...

The Black Swan - Harvard Kennedy School

by user

on
Category:

canada

8

views

Report

Comments

Transcript

The Black Swan - Harvard Kennedy School
Unforeseen Future Global
Developments
Jeffrey Frankel
Harpel Professor
Windsor Conference
Harvard Faculty Club
September 20, 2010
As a rule, it is highly probable that
something improbable will happen
Consider 3 events that occurred last spring
over the span of less than one month:



A volcano in Iceland shuts down
European air traffic (April 14, 2010)
An oil platform explosion leaks
some 30 million gallons into
the Gulf of Mexico (BP, April 20).
Some unknown glitch briefly sends
the stock market down about 7%
in 15 minutes -- some stocks
almost to zero (flash crash, May 6).
Who would have thought, beforehand,
to put a probability as large as 1% on any of these events?
2
I will make three passes at the subject:

Black Swans

Some possible developments,
each of which is highly improbable

Some more probable future paths.
3
Black Swans


We have heard a lot about “black swans”
ever since the 2007-08 financial crisis.
Unfortunately, the phrase is used
simply to mean a very unlikely event.
4

In 1998, LTCM managers said that they could
not be expected to have allowed for their crisis,
because it was a 7-standard-deviation event,
• or even a 22-std.dev. event.


In 2007, Goldman Sachs’ CFO said the subprime mortgage crisis was a 25-s.d. event.
Nonsense.
• My guess:
if our normal
distribution tables
reported numbers out
to 22 standard deviations, we would be in the realm
of the probability that 2 meteors hit Earth at once.
5

Slightly more enlightened are people who cite
Knightian uncertainty or “unknown unknowns”:
• Ignorance with humility
is better than ignorance without it.

A still better interpretation is
that distributions have fat tails.
• Why? Unconditional distributions
may have fat tails because conditional distributions,
even if normal, have time-varying variances;
• Perhaps, in crises, people’s actions become less independent.

It would be nice, however, to get beyond the Jurassic
Park lesson (“don’t be surprised if things go wrong”),
to say intelligent things about the tail events.
6
Definition of black swan?


In my view, “black swan” should refer to
an event that is considered virtually impossible
by those whose frame of reference is limited
in time span and geographical area,
but that is well within the probability
distribution for those whose data set includes
other countries and other decades/centuries.
Consider 5 examples of mistakes made
by those whose memory did not extend beyond a
few years or decades of personal experience in a
small number of countries.
7
Example #1: “All swans are white.”

The origin of the black swan metaphor was the
belief, which might have been held by a 19th
century Englishman based on induction from a
lifetime of personal experience,
that all swans were white. [1]

But ornithologists already knew that there existed
black swans, because explorers had discovered
them in Australia in 1697.
[1] The proposition that all swans are white goes back to ancient Roman philosophy,
where it was considered a truism. But induction from observation has its limits,
as pointed out by Hume (1739), followed by Mill (1843) & Popper (1959):
All it takes is a single black swan to falsify a generalization based on thousands
of observations of white swans.
8
Example #1: “All swans are white.”



An Englishman encountering a black swan might
have considered it a 7-standard deviation event,
even when the relevant information to the contrary
had already been available in ornithology books.
It seems to me a waste of a good metaphor to use
the term just to mean a highly unexpected event; [2]
a better use of it is to refer to an event that
would not have been so unexpected ex ante if
forecasters had adopted a perspective broader
than their immediate personal experience.
.
[2] Taleb (2007).
9
Example #2:
“Terrorists don’t blow up tall office buildings”

Before Sept. 11, 2001, some terrorist experts warned that
terrorists might try to blow up big US office buildings. [1]

The warnings were not taken seriously
by those in power at the time, nor by the public.
•
•

“It’s never happened before.”
Most Americans probably did not know the history
of terrorist events in other countries & in other decades.
Still today, there is a large gap between the probability
of a nuclear event as perceived by the public and
the probability as perceived by terrorism experts
(Graham Allison, 2005: 50% ). [2]
[1] They included the anti-terrorism director at the National Security Council in the Clinton & Bush
Administrations. Clarke (2004): Against All Enemies.
[2] “…on the current trend line, the chances of a nuclear terrorist attack in the next decade are
10
greater than 50 percent.”
Example #3: “Housing prices don’t fall.”


