Dave Ingram, MAAA, FSA, CERA, FRM, PRM Chair of IAA Enterprise & Financial Risk Committee November, 2011 Stress and Scenario Testing in a Risk-based Solvency Regime 1
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Dave Ingram, MAAA, FSA, CERA, FRM, PRM Chair of IAA Enterprise & Financial Risk Committee November, 2011
Stress and Scenario Testing in a Risk-based Solvency Regime
1
DISCLAIMER
• This presentation represents the opinions of the presenter, David Ingram. These are not necessarily the views of the International Actuarial Association or his employer, Willis Re.
• In the area of Risk Management it is always dangerous to rely on the opinions and analysis of a third party without verifying with your own research, thought, and analysis.
• This presentation is no exception to that rule.
2
A word from our lawyers • Willis Re Inc. is a reinsurance broker. Willis Re Inc. is not a law firm,
investment advisor or tax advisor. We do not give legal, investment or tax advice and nothing herein constitutes nor should be construed as such. Such ideas are offered for discussion purposes only and do not constitute advice of any kind. It is believed that the information used in creating this presentation is correct, but no representations are made as to its completeness or accuracy, nor are any warranties made as to its fitness for any purpose. You and your advisors must make an independent assessment regarding all such matters.
• Any comments or observations made herein are for academic purposes only and are not for the purposes of reliance. Any such comments do not reflect the views of Willis Re Inc. or its clients.
3
Humanity
“The real problem with humanity is the following: We have Paleolithic emotions; medieval institutions; and god like technology.”
2009 Dr. E.O. Wilson, Harvard University
Source: The Watchman’s Rattle, Thinking Our Way Out of Extinction, Rebecca D. Costa, 2009 Vanguard Press, Kindle Edition
4
Low frequency / High Severity Earthquake + Tsunami + Nuclear
5
False probabilities & squandered resources
0
100,000
200,000
300,000
400,000
500,000
600,000
700,000
2009
Car CrashesHeart DiseaseTerrorismFirearms
Source: CDC 2007 Heart Disease & Firearms, FBI terrorism 2009, NHTSA 2009 Car crashes
Death
6
Decision Biases
Source: Wikipedia, Decision Making and behavioral biases
7
Additional types of decision biases
8
It is so darn frustrating • You note trends that reveal an
emerging uncertainty or you spot an emerging risk
• You can’t persuade anyone – Competitive market will not let you
condition the risk – The BOD is indifferent – The CFO’s vision is limited by the
needs of the quarter
9
As a consequence of the Financial Crisis
• Many different groups are coming to grips with this issue
• The only solution put forward so far is
10 10
Picture of a Stress Test
11 11
Stress Testing
With Stress Testing only one thing is different
IAJ November 2011 ERM NIRVANA 12
Scenarios
With Scenario testing, you imagine an entire world
!"#$%&'(() *"+,&-.$,/"#012!"#$%&3)4&'(()&
!"##$%&'(&)&$'*+'&,$'-*#./
0.&$#%)&$'(&)&$('*+'&,$'-*#./
!"#$%&'%#%()*+#,&-.
.$,/"#012&$"/&5,&2,,/&"2&6%17+%6&,89,#0:,/62&"5176&912205;,&<767#,&26"6,2&1<&6%,&=1#;>?&.$,/"#012&"#,&/16&<1#,$"2624&0/&6%"6&6%,@&/,,>&/16&9#,>0$6&6%,&<767#,&>,A,;19:,/64&576&#"6%,#&2%17;>&0;;7:0/"6,&912205;,&576&9,#%"92&,86#,:,&2067"601/2?&.$,/"#012&"#,&";21&>0<<,#,/6&<#1:&2,/2060A06@&"/";@202&=%,#,&6%,&0:9"$6&1<&"&B2:";;C&$%"/+,&1<&"&20/+;,&A"#0"5;,&02&,A";7"6,>?
