John Roe - The Sting in the tail · Peasants’ Revolt, 1381 Battle of Kosovo, 1389 24 46 Great...

© 2010 The Actuarial Profession � www.actuaries.org.uk

Life Conference and Exhibition 2011John Roe

The Sting in the tail

1

A Foreword

• The views represented are my own as are any errors

• I work for Legal and General Investment Management onstrategic investment analysis and solutions

• I am not a regulatory actuary focused on Solvency II

• Analysis is approximate and to provide context, rather thanto challenge others’ more granular analysis

• Solvency II is not finalised

• New draft text is expected to introduce a matching premiumfor annuities and reduce capital volatility

2

13The Hook – “1-in-200 year” risk calibrations3

26The Tale – taming the business cycle4

29The Wire – Great Moderation in action5

36The Shut Out – when stabilisation failed6

7

2

1

Section Slide

The Players – defining tail events 2

The Set-up – insurers’ approach to tail risks 7

The Sting – where to go from here 39

3

The Players

Defining tail events

4

Naseem Taleb, The Black Swan,2007

Swans of every colour

• High-impact, hard-to-predict, rare

• Non-computable

• Hindsight

“Contrary to conventional wisdom, crises are not black swansbut white swans: the elements of boom and bust areremarkably predictable”

Nouriel Roubini, Crisis Economics, 2011

5

A simple definition of tail risks

For senior managementAn event which defines their lasting legacy

For insurers and pension fundsAn event which recasts their future

6

Franchise value destruction and derisking

Free capital + embedded value

Ris

k to

lera

nce

Institution‘s risk tolerance

Reckless gamblingUnconstrained strategy

Reduced abilityto leverage

brand

Fullderisking

7

Pension funds have the opposite problem

Time

Ris

k to

lera

nce

Liability driveninvestment

Full derisking/ buyout

Return seeking assets

Extended derisk path

Increased employercontributions

8

The Set up

Insurers’ approach

9

Risk management – rapid progress

• GAO hedging c.25 years after Black-Scholes• Rapid progress since

– Realistic Balance Sheet– Individual Capital Assessment– Solvency II

• Stochastic and stresses• Principles and rules

10

Draft Solvency II SCR – the year’s must haveslide?

• Standard QIS 5 formula• Equities: 39% for Global, 49% for Other• Property: 25%• FX: 25%• Liabilities discounted based on swaps + illiquidity premium

4.52.52.51.4BBB

2.61.41.41.1A

1.51.11.10.0AA

1.30.90.90.0AAA

CreditDerivs

(%)

StructureProducts

(%)

Corporates

(%)

Non-EEAGovs

(%)

11

Draft (QIS 5) SCR correlations

10.50.750 (0.5)MKTsp

0.510.750 (0.5)MKTprop

0.750.7510 (0.5)MKTeq

0 (0.5)00 (0.5)1MKTint

MKTspMKTpropMKTeqMKTintUp (Down)

• Solo entity diversification reduces SCR by c.35%• Owning multiple asset classes reduces benefit of the

illiquidity premium

12

Insurers financial strength through recent crisis

Source: Bloomberg LP

0.0

20.0

40.0

60.0

80.0

100.0

120.0

140.0

160.0

Jan-06 Jul-06 Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09 Jan-10 Jul-10 Jan-11

Pric

e (G

Bp)

Prudential Old Mutual Legal & General Aviva Standard Life

13

European insurers’ sovereign exposures

Source: Societe Generale – Cross Asset Research as at 07/07/2011

• Market storms are a long way from being weathered

Gross Sovereign debt (gross of policyholder participation)

0%

100%

200%

300%

400%

500%

600%

700%

Insurer A Insurer B Insurer C Insurer D Insurer E Insurer F Insurer G Insurer H Insurer I

as %

of S

hare

hold

er e

quity

Italy Spain Ireland Portugal Greece

14

The Hook

Enough capital for a 1-in-200 year event ?

