RISK MANAGEMENT

PROFESSIONAL CERTIFICATE

DAY 1
Introduction to Risk Management
 Module 1
Introduction

A FINANCIAL TIMES COMPANY 2

RISK

“The possibility that something bad

or unpleasant (such as an injury or a
loss) will happen”

Merriam-Webster

“. . . any event or action that may

adversely affect an organization’s ability
to achieve its objectives and execute its
strategies or, alternatively, the
quantifiable likelihood of loss or
less-than-expected returns.”

McNeill, Frey, Embrechts, Quantitative

Risk Management, 2005.
A FINANCIAL TIMES COMPANY 3
RISK MANAGEMENT

“Risk management is the discipline

that clearly shows management the S&P/Case-Shiller 10-City Composite Home Price Index©

risks and returns of every major 240

strategic decision at both the 220

200
institutional level and the
180
transaction level. Moreover, the risk

(Index Jan 2000=100)

160
management discipline shows how to
140
change strategy in order to bring the 120
risk return trade-off into line with 100

the best long- and short-term 80

interests of the institution.” 60

1990 1995 2000 2005 2010 2015

Source: S&P Dow Jones Indices LLC

[Link]
Van Deventer, Imai and Mesler,
Advanced Financial Risk Management,
2013.
A FINANCIAL TIMES COMPANY 4
THE STANDARD MODEL

I Perfect capital markets (Modigliani and Miller, 1958)

I No transaction costs
I No taxes
I No indivisible assets
I Borrowing and lending at the same rate of interest
I Unlimited short-sales of assets
I No informational asymmetries
I Financial agents are price takers
I Implications
I Capital structure of the firm has no bearing on its value
I Firm’s dividend policy has no bearing on its value
I There is no ‘room’ for risk management
I ‘Costless’ risk management does not change firm value
I Costly risk management reduces firm value!

A FINANCIAL TIMES COMPANY 5

THE STANDARD MODEL: DEBT AND EQUITY

B(T )
I Assets, A(t)
I Debt, B(t): D
Zero coupon bond, face value,
D, maturity, T
A(T )
B (T ) = min [D, A(T )] D
I Equity, S(t):
S(T )
S (T ) = max [0, A(T ) − D]

I Balance Sheet:

A(t) = S(t) + B(t)

A(T )
D
A FINANCIAL TIMES COMPANY 6
 16 2. Economic Principles of Risk Management

RATIONALE FOR RISK MANAGEMENT protecting shareholder equity value from the losses represented directly
by changes in market prices, but is rather for the purpose of reducing
the frictional costs that are sometimes associated with changes in market
value, such as financial distress costs. Shareholders can, on their own, adjust
their overall exposures to market risks, one of the key points of the theory
of Modigliani and Miller (1958) that, to some degree, holds true even in
imperfect capital markets.

2.2.1. Profit-Loss Asymmetries

An operating loss of a given amount x may reduce the market value of
the firm by an amount that is greater than the increase in market value
“ . . . one has to turn the Modigliani-Miller Theorem upside down and
caused by an operating gain of the same size x . An obvious example of
this asymmetry arises from taxation. With a progressive tax schedule, as
identify situations where risk management enhances the value of a firm
illustrated in an exaggerated form in Figure 2.2., the expected after-tax
profit generated by equally likely before-tax earnings outcomes of X 1 =
X + x and X 2 = X − x is less than the after-tax profit associated with the
by deviating from the unrealistically strong assumptions of the theorem.”
average level X of before-tax earnings. Reducing risk therefore increases
the market value of the firm merely by reducing the expected present value
of its tax liability. On average, with a progressive tax scheme such as that
McNeill, Frey, Embrechts, Quantitative Risk Management, 2005.
illustrated in Figure 2.2., a firm would prefer to have a before-tax profit of
X for sure than uncertain profits with a mean of X .

After-Tax Profit

I Taxes: Risk management can

reduce the costs of progressive
X2 X X1
taxes by reducing the variance of Before-Tax Profit

before-tax profit

Figure 2.2. Concavity effect of a progressive tax schedule.

A FINANCIAL TIMES COMPANY 7

RATIONALE FOR RISK MANAGEMENT
18 2. Economic Principles of Risk Management

I Bankruptcy Costs: Risk

management can reduce the
costs of financial distress
(bankruptcy), by reducing the
likelihood of distress Figure 2.3. Financial distress and trading profit and loss.
Source: Duffie, Singleton, Credit Risk, 2003.

I Simple model: Fixed B(T )

hold an effective option on the total market value of the firm. For purposes
of this simple illustration, we assume that debtholders exercise a protec-
bankruptcy costs, K , are tive covenant that causes liquidation of the firm if its total market value
falls below the total principal K of the outstanding debt, in which case
incurred when firm defaults the liquidation value of the firm goes to creditors. Under this absolute-
D
priority rule, the schedule relating the liquidation value of the net assets
of the firm to the liquidation value of equity is convex, as illustrated in
Figure 2.4. Jensen’s inequality therefore operates in a direction opposite
D −K
to that illustrated in Figures 2.2. and 2.3. Equity shareholders may actually
prefer to increase the risk of the firm, perhaps by substituting low-risk po-
sitions with high-risk positions or by increasing leverage. In the corporate-
finance literature, this is called asset substitution. Unless restricted by other
A(T )
debt covenants or by regulation, equity shareholders can play a “heads-I-
K D
win, tails-I-don’t-lose” strategy of increasing risk in order to increase the
market value of their share of the total value of the firm. This effect is il-
A FINANCIAL TIMES COMPANY lustrated in Figure 2.5, which shows the market value of equity as an option 8
on net assets struck at the liability level K for two levels of asset volatility,
RATIONALE FOR RISK MANAGEMENT

20 2. Economic Principles of Risk Management

I Asymmetric Information: Risk
management can reduce the
impact of costly external
funding (as a result of
information asymmetries) by
reducing the likelihood of
seeking external funding
I Stakeholder Conflicts: Risk
management may mitigate
conflicts of interest between
shareholders and bondholders
that increase firm value
Source: Figure Exposure ofCredit
[Link],
Duffie, market value
Risk,of2003.
equity to low and high risks.

2.2.3. Principal-Agent Effects

A FINANCIAL TIMES COMPANY 9
 50.0%

0.0%
1 13 25 37 49 61 73 85 97 109 121 133 145 157 169 181 193 205 217 229 241 253 265 277 289
Month Number
MANAGEMENT VS MEASUREMENT
EXHIBIT I.1 Feeling Homesick?

(U.S. Jan. 2000 ! 100, Japan: Dec. 1985 ! 100) Futures

260

240 U.S. 10 Cities Composite Home Price Index

220 Japan: Tokyo Area Condo Price1

Composite
200 Index Futures
“I think risk manager is a misnomer.
180
I dont manage the risks, its up to 160

the businesses to manage the risks. 140

Japan: Osaka Area Condo Price1
What we’re here to do is to provide 120

100
assurance that risks are being 80
monitored and managed.” 60

40
92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 US
77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 Japan
Kate Boothroyd, Treasury and Risk 1
Per m2, 5-month moving average.
Management, 2004. EXHIBIT I.2 U.S. and Japan—Collapse in Home Prices
Sources:
Source: Bloomberg, Real Estate Economic
Van Deventer, Institute,
Imai and Japan, S&P,
Mesler, S&P/Case Shiller
Advanced
s
Financial HomeRisk
Price Indices,
asManagement,
of October 5, 2011.2013.

A FINANCIAL TIMES COMPANY 10

RISK MANAGEMENT: GOLDMAN SACHS

Risk Mismanagement, Joe Nocera, NY Times, January 2. 2009:

How then do we account for that story that made the rounds in
the summer of 2007? It concerns Goldman Sachs, the one Wall
Street firm that was not, at that time, taking a hit for billions
of dollars of suddenly devalued mortgage-backed securities.
Goldman had somehow sidestepped the disaster that had
befallen everyone else. . . . in December 2006, Goldman’s
various indicators, including VaR and other risk models, began
suggesting that something was wrong.

A FINANCIAL TIMES COMPANY 11

RISK MANAGEMENT: GOLDMAN SACHS

Risk Mismanagement, Joe Nocera, NY Times, January 2. 2009:

“We look at the P&L of our businesses every day,” said

Goldman Sachs’ chief financial officer, David Viniar, . . . “We
have lots of models here . . . , but none are more important than
the P&L, and we check every day to make sure our P&L is
consistent with where our risk models say it should be. In
December our mortgage business lost money for 10 days in a
row. It wasnt a lot of money, but by the 10th day we thought
that we should sit down and talk about it.”

A FINANCIAL TIMES COMPANY 12

RISK MANAGEMENT: GOLDMAN SACHS

Risk Mismanagement, Joe Nocera, NY Times, January 2. 2009:

So Goldman called a meeting of about 15 people, including

several risk managers and the senior people on the various
trading desks. They examined . . . every trading position the
firm held. . . . They examined their VaR numbers and their
other risk models. They talked about how the mortgage-backed
securities market “felt.” “Our guys said that it felt like it was
going to get worse before it got better,” Viniar recalled. “So we
made a decision: lets get closer to home.”

A FINANCIAL TIMES COMPANY 13

RISK MANAGEMENT: GOLDMAN SACHS

Risk Mismanagement, Joe Nocera, NY Times, January 2. 2009:

In trading parlance,“getting closer to home” means reining in

the risk, which in this case meant either getting rid of the
mortgage-backed securities or hedging the positions so that if
they declined in value, the hedges would counteract the loss
with an equivalent gain. Goldman did both. And that’s why,
back in the summer of 2007, Goldman Sachs avoided the pain
that was being suffered by Bear Stearns, Merrill Lynch, Lehman
Brothers and the rest of Wall Street.