()*+#,&-'%#,*%#+%*''*+$&#/%*/*0*+$%-1%,&'2%0#+#3*0*+$%#+4%,*35/#$-,6%,*75&,*0*+$'8%9"*6%#//-:%'*+&-,%0#+#3*0*+$%#+4%$"*%;-#,4%$-%0#2*%&+1-,0*4%4*)&'&-+'%#+4%#+%#+#/6'&'%-1%
$"*%&+'5,*,<'%,&'2%)-+)*+$,#$&-+'13
Types of Scenarios
Reverse Scenarios
Historical Scenarios
Synthetic Scenarios
Company Specific
Scenarios Single-Event
Scenarios Multi-Event Scenarios
Global Scenarios Top Down Stress (Balance Sheet) vs. Bottom Up
(transaction level)
14
Uses of Stress Tests
Emerging Risk Evaluation
Primary Risk Metric
Communicating Risk
Supervisors and Rating Agencies
Risk Capital Determination
15
STRESS TESTING AND EMERGING RISKS
16
Emerging Risk Process • Environmental Scanning
– To provide advance signals of potential Crisis developments • Process for Anticipating Emerging Risks
– Development of Emerging Risk Scenarios • Process for Envisioning Significance of Emerging Risks
– Stress Testing & Liquidity Risk Analysis • Process for Preparing Response to Emerging Risk Solutions
– Contingency Planning • Execution of Company in Emerging Risk Solutions
– Changes to company business and risk management practices • Company learning process from Emerging Risk Situation
Source: Standard & Poor’s
17
World Economic Forum Emerging Risks Scenarios
• Oil price shock • Fall in value of US$ • Chinese economic hard
landing • Demographic shift • Blow up in asset prices
Economic Environmental • Climate change • Freshwater loss • Tropical storms • Earthquakes • Inland flooding
18 18
World Economic Forum Emerging Risk Scenarios
• International terrorism • Weapons of mass
destruction • Interstate/civil wars • Failed states • Transnational crime • Globalization fallback • Regional instability
Geopolitical Societal • Pandemics • Infectious diseases • Chronic diseases • Liability regimes
Technological • Critical information
infrastructure • Nanotechnology
19 19
Emerging Risks Survey
Source: Society of Actuaries
20
!"#$%&' !()*+
• !,&-. /001
– 234 5-. ,&-#6 '$%#7
– 804 9.-:);6 #$)"<6
– 804 =.%( >, -" )''6;,&-#6'
• ?6#6:@6& /0A0
– 8B4 C).. -" D).>6 %EFGH
– 8I4 J";6&");-%").;6&&%&-':
– 8A4 9$-"6'6 6#%"%:-#$)&K .)"K-"<
– 804 5-. ,&-#6 '$%#7
B
20
Other Emerging Risks • Depression • Sovereign Default • Hyperinflation • Currency Crisis • Break up of the Euro
• Solar Weather
21
Emerging Risks Types of Tests
Types of Tests – Stress Tests – Reverse Scenarios – Macro Scenarios – Historical Scenarios – Synthetic Scenarios – Company Specific Scenarios – Narrowly Specified Scenarios – Broadly Specified Scenarios
22
More on Emerging Risks Process
• May want to consider setting action triggers depending upon future changes in stress test results – As well as on indicators of immanence
23
STRESS TESTS AS A PRIMARY RISK METRIC
24
Stress/Scenario Tests as Primary Metrics
• Want to measure risk over a long time scale – Initial situation is not important – Long term risk is not thought to change much – What if the immediate future is the next time that the risk event
hits?
• Don’t want to spend the money to do stochastic model – Business is not that volatile and/or not that important
• Have been using RBC/Rating Agency risk estimates – And this provides a first order improvement
• Want to include a risk that you do not know how to model frequency of major loss events
25
Primary Risk Metric
• Test is repeated regularly • Company sets limits and control process around
the Test outcome • Company looks at Test to determine
effectiveness of mitigation activity
• Beware of Unintended Consequences – Risk management tailored to the Test
26
Lloyd’s RDS
• Florida Windstorm • Gulf of Mexico Windstorm • European Windstorm • Japanese Windstorm • California Earthquake • New Madrid Earthquake • Japanese Earthquake • UK Flood • Terrorism
• Marine • Loss of Major Complex • Aviation Collision • Satellite Risks • Liability Risks • Political Risks
27 27
Lloyd’s RDS Terrorism Scenario
28 28
Secondary Risk Metric
• Stress tests are sometimes used by banks as a secondary risk metric – Will have stress test limits in addition to their
VaR limits
29
Primary Risk Metric
Reverse Scenarios
Historical Scenarios
Synthetic Scenarios
Company Specific
Scenarios Single-Event
Scenarios Multi-Event Scenarios
Global Scenarios Top Down Stress (Balance Sheet) vs. Bottom Up
(transaction level)
30
STRESS TESTS FOR COMMUNICATING WITH SENIOR MANAGEMENT AND BOARD
31
Problem with Statistical Approach
• Many people are not conversant with statistics “The average loss for stochastic scenarios with likelihood less than
1%”
– is just not something that resonates
• Stress Tests and especially Scenario Tests are stories – People all relate to stories
32
Stories
• Might include: – Reason for the choice of scenario
• Relevance of Scenario to the company
– Back story on the scenario – Description of the scenario – Description of Post event happenings – Other assumptions
Story must be believable!
