structural but a cautious interpretation

the missing defaults is that the protracted nature of the

downturn in international conditions has yet to take its

cumulative toll or lingering weaknesses will only become

evidence once the major central banks move further along in

renormalizing the stances of their policies.

Reinhart