Supervisory Formula Approach and Internal Assessment Approach

The risk weight of securitization positions subject to the supervisory formula approach (“SFA”) are determined based on a formula which takes as input the capital requirement of the securitized portfolio and the seniority of the securitization position in the waterfall, amongst others. When applying the SFA, we estimate the risk parameters PD and LGD for the assets included in the securitized portfolio, by using internally developed rating systems approved for such assets. We continue to develop new rating systems for homogenous pools of assets to be applied to assets that have not been originated by us. The rating systems are based on historical default and loss information from comparable assets. The risk parameters PD and LGD are derived on risk pool level.

Approximately 72 % of the total banking book securitization positions are subject to the SFA. This approach is predominantly used to rate positions backed by corporate loans, auto-related receivables and commercial real estates.

For unrated IRBA-securitization positions which are related to ABCP programs and which are not asset backed commercial paper, the risk weight is calculated based on the internal assessment approach (“IAA”). Apart from using this concept for regulatory purposes, the internal rating is used for expected loss and economic capital calculations and plays a significant role in the credit decision and monitoring process.

We have received approval from BaFin to apply the IAA to approximately 71 % of our ABCP conduit securitization exposure.

Asset classes subject to IAA are governed by a specific and detailed set of rules per asset class. These asset class write-ups (“ACW”) have been established by Credit Risk Management, Risk Analytics and Living Wills and the Front Office. They are reviewed and approved in a formal internal process, and subject to an at least annual review. For BaFin approved asset classes, the ACW require re-approval by the regulator in case of significant changes during the review process.

BaFin approval for IAA has been received for currently 14 different asset classes in both consumer and commercial assets. The stress factors are different per asset class and rating level; they are established based on criteria set by the “dominant” external rating agency which forms the basis of the internal qualitative and quantitative rating analysis. The stress factor multiples generally indicate how much credit enhancement is required to obtain a specific rating compared to the level of pool expected loss.

The following tables summarize (a) the stress factor multiples per rating level, or (b) key stress testing methodology for those without defined Stress Factor Multiples, based on the methodology published by the respective dominant rating agencies:

Stress Factor Multiples per Rating Level by dominant Rating Agencies

Asset Class

Auto Loans

CDO

Comm. Lease & Loan

Consumer Loans

Credit Cards

Trade Receivable

Dominant Rating Agency

S&P

Moody’s

S&P

S&P

S&P

S&P

AAA

3.75–5

1.95

5

4–5

3–6.6

2.5

AA

3–4

1.8–1.76

4

3–4

2.5–5

2.25

A

2–3

1.73–1.69

3

2–3

2–3.75

2

BBB

1.75–2

1.67–1.63

2

1.5–2

1.5–2.5

1.5

BB

1.5–1.75

1.5–1.2

 

1–1.5

1.25–1.5

 

Summary of Rating agency Stress Factor Methodologies without defined Stress Factor Multiples

Asset Class

Capital Calls

MBS Servicer Advances

Market Values CDO

RMBS Australia

RMBS Europe

RMBS US

Structured Settlements

Student Loans FFELP

Information based on methodology published by the respective Dominant Rating Agencies, which may be amended from time to time.

Dominant Rating Agency

Moody’s

Moody’s

S&P

S&P

S&P

S&P

S&P

Moody’s

Comment

Methodology relies on con­servative as­sumptions regarding debtor ratings and recovery rates; supported by conservative correlation criteria

Methodology applies rating-specific stressed recovery curves for each type of servicer ad­vances. Cash flows are run under multiple interest rate stress scenarios

Methodology of both S&P and Moody’s is using Advance rates instead of Stress Factor Multiples, which are available on their respective websites

Methodology uses default & loss assump­tions per rating level, based on benchmark pools with adjustments appropriate for the respective pool being compared

Methodology uses default & loss assump­tions per rating level, based on benchmark pools with adjustments appropriate for the respective pool being compared

Stress­testing by applying S&P default and loss assumptions per rating level on each individual loan in the pool

Generating a probability distribution of potential default rates at each rating level for the portfolio using industry-specific recovery rates. Additional stress tests regarding Largest Obligor and Largest Industry Defaults

Applying rating-level specific stresses inclu­ding defined cumulative default rates, voluntary pre-payment rates, servicer reject rates and bor­rower benefit rates

The underlying cash flow models per asset class are also subject to the regular review process described above. For securitizations in these asset classes we utilize external credit assessment institutions, namely Standard & Poor’s and Moody’s as outlined in the tables above.


Key figures comparison

Compare key figures of the past years. more