33
Historical Scenarios
34
!"#$"##%
$%
&'()*+,-%./%0,1-234256%(17%08,1(92.%:1(+6-2-%
!"#$%&'()*$%+,-'-).'-/(&'"%+%-8,1(92.-%%;,<=<%05.8>%)(9>,5%89(-?%./%#!@AB%0,*5,)C,9%##5?%
DEE#B%FG+/%H(9B%DEEA"DEE@%0GC*92),%89,725%892-,-B<<<I%
.,0(/1#2"%+%-8,1(92.-%;7,4,+.*),15%./%*.5,13(+%-8,1(92.-%5?(5%8.G+7%
.88G9%%-G8?%(-%(%*(17,)28%I%
.,3&'4)-8,1(92.-%;?6*.5?,38(+%-8,1(92.-%21/.9),7%C6%?2-5.928(+%)(9>,5%
).4,-",4,15-%CG5%1.5%1,8,--(92+6%+21>,7%5.%(%-*,82J8%892-2-<I%
!#$-'25'/,)%$%+,-'-)K2,+7%LG94,%0?2M-%;,<=<%N"O%#"DP"PE"#EEC*I%
059,--%Q(5928,-%R%-2)G+5(1,.G-%).4,%21%92->%/(85.9%*928,"9(5,-%(17%
4.+(3+23,-%%
S(9=,%TG85G(3.1-%21%,'8?(1=,%9(5,-%
5
U%
Example: Most historical scenarios run by banks by type of assets (BCBS, 2005) BCBS= Basel Committee on Banking Supervision Asset type Historical scenarios Interest rates 1994 – bond market sell-off 1997 – Asian financial crisis 1998 – Combined Russian debt default and LTCM failure 2001 – 9/11 terrorist attacks in the U.S. 2003 – bond market sell-off Equities 1987 – October Black Monday
1997 – Asian financial crisis 2000 – bursting of the IT bubble 2001 - 9/11 terrorist attacks in the U.S.
FX 1992 – EMS (European Monetary System) crisis
1997 – Asian financial crisis 1998 - Russian debt default
Commodities 1973-1974 – Oil crisis Credit 1997 – Asian financial crisis
1998 – Combined Russian debt default and LTCM failure 2001 - 9/11 terrorist attacks in the U.S.
34
Synthetic Scenario Story
35
!"#$"##%
#&%
'()*+,(-(./0%12,%3%4225%6/,(00%67(.3,+2%
• '(8(93.7(%! :;(%0/,(00%07(.3,+2%0;2*85%<(%,(8(93./%/2%3.38=>(%/;(%7*,,(./%?20+@2.%+.%/;(%A,-B0%?2,C28+2%(D3-?8(%
• '(38+0@7%! :;(%0/,(00%07(.3,+2%0;2*85%<(%72.0+0/(./%E+/;%/;(%7*,,(./%(72.2-+7%(.9+,2.-(./F%
• G./(,.38%H2.0+0/(.7=%! :;(,(%0;2*85%<(%.2%+.72.0+0/(.7+(0%+.%/(,-0%21%/;(%-3,I(/%,+0I%0;27I0%+.%/;(%0/,(00%/(0/%J(F4F%.(43@9(%9283@8+@(0%2,%+./(,(0/%,3/(0KF%%
• L,3.*83,+/=%! :;(%0/,(00%07(.3,+2%0;2*85%<(%72.5*7/(5%E+/;%0*M7+(./%5(/3+8%/2%?,29+5(%-(3.+.41*8%,(0*8/0%3/%93,+2*0%0*<N?2,C28+2%8(9(80F%
31
%%%%%%67(.3,+2%O.38=0+0%%PD3-?8(Q%R+8%H,+0+0%67(.3,+2%O.38=0+0%%
• !"#$%"#&&'()*+&$'&+,--./&*(+0*%12$"+&
• 340*(5(&6()07(*&1$"8%2$"+&0*%##(*&)&9:;&%"1*()+(&%"&<*,8(&-*%1(+&=6%07&
)&7%#7&$'&>:;&)*$,"8&8)/&9:?&
• @A&*(.()+(+&$%.&'*$5&07(&+0*)0(#%1&*(+(*B(&
<$"+0*)%"(8&1)-)1%0/&)0&*(C"(*%(+&5%2#)0(+&07(&*(.%('&+$5(67)0&
• D(8&*)%+(+&*)0(+&0$&*(+-$"8&0$&-$0("2).&%"E)2$"&
F$6)*8+&07(&("8&$'&07(&G,)*0(*H&+)'(I7)B("&J,/%"#&$'&+7$*0I0(*5&
F*()+,*/&+(1,*%2(+&*(B(*+(+&+$5(&$'&07(&%5-)10&
• <*(8%0&A-*()8+&6%8("&+7)*-./&)+&07(&@A&(1$"$5/&8(0(*%$*)0(+&)"8&
8(1.%"(+&J/&K;&=L&J/&9::&J-+H&LLL&J/&KKMJ-+?&
• @A&N$..)*&#)%"+&5$8(*)0(./&=9;&B+O&PQRH&S;&B+O&3@TH&U;&B+O&<VN?&
N$..)*&+0*("#07("%"#&%+&.%5%0(8&J/&07(&-$++%J%.%0/&07)0&$%.&-*$8,1(*+&6%..&
+((W&0$&-*%1(&$%.&%"&3,*$+&*)07(*&07)"&N$..)*+&
• T)0(+&*%+(&5$8(+0./&6%07&)&E)X("%"#&J%)+&=9:J-+&%"&07(&.$"#&("8H&K::&
J-+&%"&07(&+7$*0&("8?&
Y"%2).&7%#7(*&%"1*()+(+&J(%"#&$Z+(0&J/&+)'(I7)B("&J,/%"#&%"&07(&.)0(*&
+0)#(&$'&07(&G,)*0(*&
32 35
Soverign Default Story
36
4
Scenario: Narrative
Central banks and finance ministers, trying to support banks that heavily invested in defaulting sovereigns and that have to satisfy the more strict Basel III liquidity rules, push for regulation that make the transfer of high-quality assets from insurers and pension funds to banks easier.