15

People judge books by covers

Arbitrary coherence“1-in-200” encourages risk measurement not management

Very hard to shift mindset once anchored

The impact can be extremeIt can double the price of (Neuhaus) chocolates1

So what if the risk calibration and time horizons arewrong?

1 Source: Dan Ariely, George Lowenstein, and Drazen Prelec, “Tom Sawyer and the Construction of Value”, Journal of Economic Behaviour and Organization (2006)

16

Time horizons

• Ever dwindling• Mark to market a guiding Solvency II principle• Longevity recognised as emerging over the long term• Credit remains controversial• Liability illiquidity is the key to the debate

Solvency II isn’t really a one year time horizonMTM only works if assets and liabilities share similar liquidity

levels

17

Experiments in mark to market liabilities – withprofits

1 Source: Asset and Liability Studies on a With Profit Fund, Tim Roff, Presented to the Staple Inn Actuarial Society, October 1992

2 Source: Smashing With-Profits Business, Howard Froggatt and Icki Iqbal, Staple Inn Actuarial Society, October 2002

– Average EBR around 75% at end of 19911

– Average EBR constant around 70% between 1990 and 20002

– Average EBR rose steadily upwards from c.35% to c.70%between 1970 and 19902

Equity backing ratios (2005 – 2010)

Source: 2010 – 2005 FSA Returns, Form 48, Column 2

30%

40%

50%

60%

2005 2006 2007 2008 2009 2010

Average EBR (Companies >500m) Average EBR (Companies <500m)

18

-350-300

-250-200

-150-100

-50

0

50

2006 2007 2008

Prof

it/Lo

ss

Pillar 1 Pillar 2

Experiments in mark to market liabilities –annuities

-300

-250

-200

-150

-100

-50

0

50

100

150

Gilts Credit Credit + Swaps

Prof

it/Lo

ss

Pillar 1 Pillar 2

• 2009 Institute of Actuaries Working Party1

• £2bn starting liability portfolio

Corporate bond investment - net surplus emerging 2006-2008Net surplus emerging 2009

1 Source: Unintended consequences and the avoidance of self-fulfilling prophecies, Impacts of regulation and market turbulence on annuity fund investment strategies working party, June 2010

19

Calibration

By necessity standard formula simplifiesFSA paper

Calibration of the Enhanced Capital Requirement for with-profit life insurers, June 2004

Test the calibration (and that of Solvency II)

20

0 %

10 %

20 %

30 %

40 %

50 %

60 %

70 %

1871 1883 1896 1908 1921 1933 1946 1958 1971 1983 1996 2008

Stress based on S&P Returns

Realised historical equity volatility

Great Depression

World War IIOil Crisis

Credit Crunch

Source: Reuters Ecowin

21

Credit Stresses

21© 2010 The Actuarial Profession � www.actuaries.org.uk

1. Source: Bloomberg for data from 1999 onwards,.Based on iBoxx spreads. Prior to Jan 1999 only proxy BBB spreads available - these were dampened to allow for greater volatility of BBB relative to overall portfolio. From October 2007 BarCapdata used to infer impact on credit-spreads of re-ratings.

2. Threshold represents 1 in 200 event as implied by table in paragraph 8.8 of the June 2004 FSA paper ‘Calibration of the Enhanced Capital Requirement for with-profit life insurers’ by Watson-Wyatt and fitting a Gaussian distribution to extrapolatethe tail.

Year Change in Credit Spread1,2

-5%-4%-3%-2%-1%0%1%2%3%4%5%

1973 1978 1983 1988 1993 1998 2003 2008

Yearly change in credit spread Threshold Yearly change in credit spread allowing for re-ratings