A FINANCIAL TIMES COMPANY 14

 Module 2
A Taxonomy of
Risks

A FINANCIAL TIMES COMPANY 15

RISKS
I Market risk
I Risk of unexpected changes in prices, rates and volatilities
I Credit risk
I Risk of changes in value associated with unexpected changes in
credit quality
I Liquidity risk
I Risk that the costs of adjusting financial positions will increase
substantially or that a firm will lose access to financing
I Operational risk
I Risk of fraud, systems failures, trading errors (e.g., deal mispricing),
and internal organizational risks.
I Systemic risk
I Risk of breakdowns in marketwide liquidity or ‘knock-on’ default

A FINANCIAL TIMES COMPANY 16

CREDIT RISK
I Risk that a party to a (financial) contract violates one or more
contractual obligations
I Obligation is typically a payment obligation
I Violation may be voluntary (by choice, e.g. ‘strategic’ mortgage
defaults) or involuntary
I Violation = Default
I Remedies to default include foreclosure, bankruptcy
I Best response to default may be to do nothing (forgive) or very little!
I Default is referred to as a ‘credit event’ in credit derivatives markets
I Credit risk is ubiquitous: corporate and sovereign bonds/loans,
mortgages, consumer credit, counterparty exposure in derivative
contracts, etc.
Credit risk can have profound and often unexpected effects on the
value of securities
A FINANCIAL TIMES COMPANY 17
 The Credit Default Swap 87
CREDIT EVENTS
Table 5.1 Description of the most commonly used credit events

Credit event Hard or soft Description

Bankruptcy Hard Corporate becomes insolvent or is unable to pay its

debts. The bankruptcy event is not relevant for
sovereign issuers.
Failure to pay Hard Failure of the reference entity to make due
payments, taking into account some grace period.
Obligation acceleration Hard Obligations have become due and payable earlier
than they would have been due to default or other
and have been accelerated. This event is used mostly
in certain emerging market contracts.
Obligation default Hard Obligations have become due and payable prior to
maturity. This event is hardly ever used.
Repudiation/moratorium Hard A reference entity or government authority rejects or
challenges the validity of the obligations. Used in
emerging market sovereign CDS.
Restructuring Soft Changes in the debt obligations of the reference
creditor but excluding those that are not associated
with credit deterioration such as a renegotiation of
more favourable terms.

ISDA Credit Events

A FINANCIAL TIMES COMPANY 18
 The drivers
CREDIT of Credit Losses
MODELING

Probability of Default (PD) Loss Given Default (LGD)

Will an asset become a What proportion of the value of

defaulted asset? a defaulted asset will we lose?

Exposure At Default (EAD) Maturity

What is the expected value of The effective remaining term of

the defaulted asset at the time a facility.
of default?

Source: Abesida (2012) Corporate Governance and Risk Management, GARP.

A FINANCIAL TIMES COMPANY 19

APPROACHES TO CREDIT MODELING

1 2 3 4 5 7 10
I Empirical Aaa
Aa
0.000
0.022
0.013
0.069
0.013
0.139
0.037
0.256
0.106
0.383
0.247
0.621
0.503
0.922
A 0.063 0.203 0.414 0.625 0.870 1.441 2.480
I Rating agencies Baa 0.177 0.495 0.894 1.369 1.877 2.927 4.740
I Historical default rates Ba 1.112 3.083 5.424 7.934 10.189 14.117 19.708
B 4.051 9.608 15.216 20.134 24.613 32.747 41.947
I Rating transition frequencies Caa-C 16.448 27.867 36.908 44.128 50.366 58.302 69.483

I Structural Cumulative average default rates, 1970-2012, Moodys

I Option theoretic
I Debt and equity are 17.5
Moody's Seasoned Baa Corporate Bond Yield©

derivatives on firm assets 15.0

I Reduced form (Percent)

12.5

10.0
I Default probabilities implied
7.5
by market prices
I No-arbitrage (almost 5.0

atheoretical) approach 2.5

1950 1960 1970 1980 1990 2000 2010

Source: Board of Governors of the Federal Reserve System

Shaded areas indicate US recessions - 2014 [Link]

A FINANCIAL TIMES COMPANY 20

LIQUIDITY RISK

“We use stress testing and scenario analysis to evaluate the

impact of sudden and severe stress events on our liquidity
position. The scenarios we apply have been based on historic
events, such as the 1987 stock market crash, the 1990 U.S.
liquidity crunch and the September 2001 terrorist attacks,
liquidity crisis case studies and hypothetical events, as well as
the lessons learned from the latest financial markets crisis.
They include a prolonged term money-market and secured
funding freeze, collateral repudiation, reduced fungibility of
currencies, stranded syndications as well as other systemic
knock-on effects.”
Deutsche Bank (2013) Risk Report.

A FINANCIAL TIMES COMPANY 21

REPURCHASE AGREEMENTS
I Repurchase agreement (Repo) is an agreement to sell a security
16
(today) and a simultaneous commitment to repurchase the security
AN INTRODUCTION TO FIXED INCOME MARKETS

at some future (fixed) date, at a price determined today.

Figure 1.4 Schematic Repo Transaction

time t

buy bond at P t deliver bond

=⇒ =⇒
MARKET TRADER REPO DEALER
⇐= ⇐=
pay P t get Pt − haircut

time T = t + n days

sell bond at P T get the bond

⇐= ⇐=
MARKET TRADER REPO DEALER
=⇒ =⇒
n
get PT pay (P t − haircut)× (1+repo rate × 360 )

A FINANCIAL TIMES COMPANY 22

REVERSE REPURCHASE AGREEMENTS
I Reverse repurchase agreement is an agreement to buy a security
(today)
18 and a simultaneous
AN INTRODUCTION agreement
to sell the security at some
TO FIXED INCOME MARKETS

future (fixed) date, at a price determined today.

Figure 1.5 Reverse Repo Transaction

time t

sell bond at P t borrow bond

⇐= ⇐=
MARKET TRADER REPO DEALER
=⇒ =⇒
get Pt use P t as cash collateral

time T = t + n days

buy bond at P T give bond back

=⇒ =⇒
MARKET TRADER REPO DEALER
⇐= ⇐=
n
pay P T get back P t ×(1+ repo rate × 360 )

A FINANCIAL TIMES COMPANY 23

LIQUIDITY SCENARIOS

Repo Repo 1-year Sale 1-year Sale

Haircut Haircut Repo Haircut Haircut 1-year Sale
Severe Moderate Haircut Severe Moderate Haircut No
Stress Stress No Stress Stress Stress Stress
Asset category
Reverse repo and excess collateral 0% 0% 0% 0% 0% 0%
Cash 0% 0% 0% 0% 0% 0%
G7 and Swiss government 3% 2% 1% 0% 0% 0%
Other developed market government and supranational bonds 5% 3% 2% 0% 0% 0%
Agencies and municipal bonds 5% 3% 2% 0% 0% 0%
Emerging market government bonds 20% 10% 5% 50% 25% 0%
Bonds corporate - investment grade 15% 10% 5% 20% 10% 0%
Bonds corporate - below investment grade 30% 20% 10% 30% 15% 0%
Bonds corporate - non rated 40% 30% 20% 40% 20% 0%
Bonds structured - AAA rated 30% 25% 10% 30% 15% 0%
Bonds structured - AA rated 35% 25% 10% 40% 20% 0%
Bonds structured - other investment grade 50% 30% 15% 50% 25% 0%
Bonds structured -below investment grade 50% 30% 15% 50% 25% 0%
Bonds structured - non-rated 100% 100% 50% 100% 50% 0%
Bonds (non-traded) 100% 100% 100% 100% 50% 0%
Listed equities 35% 20% 15% 0% 0% 0%
Hedge funds, listed direct investments, prime direct real estate 100% 100% 100% 50% 25% 0%
Private equity, strategic holdings, other assets 100% 100% 100% 100% 50% 0% !

A FINANCIAL TIMES COMPANY 24

LIQUIDITY RISK: NORTHERN ROCK
Shin (2009) “Reflections on Northern Rock,” Journal of Economic
Perspectives, Vol 23, No. 1
Reflections on Northern Rock 105
I Established in 1965 by the
merger of two building societies
Figure 1
(mutually owned savings and Composition of Northern Rock’s Liabilities, June 1998–June 2007

mortgage banks) 120

Equity
I Goes public in 1997 100

Other liabilities
I Total assets grow at an annual 80

Billion pounds
rate of 23.2% from GBP 17.4 60
Securitized notes
billion (June, 1998) to GBP 40

113.5 billion (June, 2007) 20

Retail deposits
I Retail funding (deposits) falls 0

from 60% of liabilities (June,

Ju

Ju 98

Ju 99

Ju 00

Ju 1

Ju 02

Ju 03

Ju 04

Ju 5

Ju 06
De 98

De 99

De 00

De 01

De 02

De 03

De 04

De 05

De 06
n-

n-

n-

n-

n-

n-

n-

n-

n-

n-
c-

c-

c-

c-0

c-

c-

c-

c-0

c-
07
1998) to 23% of liabilities Source: Northern Rock, annual and interim reports, 1998–2007.

(June, 2007) model unusual, its balance sheet less traditional, and that securitization was some-
how responsible in Northern Rock’s downfall (for example, see Mayes and Wood,
A FINANCIAL TIMES COMPANY 2008; Milne and Wood, 2008; and others). However, I will argue that the 25 role of
LIQUIDITY RISK: NORTHERN ROCK
I Covered bonds, securitized
notes and “wholesale funding”
account for more than 70% of
liabilities
I During Q3, 2007 wholesale
funding markets sieze
I Northern Rock is unable to
refinance maturing money
market obligations
I On September 12, 2007 the
bank asks the Bank of England,
as lender of last resort, for
emergency liquidity support
I Northern Rock is nationalized
on February 17, 2008
A FINANCIAL TIMES COMPANY 26
LIQUIDITY RISK: NORTHERN ROCK
106 Journal of Economic Perspectives

Figure 2
Structural Diagram of the Securitization Transaction for Northern Rock’s Granite
Master Issuer Series 2005-2

Assignment
of mortgage
portfolio
Northern Rock PLC Granite Finance Trustees Ltd
(Originator) (Mortgage Trustee)
Proceeds

Funding 2 Ltd
Class A Notes (Special Purpose Entity)

Class B Notes
Note proceeds

Class M Notes Granite Master Issuer PLC

(Note Issuer)
Principal
Class C Notes and interest

Class D Notes

Source: Supplement to Prospectus, May 23, 2005. At ![Link]

securitisation/prospectus%20&%20us%20supplement%[Link]".