The economies in the EU struggle, as some sovereigns are being bailed out and taxes rise. Central banks keep interest rates low, and investors are hunting for yields. Rising commodity and oil prices fuel a bubble. Analysts and business journalists insist that this time it is different since the booming Asian economies and declining oil reserves imply ever-rising prices.
The Build-Up 2012
European and US governments further fuel the bubble by supporting industries that compete against Asian resource extraction firms in Africa.
Investing in any commodity related assets becomes artificially cheaper, as they are being seen as having government support and are becoming national champions. 7
Scenario: Narrative
8
A number of banks default that were heavily exposed to speculative investments, impacting directly those insurers that lent liquid bonds against illiquid collateral
Insurers holding illiquid assets are savagely punished, since the market does not know the value of their portfolios and their share prices collapse..
In 2015, the commodity bubble bursts, the Asian and African economies which propped up the financial markets go into recession and a flight to safety results.
The Crash 2015
Speculators and day traders go into commodities, throwing further gasoline into the bubble. Africa is becoming the new China and is taking up the slack of Asia, that sees growth declining.
The Mania 2013-2014 Some insurers are buying up long-dated bonds issued by start-ups that try to benefit from the commodities bubble. They outperform their more conservative competitors. The CEOs of the underperforming insurers are replaced and exposures to the speculative bubble in the insurance market increases
36
Soverign Default Story
37
4
Scenario: Narrative
Central banks and finance ministers, trying to support banks that heavily invested in defaulting sovereigns and that have to satisfy the more strict Basel III liquidity rules, push for regulation that make the transfer of high-quality assets from insurers and pension funds to banks easier.
The economies in the EU struggle, as some sovereigns are being bailed out and taxes rise. Central banks keep interest rates low, and investors are hunting for yields. Rising commodity and oil prices fuel a bubble. Analysts and business journalists insist that this time it is different since the booming Asian economies and declining oil reserves imply ever-rising prices.
The Build-Up 2012
European and US governments further fuel the bubble by supporting industries that compete against Asian resource extraction firms in Africa.
Investing in any commodity related assets becomes artificially cheaper, as they are being seen as having government support and are becoming national champions. 7
Scenario: Narrative
8
A number of banks default that were heavily exposed to speculative investments, impacting directly those insurers that lent liquid bonds against illiquid collateral
Insurers holding illiquid assets are savagely punished, since the market does not know the value of their portfolios and their share prices collapse..
In 2015, the commodity bubble bursts, the Asian and African economies which propped up the financial markets go into recession and a flight to safety results.
The Crash 2015
Speculators and day traders go into commodities, throwing further gasoline into the bubble. Africa is becoming the new China and is taking up the slack of Asia, that sees growth declining.
The Mania 2013-2014 Some insurers are buying up long-dated bonds issued by start-ups that try to benefit from the commodities bubble. They outperform their more conservative competitors. The CEOs of the underperforming insurers are replaced and exposures to the speculative bubble in the insurance market increases
37
Learning from Scenarios
• If you have other risk metrics, should regularly include them along with the scenarios – Shift discussion emphasis over time if other measures
are primary
• Be careful not use scenarios with the Board that are not a regular part of your risk management – They may correctly reach very different conclusions
about risk decisions
38
Black Swan Calibration
For communicating results of Reverse Stress Tests
• 1 Black Swan (BLS) = Worst Historical Scenario
• So a company could report that their failure point was: – 1.5 BLS for Equities; 2.5 BLS for Hurricane – 3.0 BLS for Credit; 3.5 BLS for Pandemic
39
Communicating Risk
Reverse Scenarios
Historical Scenarios
Synthetic Scenarios
Company Specific
Scenarios Single-Event
Scenarios Multi-Event Scenarios
Global Scenarios Top Down Stress (Balance Sheet) vs. Bottom Up
(transaction level)
40
STRESS TESTS FOR REGULATORS AND RATING AGENCIES
41
Recent Developments
• US and European Bank Stress Tests • European Insurer Stress Tests • IMF / IAA Stress Testing Workshop • AM Best SRQ
42
AM Best 2011 SRQ
• Requires Stress Tests for: – Market Risk – Underwriting Risk – Credit Risk – Operational Risk – Liquidity Risk – Strategic Risk – Inflation – Interest Rate spike
43 43
44 44
Outcome of EIOPA Stress Tests
45
!!"
