John Roe - The Sting in the tail · Peasants’ Revolt, 1381 Battle of Kosovo, 1389 24 46 Great...
Transcript of 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
22© 2010 The Actuarial Profession � www.actuaries.org.uk
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
23© 2010 The Actuarial Profession � www.actuaries.org.uk
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
American CivilWar, 1861
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
American CivilWar, 1861
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
24© 2010 The Actuarial Profession � www.actuaries.org.uk
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
25© 2010 The Actuarial Profession � www.actuaries.org.uk
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
26© 2010 The Actuarial Profession � www.actuaries.org.uk
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
Source: Bloomberg LP
-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
Source: Bloomberg LP
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
34© 2010 The Actuarial Profession � www.actuaries.org.uk
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
35© 2010 The Actuarial Profession � www.actuaries.org.uk
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
43© 2010 The Actuarial Profession � www.actuaries.org.uk
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.
55© 2010 The Actuarial Profession � www.actuaries.org.uk