There is another contrast between Northern Rock and the U.S. and European
banks caught up in the subprime crisis. The latter banks sponsored off-balance-
A FINANCIAL TIMES COMPANY 27
sheet entities (such as “conduits” and “structured investment vehicles”) that held
LIQUIDITY RISK: NORTHERN ROCK

I 2006 annual report:

“During the year, we raised GBP 3.2 billion medium term wholesale
funds from a variety of globally spread sources . . . This included two
transactions sold to domestic US investors totalling US$3.5 billion.
In January 2007, we raised a further US$2.0 billion under our US
MTN [medium term notes] programme. Key developments during
2006 included the establishment of an Australian debt programme,
raising A$1.2 billion from our inaugural issue. This transaction was
the largest debut deal in that market for a single A rated financial
institution targeted at both domestic Australian investors and the
Far East.”

A FINANCIAL TIMES COMPANY 28

LIQUIDITY RISK: NORTHERN ROCK
108 Journal of Economic Perspectives Reflections on Northern Rock 109

Figure 3 Figure 4
Composition of Northern Rock’s Liabilities Before and After the Run Composition of Retail Deposits of Northern Rock Before and After Run
(million pounds)
(millions of pounds)

26,710
28,473
10,201

Loan from Bank of England

24,350 11,472
Wholesale
10,469 Retail
8,105 4,105
8,938 Covered bonds
4,351
Securitized notes
2,752 Postal accounts
1,371 Offshore and other accounts
45,698 43,070 1,712 Internet and telephone accounts
5,573
Branch accounts
3,035

June 2007 Dec 2007 Dec 2006 Dec 2007

Source: Northern Rock annual report for 2007. Source: Northern Rock, annual report for 2007.

funding with increasing diversification of our global investor.” Medium-term fund- “substantial outflows of wholesale funds, as maturing loans and deposits were not
ing refers to term funding of six months or longer, while short-term funding has a renewed. This resulted in a full year net outflow of £11.7 billion.” Thus, the key to
maturity less than six months. The 2006 annual report (p. 41) is worth quoting the initial “run” on Northern Rock was the nonrenewal of Northern Rock’s short-
verbatim for an insight into the nature of this short-term funding: and medium-term paper. This was the run that led to the demise of Northern
A FINANCIAL TIMES COMPANY Rock—a run that happened out of sight of the television cameras. 29
 112 Journal of Economic Perspectives

LIQUIDITY RISK: NORTHERN ROCK

Table 1
Haircuts for Repos during March 2008

Security Typical haircuts March 2008 haircuts

Treasuries ! 0.5% 0.25% " 3%

Corporate bonds 5% 10%
AAA asset-backed securities 3% 15%
AAA residential mortgage-backed securities 2%
Reflections on Northern
20%
Rock 113
AAA jumbo prime mortgages 5% 30%

Source: Bloomberg.
Note: In a repurchase agreement, the borrower sells a security today for a price below the current market
price on the understanding that it will buy back the security in the future at a pre-agreed price. The
Figure 5difference between the current market price of the security and the price at which it is sold is called the
“haircut” in the repo.
Northern Rock’s Leverage, June 1998 –December 2007

90 assets to equity—is 50 (the reciprocal of the haircut ratio). In other words, to hold
$100 worth of securities, the borrower must come up with $2 of Leverage
[Link] common equity
80 Suppose that a borrower leverages up to the maximum permitted level and has
a highly leveraged balance sheet with a leverage of 50. If at this time a shock to the
70 financial system raises the market haircut to 4 percent, then the permitted leverage
Leverage
halves to 25, from 50. In fact, times of financial stress are associated on shareholder
with sharply equity
Leverage ratio

60 higher haircuts. Table 1 show the haircuts that were being applied during the peak
of the market disruptions in March 2008 compared to the haircuts prevailing
50
during normal times. For instance, a borrower holding AAA-rated residential
40 mortgage-backed securities would have seen a ten-fold increaseLeveragein haircuts, mean-
on total equity
ing that its leverage must fall from 50 to just 5.
30 Clearly, an increase in haircuts entails very substantial reductions in leverage,
which creates hard choices. Imagine a borrower who sees the extent of its possible
20 leverage fall by half. Either the borrower must raise new equity, so that its equity
doubles from its previous level, or the borrower must sell half its assets, or some
10
combination of both. Either raising new equity or cutting assets will entail painful
Ju

J u -98

J u -99

Ju -00

Ju -01

J u -02

Ju -03

Ju -04

J u -0 5

Ju -06
D -98

D -99

D -00

D -01

D -02

D -03

D -04

D -05

D -06

D -07

adjustments. Raising new equity is notoriously difficult in distressed market condi-

ec

ec

ec

ec

ec

ec

ec

ec

ec

ec
n

n
-0

tions— but selling assets in a depressed market is not much better. For financial
7

institutions that have assets which are very short-term and liquid—such as short-
Source: Northern Rock, annual and interim reports, 1998 –2007.
term collateralized lending—a common approach to this situation is to make the
Note: The leverage ratio is the ratio of total assets to equity.
A FINANCIAL TIMES COMPANYnecessary adjustment by reducing lending (which in effect is reducing assets) and 30
ABC P2: c/d QC: e/f T1: g
WBT422-Jorion November 22, 2010 8:0 Printer Name: Yet to Come

OPERATIONAL RISK
616 OPERATIONAL AND INTEGRATED RISK MANAGEMENT

TABLE 25.1 Operational Risk Classification

Internal Risks

People Processes Systems

Employee collusion/fraud Accounting error Data quality

Employee error Capacity risk Programming error
Employee misdeed Contract risk Security breach
Employer liability Misselling/unsuitability Strategic risk
Employment law Product complexity (platform/supplier)
Health and safety Project risk System capacity
Industrial action Reporting error System compatibility
Lack of knowledge/skills Settlement/payment error System delivery
Loss of key personnel Transaction error System failure
Valuation error System unsuitability

External Risks

External Physical

Legal Fire
Money laundering Natural disaster
Outsourcing Physical security
Political Terrorism
Regulatory Theft
Supplier risk
Tax

Source: British Bankers’ Association survey.

which is due to the use of wrong models for valuation and risk management. This
is TIMES
A FINANCIAL an internal risk that combines lack of knowledge (people) with product com-
COMPANY 31
 Module 3
Money and Capital
Markets

A FINANCIAL TIMES COMPANY 32

THE FINANCIAL SYSTEM

Mishkin, Financial Intermediaries

Chapter 2, Figure 1 (p.24) Indirect Finance
= Financial Institutions
Financial Intermediaries Funds
Funds

Lenders Borrowers

1. Households Funds 1. Businesses

2. Businesses 2. Governments
3. Governments 3. Households
4. Foreigners 4. Foreigners

Funds Funds

Buy Securities Issue Securities

(Assets) Financial Markets (Liabilities)
Direct Finance Securities
= Financial Instruments
= Claims on Income or Assets

Source: Mishkin, F., The Economics of Money, Banking and Financial Markets
A FINANCIAL TIMES COMPANY 33
MONEY AND CAPITAL MARKETS

I Money Market Instruments I Capital Market Instruments

I US Treasury Bills I Corporate Stocks
I Negotiable Bank Certificates I Residential, Commercial, and
of Deposit Farm Mortgages
I Commercial Paper I Corporate Bonds
I Bankers Acceptances I US Government Securities
I Repurchase Agreements (Intermediate and
I Federal Funds Long-Term)
I Eurodollars I State and Local Government
(Municipal) Bonds
I US Government Agency
Bonds
I Bank Commercial and
Consumer Loans

A FINANCIAL TIMES COMPANY 34

US REGULATORY STRUCTURE
I Dual Banking System
I National or state charters for depository institutions
I National Currency Act (1863) / National Bank Act (1864)
I Home Owners Loan Act (1933) - chartering of federal savings
associations
I Various state banking laws
I Separation of Banking and Commerce
I Bank and thrift holding companies are subject to restrictions on
activities
I Bank Holding Company Act (1956)
I Functional Regulation
I Gramm-Leach Bliley Act (1999)
I Affiliates of banks subject to regulation based upon function (i.e.
broker-dealer, insurance)
I Dodd-Frank (Wall Street Reform and Consumer Protection) Act
(2010) grants Federal Reserve Board increased authority to examine
bank holding companies and non-bank subsidiaries
A FINANCIAL TIMES COMPANY 35
PROTECTION OF DEPOSITORS AND CONSUMERS

I Federal deposit insurance for deposit accounts at depository

institutions (banks and thrifts)
I Banking Act (1933)
I Coverage of $250,000 per depositor (subject to certain aggregation
rules) made permanent by Dodd-Frank Act and retroactive to 01/08
I Dodd-Frank Act established the Consumer Financial Protection
Bureau (CFPB) within the Federal Reserve System
I CFPB assumed “consumer financial protection functions” previously
performed by the federal banking agencies (FRB, OCC, FDIC,
NCUA), as well as HUD and the FTC
I CFPB has rulemaking authority for all depository institutions and
supervision/examination authority for large depository institutions
and affiliates, as well as most non-depository institutions

A FINANCIAL TIMES COMPANY 36

 Who Regulates Whom and How? An Overview of U.S. Financial Regulatory Policy

FEDERAL REGULATORS
Table 1. Federal Financial Regulators and Organizations
(acronyms and area of authority)
Prudential Bank Securities and Other Regulators of
Regulators Derivatives Regulators Financial Activities Coordinating Forum

Office of the Comptroller Securities and Exchange Federal Housing Finance Financial Stability
of the Currency (OCC) Commission (SEC) Agency (FHFA) Oversight Council (FSOC)
Federal Deposit Insurance Commodities Futures Consumer Financial Federal Financial
Corporation (FDIC) Trading Commission Protection Bureau (CFPB) Institutions Examinations
(CFTC) Council (FFIEC)
National Credit Union President’s Working
Administration (NCUA) Group on Capital Markets
(PWG)
Federal Reserve Board
(FRB, or the Fed)

Source: The Congressional Research Service (CRS).