Many Americans up to 2006 assumed that nominal
housing prices, even if they slowed down, would not fall.
After all, “they never had before,” [1]
• which meant that they had not fallen in living U.S. memory.
• Few Americans were aware that housing prices had fallen
in other countries, and in the US before the 1940s.

Needless to say, many a decision would have been made
very differently, whether by indebted homeowners or
leveraged bank executives, if they had thought there was a
non-negligible chance of an outright decline.
[1] Shiller (2007).
11
As of 2006, 30 years of house prices seemed
to show a clear upward trend,
with no declines in nominal terms
12
But 100 years of data could have revealed
that the last decade was an aberration:
Housing prices can fall, even in nominal terms,
and in real terms tend to revert to their 1950-75 level
13
House prices had also fallen in other countries.
14
The crash
Irish home prices
15
Example #4: Volatilities are low.
“The Great Moderation”
Financial markets perceived risk as very low

•
especially during 2004-07.

This was most directly visible via the implicit
volatilities in options prices such as the VIX.

But it was also manifest in
•
•
•
Sovereign spreads, 1998-2007
junk bond spreads,
sovereign spreads,
and other financial prices.
16
Source: “The EMBI in the Global Village,” Javier Gomez May 18, 2008 juanpablofernandez.wordpress.com/2008/05/
In 2003-07, market-perceived
volatility, as measured by
options (VIX), plummeted.
So did spreads on US junk
& emerging market bonds.
In 2008, it all reversed.
17
17
Example #4: Low volatilities.
Why this historic mis-pricing of risk?

•
•
Traders plugged into their Black-Scholes formulas
variance estimates that went back only a few years,
or at most a few decades,

•

the period of the late Great Moderation.
Risk officers did the same with VaR models.
They should have gone back much farther –
or better yet, formed judgments based on
a more comprehensive assessment of what risks
might lie in wait for the world economy
.[1]
18
[1] E.g., Frankel (2008).
Example 5: “European governments don’t default.”
Greece’s recent debt troubles should
not have caught anyone by surprise,

•
least of all northern Europeans.

The same with Portugal, Spain, Italy & Ireland.

And yet from the time they joined the euro,
until 2009, these governments could borrow at
interest rates virtually as low as Germany.

Among the reasons: a perception that advanced
countries in general, and euro countries in particular,
were fundamentally different from emerging
markets and would never default.
19
Judging from spreads, 2001-07,
investors put zero odds on a default by Greece
or other Mediterranean countries
Council on Foreign Relations
20
Example 5: “European governments don’t default.”

Suddenly, in 2010, the Greek sovereign spread
shot up, exceeding 800% by June.

Even when the Greek crisis erupted,
leaders in Brussels & Frankfurt
seemed to view it as a black swan,
• instead of recognizing it as a close cousin
of the Argentine crisis of ten years earlier,

and many others in history,
• including among European countries.
21
Are these examples hindsight?
How can one gauge risk,
other than time series estimation?

Consider a talk I used to give, 2005-07

This particular slide is taken from a version
I gave to State Street Global Investors, May 24, 2006.

Did I predict the financial crises? No. But…
22
Medium-term global risks
Odds of each alone small; but cumulative odds > 1/2

Hard landing of $: foreigners pull out =>
$↓ & i↑

=> possible return of stagflation in US.
Bursting bubbles => consumption would fall .
• Bond market: conundrum of low LT rates 2004-05.
• Housing market: valuations very high . This bubble indeed burst

New oil shocks,
•

e.g., from Russia, Venezuela, Nigeria, Iran, S.Arabia…
New security setbacks
•
•
•
soon after this presentation
The other
risks remain.
Big new terrorist attack, perhaps with WMD
Korea or Iran go nuclear/and or to war
Islamic radicals take over Pakistan, S.A. or Egypt
23
Some possible events
-- each of which is improbable taken alone -in four categories

Financial markets

Macroeconomic

Geopolitical

Environmental & miscellaneous
24
Financial markets



Housing prices fall another 1/3
Equity prices fall by ½
Bond prices fall by ½
• US i 10 yr. rises from 2 ½ % to 5%+


Hard landing: Dollar falls by ½
Repeat of financial crisis:
interbank spreads in major countries
go back to Sept. 2008 high (350 basis pts.)
25
Financial markets, continued

Sovereign default crisis,
• beginning with debt
restructuring in Greece --
P=½
• spreading through Europe -- P= ½ ½ = ¼
• and to the US
--
P = ½ ½ ½ = 1/8
by contagion
26
Desmond Lachman, Euro Will Unravel, and Soon, American Enterprise Institute, Sept.2010
Debt dynamics:
Greece can’t achieve primary surplus & keep its GDP growth rate above its interest rate.
27
=> Its debt/GDP could rise to 175% over the next few years.
Macroeconomy

Return of high inflation
• the monetary base more-than-doubled in 2008,


and that money is still out there.
Deflationary trap
• central bank loses ability to expand monetary policy:

higher real interest rates
• <= negative inflation
•
+ zero-lower-bound
on nominal interest rates.


