Identifying At-Risk Names in Your Credit PortfolioDavid T. Hamilton, PhD, Managing Director, Capital Markets Research Group

Irina Makarova, Assistant Director, Capital Markets Research Group

October 2, 2013

Originally presented as a part of a Moody’s Analytics webinar

2

“Yes, risk taking is inherently failure-prone. Otherwise, it would be called sure-thing-taking.”

− Jim McMahon

3

Strategies for identifying at-risk names in your portfolio1

4

Developing a Monitoring and Early Warning Toolkit

Public Expected Default Frequency (EDF) measures can be utilized in four ways to more effectively monitor risk and signal early warning:

1. EDF level

– EDFs have been shown to be powerful at rank ordering risk

– How can we choose the right EDF warning trigger level?

2. Relative EDF level

– Peer group or industry sector comparisons add incremental predictive power to detect credit risk

3. EDF change

– EDF measures exhibit momentum

4. EDF relative change

– Like relative level, comparisons to a peer or industry group add incremental predictive power

5

Challenges to setting up an effective monitoring and early warning process» System must be set up to detect outcome(s) of

interest/importance

– Default

– Rating change

– Spread movement

» Filters are likely to be very sensitive to time period, peer group, time horizon, etc.

» Triggers should also be a function of the economic importance of the exposure

» Process can only filter results, cannot dictate actions

6

Effective credit risk monitoring using EDF measures

» Monitoring and early warning are problems of classification: which firms in a portfolio should be considered relatively more risky, and therefore merit deeper investigation?

» Estimating the benefits and costs of classification may be difficult, imprecise, or impossible

» We may therefore attempt to maximize predictive accuracy of classification instead

7

Assessing risk classification accuracy

Actual Outcome

Low Risk High Risk

Predicted by Credit Measure

Low Risk True Negative False Negative

High Risk False Positive True Positive

A credit risk measure, like an EDF or a rating, makes a prediction that will turn out to be correct or not. There are thus four types of classification error.

FNTP

TPTPR

FPTN

FPFPR

We can also define two useful measures of accuracy from this table:

The overall predictive power of any risk measure is determined by both the TPR and FPR

8

Developing an optimal early warning signal

We can calibrate an optimal early warning trigger level by measuring the TPR and FPR for a given trigger value.

Finding the optimal trigger levels entails three steps:

1. Choose a candidate EDF trigger level

2. Measure the TPR and FPR associated with the EDF trigger level

3. The optimal EDF trigger level is the one that maximizes (TPR – FPR)

t *Percentile Rank

Median25th Percentile

75th Percentile

2.673% 82.701% 0.251% 0.084% 1.224%

Optimal EDF early trigger level, global public companies, 1997-2012

9

Optimal EDF triggers for select industry sectors

9

Sectort *

(Pctile)

Banks0.281%

(75.464%)

Consumer Products0.247%

(65.045%)

High Tech13.715%

(95.489%)

Transportation4.344%

(83.923%)

Global public companies, 1997-2011

10

Static EDF early warning triggers are not optimal

Box plots of EDF measures for US Firms, January 2008 vs. January 2009

0

10

20

30

40

ED

F %

Jan 2008 Jan 2009

Review Trigger Level

11

The optimal trigger will vary significantly over the cycle

12

Firms whose EDF measures are high relative to their industry sectors experience higher default rates

Relative EDF Level

Quartile 1 2 3 4

1 0.02% 0.10% 0.35% 0.72%

EDF 2 0.05% 0.31% 0.45% 1.71%

Level 3 0.00% 0.29% 0.77% 2.78%

4 NA 2.64% 2.68% 6.65%

One-year default rates conditioned on EDF quartile and relative EDF quartile

Default risk increases when EDFs are relatively high

Default risk rises with EDF level

13

Negative EDF momentum signals higher default riskOne-year default rates conditioned on EDF momentum

14

Default rates are sensitive to EDF momentum vs. sector

One-year default rates conditioned on EDF decile and EDF change vs. sector change