Source: Congressional Research Service

The policy problems and regulatory approaches of the agencies listed in Table 1 vary
considerably. Before providing a detailed analysis of each agency, it may be useful to consider
how the agencies are related to each other and briefly sketch the types of policies they generally
A FINANCIAL TIMES COMPANY 37
 Who Regulates Whom and How? An Overview of U.S. Financial Regulatory Policy

FEDERAL REGULATORS Table 3. Federal Financial Regulators and Who They Supervise
Emergency/Systemic Other Notable
Regulatory Agency Institutions Regulated Risk Powers Authority

Federal Reserve Bank holding companies Lender of last resort to Numerous market-level
and certain subsidiaries, member banks (through regulatory authorities,
financial holding discount window lending) such as checking services,
companies, securities lending markets, and other
holding companies, savings In “unusual and exigent banking-related activities.
and loan holding circumstances,” the Fed
companies, and any firm may extend credit beyond
designated as systemically member banks, to provide
significant by the FSOC. liquidity to the financial
system, but not to aid
State banks that are failing financial firms.
members of the Federal
Reserve System, U.S. May initiate resolution
branches of foreign banks, process to shut down
and foreign branches of firms that pose a grave
U.S. banks. threat to financial stability
(requires concurrence of
Payment, clearing, and two-thirds of the FSOC).
settlement systems The FDIC and the
designated as systemically Treasury Secretary have
significant by the FSOC, similar powers.
unless regulated by SEC or
CFTC.
Office of the Comptroller National banks, federally
of the Currency (OCC) chartered thrift institutions
Federal Deposit Insurance Federally insured After making a Operates a deposit
Corporation (FDIC) depository institutions, determination of systemic insurance fund for
including state banks and risk, the FDIC may invoke federally and state
thrifts that are not broad authority to use the chartered banks and
members of the Federal deposit insurance funds to thrifts.
Reserve System. provide an array of
assistance to depository
institutions, including debt
guarantees.
National Credit Union Federally chartered or Serves as a liquidity lender Operates a deposit
Administration (NCUA) insured credit unions to credit unions insurance fund for credit
experiencing liquidity unions, known as the
shortfalls through the National Credit Union
Central Liquidity Facility. Share Insurance Fund
(NCUSIF).

Source: Congressional Research Service

A FINANCIAL TIMES COMPANY 38

FEDERAL REGULATORS Who Regulates Whom and How? An Overview of U.S. Financial Regulatory Policy

Emergency/Systemic Other Notable

Regulatory Agency Institutions Regulated Risk Powers Authority

Securities and Exchange Securities exchanges, May unilaterally close Authorized to set financial
Commission (SEC) brokers, and dealers; markets or suspend accounting standards in
clearing agencies; mutual trading strategies for which all publicly traded
funds; investment advisers limited periods. firms must use.
(including hedge funds with
assets over $150 million)
Nationally recognized
statistical rating
organizations
Security-based swap (SBS)
dealers, major SBS
participants, and SBS
execution facilities
Corporations selling
securities to the public
must register and make
financial disclosures.
Commodity Futures Futures exchanges, May suspend trading,
Trading Commission brokers, commodity pool order liquidation of
(CFTC) operators, and commodity positions during market
trading advisors emergencies.
Swap dealers, major swap
participants, and swap
execution facilities
Federal Housing Finance Fannie Mae, Freddie Mac, Acting as conservator
Agency (FHFA) and the Federal Home (since Sept. 2008) for
Source: Congressional Research Service Loan Banks Fannie Mae and Freddie
Mac
Bureau of Consumer Nonbank mortgage-related Writes rules to carry out
Financial Protection firms, private student the federal consumer
A FINANCIAL TIMES COMPANY lenders, payday lenders, financial protection laws 39
and larger “consumer
 execution facilities
Corporations selling
securities to the public
must register and make
financial disclosures.
Commodity Futures Futures exchanges, May suspend trading,
FEDERAL REGULATORS
Trading Commission
(CFTC)
brokers, commodity pool
operators, and commodity
order liquidation of
positions during market
trading advisors emergencies.
Swap dealers, major swap
participants, and swap
execution facilities
Federal Housing Finance Fannie Mae, Freddie Mac, Acting as conservator
Agency (FHFA) and the Federal Home (since Sept. 2008) for
Loan Banks Fannie Mae and Freddie
Mac
Bureau of Consumer Nonbank mortgage-related Writes rules to carry out
Financial Protection firms, private student the federal consumer
lenders, payday lenders, financial protection laws
and larger “consumer
financial entities” to be
determined by the Bureau
Consumer businesses of
banks with over $10 billion
in assets
Does not supervise
insurers, SEC and CFTC
registrants, auto dealers,
sellers of nonfinancial
goods, real estate brokers
and agents, and banks with
assets less than $10 billion

Source: The Congressional Research Service (CRS), with information drawn from agency websites, and financial
regulatory
Source: Congressional legislation.
Research Service
a. See Appendix B.

A FINANCIAL TIMES COMPANY 40

 OCC at the depository level, and by the Federal Reserve on a consolidated basis at the holding
company level. As a public company, JPMorgan’s disclosures of the trades to its stockholders
were regulated by the SEC. As a participant in derivatives markets, JPMorgan’s transactions were
subject to CFTC regulation. As an insured depository institution, JPMorgan’s safety and
soundness was also subject to the FDIC.
JPM OVERSIGHT
Figure 1. An Example of Regulation of JPMorgan Derivatives Trades

Source: CRS.
Source: Congressional Research Service
Table 2 compares the general policy options and approaches of the banking regulators to the
A FINANCIAL TIMES COMPANY 41
 UK REGULATORY STRUCTURE
egulatory landscape April 2013
I Financial Services Act (2012) established a new regulatory structure
Figure 1: The New Regulatory Structure

ority (FCA) and

ficially came
inancial
ces Act which

rnment had
ory framework.
ded and
to the Bank of
cial Policy
anning for
Authority (PRA)
of systemically
Authority (FCA),
ction and
pervision of

ces Act in
ority and
orce and
Financial
Source: Chartered Insurance InstituteConduct Authority Diagram, from FCA Business Plan
2013/14, p.58
A) A FINANCIAL TIMES COMPANY 42
UK REGULATORY STRUCTURE

I Financial Conduct Authority (FCA):

I Operational objectives - consumer protection, integrity of UK
financial system, promoting competition in interests of consumers
I Prudential Regulation Authority (PRA):
I Promote safety and soundness of systemically important firms,
including insurers, and ensuring policyholders are protected in the
event of failure
I Financial Policy Committee (FPC):
I Committee within the Bank of England responsible for identifying
emerging risks to the financial system and providing strategic
direction for the entire regulatory regime
I FPC has the power to use “macro-prudential tools” to counteract
systemic risk. Tools could include leverage limits on banks or
enforcing particular capital requirements for specific asset classes

A FINANCIAL TIMES COMPANY 43

 EU

EU REGULATORY STRUCTURE N/A

I European System of Financial Supervision (ESFS), network of

national and EU supervisors, created by EU, January 1, 2011.

European European Central

European
European
and
and Markets

European

European

Source: Grant Thornton

A FINANCIAL TIMES COMPANY 44
EU REGULATORY STRUCTURE

I European Systemic Risk Board (ESRB)

I Independent body responsible for the macro-prudential oversight of
the EU financial system.
I ESRBs day-to-day business entrusted to the European Central Bank
I European Supervisory Authorities (ESAs):
I European Banking Authority (EBA)
I European Securities and Markets Authority (ESMA)
I European Insurance and Occupational Pensions Authority (EIOPA)
I Joint Committee of the ESA’s
I Deals with “cross-sectoral” issues
I EU Member State national supervisors (28), carry out day-to-day
supervision of financial institutions

A FINANCIAL TIMES COMPANY 45

 Module 4
Risk Concepts

A FINANCIAL TIMES COMPANY 46

 other).

Collect the portfolio positions and map them onto the risk factors.

Use the risk engine to construct the distribution of portfolio profit and losses over the
GETTING TO THE LOSS DISTRIBUTION selected period. This can be summarized by a Value-at-Risk (VAR) number, which

represents the worst loss that will not be exceeded at the pre-specified confidence

level.

I Non-parametric
I Historical simulation / Positions
Trades from
Risk Factors
front office
scenarios Global Historical
Repository Data feed with Market Data
I Assume that history contains current prices

all relevant risk information

Mapping Model
I Parametric Risk Engine3a

I Fit an assumed probability Portfolio

Distribution Distribution
Positions
model to historical loss data of
Risk Factors
and compute risk measures Value at Risk
Reports
Data Warehouse Risk Warehouse
I Assume / fit stochastic
Fig. 1. Components of a risk measurement system
processes for risk factors and
generate loss distributions by The key feature
Source: Jorion of this system
(2009) is that it is position-based.
“Risk Management Lessons from Traditionally,
the Credit risk

Monte Carlo simulation measuresCrisis,”

have been built from
European returns-based
Financial information. The latter is easy and cheap t
Management

implement. It also accounts for dynamic trading of the portfolio. On the other hand,

A FINANCIAL TIMES COMPANY 47

PORTFOLIO VALUE

I V0 = value of portfolio today (t = 0)

I V (t) = value of the portfolio at some time in the future, t > 0
I V (t) is a function of one or more risk factors, si (t), i ≥ 1:
 
V (t) = f s1 (t), s2 (t), . . . , sn (t)

I Risk factor values at t > 0 are not known today (t = 0)

I Future values of risk factors are modeled as random variables

A FINANCIAL TIMES COMPANY 48

RISK FACTORS
DJIA#

14000#
I Stock prices / Equity index
levels 12000#
I Sovereign bond prices / Interest
10000#
rates 8/7/06# 2/3/07# 8/2/07# 1/29/08# 7/27/08#
!
I Corporate bond prices / Credit 18
16
spreads 14

Spread (bps, hundreds)

12
I Exchange Rates 10
8

I Volatility Equity /Rates / 6

4

Spreads / Exchange Rates 2

0

Jul-06

Jan-07

Jul-07

Jan-08

Jul-08

Jan-09

Jul-09

Jan-10

Jul-10
I Term structures - Rates /
USD Industrial AA USD Industrial BBB
Spreads / Forwards USD Financials AA USD Financials BBB

A FINANCIAL TIMES COMPANY 49

PORTFOLIO VALUE
F(v,t)  
1.00  
I V (t) is a random variable from
0.75  
the perspective of today.
0.50  
I Distributional properties of risk
factors determine distribution 0.25  

of V (t)
0.00  
0.00   0.50   1.00   1.50   2.00   2.50  
Distribution function FV (v , t):
f(v,t)'
FV (v , t) = p[V (t) ≤ v ] 1.6#

1.2#
Density function fV (v , t)
0.8#

∂FV (v , t)
fV (v , t) = 0.4#

∂p
0.0#
0.00# 0.50# 1.00# 1.50# 2.00# 2.50#
!