Like Japan.
Unemployment stays very high for a decade,
as the long-term unemployed lose skills
28
Geopolitics

Military conflict
•
•
•
•

between Iran and US or Israel
between North Korea and US
Between India and Pakistan
Other new serious military conflict
Islamist extremists take over in:
• Saudi Arabia (oil)
• Egypt (population & cultural center of Arab world)
• Or Pakistan (nuclear weapons)

Nuclear or biological or radiological terrorism
29
Environment & miscellaneous

Severe contagion,
• like SARS, Avian flu, H1N1,… but worse

Price of oil goes above $200/barrel

Price of wheat goes above $500/metric ton

Severe drought, probably brought
on by global climate change

Some other environmental catastrophe

Some IT disaster
30
More-probable scenarios




We are seeing
a historic reversal of roles between those countries
traditionally viewed as advanced/industrialized
and those viewed as emerging/developing.
No longer do industrialized countries automatically
have higher creditworthiness than the latecomers.
Indeed, China, Chile, and others have been able to
follow fiscal countercyclical policy over the last
decade, while the US & UK have forgotten how to.
The advanced countries are now the debtors.
31
A remarkable role-reversal:
• Debt/GDP of the top 20 rich countries
(≈ 80%) is already twice that
of the top 20 emerging markets;
• and rising rapidly.
• By 2014 (at ≈ 120%), it could be triple.
32
1. Fairly likely scenario.
Muddling through in short-term:
• Recovery is long & slow among advanced countries,

especially with no further fiscal stimulus, because politicians
• are either unwilling (Hoover economics)
• or unable (as a legacy of Bush economics).

Growth barely above potential
• => unemployment comes down only slowly.
• No more major disruptions for a while.
• Long-term debt problems lead to eventual crises.
33
World Economic Outlook Projections, IMF, July 2010
Growth rates (% change)
Year over Year
Projections
2008
World Output
3.0
Advanced Economies
0.5
United States
0.4
Euro Area
0.6
Japan
–1.2
United Kingdom
0.5
Canada
0.5
Other Advanced Econ.s
1.7
Newly Ind.Asian Economies 1.8
Emerging & Developg. Ec.s 6.1
2009
–0.6
–3.2
–2.4
–4.1
–5.2
–4.9
–2.5
–1.2
–0.9
2.5
2010
4.6
2.6
3.3
1.0
2.4
1.2
3.6
4.6
6.7
6.8
Q4 over Q4
Estimates
2009
4.3
2.4
0.1
–2.1
–1.4
–3.1
–1.1
3.1
6.1
5.7
Projections
2010
4.2
2.3
3.2
1.1
1.1
2.1
4.0
3.4
4.3
6.9
2011
4.3
2.6
2.6
1.6
3.0
1.9
2.6
4.6
6.3
34
6.8
2. Hard landing for the dollar:
Definition: foreigners lose willingness
to continue accumulating US assets
=> $ ↓ & i ↑
=> possible return of stagflation in US
• Many have warned of this possibility for some time.
• So far, foreigners have happily financed US deficits.
• But just because the $ has been the leading international
currency for 70 years, doesn’t mean it always will.
• Cautionary tale: £ (1914-1956)
With a lag after US-UK reversal
of economic size & net debt,
$ passed £ as #1 international currency.

35
Global current account imbalances
– China’s surplus and America’s deficit –
are now widening again,
with recent recovery in US income & the $.
36
Pre-2007, economists were split between
those who saw the US deficit as
unsustainable, requiring a $ fall,
and those who saw
no problem.