22

Correlation between US equities and Treasuries


1891 1931 1971 2011

Source: DataStream

-0.8

-0.4

0

0.4

0.8

1.2

Correlation between S&P 500 Returns and 10yr US Treasuries

1891 1931 19711891 1931 1971

23

“Tail Events” more common than consensus


American CivilWar, 1861

InfluenzaEpidemic, 1918

GreatDepression, 1929World War I, 1914

InternetRevolution

Invention ofcomputer

Invention ofaeroplane

World War II,1939

South Asianearthquake andtsunami, 2004

Invention of telephone

1811 2011


Spanish CivilWar, 1936

Korean War, 1950Global FluEpidemic, 1890

Formation People’sRepublic of China,

1949

World War II, 1939

Crimean War, 1853

20111811

Vietnam War,1955

September 11,2001

Lehman fall,2008

Black Monday,1987

1011 1211

East-WestSchism, 1054

First Crusade, 1099

Foundation ofHanseatic League,

1158NormanInvasion, 1066

First knownMerchant Guildformed, 1193

Third Crusade, 1190Second Crusade,

1147

Peace ofConstance

signed, 1183

Fourth Crusade,1202

Establishment ofMongol Empire,

1206

Spanish-Christiandefeat of Moors,

1212

Great Famine,1311

Black Plague, 1347Magna Carta,1215

Russian victoryover Mongols,

1380

Western Schism,1378

Beginning ofHundred Years

War, 1337

Mongol Empirebreakup, 1361

Peasants’ Revolt,1381

Battle ofKosovo, 1389

1211 14111411 1611

Great Plague inEngland, 1509

Fall ofConstantinople,

1453

Sack of Rome, 1527

German Peasants’War, 1524

Discovery ofAmerica, 1492

ProtestantReformation,

1517

Shaanxiearthquake, 1556Spain declared

bankruptcy, 1557

Printing pressinvention, 1439

Unification ofSpain, 1469

Vesuvius erupts, 1631

Tulip Mania, 1637

Great TurkishWar, 1667Manchurians

conquer China,1644

LisbonEarthquake, 1755

MississippiBubble, 1720

EnglishRevolution, 1649

South SeaBubble, 1719

French Revolution,1789

American War ofIndependence,

1776

1611 18111811 2011

East-West Schism,1054

man Invasion, 1066

First Crusade, 1099

Second Crusade,1147

Foundation ofHanseatic

League, 1158

Peace of Constancesigned, 1183

Third Crusade, 1190

Formation of firstknown merchant

guilds, 1193

Fourth Crusade,1202

Establishment ofMongol Empire,

1206

Spanish-Christiandefeat of Moors,

1212

Magna Carta, 1215

Great Famine, 1311

Beginning ofHundred Years War,

1337

Black Plague, 1347

Mongol Empirebreakup, 1361

Western Schism,1378

Russian victory overMongols, 1380

Peasants Revolt,1381

Battle of Kosovo,1389

Printing pressinvention, 1439

Constantinoplefalls, 1453

Unification of Spain,1469

Discovery ofAmerica, 1492

Great Plague inEngland, 1509

ProtestantReformation, 1517

German Peasants’War, 1524

Sack of Rome, 1527

Shaanxi earthquake,1556

Spain declaredbankruptcy, 1557

Vesuvius erupts,1631

Manchurian conquerof China, 1644

Tulip Mania, 1637

English Revolution,1649

Great Turkish War,1667

South SeaBubble, 1719

Mississippi Bubble,1720

Lisbon earthquake,1755

French Revolution,1789

American War ofIndependence, 1776

American Civil War,1861

World War I, 1914

Worldwide InfluenzaEpidemic, 1918

Invention ofaeroplane

Great Depression,1929

World War II, 1939

Invention ofcomputer

Internet revolution

Gulf War, 1990

South Asianearthquake andtsunami, 2004

1211 1411 1611 1811


Spanish CivilWar, 1936

Korean War, 1950Global FluEpidemic, 1890

Formation People’sRepublic of China,

1949

World War II, 1939

Crimean War, 1853

20111811

Vietnam War,1955

September 11,2001

Lehman fall,2008

Black Monday,1987

24

The UK annuity and DB pension market


Standard 65 year old male/female fixed annuity rate2:Male: 6.1%, Female: 5.7%

Total: £103.4bn

Other25%

Govts18%

Corps51%

Equity1%

Var.Int.