A FINANCIAL TIMES COMPANY 50

PORTFOLIO VALUE

I Expected value:
Z +∞

E0 [V (t)] = v (t)fV v (t) dv (t)
−∞

I Variance:
V0 [V (t)] = E0 [V (t)2 ] − E0 [V (t)]2
I Standard deviation:
p
SD0 [V (t)] = V0 [P(t)]

A FINANCIAL TIMES COMPANY 51

PROFIT
f(v,t)'
1.6#

1.2#

I Profit realized at t
0.8#

P(t) = V (t) − V0 0.4#

0.0#
I P(t) is a random variable from 0.00# 0.50# 1.00# 1.50# 2.00# 2.50#
!
the perspective of today, t = 0 f(p,t)'
I Expected value 1.6#

1.2#
E0 [P(t)] = E0 [V (t)] − V0
0.8#

I Variance 0.4#

0.0#
V0 [P(t)] = V0 [V (t)] )1.00# )0.50# 0.00# 0.50# 1.00# 1.50#
!

A FINANCIAL TIMES COMPANY 52

PROFIT PROBABILITY FUNCTIONS

F(p,t)  
1.0  

0.8  

Profit distribution function

0.5  
FP (p, t):
0.3  

FP (p, t) = p[P(t) ≤ p] 0.0  

-­‐1.00   -­‐0.50   0.00   0.50   1.00   1.50  

Profit density function fP (p, t) f(p,t)'

1.6#
∂FP (p, t)
fP (p, t) = 1.2#
∂p
0.8#

0.4#

0.0#
)1.00# )0.50# 0.00# 0.50# 1.00# 1.50#
!
A FINANCIAL TIMES COMPANY 53
LOSS
f(v,t)'
1.6#

1.2#

I Loss realized at t
0.8#

L(t) = V0 − V (t) 0.4#

0.0#
I L(t) is a random variable from 0.00# 0.50# 1.00# 1.50# 2.00# 2.50#
!
the perspective of today, t = 0 f(l,t)'
I Expected value 1.6#

1.2#
E0 [L(t)] = V0 − E0 [V (t)]
0.8#

I Variance 0.4#

0.0#
V0 [L(t)] = V0 [V (t)] )1.00# )0.50# 0.00# 0.50# 1.00# 1.50#
!

A FINANCIAL TIMES COMPANY 54

LOSS PROBABILITY FUNCTIONS

Loss distribution function

FL (l, t):

FL (l, t) = p[L(t) ≤ l]

Loss density function fL (l): f(l,t)'

1.6#
∂FL (l, t)
fL (l, t) = 1.2#
∂l
0.8#

0.4#

0.0#
)1.00# )0.50# 0.00# 0.50# 1.00# 1.50#
!
A FINANCIAL TIMES COMPANY 55
‘EXCESS’ PROFIT

I Excess profit realized at t

P ∗ (t) = V (t) − E0 [V (t)]

I P ∗ (t) is a random variable from the perspective of today, t = 0

I Expected value

E0 [P ∗ (t)] = E0 [V (t)] − E0 [V (t)] = 0

I Variance
V0 [P ∗ (t)] = V0 [V (t)]

A FINANCIAL TIMES COMPANY 56

EXCESS PROFIT DISTRIBUTION
f(p,t)'
1.6#

1.2#

I Excess profit density function is 0.8#

just the profit density function
0.4#
function shifted to the left by
the amount: 0.0#
)1.00# )0.50# 0.00# 0.50# 1.00# 1.50#
!
∗
P(t) − P (t) = E0 [V (t)] − V0 f(p*,t)(
1.6#

I
1.2#
P(t) − P ∗ (t) > 0
0.8#

I if 0.4#
E0 [V (t)] > V0
0.0#
)1.00# )0.50# 0.00# 0.50# 1.00# 1.50#
!

A FINANCIAL TIMES COMPANY 57

‘EXCESS’ LOSS

I Excess loss realized at t

L∗ (t) = E0 [V (t)] − V (t) = −P ∗ (t)

I L∗ (t) is a random variable from the perspective of today, t = 0

I Expected value

E0 [L∗ (t)] = E0 [V (t)] − E0 [V (t)] = 0

I Variance
V0 [L∗ (t)] = V0 [V (t)]

A FINANCIAL TIMES COMPANY 58

EXCESS LOSS DISTRIBUTION
f(l,t)'
1.6#

1.2#

I Excess loss density function is 0.8#

just the loss density function

0.4#
function shifted to the right by
the amount: )1.00# )0.50#
0.0#
0.00# 0.50# 1.00# 1.50#
!

∗ f(l*,t)(
L(t) − L (t) = V0 − E0 [V (t)]
1.6#

I 1.2#

L(t) − L∗ (t) < 0 0.8#

I if 0.4#

E0 [V (t)] > V0
0.0#
)1.00# )0.50# 0.00# 0.50# 1.00# 1.50#
!

A FINANCIAL TIMES COMPANY 59

RISK MEASURES: LOWER PARTIAL MOMENTS

I Semi-Variance
h  2 i
SV0 [P(t)] = E0 min P(t) − E0 [P(t)], 0

I Semi-Standard Deviation
p
SSD0 [P(t)] = SV0 [P(t)]

I ‘Down-side’ risk measure

I Lower Partial Moments
h  k i1/k
LPMk = E0 max E0 [P(t)] − P(t), 0

A FINANCIAL TIMES COMPANY 60

QUANTILES
F(p,t)'
1.00#

0.75#
I α-quantile for FP (p), pα :
0.50#

p[P ≤ pα ] = FP (pα ) = α 0.25#

0.00#
(1.0# (0.5# 0.0# 0.5# 1.0# 1.5# 2.0#
FP−1 (α) = pα !

f(p,t)'
I Common quantiles: 1.4#

1.2#
I Percentiles 1.0#
I Deciles 0.8#
I Quintiles 0.6#

I Quartiles 0.4#

0.2#

0.0#
)1.0# )0.5# 0.0# 0.5# 1.0# 1.5# 2.0#
!

A FINANCIAL TIMES COMPANY 61

VALUE AT RISK (VaR)
VaR$%$P$Density$Func1on$
1.6#

1.2#

0.8#
I One version - ‘Zero’ VaR
α$ 0.4#
I VaR is a quantile on a profit
(loss) distribution function )1.00# )0.80# )0.60# )0.40# )0.20#
0.0#
0.00# 0.20# 0.40# 0.60# 0.80# 1.00#

VaR1−α = −FP−1 (α) Var$%$L$Density$Func1on$

1.6#

VaR1−α = FL−1 (1 − α) 1.2#

I α is the level of significance 0.8#

I 1 − α is the level of confidence 0.4# α$

0.0#
)1.00# )0.80# )0.60# )0.40# )0.20# 0.00# 0.20# 0.40# 0.60# 0.80# 1.00#

A FINANCIAL TIMES COMPANY 62

VALUE AT RISK (VaR)
Var$%$P*$Density$Func2on$
1.6#

1.2#

I Another version - ‘Mean’ VaR 0.8#

I VaR is a quantile on an excess α$ 0.4#

profit (loss) distribution 0.0#

function )1.00# )0.80# )0.60# )0.40# )0.20# 0.00# 0.20# 0.40# 0.60# 0.80# 1.00#

Var$%$L*$Density$Func2on$
VaR1−α = −FP−1
1.6#
∗ (α)

1.2#
VaR1−α = FL−1
∗ (1 − α)
0.8#
I α is the level of significance
α$
0.4#
I 1 − α is the level of confidnece
0.0#
)1.00# )0.80# )0.60# )0.40# )0.20# 0.00# 0.20# 0.40# 0.60# 0.80# 1.00#
!

A FINANCIAL TIMES COMPANY 63

VaR PROBLEM

I All outcomes between a loss of $10 million and a profit of $10

million are equally likely for a fixed income trading desk over a
one-week period
I What is the VaR for a one-week time horizon at a 95% confidence
level?
I Answer: $9 million (loss)

A FINANCIAL TIMES COMPANY 64

VaR EXAMPLE

I Simple portfolio, e.g. a single

f(s,1)"
stock 1.6#

I Risk factor is stock price 1.2#

0.8#
V (t) = s(t)
0.4#

I Risk factor process: 0.0#

0.00# 0.50# 1.00# 1.50# 2.00# 2.50#
E"["s(t)"]"="1.15"
s(t) = s0 + µt + σz(t)
√
z(t) ∼ N(0, t)

A FINANCIAL TIMES COMPANY 65

VaR EXAMPLE

I Let:
V0 = s0 = 1
µ = 0.15 f(V,1)"
1.6#
σ = 0.3
1.2#

t=1 0.8#

I Portfolio value at t = 1: 0.4#

0.0#
0.00# 0.50# 1.00# 1.50# 2.00# 2.50#
E0 [V (t)] = V0 + µt = 1.15 E"["V(1)"]"="1.15"

V0 [V (t)] = σ 2 t = 0.09
√
SD0 [V (t)] = σ t = 0.3
V (t) ∼ N(1.15, 0.3)

A FINANCIAL TIMES COMPANY 66

VaR EXAMPLE

I Profit
f(p,1)"

P(t) = V (t) − V0 1.6#

1.2#
I Profit at t = 1: 0.8#

0.4#
E0 [P(t)] = µt = 0.15
0.0#
)1.00# )0.50# 0.00# 0.50# 1.00#
2
V0 [P(t)] = σ t = 0.09 E"["P(t)"]"=".15"

√
SD0 [V (t)] = σ t = 0.3
P(t) ∼ N(0.15, 0.3)

A FINANCIAL TIMES COMPANY 67

VaR EXAMPLE

I (Zero) VaR with α = 0.05 and

VaR$%$P$Density$Func1on$
t=1 1.6#

VaR1−α = −FP−1 (α)

1.2#

0.8#

VaR.95 = −FP−1 (.05)

α$ 0.4#

I Excel
0.0#
)1.00# )0.80# )0.60# )0.40# )0.20# 0.00# 0.20# 0.40# 0.60# 0.80# 1.00#

= −[Link](0.05, 0.15, 0.3)

≈ 0.3435

A FINANCIAL TIMES COMPANY 68

VaR EXAMPLE

Var$%$P*$Density$Func2on$
1.6#
I Show that the ‘mean’ VaR is
1.2#

VaR.95 = −FP−1
∗ (.05) 0.8#

≈ 0.4935 α$ 0.4#

I What is the 95% VaR over 6 )1.00# )0.80# )0.60# )0.40# )0.20#
0.0#
0.00# 0.20# 0.40# 0.60# 0.80# 1.00#
months?