1 2 3 4 5 6 7 8 9 10 ALL1 0.05% 0.03% 0.02% 0.00% 0.00% 0.01% 0.03% 0.00% 0.00% 0.00% 0.02%2 0.10% 0.05% 0.06% 0.06% 0.00% 0.00% 0.02% 0.07% 0.11% 0.27% 0.05%3 0.10% 0.06% 0.01% 0.03% 0.01% 0.03% 0.07% 0.06% 0.03% 0.18% 0.05%4 0.28% 0.12% 0.17% 0.15% 0.09% 0.10% 0.08% 0.09% 0.17% 0.30% 0.15%5 0.32% 0.23% 0.24% 0.32% 0.22% 0.24% 0.21% 0.27% 0.22% 0.46% 0.27%6 0.62% 0.44% 0.45% 0.34% 0.44% 0.56% 0.44% 0.72% 0.51% 0.97% 0.55%7 0.71% 0.56% 0.66% 0.80% 0.64% 0.72% 0.73% 1.06% 1.18% 1.63% 0.89%8 1.01% 1.01% 1.19% 1.25% 1.27% 1.44% 1.58% 1.65% 2.05% 3.10% 1.68%9 3.14% 2.22% 4.83% 5.16% 5.25% 4.34% 4.87% 5.75% 6.37% 8.39% 5.60%10 6.43% 4.68% 5.76% 7.70% 7.70% 6.96% 7.67% 9.31% 9.99% 13.70% 8.94%All 0.66% 0.63% 1.08% 1.73% 1.73% 1.83% 2.24% 2.92% 3.13% 5.96% 2.16%

Firm

EDF

Leve

l

EDF Change Relative to Industry Peer Group Change

Default risk increases with poor performance vs. industry

Default risk rises with EDF level

15

Developing a monitoring and early warning playbook

Our research suggests a general approach to effective early warning using EDF measures:

1. Calibrate an appropriate EDF trigger level for your portfolio. Names in excess of the trigger are statistically more likely to default and should be analyzed further

2. For names that exceed the trigger EDF, check their EDF levels relative to their industry sectors. Names in the 75th or worse percentile are of particular concern.

3. Also check the EDF momentum for these names. Firms whose EDFs have worsened in the past are likely to continue to worsen, and are more likely to default than other firms with the same EDF level

4. Last, when an firm’s EDF momentum is underperforming its industry sector, it is more likely to default than other firms with the same EDF level

16

Strategies in practice: Case study2

17

JC Penney’s EDF metric has increased sharply over the past year, but how concerned should we be?

» Coming out of the financial crisis and recession, JC Penney’s EDF measure fell from 3.5% to 0.5% in 1Q2012

» Since 1Q2012, its EDF measure has increased sharply, reaching an all-time high of 7.9% in September 2013

» The optimal EDF trigger for the US department stores group is 0.9% on average

» Compared to Sears, which one is riskier?

Sep-0

8

Dec-0

8

Mar

-09

Jun-

09

Sep-0

9

Nov-0

9

Feb-1

0

May

-10

Aug-1

0

Oct

-10

Jan-

11

Apr-1

1

Jul-1

1

Sep-1

1

Dec-1

1

Mar

-12

Jun-

12

Aug-1

2

Nov-1

2

Feb-1

3

May

-13

Jul-1

30

1

2

3

4

5

6

7

8

STRATEGY 1: EDF LEVEL

Sep-0

8

Dec-0

8

Mar

-09

Jun-

09

Sep-0

9

Nov-0

9

Feb-1

0

May

-10

Aug-1

0

Oct

-10

Jan-

11

Apr-1

1

Jul-1

1

Sep-1

1

Dec-1

1

Mar

-12

Jun-

12

Aug-1

2

Nov-1

2

Feb-1

3

May

-13

Jul-1

30

1

2

3

4

5

6

7

8

Sep-0

8

Dec-0

8

Mar

-09

Jun-

09

Sep-0

9

Nov-0

9

Feb-1

0

May

-10

Aug-1

0

Oct

-10

Jan-

11

Apr-1

1

Jul-1

1

Sep-1

1

Dec-1

1

Mar

-12

Jun-

12

Aug-1

2

Nov-1

2

Feb-1

3

May

-13

Jul-1

30

1

2

3

4

5

6

7

8

Sears JC Penney

Source: CreditEdge

1-Y

ea

r E

DF

(%

)

18

Is a 6% EDF high? How risky is JC Penney?