#$$%&$'(&")&*+,(*""
"" !""#$"%&$'()*"+#$,
( *-(./0(1*22*3- *-(4 (
-,.$./,&"0'1.(',"/&23%&"*(%&** 455 " (
6.7.8+8")&9+.%&:";'1.(',"<6;)="/&23%&"*(%&** !4>
" (
?3,@&70A"?+%1,+*" B>4
" (
6;)"03@&%'$&"%'(.3"" "
CDEF (
6;)"03@&%'$&"%'(.3".7"/'*&,.7&"*0&7'%.3 " C>EF (
6;)"03@&%'$&"%'(.3".7"':@&%*&"*0&7'%.3 " >D!F (
6;)"03@&%'$&"%'(.3".7".72,'(.37"*0&7'%.3 " CB>F (
567%8&(3-($2*"*12$((8%7*&%2( *-(./0(1*22*3-( %,(%(7$#8$-&%"$(39($2*"*12$(8%7*&%2
:32;$-8<('$9*8*&(39(+-'$#&%=*-",((-3&(6$$&*-"(
&>$(?@0((A*-(./0(1*22*3-B
G'*&,.7&"*0&7'%.3 H> !4IH >IJ
#:@&%*&"*0&7'%.3 !4E >JIE BIB
K72,'(.37"*0&7'%.3 4D !EIE >I4
?3@&%&.$7"?(%&**"?0&7'%.3 CC 4IJ CIB
Capital
45
Reactions
46 46
Another Reaction
47 47
NAIC on Stress Tests • Utilizing the extensive NAIC database, the NAIC can perform stress
tests on both micro and macro levels. At present, insurance companies in the U.S. are not required to perform nor report stress test results to the regulators.
• The G20 has stated, “We commit to conduct robust, transparent stress tests as needed.”
• IAIS Insurance Core Principles, upon which supervisory regimes are assessed in the Financial Sector Assessment Program (FSAP), state, “The supervisory authority requires insurers to recognise the range of risks that they face and assess and manage them effectively.” There is also the following advanced criteria: “The supervisory authority requires that insurers undertake regular stress testing for a range of adverse scenarios in order to assess the adequacy of capital resources in case technical provisions have to be increased.”
48 48
NAIC Consultation Questions:
• What stress tests should be performed by the NAIC? • What stress tests and reverse stress tests should be
performed by companies? What should be required to be reported to the regulator?
• Should the regulator specify stress test scenarios to run? If so, which ones? How often should they be done?
49 49
International Monetary Fund
50
!"#$%&&'()*!"#'#+,-+.)#/,$,0.1#1"/,(2"&1"34#'%3+2(,3,1$4#,"#"115#26#(17%3'3,2"'.#022$34#'77.)#$3(1$$#31$3$#32#$13#8'7,3'.#(19%,(1&1"3$4#(1$3(',","-#3+1#%$1#26#$3(1$$#31$3,"-#'$#'#/%."1('0,.,3,1$#$%(/1,..'"81#322.:!"$%('"81#,"5%$3()#,$#2/1(;+1.&15#;,3+#"1;#'"5#$2&13,&1$#%"822(5,"'315#$3(1$$#31$3#(19%,(1&1"3$:#<+1#$)$31&,8#'"5#8(2$$02(51(#,$$%1$#'"5#,&7.,8'3,2"$#26#3+1#$2.%3,2"$#62%"5#62(#3+1#0'"=,"-#$1832(#+'/1#'#31"51"8)#32#01#3('"$61((15#32#3+1#,"$%('"81#$1832(#;,3+2%3#5%1#(1-'(5#26#5,661(1"3,'3,2"#%"51(#+,-+#3,&1#7(1$$%(1:
>1'51($+,7#(2.1#,"#'55(1$$,"-#$)$31&,8#'"5#8(2$$02(51(#,$$%1$#,"#3+1#0'"=,"-#$1832(:<+1(1#,$#'"#1?7183'3,2"#32#$11#3+1#!@A#1?7'"5,"-#3+,$#.1'51($+,7#32#3+1#,"$%('"81#'"5#23+1(#6,"'"8,'.#$1832($:B1/1.27#&'8(2182"2&,8#322.$#32#'55(1$$#$)$31&,8#(,$=#,"#3+1#;+2.1#6,"'"8,'.#$1832(4#,"8.%5,"-#,"$%('"81:C2,"3#2%3##72$$,0.1#62(&$#26#(1-%.'32()#'(0,3('-1:
50
IMF Proposal
51
!"#"$%&'(')*+"))'*")*',%+'-$%.($'(/*0#"'01)2+(1/"'-+%2&)'()'('*%%$'*%'())"))'+"-2$(*%+3'()'4"$$'()'%&"+(*0%1($'()&"/*)'%,'/+%)).%+5"+'01)2+(1/"'(/*0#0*3'
40*6'*6"'2$*07(*"'"8&"/*"5'%2*/%7"*%'07&+%#"')2+#"0$$(1/"'%,'01*"+1(*0%1($'(/*0#"'$(+-"'01)2+"+)'(15'+"01)2+"+)(#%0501-'52&$0/(*0%1'(15'(&&$301-'*6"'+0-6*'$"#"$'%,'+"-2$(*0%1'(15')2&"+#0)0%19
:6"')*+"))'*")*'40$$'."'2)"5'*%;<))"))'*6"'07&(/*'%,'/(&0*($'7%.0$0*3'01'*07")'%,',01(1/0($')*+"))<))"))7"1*'%,'-+%2&)= 01*"+1($'7%5"$)'(15'>2($0*3'%,'+0)?'7(1(-"7"1*@%1*+0.2*"'*%'*6"'(1($3)0)'%,'/%1*(-0%1'40*601'(15'(7%1-'-+%2&)A1*"+'*6"'1""5"5'B%C
51
IMF Proposal
52
!"#"$%&'(')*+"))'*")*',%+'-$%.($'(/*0#"'01)2+(1/"'-+%2&)'()'('*%%$'*%'())"))'+"-2$(*%+3'()'4"$$'()'%&"+(*0%1($'()&"/*)'%,'/+%)).%+5"+'01)2+(1/"'(/*0#0*3'
40*6'*6"'2$*07(*"'"8&"/*"5'%2*/%7"*%'07&+%#"')2+#"0$$(1/"'%,'01*"+1(*0%1($'(/*0#"'$(+-"'01)2+"+)'(15'+"01)2+"+)(#%0501-'52&$0/(*0%1'(15'(&&$301-'*6"'+0-6*'$"#"$'%,'+"-2$(*0%1'(15')2&"+#0)0%19
:6"')*+"))'*")*'40$$'."'2)"5'*%;<))"))'*6"'07&(/*'%,'/(&0*($'7%.0$0*3'01'*07")'%,',01(1/0($')*+"))<))"))7"1*'%,'-+%2&)= 01*"+1($'7%5"$)'(15'>2($0*3'%,'+0)?'7(1(-"7"1*@%1*+0.2*"'*%'*6"'(1($3)0)'%,'/%1*(-0%1'40*601'(15'(7%1-'-+%2&)A1*"+'*6"'1""5"5'B%C
52
IMF Next Steps
53
!"#$%&'()'*)$()&(*'"+$,)-"'./)'*)#0/#"%1)2#3#,'/)%4#)+#%4'5','-6)%4$%)&1)$//,&7$8,#)%')1%"#11)%#1%&(-)-,'8$,)-"'./19($8,#)$(5)/"'+'%#)%4#).1#)'*)%4#)1%"#11)%#1%1)*'")-,'8$,)-"'./1)$%)%4#)1./#"3&1'"6)7',,#$-.#1),#3#,:'(&%'")%4#)/"'/#").1#)$(5)7'(1%$(%)./5$%&(-)'*)%4#)%'',1;
53
IMF Insurance Stress Tests 2010
Isle of Gurnsey FSAP • The insurance sector shocks have been
designed to parallel the shocks carried out for the banks, – Market risk tests - sudden changes in interest rates, FX rates and asset
prices. For the interest rate tests, a parallel shift of the pound sterling yield curve has been simulated. Likewise, the test for foreign exchange rate shocks comprised all sterling pound tests run for the banks.
– Recommends simulate the impact of actual losses at maximum historical levels within a scenario type test or to double the worst provision levels observed in the past.
54 54
IMF Stress Test • Mortality: a permanent increase in mortality rates of insured lives
(i.e., mortality over normal prevailing rates) of 25 percent. This shock should only be applied to insured lives (i.e. not to annuitants).
• Morbidity: a permanent increase in morbidity rates (i.e., morbidity over normal prevailing rates) of 25 percent.