A FINANCIAL TIMES COMPANY 69

VaR IN PRACTICE

I In theory, VaR is precise and simple

I In practice, not so!
I In practice, VaR is at best an estimate of the loss (profit)
distribution quantile . . . more often it is closer to an ‘educated’ guess
for the quantile
I Why?
I We dont know the true process for V (t) . . . so we estimate it . . . or
guess!
I Technical term for guessing is (model) calibration
I Lots of other problems!

A FINANCIAL TIMES COMPANY 70

EXPECTED SHORTFALL (ES)
I How much will we lose if the
VaR in Visual Terms
VaR quantile is exceeded?
Profit & Loss Distribution (P&L)
I ES is the expected loss given 95% VaR = 1.6 Mean profit = 2.4

0.25
that VaR is exceeded

0.20
I Using the profit distribution

probability density
function

0.10 0.15
FP−1 (α) = pα = −VaR1−α

0.05
5% probability

0.0
 
ES1−α = E0 P(t) | P(t) ≤ pα -10 -5 0 5 10
Z pα
1
= p fP (p)dp c 2004 (McNeil, Frey & Embrechts)
⃝
α −∞
18

Z α
1
= F −1 (u)du
α 0 P
A FINANCIAL TIMES COMPANY 71
EXPECTED SHORTFALL (ES)
Losses and Profits
Loss Distribution
I Using the loss distribution Mean loss = -2.4

0.25
95% VaR = 1.6
function 95% ES = 3.3

0.20
FL−1 (1 − α) = l1−α = VaR1−α

probability density
0.10 0.15
 
ES1−α = E0 L(t) | L(t) > l1−α

0.05
Z ∞
1 5% probability
= l fL (l)dl
0.0
α l1−α -10 -5 0 5 10

Z 1
1
= F −1 (u)du c 2004 (McNeil, Frey & Embrechts)
⃝ 19
α 1−α L

A FINANCIAL TIMES COMPANY 72

ES EXAMPLE
√
I P(t) ∼ N(µt, σ t):
 
E0 P(t) | P(t) ≤ pα
VaR$%$P$Density$Func1on$
fP (pα ) 1.6#
= E0 [P(t)] − V0 [P(t)]
α
σ 2 t fP (pα ) 1.2#

= µt −
α 0.8#

I µ = 0.15, σ = 0.3, t = 1, α = 0.4#

α$
0.05, pα ≈ −0.3435:
0.0#
)1.00# )0.80# )0.60# )0.40# )0.20# 0.00# 0.20# 0.40# 0.60# 0.80# 1.00#
 
E0 P(t) | P(t) ≤ pα ≈ −0.46881

 
ES1−α = −E0 P(t) | P(t) ≤ pα
≈ 0.46881
A FINANCIAL TIMES COMPANY 73
LOG-NORMAL VaR

I Simple portfolio, e.g. a single stock

I Risk factor is stock price

V (t) = s(t)

I Risk factor process:

ln s(t) = ln s0 + µt + σz(t)

√
z(t) ∼ N(0, t)
I Stock price:
ln V (t) = ln V0 + µt + σz(t)

A FINANCIAL TIMES COMPANY 74

LOG-NORMAL VaR

I Moments:
E0 [ln V (t)] = ln V0 + µt

V0 [ln V (t)] = σ 2 t

√
ln V (t) ∼ N(ln V0 + µt, σ t)

A FINANCIAL TIMES COMPANY 75

LOG-NORMAL VaR

I Moments:
h i
E0 [V (t)] = exp E0 [ln Vt ] + 21 V0 [ln V (t)]
2
/2)t
= V0 e (µ+σ

h ih   i
V0 [V (t)] = exp 2E0 [ln Vt ] + V0 [ln V (t)] exp V0 [ln V (t)] − 1
2  2
= V02 e (2µ+σ )t e σ t − 1


 p 
V (t) ∼ LN E0 [V (t)], V0 [V (t)]

A FINANCIAL TIMES COMPANY 76

LOG-NORMAL VaR

I ln V (t)α
−1
ln V (t)α = FlnV (α)
I V (t)α
V (t)α = e ln V (t)α
I ‘Zero’ VaR
VaR1−α = V (t)α − V0
I ‘Mean’ VaR
VaR1−α = V (t)α − E0 [V (t)]

A FINANCIAL TIMES COMPANY 77

EXPECTED SHORTFALL (ES)

I Conditional expectation for V (t):

" #
1 ln Vα − E0 [ln Vt ] − V0 [ln V (t)]
E0 [V (t)|V (t) ≤ Vα ] = E0 [V (t)] Φ
α SD0 [ln V (t)]

I Expected Shortfall based on ‘Zero’ VaR:

E0 [V (t)|V (t) ≤ Vα ] − V0

I Expected Shortfall based on ‘Mean’ VaR:

E0 [V (t)|V (t) ≤ Vα ] − E0 [V (t)]

A FINANCIAL TIMES COMPANY 78

LOG-NORMAL VaR EXAMPLE

I Let:
V0 = s0 = 1
µ = 0.15
σ = 0.3
t=1
I Log portfolio value at t = 1

E0 [ln V (1)] = ln V0 + µ(1) = 0.15

V0 [ln V (1)] = σ 2 (1) = 0.09

ln V (1) ∼ N(0.15, 0.3)

A FINANCIAL TIMES COMPANY 79

LOG-NORMAL VaR EXAMPLE

I Portfolio value at t = 1
2
/2)
E0 [V (1)] = V0 e (µ+σ
≈ 1.2153

2 2
V0 [V (1)] = V02 e (2µ+σ )
eσ − 1
 

≈ 0.1391

V (1) ∼ LN(1.2153, 0.3730)

A FINANCIAL TIMES COMPANY 80

LOG-NORMAL VaR EXAMPLE

I α = 0.05
−1
ln V (1).05 = FlnV (0.05)
≈ −0.3435

I Excel:
[Link](0.05, [Link])
I

V (1).05 = e ln V (1).05
≈ 0.7093

A FINANCIAL TIMES COMPANY 81

LOG-NORMAL VaR EXAMPLE

I ‘Zero’ VaR:
VaR.95 = V (1).05 − V0 ≈ −0.2907
I ‘Mean’ VaR:

VaR.95 = V (1).05 − E0 [V (1)] ≈ −0.5060

A FINANCIAL TIMES COMPANY 82

EXPECTED SHORTFALL (ES)

I Conditional expectation for V (t):

E0 [V (1)|V (1) ≤ V.05 ]

" #
1 ln V.05 − E0 [ln V (1)] − V0 [ln V (1)]
= E0 [V (1)] Φ
0.05 SD0 [ln V (1)]
≈ 0.0315

I Expected Shortfall based on ‘Zero’ VaR:

E0 [V (1)|V (1) ≤ V.05 ] − V0 ≈ −0.3706

I Expected Shortfall based on ‘Mean’ VaR:

E0 [V (t)|V (t) ≤ Vα ] − E0 [V (t)] ≈ −0.5859

A FINANCIAL TIMES COMPANY 83

HISTORICAL LOSS DISTRIBUTION

I Loss statistics Loss  

30.00%  
Scenarios 500
25.00%  
Min -555.7954
20.00%  

Max 477.8410 15.00%  

Mean 0.8701 10.00%  

Std Dev 93.6984 5.00%  

0.00%  
-­‐600  -­‐550  -­‐500  -­‐450  -­‐400  -­‐350  -­‐300  -­‐250  -­‐200  -­‐150  -­‐100  -­‐50   0   50   100  150  200  250  300  350  400  450  500  

A FINANCIAL TIMES COMPANY 84

HISTORICAL VALUE AT RISK

I VaR confidence / significance

Ranked
level
Scenario Loss
1 − α = 0.99 α = 0.01 494 477.84
339 345.44
I Select the 5th largest loss level 349 282.20
329 277.04
VaR1−α ≈ L487 (t + 1) 487 253.38
≈ 253.38 227 217.97
131 202.26
I MS Excel 238 201.39
473 191.27
[Link](L1 : L500 , 0.99) 306 191.05
≈ 253.03

A FINANCIAL TIMES COMPANY 85

PARAMETRIC VaR: t APPROXIMATION

I t distribution with matched

mean, std dev and kurtosis 30.00%  

25.00%  

t Loss 20.00%  

Mean 0.8701 15.00%   Loss  

Std Dev 93.6984 10.00%  

T  

Skewness 0 5.00%  

Kurtosis 4.2197
0.00%  

50  
0  
-­‐600  
-­‐550  
-­‐500  
-­‐450  
-­‐400  
-­‐350  
-­‐300  
-­‐250  
-­‐200  
-­‐150  
-­‐100  
-­‐50  

100  
150  
200  
250  
300  
350  
400  
450  
500  
A FINANCIAL TIMES COMPANY 86
t VaR and ES

I Assume
L(t + 1) ∼ τv (µL , σL )
I Degrees of freedom are chosen to match the kurtosis, k, of the
sample loss distribution
4k − 6
v=
k −3
I t VaR 1/2
v −2