» Mapping PDs into ratings can be useful rules of thumb, to grasp levels and make relative comparisons

» Here, we map EDFs into ratings using historical default rates

» JC Penney’s current 6% EDF maps to a B3 rating, suggesting a relatively high risk of default

» Sears’ EDF has improved to 1.5% as of Sept., equivalent to a Ba3 rating Se

p08

Dec08

Mar09

Jun0

9

Sep0

9

Dec09

Mar10

Jun1

0

Sep1

0

Dec10

Mar11

Jun1

1

Sep1

1

Dec11

Mar12

Jun1

2

Sep1

2

Dec12

Mar13

Jun1

3

Sep1

30

1

2

3

4

5

6

7

8

1-Y

ear

ED

F (

%)

Caa1

B3

B2

B1

Ba3

Ba2

Baa3Ba1

STRATEGY 1: EDF LEVEL

Sep0

8

Dec08

Mar09

Jun0

9

Sep0

9

Dec09

Mar10

Jun1

0

Sep1

0

Dec10

Mar11

Jun1

1

Sep1

1

Dec11

Mar12

Jun1

2

Sep1

2

Dec12

Mar13

Jun1

3

Sep1

30

1

2

3

4

5

6

7

8

1-Y

ear

ED

F (

%)

Caa1

B3

B2

B1

Ba3

Ba2

Baa3Ba1

Sears JC Penney

Source: CreditEdge

19

As JC Penney’s EDF has risen, its relative EDF level shows that it is risky for its industry sector

» Between 2008 and 2012, JCP’s EDF has either outperformed or tracked the median EDF of its industry sector

» Over the past year its EDF has deteriorated in both absolute and relative terms, and is now trending in the 75th percentile of its industry sector

» Firms whose EDF levels are in the 75th percentile of their sectors are 4.5X more likely to default than those below the 75th percentile

Sep0

8

Jan0

9

May

09

Sep0

9

Jan1

0

May

10

Sep1

0

Jan1

1

May

11

Sep1

1

Jan1

2

May

12

Sep1

2

Jan1

3

May

13

Sep1

30.0

0.1

1.0

10.0

100.0

US Department Stores Group EDF

1-Y

ear

ED

F (

%,

log s

cale

)

75%

Median

25%

STRATEGY 2: RELATIVE LEVEL

Source: CreditEdge

20

The sharp jump in JCP’s EDF suggests that risk is more likely to increase than to decrease in the future

STRATEGY 3: EDF CHANGE

» JCP’s EDF has increased at an accelerating rate, doubling roughly every six months since the start of 2012

» Although JC Penney’s EDF has fallen recently, its EDF is more likely to continue to increase than to decrease over the next year

» Sears is an interesting contrast: its EDF has been higher longer, but has basically moved sideways since its 2011 spike

Jan-

11

Mar

-11

Apr-1

1

Jun-

11

Jul-1

1

Sep-1

1

Oct

-11

Dec-1

1

Jan-

12

Mar

-12

Apr-1

2

May

-12

Jul-1

2

Aug-1

2

Oct

-12

Nov-1

2

Jan-

13

Feb-1

3

Apr-1

3

May

-13

Jun-

13

Aug-1

3

Sep-1

30

1

2

3

4

5

6

7

8

1-Y

ear

ED

F (

%)

Source: CreditEdge

JC Penney

Sears

21

The deterioration in JCP’s EDF was also in sharp contrast with general improvement in its sector’s EDFs

» Since 2012, the whole distribution of EDF levels for the US department stores sector has shifted lower

» JCP’s EDF has underperformed its sector since 2Q2012

» The change in its EDF relative to its sector places it in the 90th percentile, making it twice as likely to default than companies at the median

STRATEGY 4: RELATIVE CHANGE

Jan1

1

May

11

Oct11

Feb1

2

Jun1

2

Oct12

Feb1

3

Jun1

30.0

0.1

1.0

10.0

US Department Stores Group

1-Y

ear

ED

F (

%,

log s

cale

)

75%

Median

25%

Source: CreditEdge

22

Case study key take-aways

JC Penney Sears

EDF Level 6% 1.5%

Relative Level 75th pctile 70th pctile

Momentum +1,227% -17%

Relative Momentum 90th pctile 55th pctile

» The four EDF-based early warning strategies add powerful tools for monitoring credit risk; each adds incremental information over EDF level alone

» The strategies can be implemented as part of a formal early warning process or used in name-by-name judgmental analysis

» So, who is riskier? JC Penney or Sears?