Martin Feldstein


Ken Rogoff *
Maury Obstfeld
Nouriel Roubini
Larry Summers
Menzie Chinn
Me
Lots more













Ben Bernanke
Ricardo Caballero *
Richard Cooper
Michael Dooley
Pierre-O. Gourinchas
Alan Greenspan
Ricardo Hausmann
Lots more
37
* Some claim that the financial crisis of 2007-09 fits their theories.
37
The 2007-09 financial crisis did not help resolve
the debate over current account imbalances

Those of us who predicted an unsustainable
US current account deficit and a $ hard landing
were proven wrong by the 2008 movement into $.

Meanwhile, those who said the US CA deficit was
sustainable because of the superior quality &
desirability of US assets were also proven wrong.



corporate governance, rule of law,
accounting system, rating agencies,
securities markets, …
38
MSN Money & Forbes
Projection of $ vs €
as shares of central banks’ foreign exchange reserves:
a function of country size, financial market depth, & rate of return,
with parameters estimated on 1973-98 data.
1.0
Simulation assumes $ depreciation continues at 2001-04 rate.
USD
Chinn & Frankel
(2005)
0.8
0.6
0.4
0.2
0.0
birth
of €
DEM
EUR
This scenario showed €
overtaking $ as top international
reserve currency in 2022.
39
75 80 85 90 95 00 05 10 15 20 25 30 35 40
Multiple International
Currency System

The € is now a rival international currency.

The ¥ & Swiss franc have gained a bit too.

The SDR came back
from the dead in 2009.

Gold has also made a comeback
as an international reserve asset.

Someday the RMB will join the roster.

= a multiple international reserve currency system.
40
3: Global crisis in
sovereign debt
• EU has long-term debt problems as bad as the US,
• esp. future liabilities related to retirement & health.
• Not just the PIIGS countries.
• Japan is perhaps even worse.
• Only Canada has done things better.
• Worldwide savings shortage is in the offing.
• I now think that a sharp fall in bond markets among
advanced countries generally is somewhat more likely
than a hard landing for the US per se, relative to the others.
41
4. The best of the four scenarios,
is the least likely
• Recovery becomes firmly established, and then
• the US and other major advanced countries
seriously address their long-term fiscal situations,
to reduce chance of long-term crisis.
• Good economic policy would combine, in my view,

Fiscal stimulus today, designed with high bang-for-buck:
• More money for states, to avoid lay-offs
• Infrastructure investment, for the future
• Among tax cuts, extending Bush benefits for super-rich
should be last priority (e.g., 2010 abolition of estate tax)

together with steps today to lock in
long-term return to fiscal discipline
• including social security reform, e.g., raising retirement age
42
Other possible events around the world
-- each of which is improbable taken alone -
The euro-zone breaks up

China has a financial crisis at some stage

China-US military conflict
• over Taiwan, or South China Sea

Mexico collapses (narcotics violence)
43
China could suffer the rite-of-passage
of any new trade+manufacturing power:
an asset market bubble & collapse
Real Beijing land prices
44
45
References










Allison, Graham, 2005, Nuclear Terrorism: The Ultimate Preventable
Catastrophe (Henry Holt)
Clarke, Richard, 2004, Against All Enemies: Inside America's War on Terror
(Free Press).
Frankel, Jeffrey, “Responding to Crises,” Cato Journal, vol. 27, no. 2,
Spring/Summer 2007, pp.165-178.
Frankel, & George Saravelos, 2010, "Are Leading Indicators of Financial
Crises Useful for Assessing Country Vulnerability? Evidence from the 2008-09
Global Crisis" NBER WP 16047.
Hume, David, 1888, Hume's Treatise of Human Nature (Oxford, Clarendon
Press). Originally published 1739–40.
Mill, John Stuart, 1843, A System of Logic: Ratiocinative and Inductive. (Harper
& Bros.).
Popper, Karl, 1959, The Logic of Scientific Discovery (Basic Books).
Reinhart, Carmen, and Kenneth Rogoff, 2009, This Time is Different: Eight
Centuries of Financial Folly (Princeton University Press).
Shiller, Robert, 2007, Understanding Recent Trends in House Prices and Home
Ownership Cowles Foundation Discussion Paper No. 1630 .
Taleb, Nassim Nicholas, 2007, The Black Swan: The Impact of the Highly
46
Improbable (Penguin).
Appendices

Doubling of US monetary base

Creditworthiness: Some advanced economics
have fallen as emerging markets have risen

Competing international currencies
47
Federal Reserve Assets ($ billions)
have more-than-doubled, through new
facilities, rather than conventional T bill purchases
48
Source: Federal Reserve H.4.1 report
48
Sovereign spreads for 5 euro countries
shot up in the 1st half of 2010
Creditworthiness: Some advanced economics
have fallen, as emerging markets have risen.
49
Ratings for “Advanced Economies”
Ratings for “Emerging Economies”
50
When will the US day of reckoning come?