VaR1−α = µL + σL
v
I t ES
2
v − 2 + t1−α
  
fT (t1−α )
E[L(t + 1)|L(t + 1) > l1−α ] = µL + σL
α 1−v

A FINANCIAL TIMES COMPANY 87

t VaR AND ES
350  
 $  VaR   t  VaR  

300  

250  
30.00%  

200  
25.00%  

150  
20.00%  

100  
Loss   0.90   0.92   0.94   0.96   0.98   1.00  
15.00%  
T   450  
ES   t  ES  
10.00%   400  

350  
5.00%   300  

250  
0.00%   200  
50  
0  
-­‐600  
-­‐550  
-­‐500  
-­‐450  
-­‐400  
-­‐350  
-­‐300  
-­‐250  
-­‐200  
-­‐150  
-­‐100  
-­‐50  

100  
150  
200  
250  
300  
350  
400  
450  
500  

150  

100  

50  

0  
0.90   0.92   0.94   0.96   0.98   1.00  

A FINANCIAL TIMES COMPANY 88

COHERENT RISK MEASURES

I Artzner, Delbaen, Eber and Heath (1999) “Coherent Measures of

Risk,” Mathematical Finance
I Monotonicity: If a portfolio produces a greater loss than another
portfolio for every state of the world, its risk measure should be
greater.
I Translation Invariance: If a riskless asset (e.g. cash) with value K is
added to a portfolio, its risk measure should be reduced by K .
I Homogeneity: Changing the size of a portfolio by a factor λ while
keeping the relative amounts of different items in the portfolio the
same, should result in the risk measure being multiplied by λ.
I Subadditivity: The risk measure for two portfolios combined should
be no greater than the sum of their individual risk measures.

A FINANCIAL TIMES COMPANY 89

CRITIQUE OF VaR

“To put it in blunter terms, could VaR and the other risk
models Wall Street relies on have helped prevent the financial
crisis if only Wall Street paid better attention to them? Or did
Wall Street’s reliance on them help lead us into the abyss?”
Nocera (2009).

“The fact that you are not likely to lose more than a certain
amount 99 percent of the time tells you absolutely nothing
about what could happen the other 1 percent of the time. You
could lose $51 million instead of $50 million - no big deal. That
happens two or three times a year, and no one blinks an eye.
You could also lose billions and go out of business. VaR has no
way of measuring which it will be.”
Taleb quoted in Nocera (2009).

A FINANCIAL TIMES COMPANY 90

IN DEFENSE OF MODELS

“Models, if used properly, help decision making. Models, if used

improperly, generate bad decisions and lead to losses. This
doesnt mean that we should not use models. Quite the contrary.
It only means that we need more educated use of models.”
Jarrow (2011) Risk Management Models, Journal of Derivatives.

A FINANCIAL TIMES COMPANY 91

STRESS TESTING

“Stress testing is a risk management tool for quantifying the

size of potential losses under stress events, and for quantifying
the scenarios under which such losses might occur. A
traditional definition of a stress event is an exceptional but
credible event in the market to which the portfolio is exposed.”
Alexander (2008) Value-at-Risk Models, Wiley.

“A method for the quantification of potential future extreme,

adverse outcomes in a portfolio of financial instruments.”
Dowd (2005) Measuring Market Risk, Wiley.

A FINANCIAL TIMES COMPANY 92

STRESS TESTING

“It is common belief that VaR does not provide a complete

picture of portfolio risk and that stress testing is a means of
addressing that. . . ”
Schachter (2004) Stress Testing in PRM Handbook, PRMIA.

“. . . stress testing is also a natural complement to

probability-based risk measures such as VaR and ES. . . . VaR
gives us the maximum likely loss at a certain probability, but
gives us no idea of the loss we might suffer if we experience a
loss in excess of VaR. ES is a little better because it gives us
the expected value of a loss in excess of VaR, but even ES tells
us nothing else about the distribution of ‘tail losses’ other than
its expected value.”
Dowd (2005) Measuring Market Risk, Wiley.

A FINANCIAL TIMES COMPANY 93

STRESS TESTING

I VaR and ES are probabilistic approaches to risk measurement -

typically under ‘normal’ market conditions
I Scenario analysis and stress testing represent ‘what-if’ analysis, e.g.
what if Black Monday (stock market crash, October 1987) happens
again?
I Probabilities of scenarios are generally not known
I Reverse stress-tests require a firm to consider scenarios that would
render its business model unviable, thereby identifying potential
business vulnerabilities.
I Reverse stress-testing starts from an outcome of business failure and
identifies circumstances where this might occur.

A FINANCIAL TIMES COMPANY 94

 The PRM Handbook – III.A.4 Stress Testing

portfolio, the frequency with which it is traded, the liquidity of the instruments in the portfolio,

APPROACHES TO STRESS TESTING

the volatility of the markets in which the instruments are traded, and the strategies employed. 4

Table III.A.4.1: Typology of stress tests

Approach Description Pros Cons

Historical Replay crisis event It actually happened that Proxy shocks may be numerous
scenarios way
No probabilistic interpretation

No guarantee of ‘worst case’

Hypothetical 1. Covariance matrix 1. Relatively easy 1. Empirical support mixed

scenarios
2. Create event 2. Very flexible 2. No guarantee of ‘worst case’

3. Sensitivity analysis 3. Can be detailed 3. Limited risk information

Algorithmic 1. Factor push 1. Minimal qualitative 1. No guarantee of ‘worst case’

elements
2. Maximum loss 1. Ignores correlations
2. Identifies ‘worst case’ in
feasible set (maybe) 2. Assumes data from normal periods
are relevant

2. Computationally intensive

Most
Source: Schachter, ofHandbook,
PRM the regulatory
2004 attention has focused on stress testing at the portfolio level. For the
regulators it is the aggregated impact of stressed market environments that poses risks that
interest them. For some time international organisations have pursued the idea of aggregating
A FINANCIAL TIMES COMPANY 95
 Scenario data can be obtained from historical data sets or economic scenario generator
engines; it can also be provided by a third party. For certain types of stress tests, such
as SCAP and CCAR, the regulator will provide some of the data. In cases where all
the risk factors are not present, models can be used to complete the required data
COMMON HISTORICAL SCENARIOS
for analysis. For reference, documented below (Table 2) is a collection of recent crises
starting with the 1987 Black Monday and ending in with the recent 2007-2011 global
financial crisis.

Russell 1000® Total Market Index

Year Crisis 7,000

1987 Black Monday

6,000
1989-91 US Savings and Loan Crisis
1990 Japanese Asset Bubble, 5,000

Swedish and Finnish Banking Crisis

4,000
1992-93 Black Wednesday

(Index)
1994-95 Mexican Peso Crisis 3,000

1997-98 Asian Financial Crisis

2,000
1998 Russian Ruble Crisis
1,000
2001 Argentine Crisis
2001 Dot-Com Bubble Bursting 0
1980 1985 1990 1995 2000 2005 2010 2015
2007-11 Global Financial Crisis Source: Russell Investments
Shaded areas indicate US recessions - 2015 [Link]

Table 2: Wikipedia list of recent financial crises.

Source SAS, Firmwide Stress Testing
Once the scenario data is obtained, it can be fed into appropriate models to determine
additional risk factors, forecasts, risk metrics calculations and balance sheet/income
statement projections.
A FINANCIAL TIMES COMPANY 96
HISTORICAL SCENARIO ANALYSIS EXAMPLE
Price history
Credit Spread history
RMBS SubPrime Credit Spreads: Spain USD 5Y
Financial crisis, 2007 - 2011
DRAFT DRAFT

AA Greatest Increase: 193

bps on 6 May 2010
400
105 Greatest Decrease: (105)
350 bps on 13 Aug 2009
90 Greatest negative
300
move over 12 mo.
75 period: 90 percent on 250

60 07/23/2008
200

45 150

30 100
Price history Credit Spread history
15 50
CMBX Credit Spreads: Italy USD 5Y
0 0
Jul-07

Jan-08

Jul-08

Jan-09

Jul-09

Jan-10

Jul-10

Jan-11
DRAFT

Feb-05

Aug-05

Feb-06

Aug-06

Feb-07

Aug-07

Feb-08

Aug-08

Feb-09

Aug-09

Feb-10

Aug-10

Feb-11
DRAFT

AA, 3yr, Price Greatest Increase: 170

Greatest negative bps on 5 Dec 2008
move over 12 mo. 300
110
period: 80 percent on Greatest Decrease: (122)
100 Source: Bloomberg 5Y Generic CDS Spreads
bps on 1 Sep 2009
4/14/2009 250
Source:90Bloomberg Pricing Source; [Link].07-2 [2006 and 2007 vintage AA SubPrime deals] 11
80
200
70
7
60
150
50
40 100
30
20 50
10
0 0
Feb-07

Aug-07

Feb-08

Aug-08

Feb-09

Aug-09

Feb-10

Aug-10

Feb-11

Feb-05

Aug-05

Feb-06

Aug-06

Feb-07

Aug-07

Feb-08

Aug-08

Feb-09

Aug-09

Feb-10

Aug-10

Feb-11
Top: RMBS Index Levels ([Link].07-2), Bottom: CMBS Index Top: 5Y CDS Spain, Bottom: 5Y CDS Italy
Source: Bloomberg Pricing Source; [Link].2 [2005 and 2006 vintage 3yr AAA deals]
Levels ([Link].2)
Source: Bloomberg 5Y Generic CDS Spreads
4
A FINANCIAL TIMES COMPANY 9 97
BIS REQUIREMENTS

I 1996 Market Risk Amendment to 1988 Basel Accord (Basel I)

requires banks using internal models for market risk capital to use
stress testing to identify events (scenarios) that could have a
significant impact on banks’ capitalization
I Basel II (1999) requires banks to have sufficient capital to cover the
results of stress tests required by the Market Risk Amendment
(1996) and specifies additional stress tests relating to a number of
scenarios

A FINANCIAL TIMES COMPANY 98

REGULATORY STRESS TESTS

I Supervisory Capital Assessment

Program (SCAP)  
 