23

Conclusion3

24

Summary

» Using EDF measures as the key metric, we showed how to develop an early warning “playbook” using four key monitoring strategies

» These strategies have the advantage of being intuitive and relatively easy to calibrate and implement with available data

» In the JC Penney case study we applied these strategies to show how to differentiate between two companies that, by appearance, seem equally risky

» The strategies can be used a basis for a synthetic early warning signal, or as part of an automated early warning process

25

Putting It All Together: A Traffic Light Approach

EDF Level

ED

F R

elat

ive

Leve

l EDF Change

26

Q & A

27

Contact

Join the conversation on LinkedIn: Credit Edge – Best Practices in Credit Risk

David T. HamiltonManaging DirectorQuantitative Credit ResearchCapital Markets Research Group

Moody’s Analytics7 World Trade CenterNew York, NY 10007

+1 212 553-1695 tel+1 212 298-6934 faxdavid.hamilton@moodys.com

Irina MakarovaAssistant DirectorQuantitative Credit ResearchCapital Markets Research Group

Moody’s Analytics7 World Trade CenterNew York, NY 10007

+1 212 553-4307 tel+1 212 298-6934 faxirina.makarova@moodys.com

28

© 2013 Moody’s Analytics, Inc. and/or its licensors and affiliates (collectively, “MOODY’S”). All rights reserved. ALL INFORMATION CONTAINED HEREIN IS PROTECTED BY COPYRIGHT LAW AND NONE OF SUCH INFORMATION MAY BE COPIED OR OTHERWISE REPRODUCED, REPACKAGED, FURTHER TRANSMITTED, TRANSFERRED, DISSEMINATED, REDISTRIBUTED OR RESOLD, OR STORED FOR SUBSEQUENT USE FOR ANY SUCH PURPOSE, IN WHOLE OR IN PART, IN ANY FORM OR MANNER OR BY ANY MEANS WHATSOEVER, BY ANY PERSON WITHOUT MOODY’S PRIOR WRITTEN CONSENT. All information contained herein is obtained by MOODY’S from sources believed by it to be accurate and reliable. Because of the possibility of human or mechanical error as well as other factors, however, all information contained herein is provided “AS IS” without warranty of any kind. Under no circumstances shall MOODY’S have any liability to any person or entity for (a) any loss or damage in whole or in part caused by, resulting from, or relating to, any error (negligent or otherwise) or other circumstance or contingency within or outside the control of MOODY’S or any of its directors, officers, employees or agents in connection with the procurement, collection, compilation, analysis, interpretation, communication, publication or delivery of any such information, or (b) any direct, indirect, special, consequential, compensatory or incidental damages whatsoever (including without limitation, lost profits), even if MOODY’S is advised in advance of the possibility of such damages, resulting from the use of or inability to use, any such information. The credit ratings, financial reporting analysis, projections, and other observations, if any, constituting part of the information contained herein are, and must be construed solely as, statements of opinion and not statements of fact or recommendations to purchase, sell or hold any securities. NO WARRANTY, EXPRESS OR IMPLIED, AS TO THE ACCURACY, TIMELINESS, COMPLETENESS, MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE OF ANY SUCH RATING OR OTHER OPINION OR INFORMATION IS GIVEN OR MADE BY MOODY’S IN ANY FORM OR MANNER WHATSOEVER. Each rating or other opinion must be weighed solely as one factor in any investment decision made by or on behalf of any user of the information contained herein, and each such user must accordingly make its own study and evaluation of each security and of each issuer and guarantor of, and each provider of credit support for, each security that it may consider purchasing, holding, or selling.