Bond3%

Prop.3%

Top ten UK annuity funds Total UK DB Market

£845bn

Source: FSA Returns End 2010, From 48, 2: As at 15/1//2011

Source: Calibration of the Enhanced Capital Requirement for with-profit life insurers by Watson Wyatt as at June 2004

£103bn

25

Extreme risk aversion in annuities – QIS 5 vs Giltonly strategies


Longevity 13%, Operational 3%, Credit 11%,Total Expenses 50bps6% Cost of Capital1% liquidity premium, additional 0.75% credit return post haircuts

0%

10%

20%

2011 2020 2030 2040 2050

Longevity Capital Operational CapitalCredit Capital Total Capital (with Spread Risk)Total Capital (Risk Free)

26

Potential impacts of Gilt only investments


All results vs a QIS 5 capital treatmentAnnuity pricing worsens by c.10-15%Impacts of government substitution

Insurer shareholders lose c.60%Corporate funding costs rise, c.£80bn Sterling corporate debtIncreased borrower reliance on short term financing

Liabilities affected MTM impact on UK public debt (£bn:%)1 year of annuities £1.4bn:0.1%All existing annuities £17bn:1.8%All DB pensions £110bn:11.3%

27

The Tale

Risk had been tamed

28

The Great Moderation 1985 - 2007

U.S. GDP change and volatility


-15

-10

-5

0

5

10

15

20

25

30

Dec-71 Dec-76 Dec-81 Dec-86 Dec-91 Dec-96 Dec-01 Dec-060

1

2

3

4

5

6

7

8

9

GDP U.S. Nominal QoQ Rolling 2 year volatility (rhs)

Pre Period Great Moderation Years Post

“potential gains fromimproved stabilization

policies are on the orderof hundredths of a

percent of consumption”

Robert Lucas,presidential address tothe American Economic

Association, January2003

29

The monetary policy tools

Unemployment

Wage andprice inflation

Riserates

Lowerrates

Real GDP

Time

30

The Wire

Realised volatility was lower

31

0

1

2

3

4

5

6

7

8

9

10

Jun-87 Jun-89 Jun-91 Jun-93 Jun-95 Jun-97 Jun-99 Jun-01 Jun-03 Jun-05 Jun-07 Jun-09 Jun-11

Fed Fund's Rate

Moderation in action – the Greenspan Put

The fed funds rate and identifiers of Greenspan put

Black Monday

Savings & Loan Crisis

Gulf War

Peso Crisis

LTCM

Dot-com Bubble

9/11

Financial Crisis


BernankePut

32

0

10,000

20,000

30,000

40,000

1996 1998 2000 2002 2004 2006 2008 2010-6

-2

2

6

10

US ImportsFrom China

ChinaProducerPrices ofConsumerGoodsYoY (rhs)

Imported deflation

• Improving Emerging market labour productivity• Controlled exchange rates• China lowered U.S. import inflation by c.80bp p.a between

1993 to 20041

1 Source: Board of governors of the federal reserve discussion paper, Is China “Exporting Deflation”?, 20042 Source: Data source Bloomberg L.P. , National bureau of statistics of China

Chinese imports and declining prices2

Great Moderation Years

33

Focus on full employment through cycles

U.S. annualised QoQ inflation and unemployment3

• Supercharged economy

3 Source: Data source Bloomberg L.P.

Great Moderation Years

-4

0

4

8

12

16

1975 1980 1985 1990 1995 2000 2005 20100

2

4

6

8

10

12

U.S. CPI Urban Consumers U.S. Unemployment Rate (rhs)

34

• Bank bonuses encourage short-term outlook.• Principal-agent problem: downside consequences not

fully passed to traders.• Example

A – good/usual year, probability 95%

B – very bad year (tail-risk), probability 5%

Payoff for Trader (remuneration) is 10% of profit madeby Bank but with a lower limit of zero.

Traders – set up


35

Traders – payoffs

Payoffs for Bank and for Trader (arbitrary units):

24.0353

-1,515

105

Bets on A

Bank’s Profit

24.0453

2,000

-80

Bets on B Bets on BBets on AProbabilityCase

20005%B

10.044

10.02.3

Expected payoff:Volatility of payoff:

010.595%A

Trader’sremuneration


36

Bank best1 strategy : c.55% bet on A, 45% bet on B→ Bank has expected payoff of 24.0 and volatility of 9.8.