• Longevity: a permanent reduction in annuitant mortality of 25 percent across all current and deferred annuitants. This shock should only be applied to annuitants, i.e., not to insured lives.
• Persistency: 30 Percent of policyholders where there is a right to surrender, surrendered their contracts at the date of the investigation.
• asset prices, a potential fall of property prices by 20 percent.
55 55
Canadian DCAT
!
!"#$%&'()*+,-#'"(.-/0
" !"#"$%&"'()*(+,"(-./.'0./(1/2+0+3+"(%4(56+3.70"2(8-159(0/(+,"(&"70%'(:;<=>:;;?
" @"A307"'(.//3.$(+"2+(4%7(.$$(-./.'0./($04"(0/237"72(20/6"(:;;B
" @"A307"'(%4(CD-(8E"/"7.$(0/237./6"9(6%F&./0"2(20/6"(:;;<
56 56
Canadian DCAT
57
!"
!"#$%&'$#(#&) *$+,&-.#/0,0#
# !"#$%&'$(# !"#)'*'$(# +,#-'-$,./(# 0%-123&"425'-5%$/1# 6,$,#'"#%$'".2"32%--,$27%&8,-29/#,*'$2#'-:;
# <,42)8-'.,--# =>?,.-,# @,'.-8#%./,# A"7,#.5,.$2%.*2?"&'$'/%&2%/$'".
# B332)%&%./,2-1,,$
57
Canadian DCAT
58
!"
!"#$%&'$#(#&) *+,+-".&/,#0-+-#
# !"#$%#&'()*&+),#-#"./(
# 0".'.&1# 2.,#,/.3*/.4& 45)647.'()7.*8.7./.#,
# 9&57*/.4&# 9&/#"#,/)"*/#
# 0"#3.%3)-47%3## :;6#&,## <#.&,%"*&'## =#/#".4"*/.4&)45)*,,#/)-*7%#,)>'"#+./)".,?@
# A4-#"&3#&/)*&+)647./.'*7)*'/.4&
# B55)8*7*&'#),C##/
58
Canadian DCAT
59
!"
!"#"$%&'()*$"'+,&-*
# !"#$%#&$'"()*%+#+",,-%.$()/$.%0(*1%23$%)'#+($(4#%5+#)'$##%6,"'%7 6(*/).$#%"%#2"'."(.%*0%&*16"()#*'
# 8*.$,%1"'"9$1$'2%($"&2)*'%2*%&3"'9$#%)'%$:6$()$'&$$ %&'(&)*+,-+.')/01-'-2+,
# ;*'#).$(%($9+,"2*(-%($"&2)*'
59
Canadian DCAT
!"
!"#$%&'()*+,-$.#/0*'
# !"#$%&$'(&)(*$+")#(,-+-.$,$+&(-+'(/)-#'*()0('"#$%&)#*$ %&'()*+,'-.'(+/'0(-/1(./-0,01(/2().1*)3,'04('51'3-)6(7/)3+(8'87'30(8)9(-/1(&):'()(0/*-+(*-+'301)-+,-;(/2(1&'(,-0*3)-.'(7*0,-'00
# 1)+2$3(,$-+"+.045("+0)#,-&")+("+(-(,-++$#(&6-&("*("+&$#$*&"+.(&)(&6$(#$-'$#
60 60
Canadian DCAT Example: Recession and Stagflation scenario
• Driven by rapid rise in price of oil caused by industrialization of India and China
• US$ falls 25% over five years due to budget and trade deficits
• Inflation driven upwards 300 points in one year by lower dollar and oil price
• Interest rates follow suit
From 2006 DCAT Report 61
62
ORSA October 2011
• International Association of Insurance Supervisors agree to standards – ERM – ORSA
• NAIC (US) had exposed those standards for comment – Plans to implement the international standards
• Financial Stability Board – Requires all G20 countries to adopt – And self assess regarding implementation by 2012
62
63
ORSA standard
• Own Risk and Solvency Assessment – An opinion on the adequacy of the adequacy
of the ERM system and the Capital of the firm – Made by the management and the board
63
64
ORSA standard
• Three elements – Measurement of resources needed for
solvency – Determination of capital quality – Determination of ERM adequacy
• Conclusion: Statement of Capital & ERM Adequacy
64
Regulatory and Rating Agency
Reverse Scenarios
Historical Scenarios
Synthetic Scenarios
Company Specific
Scenarios Single-Event
Scenarios Multi-Event Scenarios
Global Scenarios Top Down Stress (Balance Sheet) vs. Bottom Up
(transaction level)
65
Developing Stress Tests
9/13/11
3
!"#$!%%&'()*++&,*+-./&01.234&53(*+&
• 61.7+&23&"1(8)9(:&,)1.+;3)41-3.&<&*8=>*49+4&;3)&2)1+-?&28)1-3.&49+41(?>&&
– !+&(>1(&1&@1A&3)&1&;*1(8)*&3;&B1.7+&
• 5**2&(3&23&+948C(1.*38+&+()*++&(*+-./&<&1CC&3;&(>*&?34=1.9*+&1(&3.?*&– %CC3A+&?34=1)9+3.&3.&(>*&+14*&B1+9+&<&+14*&=39.(&9.&-4*D&
• 5*E(&41)7*(&)*=3)(&;)34&(>*&!%!'&A9CC&9.?C82*&+()*++&(*+(&9.;3)41-3.&• 6*+(&+()*++&(*+(&A38C2&B*&+?*.1)93+&3;&B3(>&1++*(&1.2&C91B9C9(:&+()*++*+&(>1(&