I February to May 7, 2009

I Stress test to determine if 19
largest bank holding
companies (BHCs) had
sufficient capital to withstand
the financial (subprime) crisis  
 

I Two scenarios over two

years: baseline and adverse
I 10 of 19 BHCs required
additional capital
 

7 
 

A FINANCIAL TIMES COMPANY 99

REGULATORY STRESS TESTS

I Dodd-Frank Stress Test (DFAST)

I Dodd-Frank Act requires Federal Reserve to conduct annual stress
tests of large BHCs and all nonbank financial companies designated
by the FSOC for Fed supervision
I Fed provides hypothetical scenarios - base case, adverse and severely
adverse
I DFAST assumes that firms maintain current dividends and will not
execute share buybacks apart from preventing dilution related to
employee compensation during the 9 quarters tested

A FINANCIAL TIMES COMPANY 100

a negative reaction from the corporates market. Looking at the banking sector from a 5-years-later
vantage point, we see the industry as stronger as a result of the additional Fed oversight that did
not exist before 2009. As part of the broader regulatory changes, banks are more limited now in
terms of their risk-taking.
REGULATORY STRESS TESTS
QUICK STRESS TEST RECAPS
t I
DFAST: The DFAST assesses a firm’s capital ad- The Fed Looks at 5 Capital Ratios
Comprehensive
equacy under base case, Capital
adverse,Analysis
and severely
and Review (CCAR)
adverse scenarios for 4Q 2013 to 4Q 2015. 2015
The DFAST assumes each firm will main- Minimum CCAR Ratio
thatscenarios
I Uses DFAST Minimum
tain current dividends and will not complete
I Takes into consideration
share buybacks apart from preventing dilu- Tier 1 common 5%
firms’
tion related capital plans,
to employee such asduring
compensation
dividends,
Common Equity Tier 1 5%
the 9 quarters [Link]
Of the 30and
firms tested,
only Zionslarge Bancorp’s (Ba1/BBB-/BBB-) Tier
acquisitions Tier 1 Risk-based Capital 6%
1 common the 5%ofthreshold dur- Total Risk-based Capital
fell below review
I Qualitative 8%
ing severely adverse governance,
corporate conditions. Other firms’ Tier 1 Leverage 4%
capital metrics
internalcame in belowpotential
controls, the median, but
above therisks
minimum levels.
related For a quick
to capital plans,recap
Source: Janney FISR; Federal Reserve
of DFAST, see our March 24 Fixed Income Weekly.
and “robustness” of the
t CCAR: The CCAR planning
capital also assesses a firm’s capital adequacy under the three scenarios. Unlike the
processes.
DFAST, the CCAR takes into consideration firms’ capital plans, such as shareholder remunera-
tions and large acquisitions. The CCAR reviews the BHCs from both quantitative and qualita-
tive perspectives. The quantitative is straightforward, based on capital cushion metrics during
the 9 quarters
A FINANCIAL TIMES tested.
COMPANYThe qualitative is more subjective, with the Fed analyzing firms’ corpo-101
 CMA 10.6
FITB 9.6
HUN 10.3
Regional KEY 11.3
Domestic P
MT 9.8
2015 REGULATORY STRESS TESTS PNC 11.0
•

RF 11.8
SUN 9.6
Tier 1 common ratio 3Q 2014 actual to minimum RWA
for CCAR and DFAST results by BHC USB 9.5
• Aggregate risk-weighted assets (RWAs) under the A
ZION
current general approach (Basel I) are projected to11.9
0 2 4 6 8 10 12 14 36 38
increase 4% over the stress horizon through business
All BHC 11.9 and risk profile changes, and increase a further 9% as a
BNY
result of the regime change to the Basel III standardized 13.9
approach. Significant
DBTC drivers of RWA change are the 36.6
BAC 11.3 •
BaselCustody
III treatments of off balance sheet exposures by
CITI 13.4 which custody and NTuniversal and investment banks are 12.8
GS* 15.2 disproportionally impacted.
STT 13.9
Universal
Bank + IB JPM* 10.9
4Q 2016
MS* 15.0
Credit AXP
Current 4Q 2016 13.2
Risk-weighted general Basel III Regime
WF 10.8
assets in Card
Actual Q3DFS
approach standardized change % 14.8
$ billions 2014 (Basel I) approach difference Difference
All BHCs 8790 9103 9948 845 9%
ALLY 9.7
5907BBVA
Universal 11.0
6101 6808 707 12%
BBT 10.5 bank & IB
Regional BMO 11.5
COF 12.7 1787 1878 1927 49 3%
domestic
CMA 10.6 HSBC 14.0
Regional301
Custody 309 364 54 18%
FBO
FITB 9.6 Credit card 199 MUFG204 212 9 12.7
4%

HUN 10.3 Regional FBO 596 RBS612 638 26 4%

12.9
Regional KEY 11.3
Domestic SAN 11.0
PPNR
MT 9.8
• Aggregate 9Q net revenueCCAR before provisions for DFAST
Minimum loan and
Minimum
PNC 11.0
lease losses is projected to be $310b, or 2.1% of assets,
RF 11.8 3Q 2014 Actual
compared to $316b, or 2.3%, in DFAST 2014.
*Denotes three largest banks that adjust their originally proposed capital actions
SUN 9.6 Aggregate
downward
Results by BHC type5 results
USB 9.5 % of
Source: Ernst & Young, 2015Credit
Average Universal Regional
CCAR/DFAST
Regional
Results 5

ZION 11.9 Custody 2015 2014

assets bank & IB domestic card FBO Trend
PPNR 1.4 3.3 2.7 15.2 1.2 2.1 2.3

BNY 13.9
2015 CCAR/DFAST results

DBTC 36.6
A Custody
FINANCIAL TIMES COMPANY • The combination of high losses and projected stressed 102
 tests conducted during the crisis include the US Supervisory
non-bank financial companies, as required under the
Capital Assessment Program (SCAP) in 2009(1) and the
Dodd-Frank Act.
EU-wide banking sector stress tests in 2009–11.(2) A number of
countries have also put in place frameworks for regular stress The majority of countries conduct stress tests on an annual
testing of their respective banking systems. In seeking to draw basis: Most countries see benefits in conducting annual
lessons for the design of the UK framework, this box reviews stress-testing exercises that are in line with banks’ regular
the international experience with stress testing. Table 1 capital planning cycle. Other countries conduct semi-annual
summarises the approach taken by selected jurisdictions. stress tests, the results of which are published in Financial
REGULATORY STRESS TESTING Stability Reports.

Table 1 Stress-testing in selected jurisdictions

European Banking Hong Kong(a) Ireland(a) Japan(a) Sweden(a) United States (CCAR)
Authority (EU)
Coverage Largest EU banks. All retail banks.(b)(c) Largest Irish banks. Eleven major banks and Four largest Swedish Bank holding
105 regional banks for banks. companies (BHC) with
Bank of Japan (BoJ) assets greater than
macroprudential stress US$50 billion.(d)
tests. All banks for
Financial Services
Agency (FSA) stress
tests.
Frequency Annual between 2009 Semi-annual.(b) As per Economic Semi-annual. Semi-annual Riksbank Annual regulator-led and
and 2011. Next exercise Adjustment Programme and annual semi-annual bank-led.
expected in 2014. commitments. Finansinspektionen (FI)
stress tests.
Data requirements Private data. Public and private data. Private loan-level data. Public and private data. Public data. Private loan and
account-level data.
Modelling approach:
(1) Banks’ own models.
(2) Granular (1)
(1) (1) (1) (1)
microprudential (1) (2)
models. (2) (3) (2) (2)
(3)
(3) System-wide
macroprudential
models.
Scenarios used Regulatory baseline and For (1): Banks’ own Regulatory baseline and Baseline and two stress One stress scenario in Six scenarios in total:
one regulatory stress scenarios and a common one stress scenario. scenarios for BoJ Riksbank stress test. FI regulatory baseline;
scenario. regulatory scenario. system-wide does not have an explicit adverse; and severely
macroprudential stress stress scenario; adverse as well as bank
For (2): Two regulatory tests. Several exogenous increase in equivalents of those.
stress scenarios. supervisory scenarios for loan losses is assumed.
For (3): Multiple FSA microprudential
regulatory stress stress tests.
scenarios.
Disclosure Granular Only results of Granular Only results of Individual-institution Individual-institution
individual-institution system-wide individual-institution system-wide disclosures and disclosures.
disclosures. macroprudential stress disclosures. macroprudential stress system-wide estimates
Supplemented in 2011 tests are disclosed. tests are disclosed. of credit losses.
by detailed disclosures of Disclosures made on Disclosures made on
sovereign and loan book aggregate basis. aggregate basis.
exposures.
Use of outputs Stressed capital ratios Input into Pillar 2 Input into Pillar 2 Input into Pillar 2 Input into Pillar 2 Results feed into
relative to hurdle rate assessment. Results assessment. In 2011, assessment and to assessment. Results assessment of banks’
implied a capital used to inform used to inform the inform FSA micro and used to inform capital plans, which are
shortfall which had to be supervisory strategy and required capital macroprudential supervisory strategy and subject to approval by
met. Banks and as one of the inputs to injections into Irish supervisory policy. as one of the inputs to the Federal Reserve.
supervisors required to inform micro and banks and their BoJ system-wide inform micro and
present strategy for macroprudential policy. deleveraging plans under macroprudential macroprudential policy.
meeting the shortfall. the Financial Measures stress-test results used
Programme. as part of risk
surveillance.

(a) Under Basel II, all banks are required to run their own stress tests as part of the Pillar 2 process. The information in this column relates to stress-testing practices over and above those.
(b) Coverage and frequency vary across the different stress tests conducted by the Hong Kong authorities. This information refers to system-wide macroprudential stress tests only. For instance, microprudential stress tests are
conducted on a quarterly basis and cover all locally incorporated banks.
(c) Retail banks comprise all the locally incorporated banks plus a number of the larger foreign banks with similar operations (ie banks that operate as branches in Hong Kong and are active in retail banking).
(d) Separate regime exists for BHCs with assets of US$10 billion–US$50 billion.

A FINANCIAL TIMES COMPANY 103

A FINANCIAL TIMES COMPANY 104