It didn’t come in 2008: The financial crisis caused a flight
to quality which evidently still means a flight to US $.

Chinese warnings in 2009
may have augured a turning point:
• Premier Wen worried US T bills will lose value.
He urged the US to keep its deficit at an “appropriate size”
to ensure the “basic stability” of the $ (Nov.) .
• PBoC Gov. Zhou proposed
replacing $ as international
currency, with the SDR (March 09).
51
Multiple International
Currency System

The € has gained share as an international currency,
even if it won’t challenge the $ by 2022.

The ¥ & Swiss franc have gained a bit too.

The SDR came back from the dead in 2009.

Gold has also made a comeback
as an international reserve asset.

Someday the RMB will join the roster.

= a multiple international reserve currency system.
52
Special Drawing Rights
The SDR’s surprising comeback from near-oblivion.

The G20 & IMF decided to create new SDRs ($250b).
Shortly later, PBoC Gov. Zhou proposed replacing
the $ as lead international currency with the SDR.
The IMF is now borrowing in SDRs.



•
The proposal has been revived for an international substitution
account at the IMF,



to extinguish an unwanted $ overhang in exchange for SDRs.
The SDR has little chance of standing up as a competitor
to the € or ¥, let alone to the $.
Still, it is back in the world monetary system.
53
Gold

Until very recently, central bank holdings of
gold was considered an anachronism.
• Central banks were gradually selling it off.

Now gold is back on the list of international
reserve assets
• China bought gold in early 2009.
• India bought 200 tons in November.
• Sri Lanka…
54

A multiple reserve currency system is inefficient,
in the same sense that barter is inefficient:
money was invented in the first place to cut down
on the transactions costs of exchange.

Nevertheless, if sound macro policies
in the leader country cannot be presumed,
the existence of competitor currencies gives
the rest of the world protection against the leader
exploiting its position by running up too much
debt and then inflating/depreciating it away.
55
Historical precedent: £ ‘s loss of premier
international currency status in 20th century

By 1919, US had passed UK in
1. output (1872)
2. trade (1914)
3. net international creditor position (1914-19)

Subsequently, $ passed £ as #1 reserve
currency (1940-45)
56
US overtakes UK, by trade, 1900-1957
Data sources: UK Export Data: Department of Trade and Industry, UK; UK exchange rate (1946-1970): Global Financial Data;
US Export Data: Historical Statistics of the United States, Colonial Times To 1970; Published by the U.S. Census Bureau
Value of Exports
(in millions of dollars)
3.E+04
2.E+04
2.E+04
1.E+04
5.E+03
USA
1956
1953
1950
1947
1939
1936
1933
1930
1927
1924
1921
1918
1915
1912
1909
1906
1903
1900
0.E+00
UK
57
From the lit. on reserve currencies
Determinant:
Proxy:
1.
Size
GDP
2.
Depth of Fin.mkt.
FX turnover
3. Rate of return
inflation,
LR depreciation,
Exch. rate variance
58
From the literature, continued
Network externalities
=> Tipping
captured by:
1)
Inertia
lags
2)
Nonlinearity
in determinants
logistic functional form
or
dummy for leader GDP
59
Cautionary tale, cont.

UK loss of international currency paralleled loss of its
economic pre-eminence, military hegemony, colonies
& other trappings of international power.
• Suez crisis of 1956 is often recalled as occasion on which Britain
was forced under US pressure to abandon imperial designs.
• Often forgotten: A run on the £ was the mechanism.

Paul Kennedy’s (1989) suggestion
-- that imperial overreach hypothesis
might apply to US hegemony –
may have been essentially correct
• but 15 years premature,
• much like those in early 1990s who warned
prematurely over $’s imminent demise (Kindleberger, 1995...)
60
Jeffrey Frankel
James W. Harpel Professor of Capital Formation & Growth
Harvard Kennedy School
http://ksghome.harvard.edu/~jfrankel/index.htm
Blog: http://content.ksg.harvard.edu/blog/jeff_frankels_weblog/
Fly UP