Trader best1 strategy: c.95% bet on A and 5% bet on B→ Trader has expected payoff of 10.0 and volatility of 0.005.

If Trader bets 95% on A and 5% on B the Bank suffersvolatility of profit of 313, rather than 9.8.

Severe multi-period repercussions if risk-seeking individualsrewarded/promoted

Traders – behaviour

1. Lowest standard deviation combination of bets; in this example split doesn’t impact expected payoff. 36© 2010 The Actuarial Profession � www.actuaries.org.uk

37

The Shut-out

The problem of induction

38Source: Reuters Ecowin

Source: Monthly and quarterly GDP estimates for interwar Britain by J Mitchell, S Solomou, and M.Wale as at November 2009

How recessions compare – cumulative real GDPgrowth

-12%

-10%

-8%

-6%

-4%

-2%

0%

2%

4%

6%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Quarter from start of recession

1973-1976 1979-1983 1990-1993 2008-2011 1930-1934

and an uncertain outlook…

39

Commodity prices booming

39© 2010 The Actuarial Profession � www.actuaries.org.ukSource: Bloomberg. Rebased to 100 at 30 June 1988.

0

200

400

600

800

1000

Jun-88 Jun-90 Jun-92 Jun-94 Jun-96 Jun-98 Jun-00 Jun-02 Jun-04 Jun-06 Jun-08 Jun-10

Oil Gold

$

40

The Sting

Avoiding the pitfalls

41

UK debt outlook

Government net debt to GDP ratio

Source: LGIM Fundamentals Economic and Investment Commentary as at October 2011

50

55

60

65

70

75

80

85

2009-10 2010-11 2011-12 2012-13 2013-14

%of

GD

P

A - Austerity B - Boost C - Crisis Government

42

Volatilities of asset classes1

42© 2010 The Actuarial Profession � www.actuaries.org.uk1. Source: Bloomberg. Volatility of FTSE All Share and UK Govt bonds based on rolling one year of monthly returns.

0%

10%

20%

30%

40%

50%

60%

70%

80%

1971 1974 1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010

FTSE All Share UK Gov Bonds (8 year ZCB)

43

Correlations of asset classes1


Developed/developing economies are now more connected.1. Source: Bloomberg. UK Equity: FTSE All Share. Based on rolling 5 years of monthly data.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Mar-97 Jul-98 Dec-99 Apr-01 Sep-02 Jan-04 May-05 Oct-06 Feb-08 Jul-09 Nov-10 Apr-12

UK vs EM Bonds UK vs EM Equity

44

SCR volatility – proxy methodology

• All on standard formula from QIS 5• Assets = 65% credit, 20% Gilts, 10% property, 5% equity• Capital = 150% SCR at outset• Interest rate matching liabilities• Rebalance assets monthly• Annual dividend if capital > 150% SCR

45

Solvency II is a moving feast

• Annuity results presented are based on QIS5• They do not reflect lower and more stable capital

requirements for annuities generated by the MatchingPremium included in the latest Level 2 text

• These results do not therefore reflect the expected impactof Solvency II on UK annuity business

46

Draft Solvency II (QIS 5) applied from 1972

0

5

10

15

20

25

30

35

1972 1976 1979 1982 1986 1989 1992 1996 1999 2002 2006 2009

Dividend Own Fund SCR MCR

• EIOPA is working to ensure more appropriate SCR volatility

47

And with stable Gilt yields

0

5

10

15

20

25

1972 1976 1979 1982 1986 1989 1992 1996 1999 2002 2006 2009

Dividend Own Fund SCR MCR

48

How to analyse tail risks

Think macroMinimal assumptionsMaximise debateMitigate behavioural finance issues

Multiple angles of attack are essential – we’re building asafety net, so knit the threads from every direction

49

-15%

-10%

-5%

0%

5%

10%

15%

1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009

Ret

urn

Leve

l

Historical scenarios - recalibrated

Data source Bloomberg L.P.