A*)*&?33)29.1(*2&A9(>&*1?>&3(>*)&• !.+8)1.?*&/)38=+&23&B8+9.*++&(>1(&?1.&B*&F(&9.(3&G&B8?7*(+H&
– ,)129-3.1C&9.+8)1.?*&I.2*)A)9-./&– 53.J,)129-3.1C&I.2*)A)9-./&– '*49&,)129-3.1C&8.2*)A)9-./&&
– 53.&!.+8)1.?*&
• !(&9+&=3++9BC*&(>1(&(>*&.3.J()129-3.1C&1.2&.3.J9.+8)1.?*&1?-K9-*+&1)*&+:+(*49?1CC:&94=3)(1.(&L1+&(>*:&B*?14*&;3)&%!MN&
5
• Traditional Investment • Non-Traditional Investment • Semi-Traditional Investment
61+*C&M8921.?*&3.&'()*++&,*+-./&L0*+(1(*2&;3)&!.+8)*)+N&
• Actionable – Impacting decision making
• Range of Perspectives • Documented • Flexibly Stress Testing
System • Reassess framework
regularly • Firm Wide Risk
Assessment • Interactions & Feedback
effects • Include stress to major
counterparties
• Range of Severities • Challenge basic business
assumptions (i.e. availability of funds)
• Challenge effectiveness of mitigations
• Model the actual workings of the underlying, not just the recently experienced volatility
• Include things in transition – especially include possibility that multiple step transaction will not be completed.
• Include impact of reputation risk – perhaps as a multiplier on other problems
6
66
S&P – Example of Stress Calibration
67 67
Concern with Supervisor use of Stress Tests from Insurers
• Move towards publicly disclosed stress tests – They become bright line – May overwhelm all other uses of stress tests
• At least under that name
68 68
USING STRESS TESTS FOR COMPANY DETERMINATION OF RISK CAPITAL NEEDED
69
70
Risk & solvency models
Methods
Standard factors USP Models Stress tests Simplified stochastic
Partial internal models
Full internal models
Multi-year modeling for ORSA 70
Using Stress Tests For Risk Capital
• Allows company to include detailed information about the retained risks – That standard factors do not capture
• Much easier to validate than stochastic model
• Much easier to communicate than stochastic model
71
Use of Stress Tests for USP Models
• Undertaking Specified Parameter models – Factors can be determined by analysis of
historical experience – Or with stress test results
• Careful picking parameter • Requires Quality control of risk buckets
72
Use of Stress Tests with Stochastic Models
• Calibration and validation of stochastic models – Generally involves looking at the likelihood of the loss from a scenario
test compared to • Actual historical frequency • Perceived likelihood
– As a “sense check” • Does it make sense that the equity model says that a 2009 result
(historical scenario) is a 1 in 300 year scenario? • Develop Historical Scenarios to show losses with current exposures
to compare with stochastic model
– Determine likelihood estimates from the stochastic model of the losses from a set of scenario tests
• Every time there is a change in the stochastic model or model parameters
– Provides a test of how much the model is changing
73
Risk Capital
Reverse Scenarios
Historical Scenarios
Synthetic Scenarios
Company Specific
Scenarios Single-Event
Scenarios Multi-Event Scenarios
Global Scenarios Top Down Stress (Balance Sheet) vs. Bottom Up
(transaction level)
74
Conclusions
• Stress and Scenario Tests have many valuable uses
• Supervisors can leverage all of these – Care must be taken not to overuse
75
Types of Scenarios
Reverse Scenarios
Historical Scenarios
Synthetic Scenarios
Company Specific
Scenarios Single-Event
Scenarios Multi-Event Scenarios
Global Scenarios Top Down Stress (Balance Sheet) vs. Bottom Up
(transaction level)
77
Sources
1. IAA Solvency Committee Draft Paper on Stress Testing (In process by Philipp Keller and Alan Brender)
78
Related Documents