Rus

sian

Def

ault

Mor

tgag

e C

risis

Eur

o S

ovC

risis

Japan LostDecade

Asi

an C

risis

Tech

Bub

ble

9/11

Fina

ncia

l Cris

is

S&

P 5

00 R

etur

n Le

vel

50

Japan and U.S. commercial lending rates1

Japan and U.S. YoY lending changes2

1 Source: World Bank: World Development Indicators via Thomson Reuters DataStream2 Source: Data source Bloomberg L.P.

T = 1992 for Japan and2007 for U.S.

Learning from the past – the Japanese example

0%

2%

4%

6%

8%

10%

T-7 T-6 T-5 T-4 T-3 T-2 T-1 T T+1 T+2 T+3Japan Commercial Lending Rate %U.S. Commercial Lending Rate %

-

20,000

40,000

60,000

80,000

T-7 T-6 T-5 T-4 T-3 T-2 T-1 T T+1 T+2 T+3-

2,000,000

4,000,000

6,000,000

8,000,000

Japan Consumer Credit: Domestic Licenced Banks New Loan Accounts

U.S. Loans and Leasses by Commercial Banks (rhs)

51

Learning from the past – the Japanese example

Japan and U.S. cumulative non-performing loans2

Japan and U.S. YoY real estate pricechange1

1Source: Bank of Japan via Thomson Reuters DataStream, Data source Bloomberg L.P.2Source: FDIC (Federal Deposit Insurance Corporation) , Japan Bankers Association via Nomura Bank Research

T = 1992 for Japan and2007 for U.S.

-10%

-5%

0%

5%

10%

15%

20%

25%

T-7 T-6 T-5 T-4 T-3 T-2 T-1 T T+1 T+2 T+3

Japan Land Prices Nationwide YoY% FHFA U.S. House Price Index YoY%

-1%

2%

5%

8%

11%

14%

17%

T T+1 T+2 T+3 T+4 T+5 T+6 T+7 T+8 T+9 T+10

Japanese Bank's Credit Losses U.S. Bank Delinquent Loans

52

Separate the white swans from ugly ducklings

Worst caseJapan crisis

VaRCVar

Oil CrisisFinancial crisis

Oil spike rate rise

• Set riskthreshold• Macroperspective

Scorecard combiningscenario impacts and

hedge payouts

Scen

1

Scen

2

Scen

3

Scen

4

20% 30% 35% 15%

Rates

Equities

Scenarios

Severity weighting ofscenario

FX

Commodities

53

Hedge effectiveness differ over time

FTSE 100 volatility over timeEquity volatility surface

FORMATTING

0%

10%

20%

30%

40%

50%

60%

70%

19-Aug-2011

16-Sep-2011

21-Oct-2011

16-Dec-2011

16-Mar-2012

15-Jun-2012

21-Dec-2012

21-Jun-2013

20-Dec-2013

20-Jun-2014

19-Dec-2014

18-Dec-2015

15-Dec-2017

18-Dec-2020

50%

70%

90%

100%

110%

130%

Data source: Bloomberg L.P.

0

5

10

15

20

25

30

35

40

45

50

Aug-10 Oct-10 Dec-10 Feb-11 Apr-11 Jun-11 Aug-11

54

Summary

Insurers have advanced risk management approachesHowever, 1-in-200 year capital is a mirage

We won’t find all the tail risks, but• we might identify the white swans and mitigating those…• …may also indirectly reduce exposure to the black ones

55

Questions or comments?

Expressions of individual views bymembers of The Actuarial Professionand its staff are encouraged.The views expressed in this presentationare those of the presenter.


John Roe - The Sting in the tail · Peasants’ Revolt, 1381 Battle of Kosovo, 1389 24 46 Great...

Documents

Transcript of John Roe - The Sting in the tail · Peasants’ Revolt, 1381 Battle of Kosovo, 1389 24 46 Great...